Future Drivers of Growth in RWANDA Innovation, Integration, Agglomeration, and Competition Government of Rwanda Future Drivers of Growth in Rwanda Future Drivers of Growth in Rwanda Innovation, Integration, Agglomeration, and Competition © 2020 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved 1 2 3 4 23 22 21 20 This work is a product of the staff of The World Bank with external contributions. The findings, interpreta- tions, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http:// creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution —Please cite the work as follows: World Bank and Government of Rwanda. 2020. Future Drivers of Growth in Rwanda: Innovation, Integration, Agglomeration, and Competition. Washington, DC: World Bank. doi:10.1596/978-1-4648-1280-4. License: Creative Commons Attribution CC BY 3.0 IGO Translations —If you create a translation of this work, please add the following disclaimer along with the attribu- tion: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations —If you create an adaptation of this work, please add the following disclaimer along with the attri- bution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third-party-owned individual com- ponent or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; e-mail: pubrights@worldbank.org. ISBN (paper): 978-1-4648-1280-4 ISBN (electronic): 978-1-1236-1 DOI: 10.1596/978-1-4648-1280-4 Cover and interior design: Bill Pragluski, Critical Stages, LLC Cover illustration: © Municipality of Kigali. Used with permission; further permission required for reuse. Library of Congress Cataloging-in-Publication Data has been requested. Contents Foreword. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv Executive Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxxi Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 A Strong Start despite Initial Conditions and Emerging Concerns . . . . . . . . . . . . . . . . . . . . . 2 Future Aspirations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Opportunities and Risks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Main Elements of Rwanda’s Growth Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Six Reform Areas of Importance for Rwanda . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 1  Human Capital and Innovation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Why Is Human Capital Important for Growth?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 A Framework for Thinking about Human Capital. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Stunting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Fertility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Basic Education. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Skills Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Tertiary Education and Innovation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 2 Transformation through Trade: Using Exports and Regional Integration to Drive Future Growth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Setting the Stage: Policy, Performance, and Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 FUTURE DRIVERS OF GROWTH IN RWANDA v vi FUTURE DRIVERS OF GROWTH IN RWANDA Envisioning the Future. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Priority 1: Harness the EAC as a Platform for Transformation . . . . . . . . . . . . . . . . . . . . . 107 Priority 2: Improve Trade Connectivity by Lowering Transport Costs within the Region and with the World . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 Priority 3: Increase Service Sector Productivity, Both as a Critical Input to Other Priority Sectors and as a Source of Exports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 Priority 4: Stimulate Foreign and Domestic Investment into Tradable Sectors by Using Selective and Performance-Driven Incentives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124 Priority 5: Accelerate Industrialization through Diversification, Value Addition, and Quality Upgrading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130 3  Faster Urbanization, Greater Agglomeration. . . . . . . . . . . . . . . . . . . . . . . 135 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 The Promise of Urbanization and Agglomeration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 Constraints to Successful Urbanization in Rwanda . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144 Priorities for Faster Urbanization, Greater Agglomeration. . . . . . . . . . . . . . . . . . . . . . . . . 151 Annex 3A  The Rwanda Spatial CGE Model and Scenarios . . . . . . . . . . . . . . . . . . . . . . 167 Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 4 Competitiveness and Enterprise Development for Innovation-Led Growth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171 Current State of Rwanda’s Enterprise Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 Constraints Faced by Rwanda’s Enterprise Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177 Reform Agenda for Reaching Upper-Middle-Income Levels. . . . . . . . . . . . . . . . . . . . . . . . 188 Annex 4A  Measuring Resource Misallocation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210 Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213 5  Transitioning Agriculture and Food as an Engine of Growth. . . . . . . . . 217 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217 Policy Support for Strong Growth in Agriculture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 Emerging Policy Challenges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221 Targeting Larger Markets for Growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225 Securing the Natural Resource Base for Innovation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233 Policies to Achieve Rapid and Sustainable Growth in Agriculture . . . . . . . . . . . . . . . . . . . 236 Annex 5A  A Graphic Presentation of the TFP Story in Rwandan Agriculture. . . . . . . . . 242 Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246 6  Capable and Accountable State Institutions . . . . . . . . . . . . . . . . . . . . . . . 251 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251 Rwanda’s Record in Governance and Government Effectiveness . . . . . . . . . . . . . . . . . . . . 253 Remaining Governance Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258 C ontents vii An Agenda for Reforms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264 Annex 6A  Rwanda’s Territorial Organization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276 Annex 6B  Investment as a Share of GDP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277 Annex 6C  Overview of the Public Investment Management Assessment Questionnaire. . . . 278 Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280 Boxes O.1 Major requirements of Rwanda’s income aspirations. . . . . . . . . . . . . . . . . . . . . . . 10 O.2 The importance of regional value chains for growth and outward orientation in East Asia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 O.3 Opportunities in the region around Rwanda . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 O.4 An international comparison of Rwanda’s long-term growth trajectory. . . . . . . . . 21 1.1 What is the role of social safety nets for economic growth? . . . . . . . . . . . . . . . . . . 54 1.2 Confronting a learning crisis in basic education: Key areas for action . . . . . . . . . . 65 1.3 Increasing access to education and training in the Republic of Korea. . . . . . . . . . . 67 1.4 Achieving world-class education in Singapore . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 1.5 Lessons from China: Teacher management in Shanghai . . . . . . . . . . . . . . . . . . . . . 70 1.6 Technology for improving teachers’ English skills . . . . . . . . . . . . . . . . . . . . . . . . . . 72 1.7 A strong foundation in basic education for transforming TVET in Singapore . . . . . 75 1.8 Partnering with firms to invest in workforce skills in Malaysia . . . . . . . . . . . . . . . . 76 1.9 Developing skills for economic growth and transformation in the “Asian Tiger” economies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 1.10 Achieving high-quality tertiary education in the Republic of Korea . . . . . . . . . . . . 83 2.1 Transformation of the coffee sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 2.2 Stimulating product upgrading through a supplier development program . . . . . . 112 2.3 The benefits of a mutual recognition agreement for Rwanda’s engineering professional body. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 2.4 The strategy to transform the textiles, apparel, and leather sector . . . . . . . . . . . . 126 3.1 Seoul becomes a global city . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 3.2 Measuring urbanization in Rwanda. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140 3.3 Avoiding slums through property rights and infrastructure . . . . . . . . . . . . . . . . . . 154 3.4 The New York grid: Transformation of structures and the economy around a planned grid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 3.5 Land readjustment for urban renewal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 3.6 How will urban mobility change by 2050?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 3.7 Examples of successful engagement by large cities with the informal sector. . . . . 160 3.8 Mobilizing rural-urban links: Chengdu’s coordinated development model. . . . . . 161 3.9 Land value capture: The rail + property program in Hong Kong SAR, China . . . . 163 3.10 Land value capture in East Asia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 3.11 Updating land values in Bogotá, Colombia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 3.12 Institutions for market-responsive urban planning: Yokohama, Japan. . . . . . . . . 165 3.13 Participatory planning and budgeting: Kerala, India . . . . . . . . . . . . . . . . . . . . . . . 166 3.14 Involving stakeholders in city development: Lessons on participatory planning from global cities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 3.15 Cross-district governance: The Republic of Korea’s happy living zones . . . . . . . . 167 4.1 Is informal normal? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174 4.2 The global context and technological transformation. . . . . . . . . . . . . . . . . . . . . . 184 4.3 The national innovation system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 viii FUTURE DRIVERS OF GROWTH IN RWANDA 4.4 Rwanda’s Long-Term Saving Scheme: Opportunities and challenges . . . . . . . . . . 189 4.5 Institutional framework for PPPs: An international perspective . . . . . . . . . . . . . . 194 4.6 Targeted interventions in the Republic of Korea. . . . . . . . . . . . . . . . . . . . . . . . . . 197 4.7 “Made in Rwanda” policy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198 4.8 Developing Rwanda’s mining sector. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 4.9 The importance of firm capabilities for productivity: Examples from Brazil, India, and Singapore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205 4.10 The experience of Apps.co. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207 4.11 Kigali Innovation City . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208 5.1 High-value tea development in Yunnan Province, China . . . . . . . . . . . . . . . . . . . 226 5.2 Horticulture and floriculture exports from East Africa. . . . . . . . . . . . . . . . . . . . . 228 5.3 Successful restoration of the productive landscape on China’s Loess Plateau . . . . 235 5.4 Israeli success at being more efficient with agricultural water. . . . . . . . . . . . . . . . 237 6.1 A major turnaround in the Rwandan civil service . . . . . . . . . . . . . . . . . . . . . . . . . 256 6.2 Imihigo: A Rwandan management tool of accountability. . . . . . . . . . . . . . . . . . . 258 6.3 Relationship between trust and long-term development. . . . . . . . . . . . . . . . . . . . 259 6.4 Lessons from East Asia on strengthening public administration . . . . . . . . . . . . . 266 6.5 The role of competition policy in regulatory reform in the Republic of Korea . . . 268 Figures ES.1 Future drivers of Rwanda’s growth: Innovation, integration, agglomeration, and competition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxiii O.1 Rwanda’s global rankings on Ease of Doing Business indicators . . . . . . . . . . . . . . . . 3 O.2 Average growth of GDP per capita in Rwanda and select countries, 2006–16 . . . . . 4 O.3 Sectoral labor productivity and annual change in share of employment in Rwanda, 2000–16 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 O.4 Human capital outcomes in Rwanda and other countries, by income level, 2015. . . . . 6 O.5 Average gross domestic savings rate for Rwanda and regional comparators, 2014–16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 O.6 Rwanda’s domestic savings and investment rates, 1995–2015 . . . . . . . . . . . . . . . . . 6 O.7 Labor productivity and GDP per capita in Rwanda and other countries . . . . . . . . . 6 O.8 Total factor productivity (TFP) and GDP per capita in Rwanda and other countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 O.9 Average growth of labor productivity in Rwanda, by sector, 2001–16. . . . . . . . . . . 7 O.10 Price levels and GDP per capita in Rwanda and comparator countries and economies, 2011 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 O.11 Manufacturing value added in the East African Community, 2006–16. . . . . . . . . . . 8 O.12 Projected GDP per capita for Rwanda under alternate growth scenarios, 2016–50. . . . 9 O.13 Measures of global vertical integration in Rwanda, 1995–2014. . . . . . . . . . . . . . . 14 O.14 Share of global net outflows of foreign direct investment, by region, 1994–2016. . . . . 17 O.15 Growth scenarios for Rwanda and share of the population, by consumption category, by 2040. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 O.16 Labor productivity gaps in Rwanda, by sector, 2014 . . . . . . . . . . . . . . . . . . . . . . . 19 O.17 Future drivers of Rwanda’s growth: Innovation, integration, agglomeration, and competition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 BO.4.1 GDP per capita in Rwanda and select countries and economies. . . . . . . . . . . . . . . 21 O.18 Early-grade “bulge index” in Rwanda and select countries. . . . . . . . . . . . . . . . . . 24 O.19 Projected share of Rwanda’s workforce with tertiary education under two growth scenarios, 2010–50 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 O.20 Exports as a share of GDP and log GDP per capita in Rwanda, 2000–16 . . . . . . . 28 C ontents ix O.21 Intra-bloc goods imports as a share of GDP before and after joining the bloc . . . . 29 O.22 Changes in urbanization and income in East Asia and Pacific and in Sub-Saharan Africa, 1985–2010. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 O.23 Share of firms in Rwanda, by firm size, 2011 and 2014 . . . . . . . . . . . . . . . . . . . . . 36 O.24 Resource misallocation in formal manufacturing in Rwanda and select countries. . . . 37 O.25 Decomposition of sources of agricultural growth in Rwanda, 1961–2014. . . . . . . 39 O.26 Benchmarking of Rwanda along the various dimensions of governance. . . . . . . . . 42 O.27 A framework for assessing effectiveness of state institutions in Rwanda . . . . . . . . . 44 1.1 Human capital outcomes in Rwanda and other countries, by income level, 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 1.2 Rates of human capital in Rwanda and the Republic of Korea at similar income levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 1.3 The array of human capital investments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 1.4 The trade-off between lower returns now and higher returns later. . . . . . . . . . . . . 53 1.5 Stunting rate in Rwanda versus the world, 1995–2015. . . . . . . . . . . . . . . . . . . . . 54 1.6 Prevalence of stunting and GDP per capita, 2017. . . . . . . . . . . . . . . . . . . . . . . . . . 55 1.7 Prevalence of stunting in Rwanda, by wealth quintile, 2000–15 . . . . . . . . . . . . . . 55 1.8 Reduction in stunting in countries with less than 50 percent stunting . . . . . . . . . . . 57 1.9 Projected reduction in stunting in Rwanda, 2000–35. . . . . . . . . . . . . . . . . . . . . . . 58 1.10 School completion rates in primary and secondary school in Rwanda compared with countries at different levels of income, 2017 . . . . . . . . . . . . . . . 61 1.11 Early-grade “bulge index” in Rwanda and select countries. . . . . . . . . . . . . . . . . . 62 1.12 Gap between actual and learning-adjusted years of schooling in select countries and economies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 1.13 Key proximate determinants of learning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 1.14 Lag in teacher salaries in Rwanda, by primary and secondary levels. . . . . . . . . . . 66 1.15 Current and projected primary and lower-secondary completion in Rwanda, 2000–35 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 1.16 Enrollment in secondary TVET, by region and within Africa, circa 2000 and 2014. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 1.17 Conceptual framework outlining the links between higher education and economic growth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 1.18 Projected share of Rwanda’s workforce with tertiary education under two growth scenarios, 2010–50 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 1.19 Current and projected tertiary enrollment in Rwanda, 2000–35. . . . . . . . . . . . . . 80 1.20 Number of Rwandan scientific publications available on the Web of Science, 1970–2015. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 1.21 Tertiary enrollment rate as a function of secondary completion rates . . . . . . . . . . 81 1.22 Spending on tertiary education in Rwanda versus other countries, circa 2000 and 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 1.23 Current spending on education and health in Rwanda relative to other countries in the world. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 2.1 Rwanda’s Ease of Doing Business Ranking, 2007–19. . . . . . . . . . . . . . . . . . . . . . . 93 B2.1.1 Number of coffee-washing stations and average price of coffee cherries in Rwanda, 2002–15 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 2.2 Value of Rwandan exports, by destination, 2001–16. . . . . . . . . . . . . . . . . . . . . . . 96 2.3 Exports of Rwandan goods, by destination, 2016 . . . . . . . . . . . . . . . . . . . . . . . . . 97 2.4 Agroprocessing and other manufacturing as a share of total exports in Rwanda, 2005–16 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 2.5 Ranking of revealed comparative advantage for East African Community countries, 2013. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 x FUTURE DRIVERS OF GROWTH IN RWANDA 2.6 Value and volume index of Rwandan commodity exports, 2000–15 . . . . . . . . . . . 99 2.7 Share of export value in Rwanda, 2009–16 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 2.8 Labor productivity in Rwanda, 2005 and 2014 . . . . . . . . . . . . . . . . . . . . . . . . . . 100 2.9 Rwanda’s financing stock, by sector, 2006–15 . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 2.10 Exports as a share of GDP and log GDP per capita in Rwanda, 2000–16 . . . . . . 101 2.11 Cumulative exports as a share of GDP in Rwanda, 2005–16, and potential trajectory to upper-middle-income status by 2035 . . . . . . . . . . . . . . . . . . . . . . 106 2.12 Intra-bloc goods imports as a share of GDP before and after joining the bloc . . . . . 107 2.13 Total imports of exporting versus nonexporting firms in Rwanda, 2008–14 . . . . 108 2.14 Impact of One Network Area on mobile phone traffic between Rwanda and Kenya. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 2.15 Average transport time, by route . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 2.16 Transport rates for a 40-foot container, by route . . . . . . . . . . . . . . . . . . . . . . . . . 116 2.17 Tertiary gross enrollment rate and log services exports per capita, 2000–16 . . . . 123 2.18 Upper-secondary completion rate and log services exports per capita, 2000–16. . . . 123 2.19 Number of professionals per 100,000 inhabitants in select African countries. . . . 124 2.20 FDI stock as a share of GDP in select East African countries, 2000–16 . . . . . . . . 125 3.1 Changes in urbanization and income in East Asia and Pacific, 1985–2010. . . . . . 138 3.2 Contribution of urbanization to structural transformation and labor productivity in Rwanda, 2002–12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 3.3 Simulated changes in growth outcomes with faster urbanization in Rwanda . . . . 141 3.4 Puga Index of connectedness in Kigali and select comparators. . . . . . . . . . . . . . . 144 3.5 Share and scale of rural-urban migration in Rwanda, 2014 . . . . . . . . . . . . . . . . . 149 3.6 Remittances from Kigali (and other locations) to rural areas of Rwanda, by province. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 B3.5.1 Land readjustment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 4.1 Share of firms in Rwanda, by firm size, 2011 and 2014 . . . . . . . . . . . . . . . . . . . . 172 4.2 Share of employment in Rwanda, by firm size, 2011 and 2014 . . . . . . . . . . . . . . 172 4.3 Share of firms and employment in Rwanda, by year of establishment . . . . . . . . . . 173 4.4 Share of firms and employment in Rwanda, by sector, 2011 and 2014. . . . . . . . . 173 4.5 Informality as a share of total firms and employment in Rwanda, 2011 and 2014 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174 B4.1.1 Informal employment as a share of GDP per capita in Latin America and Southeast Asia, various years. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 4.6 Number of formal firms and total employment in Rwanda, by type of firm, 2011–16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 4.7 Number and share of exporters in Rwanda, by sector, 2015 . . . . . . . . . . . . . . . . 176 4.8 Price levels and GDP per capita in Rwanda and comparator countries and economies, 2011. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177 4.9 Average lending and deposit rates in Rwanda, 2000–17 . . . . . . . . . . . . . . . . . . . 178 4.10 Cost of electricity in Rwanda and other Sub-Saharan African countries, 2016. . . . . 179 4.11 Rwanda’s global rankings on Ease of Doing Business indicators . . . . . . . . . . . . . . 179 4.12 Labor productivity and GDP per capita in Rwanda and other countries. . . . . . . 180 4.13 Average growth of labor productivity in Rwanda, by sector, 2001–16. . . . . . . . . 181 4.14 Total factor productivity and GDP per capita in Rwanda and other countries . . . . . 181 4.15 Resource misallocation in formal manufacturing in Rwanda and select countries . . . . 182 4.16 Probability of receiving tax incentives in Rwanda, by size of firm revenue, 2013–16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 4.17 Sectoral revenue growth and tax incentives in Rwanda, 2010–16. . . . . . . . . . . . 183 4.18 Correlation between innovation capacity and total factor productivity . . . . . . . . 184 Contents xi B4.3.1 The national innovation system as an ecosystem for economywide innovation. . . . . 185 4.19 Ranking of Rwanda and other East African Community countries on the Global Innovation Index, 2013 and 2017 . . . . . . . . . . . . . . . . . . . . . . . . . 186 4.20 Frontier of innovation capacity in Rwanda, South Africa, and Switzerland . . . . . 186 4.21 Quality of research institutions and number of patent applications in Rwanda and other countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187 4.22 Number of trademark applications and registrations in Rwanda, 2001–15 . . . . . 188 4.23 The capabilities escalator: A framework to support innovation and entrepreneurial activity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204 5.1 Decomposition of sources of agricultural growth in Rwanda, 1961–2014. . . . . . 221 5A.1 Heuristic diagram of the production possibility frontier. . . . . . . . . . . . . . . . . . . . 243 6.1 Relationship between government effectiveness and level of development . . . . . . 253 6.2 Benchmarking of Rwanda along the various dimensions of governance. . . . . . . . 254 6.3 Correlation between prosecution of office abuse and GDP per capita. . . . . . . . . . 254 6.4 Revenue (tax and nontax) excluding grants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255 6.5 Rwanda’s decentralization framework for service delivery. . . . . . . . . . . . . . . . . . 257 6.6 A framework for assessing effectiveness of state institutions in Rwanda. . . . . . . . 259 6.7 Rwanda’s performance on product market regulations. . . . . . . . . . . . . . . . . . . . . 262 6.8 Public and publicly guaranteed debt as a percentage of GDP in Rwanda, 2010–17 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263 6.9 The main elements for an efficient and capable civil service. . . . . . . . . . . . . . . . . 266 6A.1 Territorial organization in Rwanda . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277 6B.1 Gross fixed capital formation as a share of GDP in Rwanda, 1965–2015. . . . . . 277 Maps 1.1 Prevalence of stunting in Rwanda, by district, 2014–15. . . . . . . . . . . . . . . . . . . . . 55 3.1 Visualizing economic concentration across five countries . . . . . . . . . . . . . . . . . . 139 3.2 Fragmentation of densities beyond the central business district of Kigali . . . . . . . 145 3.3 Built land in Kigali, 1990, 2000, and 2010. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 3.4 Building footprint in central areas of three secondary cities in Rwanda . . . . . . . . 146 3.5 Location of formal jobs in Kampala, Uganda; Lusaka, Zambia; and Kigali, Rwanda. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 3.6 Changes in building stock in Kigali, by cell, 2009–15. . . . . . . . . . . . . . . . . . . . . . 147 3.7 Simulated change in access to employment in Kigali with a 50 percent drop in bus journey speed. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148 3.8 Location of planned imidugudu (grouped villages) in 12 districts of Rwanda, 2016–17 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150 3.9 Variation in tomato prices in western Rwanda . . . . . . . . . . . . . . . . . . . . . . . . . . 151 3.10 Expressways in the Republic of Korea, 1970–2010 . . . . . . . . . . . . . . . . . . . . . . . 152 3.11 Goldilocks density: Getting density “just right” through fine-grain differentiation and adaptation over time. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 B3.4.1 Plan of Manhattan, New York City, 1811 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 3.12 Concentrated urban development and universal basic services in the Republic of Korea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162 Photos B3.5.1 Way-markers map plot boundaries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 B3.5.2 Land readjustment sites in Gahanga, Kigali . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 B5.3.1 China’s Loess Plateau, 1990 and 2012. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236 xii FUTURE DRIVERS OF GROWTH IN RWANDA Tables O.1 Social and economic indicators in Rwanda compared with low- and middle-income-country averages, 1994 and 2015. . . . . . . . . . . . . . . . . . . . . . . . . 5 BO.1.1 Aspirational high-growth scenario for Rwanda: Demand-side requirements, 2000–35. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 BO.1.2 Aspirational high-growth scenario for Rwanda: Productivity requirements . . . . . . 11 BO.1.3 Aspirational high-growth scenario for Rwanda: Sectoral share of employment, 2000–35 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 BO.1.4 Aspirational high-growth scenario for Rwanda: Sectoral share of GDP, 2000–35. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 1.1 Distribution of tertiary students in Rwanda, by field of study. . . . . . . . . . . . . . . . 82 2.1 Ease of Doing Business 2019 ranking for Rwanda and average for the East African Community. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 2.2 Rwandan exports, by sector, 2005 and 2016 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 2.3 Decomposition of merchandise export growth, 2008–13 . . . . . . . . . . . . . . . . . . . 99 2.4 Manufacturing subsectors, grouped by pro-development characteristics . . . . . . . . 103 2.5 Labor costs in manufacturing in select countries . . . . . . . . . . . . . . . . . . . . . . . . . 103 2.6 Service subsectors in Rwanda, grouped by pro-development characteristics. . . . . 104 2.7 Rwanda’s labor productivity and participation in the external sector, 2015. . . . . 108 2.8 Cost of electricity in 2016 in select East African Community countries . . . . . . . . 114 2.9 Stock of foreign direct investment in Rwanda, by sector, 2015. . . . . . . . . . . . . . . 125 3.1 Share of firms in Kampala, Uganda; Lusaka, Zambia; and Kigali, Rwanda, by size, 2011. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 4.1 Efficiency and profitability of banks in Rwanda and other East African Community countries, 2013 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178 4.2 Share of total tax revenue for firms, by return on investment with and without incentives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 4.3 Large exposures in the banking sector in Rwanda, 2015–17 . . . . . . . . . . . . . . . . 191 4.4 State-owned enterprise policy initiatives in the OECD since 2006 . . . . . . . . . . . . 202 4.5 Use of science, technology, and innovation funding initiatives in Rwanda . . . . . . 209 5.1 Rwanda’s main agricultural exports, 2012/13 to 2015/16 . . . . . . . . . . . . . . . . . . 227 B5.2.1 Exports to Europe of flowers, vegetables, and fruits from select African countries, 2017 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228 5.2 Indicators of comparative advantage for select crops in Rwanda, 2016 . . . . . . . . 230 5.3 Exports of Irish potatoes and green beans by Rwanda and select African competitors, 2003 and 2013. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231 5.4 Response of demand to a 1 percent increase in household income in Rwanda, by degree of urbanization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232 6.1 The Rwanda Governance Scorecard: Scores for 2016 compared with previous scores. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253 6.2 Components and ratings of public investment management framework. . . . . . . . 255 Foreword A widely acknowledged record of more on the future drivers of growth to inform than two decades of strong eco- Rwanda’s Vision 2050. The report identifies nomic growth—including a three- four essential and interdependent drivers: and-half-fold increase in per capita income innovation, integration, agglomeration, and since 1994—has placed Rwanda among the competition. These future drivers of growth, fastest-growing economies in Africa and in turn, would receive the necessary boost the world. Traumatic memories of the 1994 from reforms in six priority areas: (a) human genocide, still prevalent, are gradually fading, capital development; (b) export dynamism and as associations begin to take a more positive regional integration; (c) well-managed urban- form—of a nation on the rise, powered by ization; (d) competitive domestic enterprises; human resilience, a sense of common pur- (e) agricultural modernization; and (f) capable pose, and a purposeful government. and accountable public institutions. Having launched its development with The report has already helped to shape the a measure of success, Rwanda is now keen reform agenda in all of the above six prior- to forge a future of security, prosperity, and ity areas. The study has been conducted at modernity. Sustaining high rates of inclu- the same time Rwanda was developing its sive economic growth is at the heart of national strategy for transformation (2017– these ambitions. Future aspirations in this 2024) and Rwanda Vision 2050. Both have regard are high, with Vision 2050 aspiring to been informed by recommendations from achieve upper-middle-income status by 2035 the study. For example, the report’s find- and high-income status by 2050. These aspi- ings about challenges facing Rwanda in rations translate into average annual growth human capital motivated the government to rates of more than 10 percent. redouble its efforts by fighting malnutrition, Motivated by these aspirations, in May launching reforms for improving the quality 2017, the government of Rwanda and the of education at all levels, and establishing World Bank Group jointly initiated this study the National Early Childhood Development FUTURE DRIVERS OF GROWTH IN RWANDA xiii xiv FUTURE DRIVERS OF GROWTH IN RWANDA Program to coordinate all interventions on of this report. We are convinced that keeping early childhood development and fighting the momentum in implementing the recom- malnutrition. mendations offered will support Rwanda’s We are delighted by the close partnership development journey and take it closer to the forged between the experts from Rwanda transformed society it aspires to be. and the World Bank Group in preparation Hafez M. H. Ghanem Dr. Ngirente Edouard Vice President, Africa Region Prime Minister of the Republic of Rwanda The World Bank Acknowledgments F   uture Drivers of Growth in Rwanda: Transformation and Rwanda Vision 2050 Innovation, Integration, Agglomeration, provided guidance to the team and validated and Competition is a joint initiative of the report. the Government of Rwanda and the World A joint Rwanda–World Bank Group Bank Group, emanating from a March 2017 team was established to prepare the report. agreement between President Paul Kagame of Team leaders were Leonard Rugwabiza Rwanda and Jim Yong Kim, President of the Minega (Economic Advisor to the Minister World Bank Group. of Finance and Economic Planning) and The report was written by a joint Rwanda- Sandeep Mahajan (Practice Manager, World Bank Group team under the supervi- Macroeconomics, Trade and Investment sion and guidance of Uzziel Ndagijimana Global Practice, World Bank Group). (Minister of Finance and Economic Planning), Financial support from the Republic of Claver Gatete (Minister of Infrastructure and Korea is gratefully acknowledged. Former Minister of Finance and Economic William Maloney (Chief Economist, Planning), the late Jan Walliser (Former Vice Equitable Growth, Finance, and Institutions President, Equitable Growth, Finance, and Practice Group, World Bank Group), Institutions Practice Group, World Bank Richard Newfarmer (Country Director for Group), Makhtar Diop (Vice President for Uganda and Rwanda, International Growth Infrastructure and Former Vice President for Centre), and Andrew Nyamvumba (Former Africa Region, World Bank Group), Hafez Head, Strategy and Policy Unit, Office of M. H. Ghanem (Vice President for Africa the President, Republic of Rwanda) were Region, World Bank Group), Diarietou Gaye senior advisers for the report. The report (Director of Strategy and Operations of also benefitted from valuable comments from Africa Region and Former Country Director peer reviewers David Dollar (Senior Fellow, for Rwanda, World Bank Group), Ceyla John L. Thornton China Center, Brookings Pazarbasioglu-Dutz (Vice President, Equitable Institution), Mona Haddad (Director, Growth, Finance, and Institutions Practice International Finance Corporation), Steve Group, World Bank Group), and Carlos Felipe Jaffee (Lead Agriculture Economist, World Jaramillo (Country Director for Rwanda and Bank Group), Godfrey Kabera (Director Former Senior Director, Macroeconomics, General, National Planning and Research, Trade and Investment Global Practice, Ministry of Finance and Economic Planning), World Bank Group). The National Steering Michael Woolcock (Lead Social Scientist, Committee for the National Strategy for World Bank Group), and Albert Zeufack FUTURE DRIVERS OF GROWTH IN RWANDA xv xvi FUTURE DRIVERS OF GROWTH IN RWANDA (Chief Economist for Africa Region, World Finance and Economic Planning), and Bank Group), as well as the following Ariane Zingiro (Planning and Program reviewers who provided comments on the Sp e cia l ist , M i n ist r y of Fi na nce a nd concept note and the decision draft: Emre Economic Planning). Alper (Senior Economist, International The report was edited by William Shaw. Monetary Fund), Iftikhar Malik (Senior Production and publication were managed Social Protection Specialist, World Bank by Stephen Pazdan and Patricia Katayama of Group), Laura Rawlings (Lead Social the World Bank Group’s Formal Publishing Protection Specialist, World Bank Group), Program. Laure Redifer (Deputy Division Chief, The report has an overview and six chap- International Monetary Fund), and Alun ters written by individual chapter teams, Thomas (Former Resident Representative in whose members are presented in alphabetical Rwanda, International Monetary Fund). order by their surnames. We ex tend specia l t ha n ks for t he insights and recommendations received from Abebe Adugna (Practice Manager, Overview Macroeconomics, Trade and Investment, Lead authors: Sandeep Mahajan (Practice World Bank Group), Anstes Agnew (Country Manager, Macroeconomics, Trade and Head for Rwanda at Tony Blair Institute for Investment, World Bank Group) and Leonard Global Change), Yasser El-Gammal (Country Rugwabiza Minega (Economic Advisor to the Manager for Rwanda, World Bank Group), Minister of Finance and Economic Planning). Obald Hakizimana (Economist, Ministry of Finance and Economic Planning), Mary Hallward-Driemeier (Senior Economic Chapter 1: Human Capital and Adviser, World Bank Group), Emmanuel Innovation H at e g e k a ( D e put y C h i e f E x e c ut ive Lead authors: David Evans (Lead Economist, Officer, Rwanda Development Board), World Bank Group) and François Ngoboka Thierry Mihigo Kalisa (Senior Economist, (Division Manager, Former Capacit y M i n i s t r y of Fi n a nc e a nd E c onom ic Development and Employment Services Planning), Danny Leipziger (Managing Board). Director of Growth Dialogue and Professor Core team: Rose Baguma (Director of Practice of International Business, George General of Education Policy and Planning, Washington University), Norman V. Loayza M i n i s t r y of E duc at ion), D r. I g n ac e (Lead Economist, World Bank Group), Johan Gatare (Principal, College of Science and A. Mistiaen (Program Leader, World Bank Technology, University of Rwanda), Deepika Group), Geraldine Mukeshimana (Minister Ramachandran (Consultant), Jee-Peng Tan for Agriculture and Animal Resources), (Consultant), and Fei Yuan (Consultant). Ildephonse Musafiri (Head of the Strategy and Policy Unit, Office of the President, Republic of Rwanda), Richard Mushabe Chapter 2: Transformation (Division Manager of National Planning through Trade: Using Exports and Research Division, Ministry of Finance and Regional Integration to and Economic Planning), Tom O’Brien (Senior Adviser, World Bank Group), Amina Drive Future Growth Rwakunda (Chief Economist, Ministry Lead authors: Thomas R. Kigabo (Chief of Finance and E conom ic Plan ning), Economist, National Bank of Rwanda), Kampeta Pitchette Sayinzoga (Director Richard Newfarmer (Country Director, General, National Industrial Research International Growth Centre), and Victor and Development Agency of Rwanda), Steenbergen (Former Country Economist, Claudine Uwera (State Minister, Ministry of International Growth Centre). A c k now l edg m ents xvii Core team: John H. Arnold (Consultant), (Senior Economist, World Bank Group), Robert Opirah (Director General for Trade Peace Aimée Niyibizi (Economist, World and Investment, Ministry of Trade and Bank Group), Pierre Sauve (Senior Private Industry), and Cordula Rastogi (Senior Sector Specialist, World Bank Group), Economist, World Bank Group). and Marcio Jose Vargas da Cruz (Senior Economist, World Bank Group). The chapter builds on background papers Chapter 3: Faster Urbanization, and notes prepared by David Bridgman Greater Agglomeration (Lead Private Sector Specialist, World Bank Group), Joern Huenteler (Energy Specialist, Lead authors: Narae Choi (Urban Specialist, World Bank Group), Maksym Iavorskyi World Bank Group), Edward Kyazze (Manager (Operations Analyst, World Bank Group), of Urbanization and Human Settlements Danny Leipziger (Managing Director of Division, Ministry of Infrastructure), Somik Growth Dialogue and Professor of Practice of V. Lall (Lead Urban Economist, World Bank International Business, George Washington Group), and Sally Beth Murray (Former U n i ve r s i t y), H i b r e t B e l e t e M a e m i r Senior Country Economist, International (Economist, World Bank Group), William Growth Centre). Peter Mako (Consultant), Hoza Thierry The chapter builds on an International Ngoga (Consultant), Ha Minh Nguyen Food Policy Research Institute (IFPRI) (Economist, World Bank Group), Yadviga research paper by Xinshen Diao, Peixun Viktorivna Semikolenova (Senior Energy Fang, Josée Randriamamonjy, and James Economist, World Bank Group), Victor Thurlow and on background notes by David Steenbergen (Former Country Economist, Bridgman (Lead Private Sector Specialist, International Growth Centre), Desiree Van World Bank Group) and Hoza Thierry Welsum (Consultant), Jan Erik von Uexkull Ngoga (Consultant). (Senior Economist, World Bank Group), and Ezequiel Zylberberg (Consultant). Chapter 4: Competitiveness and Enterprise Development for Chapter 5: Transitioning Innovation-Led Growth Agriculture and Food as an Lead authors: Claire Honore Hollweg (Senior Engine of Growth Economist, World Bank Group), Aghassi Lead authors: Christopher L . Delgado Mkrtchyan (Senior Economist, World Bank (Consultant) and Dr. Octave Semwaga Group), Jani Moliis (Former Country Head (Director General of Planning, Ministry of for Rwanda at Tony Blair Institute for Global Agriculture and Animal Resources). Change), and Jean Louis Uwitonze (Former Core team: Winston Dawes (Senior Director General of Planning, Ministry of A g r ic u lt u re E conom ist , World B a n k Trade and Industry). Group), Aparajita Goyal (Senior Economist, Core team: Teymour Abdel Aziz (Senior World Bank Group), Christopher Jackson Financial Sector Specialist, World Bank (Senior Agriculture Specialist, World Bank Group), Brice Gakombe (Financial Sector Group), and Aimée Marie Ange Mpambara Specialist, World Bank Group), Justin (Ag ricult u ral Specialist , World Ban k Piers William Hill (Senior Private Sector Group). Specialist, World Bank Group), Louise The chapter builds on an International Kanyonga (Head, Competitiveness and Food Policy Research Institute (IFPRI) Business Reforms, Rwanda Development research paper by Xinshen Diao, Peixun Board), Hibret Belete Maemir (Economist, Fang, Josée Randriamamonjy, and James World Ban k Group), G au rav Nay ya r Thurlow. xviii FUTURE DRIVERS OF GROWTH IN RWANDA Chapter 6: Capable and Group), David Bridgman (Lead Private Sector Accountable State Institutions Specialist, World Bank Group), Maksym Iavorskyi (Operations A nalyst, World Lead authors: Shyaka Anastase (Former Bank Group), and Hoza Thierry Ngoga Chief Executive Officer, Rwanda Governance (Consultant). Board), Nicoletta Feruglio (Senior Public Administration and coordination activities Sector Specialist, World Bank Group), and were managed by Lydie Ahodehou (Program Gael J. R. F. Raballand (Lead Public Sector Assistant, World Bank Group) and Sylvie Specialist, World Bank Group). Ingabire (Executive Assistant, World Bank Core team: Monica Beuran (Consultant), Group). Deborah Hannah Isser (Lead Governance It is w it h g reat pleasu re t hat t he Specialist, World Bank Group), and Felicién Government of Rwanda and the World N s e n g u mu k i z a ( H e ad , G over n a n c e , Ban k Group introduce the repor t Research, and Monitoring Department, Future Drivers of Growth in Rwanda: Rwanda Governance Board). Innovation, Integration, Agglomeration, The chapter builds on background papers and Competition. We hope that Rwandan by Jonathan Argent (Consultant), Tania citizens and all actors of development in Begazo (Senior Economist, World Bank Rwanda will find it useful. Executive Summary R wanda: the land of a thousand hills, untold beauty, a poignant past, and boundless ambition. It has, in a way, become an important fragment of modern consciousness. Traumatic memories of the 1994 genocide against the Tutsi, still prevalent, are fading gradually, as associations begin to take a more positive form: a nation on the rise, powered by human resilience, a sense of common purpose, and a purposeful government. Past successes, combined with a sense of frailty, have fueled bold ambitions for the future. But these are early days, and Rwanda is keen to continue taking meaningful strides toward becoming a more hopeful, secure, modern, and prosperous nation. Strong and inclusive growth sustained over an extended period is at the heart of these ambitions. This report—a joint undertaking by experts from Rwanda and the World Bank—seeks to evaluate the country’s possibilities and options in this endeavor. Introduction fastest-growing economies, surpassed on the continent only by Ethiopia. Equally striking After 1994 everything was a priority and are its ambitions and approaches. In ambi- our people were completely broken. But we made three fundamental choices that guide tion, it models itself after the East Asian us to this day. One—we chose to stay Tigers, seeking a development trajectory together. Two—we chose to be accountable to take it from abject poverty to prosperity to ourselves. Three—we chose to think big. within a couple of generations. In approach, it has adopted high standards and demands —His Excellency President Paul Kagame discipline, self-reliance, and hard work of in a speech on the 20th Commemoration itself—its formula for escaping the vortex of genocide against the Tutsi of the ­ low expectations, corruption, and chaos that (April 7, 2014) it rejected early on. Having launched its development with Every country, born of unique circumstances, a measure of success, Rwanda is now keen can claim with good justification to be spe- to forge a future of security, prosperity, cial. Rwanda’s claims are self-evident in both and modernity. Sustaining high rates of what is manifest and what is being sought. ­ inclusive economic growth is at the heart of A strong and widely acknowledged record of these ambitions. Future aspirations in this economic success—­ including a three-and-a- regard have been set extremely high. Recent half-fold increase in per capita income since formulations of Vision 2050 set a target 1994—places Rwanda among the world’s of achieving upper-middle-income status FUTURE DRIVERS OF GROWTH IN RWANDA xix xx FUTURE DRIVERS OF GROWTH IN RWANDA by 2035 and high-income status by 2050. include better-targeted family planning mea- These aspirations translate into average sures and continued emphasis on female edu- annual growth rates of more than 12 ­percent cation, health, and economic empowerment (more than 10 percent in per capita terms). (United Nations Population Fund 2017). New technologies associated with indus- try 4.0—robotics, the Internet of Things, Opportunities and Risks three-dimensional printing—will remain Bold aspirations require Rwanda to play disruptive for Rwanda (as for the rest of the closely to its strengths and make the most of world), presenting more risks and opportuni- all available opportunities, while judiciously ties, with Rwanda’s net benefits depending managing risks. Rwanda’s main endowment on its responses. The immediate gains for is its people, whose determination, agaciro Rwanda are from the adoption and diffusion (dignity), and strong sense of personal and of new technologies and improved produc- social responsibility should serve it well. How tion p­ rocesses, rather than the development well the country nurtures and deploys this of sophisticated research and development resource will determine its true potential. The ­ capabilities. On the one hand, to the extent country’s record of good economic gover- that the new technologies are labor saving, nance, policy coherence, and strong focus on they risk narrowing the scope for manufac- implementation also will prove useful. Major turing as a pathway to development. On the opportunities lie in favorable demographics, other hand, the diffusion and adoption of an emergent and rapidly urbanizing middle new technologies can help to leapfrog tradi- class (in Rwanda and regionally), a fertile tional enablers of development. For ­ example, (though hilly)1 agricultural landscape, and Zipline, which uses unmanned drones untapped regional and global market poten- to deliver critical medical supplies across tial that offers opportunities for economic spe- Rwanda, has the potential to overcome geo- cialization and scale economies that Rwanda graphical barriers and increase efficiency in otherwise lacks. Counterbalancing these distribution of medical supplies. Modern opportunities are challenges associated with services, enabled by information and com- being a small, landlocked economy and risks munication technology and sharing the spe- from regional security and stability, a poten- cial characteristics (scalability, international tial widening of domestic disparities, intensi- tradability, and amenability to technological fying population pressures, and global climate upgrading) that allowed productivity conver- change. gence in manufacturing, are another area of Rwanda’s population dynamics present a opportunity, with business process outsourc- significant domestic challenge and possibly ing, banking, and health care as examples. also an opportunity. Rwanda is a youthful Future growth has to be powered by trade nation, with a median age of just 19 years, and regional integration. As a small, land- a result of rapid demographic shifts that are locked economy, Rwanda does not have the influenced by relatively high but declining scale economies to sustain high growth on fertility rates and sharp reductions in child its own. It has to stay alert to major shifts in mortality. But there also are legitimate con- ­ patterns of global trade and production. For cerns regarding the pace of rising popula- one, the global value chains are maturing and tion density, already the highest in mainland losing momentum since the “unbundling of Africa, and the growing population pressures production” has largely already happened on the country’s environment and limited (OECD 2017). At the same time, production is natural resources. Managed well, the popu- becoming increasingly concentrated in regional lation pressures can be contained, and the or local hubs closer to end markets: the share underlying demographic shifts can be har- of intra-African trade for most manufactur- nessed to achieve significantly higher growth ing industries rose between 2000 and 2014 and faster poverty reduction.2 Policy options (Hallward-Driemeier and Nayyar 2017). E x ecutive S u m m ar y xxi Fortunately for Rwanda, the region around it and innovations in the health sector, Rwanda abounds in (largely untapped) opportunities. could become a hub for modern medical Plus, with its landlocked geography as a lim- services for the region. Similarly, in higher iting factor, the ability of Rwandan firms to education, Rwanda has attracted top-flight compete globally depends on the “competitive- university subsidiaries, such as Carnegie ness” and ­ “connectedness” of its neighbors. Mellon. A recent grant from the Buffett Rwanda is thus likely to seize regional trade Foundation will enable the country to build opportunities through intraregional trade and a first-class agriculture school. Another logi- participation in regional value chains con- ­ cal services export, given Rwanda’s location, nected to the global market. could be mining services. The strongest regional opportunities The presence of new technologies and exist in commodity-based processing and business and trade opportunities bodes well, exports and services trade—what some have but it is not in itself enough to generate high called “industries without smokestacks” growth. This issue connects to the broader (Newfarmer, Page, and Tarp, forthcom- “productivity paradox,” whereby global pro- ing). Taken together, these activities hold the ductivity growth has slowed considerably promise of doing for Rwanda what manufac- since 2004 or so, even as new technologies turing did for East Asia. This promise applies have emerged and diffused at a rapid clip. particularly to agribusiness and food pro- Frustrated by a similar disconnect (between cessing, the only subsectors in which Rwanda massive investments in information technol- currently has a revealed comparative advan- ogy and productivity growth), Robert Solow, tage (Hallward-Driemeier and Nayyar 2017). a Nobel laureate in economics, famously These subsectors employ unskilled workers, said in 1987 (36), “You can see the computer have low intensity of physical capital, and age everywhere but in the productivity sta- exhibit high returns from the application of tistics.” The productivity statistics did pick basic research and development while, for up eventually, but only after a considerable now, being less exposed to automation. The lag in the mid-1990s. The missing link then, products (for example, wood products, bev- as it likely is now, was c ­ apabilities—of peo- erages) typically are traded regionally (rather ple, firms, and ­ institutions—that take time than internationally) because they are bulky to build. That missing link is also Rwanda’s to transport or require proximity to raw main challenge—and the focus of this materials (for example, food production). study—as the country seeks to harness the Rwanda could look to expand high-value new wave of technologies and trade patterns agricultural exports (such as horticulture)— for future high growth. Its investments in much like Peru has done with specialty this regard need to bring the core capabilities agricultural products such as artichokes, continually closer to the ever-moving global asparagus, and avocados, among others— frontier. even beyond the region, building on its repu- tation for having a superior regulatory and policy environment. An Agenda for Reform The region also offers major opportunities Rwanda’s aspiration for upper-middle income for trade in services, which already ­ provides by 2035 and high income by 2050 calls more than half of Rwanda’s export earn- for stretch targets for future growth rates. ings. While tourism has the most immediate Such targets may prove helpful in lifting the potential, Rwanda also could look toward nation’s vision and infusing a future orien- higher-skill services such as business process tation among the people, while mobilizing outsourcing, health care, and higher educa- citizen support and resources for the needed tion. Medical tourism is on the rise in Sub- economic reforms. This has to be supported Saharan Africa, as it is elsewhere (Dihel and with periodic and systematic review of goals Goswami 2016). With its many investments and targets, with an eye to reassessing their xxii FUTURE DRIVERS OF GROWTH IN RWANDA feasibility according to evolving trends and this will achieve the growth ambitions that with scope for readjustment if needed. Rwanda has. Any future high-growth strategy will A higher-order challenge is to boost pro- require a multisectoral approach. Relying on ductivity growth, which also has a bearing on manufacturing alone as a pathway for high Rwanda’s ability to maintain high investment growth clearly has become trickier than in rates. Rwanda’s labor productivity (output per the past. Services can offer significant scope worker) and total factor productivity (output for productivity gains (IMF 2018), but the generated by a given quantity of labor and service sector alone cannot yet absorb all capital) are low for its income level. Moreover, of the unskilled labor released by agricul- the rate of total factor productivity growth has ture. Moreover, important interdependen- slowed significantly in recent years. cies between sectors (most notably between The principal requirements for sustained services and manufacturing) prevent any one high productivity growth are scale econo- sector from growing too large without suf- mies and economic specialization in areas of ficient inputs from others—for example, in Rwanda’s comparative advantage, with com- China from 2000 to 2014, services inputs petition and technology diffusion as essential into manufacturing accounted for 38 ­ percent complements. These elements have proved of growth in services value added, while essential for sustaining high growth across manufacturing inputs into services accounted the world, but they are even more important for 30 percent. for a small, landlocked country like Rwanda. The reform agenda is complex and highly Scale and specialization require Rwanda to demanding: nothing short of an ­ extraordinary make the most of external opportunities and effort will suffice. The hard work begins in to enhance the benefits of urban agglomera- classrooms. The country needs a massive tion. But, to succeed in these areas, Rwanda effort to build human capital—its own educa- has to have a competitive domestic enterprise tion-focused Marshall Plan—if it is to realize sector, both public and private, with a strong its ambitious growth targets. With all of its potential to do well in competitive environ- achievements, Rwanda still lags the average ments. Such enterprises themselves have of low-income countries in crucial aspects of three critical requirements: a strong ecosys- human capital—for example, in stunting and tem for technological innovation, world-class primary and lower-secondary school comple- human capital, and robust institutions of tion. Moreover, the concern is as much about governance. This chain of priorities forms the quality as it is about quantity. 3 An impor- high-growth strategy for Rwanda. tant related issue is high stunting rates in Rwanda’s strategy for high growth thus early years—with implications for children’s has four essential and interdependent future learning abilities and participation in ­ d rivers—innovation, integration, agglom- the knowledge and services-led economy that eration, and competition (figure ES.1). These Rwanda envisages. future ­ d rivers of growth, in turn, would The next requirement is higher investment receive the necessary boost from reforms and savings rates. Rwanda already has a rela- in six high-­ priority areas: (1) human capi- tively high investment rate of about 26 ­percent tal development; (2) export dynamism and of GDP, well in excess of the domestic savings regional integration; (3) ­ w ell-managed rate of less than 10 percent. But double-digit urbanization; (4) competitive domestic enter- growth rates would require the investment rate prises; (5) agricultural modernization; and to be significantly higher. Achieving this level (6) ­capable and accountable public institu- of investment would require a sharp increase tions. The six reform areas are discussed in ­ ultifold in investment by the private sector; a m more detail in the next section. rise in the domestic savings rate; and higher Doing well on each of these six neces- foreign direct investment. Nothing short of sary reform areas is what separated the E x ecutive S u m m ar y xxiii FIGURE ES.1  Future drivers of Rwanda’s growth: rewarding success and punishing failure—a Innovation, integration, agglomeration, and model that the Republic of Korea closely fol- competition lowed in its early years of development. Highly developed Reform Priorities for High human Growth capital Trade and Competitive Reform Priority 1: Develop Higher- regional integration domestic Order Human Capital enterprises DRIVERS OF Building an innovation-led economy, as well GROWTH Innovation as getting the most out of greater integration Integration Agglomeration (especially services trade) and agglomera- Modern Competition Well- tion, will require much higher investments in agriculture/ managed human capital. Such investments, particu- food sector urbanization larly in the priority areas of reducing stunting Capable and and improving basic education, contribute to accountable growth only after considerable time, so mov- institutions ing on them early is essential for Rwanda’s growth aspirations. Building human capital involves a wide array of investments across the life cycle. high-growth East Asian economies from those Allocating more resources and providing bet- that achieved rapid growth for a decade or ter information on outcomes are key, together two, only to see it fizzle out. Even more chal- with making dramatic progress in the follow- lenging perhaps is to go beyond the necessary ing five policy areas: to the sufficient conditions. Rwanda sees rapid development not as a choice but as an existen- 1. Reduce stunting. A big push is needed tial matter and will be pushed to take high- to improve food security and nutritional risk strategic bets to gain high returns. These practices and to increase access to clean efforts have to be calibrated and managed water, good sanitation, and nutritional carefully. Unlike the East Asian economies, supplementation. Careful monitoring sys- Rwanda does not have the luxury of high tems are required to evaluate which poli- savings, which gives it far less scope to make cies deliver the greatest gains. costly mistakes. Policy responses first need to 2. E xpand access to basic educ ation . address key cross-cutting constraints (such as Achieving universal basic education skills, finance, infrastructure, and business requires ensuring that children are not regulation), clarify the future role of state- turned away for failure to pay incidental owned enterprises (SOEs), and then, if deemed fees, which may require increasing the per necessary, selectively pick areas for direct sup- student benefit paid to schools, and reduc- port that are aligned closely with Rwanda’s ing repetition in early grades so that chil- comparative advantages and focused on dren do not become too old for their grade export promotion rather than import sub- level. Providing information on the finan- stitution. Any direct support will have to set cial returns to schooling has been dem- clear policy objectives and performance tar- onstrated to increase parent and student gets for beneficiary firms; be coordinated investment in education, while improving closely across government ­ entities; and include the quality of schooling facilitates both a rigorous system for monitoring progress and reduced repetition and higher perceived enforcing sanctions as well as incentives for returns. xxiv FUTURE DRIVERS OF GROWTH IN RWANDA 3. Improve the qu alit y of educ ation . tradable sectors would boost scale economies Improved teaching practices are neces- and specialization and constitute an important sary to ensure that children in the first source of competition and innovation. three years are literate in Kinyarwanda Meeting Rwanda’s export objectives before transitioning to English. There is requires a comprehensive trade policy that good evidence regarding effective pro- spans services, industry, and agriculture. grams, including partially scripted lessons, Producing high-quality products for the on which Rwanda can draw. Because region (especially in manufacturing and these improvements take time, transition- agroprocessing) and developing other, less ing to English later (end of primary) may ­ location-dependent sectors (such as horti- be valuable. Rwanda could leverage its culture, tourism, professional services, and major investment in technology to pro- information and communication technology) vide upper-primary and secondary school for broader markets could do for Rwanda teachers with regular opportunities to what manufacturing did for East Asia from improve their English and to increase stu- the 1990s. These products are labor inten- dents’ learning performance. Recruiting sive, tradable, and capable of scale econo- expatriates may help to remedy the imme- mies; and they have high value added per diate shortage of teachers and train a core worker. As in traditional manufacturing, group of high-quality Rwandan teachers. technological change can occur quickly and 4. Strengthen technical and vocational train- drive rapid productivity growth. There are ing. Collecting and disseminating infor- five major priorities to meet medium-term mation on the quality of skills providers trade objectives: and the returns to different skills would improve quality and encourage participa- 1. Harness the regional blocs as platforms tion in high-return programs. for transformation. Key priorities are to 5. Build the tertiary sector and foster inno- lower and revise the common external vation . Increasing access to financing tariff within the East African Community (including private) would expand enroll- (EAC) to benefit Rwandan producers ment. Enrollment in high-return fields and consumers better, promote harmo- could be increased through financing nized standards in goods and services, incentives and higher-quality science and and reduce nontariff barriers. For key engineering instruction in earlier grades. sectors such as energy and finance, the Creating incentives for researchers to regional blocs should develop regional develop and to adapt innovations to value chains to achieve scale economies. b enefit industries and getting industries ­ Finally, Rwanda should advocate for to pay for the research are e ­ ssential stronger regional Economic Communities’ to reaping the maximum returns to secretariats to review and discuss potential innovation. violations of common market protocols. The recently agreed African Continental Free Trade Area can also be advantageous, Reform Priority 2: Establish Export once details are penned and implementa- Dynamism and Leverage Regional tion modalities agreed to. Integration 2. Closely monitor the exchange rate Although exports have grown rapidly and to maintain export competitiveness . diversified over the past decade, the sector Greater flexibility and close monitoring remains too small and narrow to achieve the of exchange rates are necessary to avoid country’s aspirations for growth. Development inadvertent episodes of overvaluation, of a more robust export sector serving both especially in regional markets. It may be regional and global economies and a renewed important to require the accommoda- effort to attract foreign direct investment in tion of import surges or the use of reserve E x ecutive S u m m ar y xxv accumulation to sterilize capital inflows. the reallocation of labor off-farm, without A first step is to undertake more analyti- the accompanying agglomeration economies cal work on the past relationship between essential for rapid growth of productivity. exchange rate policy and export growth, Policy should focus on delivering the fun- particularly for dynamic export sectors. damental drivers that link urbanization to 3. Invest in economic diplomacy. More robust economic growth. This approach investment in staff knowledge is needed means focusing on where to prioritize urban regarding how to prioritize, prepare for, development, what package of policies and and undertake external negotiations. investments to include, and who needs to 4. Improve trade connectivity by lowering be in charge of implementing solutions. On transport costs. Building railway lines where, Kigali is Rwanda’s main interface along the two East African trade corridors with the world and needs to fulfill its sig- holds promise for reducing freight costs nificant economic potential following market but is expensive. Developing Rwanda as demand. This effort would be accompanied a regional logistics hub offers another by a more calibrated approach toward a way to lower transport costs. Lowering complementary (and well-networked) set of regional road tolls, ensuring their uniform secondary cities and small towns. On what, application across all EAC vehicles, and reforms to strengthen land markets and applying a standard tax policy for truck- urban planning are essential to enhance the ers from all countries would lower costs efficiency of necessary capital investments further. An aggressive push for open skies in infrastructure, housing, and commercial within the EAC would help to improve air structures. Finally, on who, stronger inter- connectivity and reduce the costs of air governmental coordination of economic and cargo. spatial planning is needed to avoid coordi- 5. Increase service sector productivity. nation failures in urban development. The Backbone services (such as transport, reform agenda has the following three main telecommunications, and professional ser- elements: vices) need fewer policy barriers to compe- tition for both domestic and foreign firms 1. Unleash Kigali’s potential. Kigali’s urban and improved access to EAC and South fabric is fragmented by small-scale, patchy African Development Community profes- land development, which undermines the sionals. Services trade also offers promi- potential to achieve agglomeration econo- nent opportunities for direct exports, mies; and raises costs. Addressing this particularly in tourism, but also in mining fragmentation requires integrating labor services, education, and business services. markets through efficient, affordable, inte- grated public transport; prioritizing dense central infrastructure (particularly roads) Reform Priority 3: Leverage the to crowd in dense private investment in Urban Dividend firms and houses; and promoting connec- Urbanization, managed well, can be an tivity to global product markets. important driver of growth by generat- 2. Stre ng the n l an d m arket s . Rwanda ing agglomeration economies that enhance may want to consider moving from the productivity, spur innovation, and foster quantity-based regulation of land use to ­ economic diversification. More than one- price-based allocation. This effort requires quarter of Rwanda’s population lives in strengthening institutions so that they can urban areas, which contributed more than independently and reliably value and pub- one-third of national structural change and licly disseminate land values across uses half of national labor productivity growth and assign and protect property rights. over the past 15 years. However, urbaniza- Credible land valuations would also enable tion has driven economic growth through urban areas to fiscalize public investments xxvi FUTURE DRIVERS OF GROWTH IN RWANDA in land through land value capture. 2. Improve targeting and monitoring of Moreover, regulations on structures industrial incentives. The effectiveness should be used to mitigate negative exter- of the government’s extensive industrial nalities like environmental degradation. policy interventions could be improved 3. Foster institutional coordination. The by shifting from general support to tar- main priorities are first to unite spatial and geting successful enterprises. Crucially, economic planning, second to strengthen a performance-based approach (focused cross-district and municipal coordination, on firm productivity and exports) should and finally to boost district capacity for be mainstreamed in all interventions. urban planning and land valuation. Industrial incentives also need to be coor- dinated ­ b etter across government agen- cies. Incentives also could do a better job Reform Priority 4: Enable the of attracting foreign direct investment in Emergence of Competitive Domestic tradable activities. Building and rigorously Enterprises implementing a credible performance mon- Strengthening the enabling environment itoring system are critical for this effort. for both private firms and SOEs to achieve 3. Define the future role of SOEs. An assess- efficient market-led outcomes is the corner- ment of the level of competition, the com- stone of improving the four future drivers of petitiveness of SOEs versus private firms, growth—innovation, integration, agglomera- economic development goals, and social tion, and competition. Competitive firms are considerations could be used to divide essential to reap the benefits of integration key sectors into four groups: (1) sectors and agglomeration and are also at the center in which SOEs will retain a monopoly; of innovation activity. This effort calls for a (2) sectors in which SOEs will compete strong state and a strong private sector, with with private firms; (3) sectors from which complementary functions. The following are SOEs will withdraw when efforts to build the main elements of the reform agenda: up private sector capacity prove success- ful; and (4) sectors from which SOEs will 1. Address cross-cutting constraints. The withdraw immediately because the pri- first imperative is to provide affordable vate sector is already capable and there is access to finance, probably best achieved no compelling social rationale for them. through regional (rather than national) Public-private boundaries can shift as the banking institutions. Over time, capital private sector gains strength. markets and nonbank financial institu- 4. Build an effective national innovation sys- tions need to be developed (again, prefera- tem. Wide-ranging policies are needed in bly as part of regional agreements). Efforts this regard. On the demand side, a strong to reduce logistics costs are important, competitive environment would encour- as is improved access to—and reduced age firms to seek out the best available cost of—broadband connectivity, backed knowledge and strengthen managerial by effective competition in key telecom- capabilities to introduce new processes munications markets. To reduce the cost and technologies, integrate them in the and increase the reliability of electricity, production system, allocate skilled staff Rwanda needs to push forward for the to use them, and make them financially establishment of a regional energy pool, viable. Key steps to help firms to meet build capacity to procure and implement technical standards would involve a more power-purchasing agreements, strengthen effective standards board; effective, more demand forecasting, and better integrate easily available, and affordable key ser- system planning and system operations vices; greater firm awareness of the ben- functions in the utility sector. efits of following standards; and better E x ecutive S u m m ar y xxvii collaboration among firms or between 1. Strengthen research and regulatory insti- firms and knowledge providers. Stronger tutions . These institutions should be coordination, monitoring, and evaluation adapted continually to evolving opportu- of funding programs for science, technol- nities and threats. Certifications, a key ele- ogy, and innovation are also needed. The ment of Rwanda’s comparative advantage, agenda on the supply side of innovation is have to remain credible. The high techni- discussed under reform priority 1. cal expertise of the Rwanda Agriculture 5. Maximize gains from the mining sector. Board will be increasingly important as Recent discoveries of mineral deposits bode the frequency and importance of sanitary well, but they require greater efficiency and and phytosanitary issues increase. attention toward raising fiscal revenues. 2. Strengthen vertical coordination. More It is necessary to provide extensive sup- effective vertical coordination in agri- port for artisanal miners, to improve the culture requires private sector partners bargaining position of the miners, and to to provide skills, capital, and entrée into provide well-targeted extension services. international markets. Private ­ s ector Undertaking detailed geological investiga- vertical coordination arrangements tions for promising deposits and develop- and changes in horizontal coordination ing a consolidated information system are through producer groups are likely to important initial steps. Other measures grow rapidly. Government promotion include improving land management, of more standardized approaches would incorporating social impact assessments, serve the interests of both farmers and strengthening the capacity of the Rwanda aggregators. Mining Board, and developing an optimal 3. Engage the region more effectively. fiscal regime to capture an efficient and Rwanda could engage more effectively and fair share of mineral wealth. opportunistically with the EAC, the South African Development Community, and the Common Market for Eastern and Southern Reform Priority 5: Transition Agriculture Africa and bilaterally with the Democratic and Food to Be an Engine of Growth Republic of Congo on a regional division The transformation of agriculture from of labor, both in production based on com- supplying commodities for domestic use parative advantage and in knowledge gen- to producing higher-value-added goods eration and diffusion. Regional approaches through regional and global food supply could help to satisfy the demand for food chains requires continued modernization staples; improve certification of food and greater responsiveness to market signals. safety, sanitary, and phytosanitary condi- Agriculture is central to the four drivers of tions; and adopt transparent science-based growth, because it is an important source standards, product registrations, and certi- of exports and a key driver of the pace at fication of agricultural inputs. which rural resources are released for urban 4. Profit from the big-data revolution in agglomeration; the transformation itself will innovation. The benefits of big data can be built on more competition and innovation be made more accessible to smallhold- within the sector. ers. “Smart farming,” where just the right Achieving higher productivity through amounts of the right inputs are used for ­ private sector–led growth requires using each parcel on the basis of information direct public support of the private sector gathered by handheld devices with the more strategically and strengthening the pub- right sensors, offers one such opportunity. lic sector’s role as a regulator, facilitator, and Big-data approaches such as blockchain provider of public goods. The following seven technology similarly offer the potential priorities are key: for lowering the costs of small financial xxviii FUTURE DRIVERS OF GROWTH IN RWANDA transactions that require secure record endeavor where improvements are necessarily keeping and decentralized input, such as incremental. land registration and mortgages. Considerable progress in restoring trust 5. Improve targeting of public investment between the country’s leadership and citi- in infrastructure . Public spending on zens, and more broadly within the society, agriculture is low, and greater resources has been achieved from the low level fol- are needed for irrigation, especially in lowing the 1994 genocide against the Tutsi. the drier eastern regions of the country. However, further progress is essential to However, such investments are high cost maintaining stability, promoting innovation, (due to Rwanda’s hilly landscape) and and achieving rapid growth. have to be targeted on high-return prod- Key recommendations are organized ucts (for example, horticulture exports). around the three pillars of state effectiveness 6. Develop a more robust policy on high- and societal trust, all of which are essential for level human capital formation in agri- rapid growth: culture. In particular, a more consistent approach to ver tical integration is 1. Build the state’s core capabilities on the needed for training managers and tech- basis of capacity, meritocracy, coordina- nical experts working for aggregators. tion, and rule-based authority. This effort Companies might be involved in this pro- requires empowering local governments cess by instituting training programs to and civil servants, which could mean build up nationally needed skill sets in adopting more flexible imihigo (perfor- agribusiness, high-value supply chains, mance contracts) with multiyear ­ targets; and agricultural technology. improving coordination of policies on key 7. Address land degradation and the impact cross-cutting issues; expanding the train- of climate change. Although considerable ing offered by the Rwanda Management progress has been made in construct- Institute, supported by skills audits and ing robust “bench” (wide) terraces, much functional reviews to identify key gaps; more needs to be done to secure land paying higher compensation to individu- assets. Proactive promotion of adaptation als with scarce skills; and developing to climate change is also vital. Fortunately, non­fi nancial incentives to attract compe- the soil and water management interven- tencies in the civil service. tions central to halting land degrada- 2. Strengthen market economy foundations. tion support adaptation through better Strengthening corporate governance of water retention and improved soil quality. SOEs, ensuring that they are not accorded Landscape restoration and conservation an advantage over private firms, opera- require the involvement of local govern- tionalizing the Rwanda Inspectorate and ment, producer groups, and national tech- Competition Authority. Creating special nical expertise and funding. courts and fast-track procedures to adju- dicate small claims; promoting alternative means of resolving commercial disputes; Reform Priority 6: Develop boosting reliance on technology; increasing Capable and Accountable training and specialization of justice sec- Institutions of Governance tor employees, including judges, prosecu- Rapid improvements in governance, includ- tors, and investigators; and improving case ing control of corruption, rule of law, regu- management techniques would enhance the latory quality, and civil service performance, judiciary’s effectiveness in enforcing con- give Rwanda a competitive edge over its tracts. The proper enforcement of expro- peers. Even so, its ambitious aspirations priation procedures, a more efficient land are bound to test its institutions severely. management system, and full enforcement Building effective institutions is a long-term of the law on intellectual property would E x ecutive S u m m ar y xxix strengthen property rights. Further, tighter 3. Eighty-five percent of students at the end of controls on public investment allocations grade 3 were rated “below comprehension” on and prudent fiscal management to build fis- a recent reading test, and one in six students cal policy buffers are essential to developing in grade 3 could not answer a single reading comprehension question. the state’s financial capacity. 3. Strengthen accountability of the state and ensure that policies and programs References are aligned with the needs of citizens. Dihel, N., and A. G. Goswami. 2016. The Enhancing the performance of key watch- Unexplored Potential of Trade in Services dog agencies, increasing revenue-raising in Africa: From Hair Stylists and Teachers by local governments, streamlining expen- to Accountants and Doctors. Washington, diture assignments, and ensuring greater DC: World Bank. https://openknowledge​ involvement of citizens in local decision .­worldbank.org/handle/10986/24968. making (for example, in setting imihigo Hallward-Driemeier, M., and G. Nay yar. objectives), coupled with reliance on more 2017. Trouble in the Making: The Future of M a n u fa c t u r i ng - L e d D e ve l o p m e n t . qualitative information in monitoring per- Washington, DC: World Bank. formance, could make local officials more IMF (International Monetary Fund). 2018. accountable and improve the quality of World Economic Outlook: Cyclical Upswing, services. Structural Change. Washington, DC: IMF, April. Newfarmer, R., J. Page, and F. Tarp, eds. Notes Forthcoming. Industries without Smokestacks: 1. Ninety percent of cropland is on slopes of African Industrialization Revisited. Oxford, 5 percent to 55 percent. U.K.: Oxford University Press. 2. There are important gains to be had from a OECD (Organisation for Economic Co-operation demographic dividend if the decline in the fer- and Development). 2017. “The Future of Global tility rate, and thus the population growth rate, Value Chains: ‘Business as Usual’ or ‘A New could be accelerated. A recent United Nations Normal’?” Directorate for Science, Technology, Population Fund report (UNFPA 2017) found and Innovation Policy Note, OECD, Paris. that Rwanda could boost its gross domestic Solow, Robert. 1987. “We’d Better Watch Out.” product growth 1 ­ percentage point by target- New York Times Book Review, July 12, 36. ing a lower dependency ratio, which is the UNFPA (United Nations Population Fund). 2017. number of people who are under and above the “Unlocking Rwanda’s Potential to Reap the working ages of 15–64 for every 100 persons Demographic Dividend.” Policy Brief, UNFPA, within the w ­ orking-age population. New York, October. Abbreviations AfCFTA African Continental Free Trade Area ASEAN Association of Southeast Asian Nations ASM artisanal mining AVE ad valorem equivalent BNR Banque Nationale du Rwanda (National Bank of Rwanda) BRD Banque Rwandaise du Développement (Development Bank of Rwanda) CET common external tariff CGE computable general equilibrium CIP Crop Intensification Program COMESA Common Market for Eastern and Southern Africa CSO civil society organization DRC domestic resource cost EAC East African Community EAPP East Africa Power Pool EICV Integrated Household Living Conditions Survey FDI foreign direct investment GDP gross domestic product GHG greenhouse gas HACCP hazard analysis and critical control point HRDC Human Resources Development Council HRDF Human Resources Development Fund IBES Integrated Business Enterprise Survey ICT information and communication technology IECMS Integrated Electronic Case Management System IER Institute of Engineers of Rwanda IFC International Finance Corporation IFRS International Financial Reporting Standards IPA investment promotion agency ITE Institute of Technical Education (Singapore) JADF Joint Action Development Forum KIC Kigali Innovation City KSEZ Kigali Special Economic Zone LUCP Land Use Consolidation Program FUTURE DRIVERS OF GROWTH IN RWANDA xxxi xxxii FUTURE DRIVERS OF GROWTH IN RWANDA MCC milk collection center MDE million dollar exporter MICE meetings, incentives, conferences, and exhibitions MINAGRI Ministry of Agriculture and Animal Resources MINALOC Ministry of Local Government MINEAC Ministry of East African Community MINECOFIN Ministry of Finance and Economic Planning MINICOM Ministry of Trade and Industry MINILAF Ministry of Lands and Forestry MININFRA Ministry of Infrastructure, Water and Sanitation MRA mutual recognition agreement MTRC Mass Transit Railway Corporation NAEB National Agricultural Export Development Board NCIP Northern Corridor Integrated Projects NCST National Commission of Science and Technology NECDP National Early Childhood Development Program NEET not in education, employment, or training NGO nongovernmental organization NIRDA National Industrial Research and Development Agency NIS national innovation system NST National Strategy for Transformation OAG Office of the Auditor General ODA official development assistance OECD Organisation for Economic Co-operation and Development OLPC One Laptop per Child PEIEX Industrial Extension Project for Exporting PIM public investment management PIRLS Progress in International Reading Literacy Study PISA Programme for International Student Assessment PPF production possibility frontier PPP public-private partnership/purchasing power parity PSDS2 Private Sector Development Strategy PSMB Pembangunan Sumber Manusia Berhad (Registration of Employers and Payment of Levy) PSTA Programs for Support to Agriculture R&D research and development RDB Rwanda Development Board RDF Rwanda Defense Forces RICA Rwanda Inspectorate and Competition Authority RMB Rwanda Mining Board RMC Rwanda Media Commission RMI Rwanda Management Institute RPF Rwandan Patriotic Front RRA Rwandan Revenue Authority RSB Rwanda Standards Board RSSB Rwanda Social Security Board RURA Rwanda Utilities Regulatory Authority SACCO savings and credit cooperative SADC Southern African Development Community SAM social accounting matrix A b b reviations xxxiii SDF Skills Development Fund SI sensitive items SMA Seoul Metropolitan Area SME small and medium enterprise SOE state-owned enterprise STEM science, technology, engineering, and mathematics TFP total factor productivity TFPR total factor productivity revenue TIMSS Trends in Mathematics and Science Study TVET technical and vocational education and training VAT value added tax WASH water supply, sanitation, and hygiene WGI Worldwide Governance Indicators Overview Future Drivers of Growth in Rwanda Innovation, Integration, Agglomeration, and Competition Introduction 7 percent per capita, average income would reach US$2,400 (2017 prices). To become Rapid economic growth is Rwanda’s over- an upper-middle-income country by 2035, arching development goal—a strategic choice Rwanda will need to grow at more than to anchor its long-term vision. Vision 2050 10 percent per capita. In 2035, the economic encapsulates this choice with long-term, landscape of Rwanda could resemble that of income-based goals that aim for upper-middle-­ present-day Bangladesh or, alternatively, sur- income status by 2035 and high-income sta- pass that of today’s Vietnam or even Georgia tus by 2050. With this vision, Rwanda has and Indonesia. It is not surprising, therefore, aligned itself with the successful East Asian that almost 70 percent of Rwandan respon- economies that began their development jour- dents in a recent survey mentioned high ney with a similar quest for high growth. The economic growth as their first priority for prioritization of long-term growth recognizes the country (World Values Survey, Wave 6: an important truth—sustained growth does 2010–2014; Inglehart et al. 2014). not just happen, especially in a global land- But economic growth is not only a matter scape marked by forces of technology, trade, of income. Growth matters for a broad range and tremendous competition. It requires a of other development outcomes, including combination of leadership, social cohesion, poverty reduction, as shown globally and and deep investments in core capabilities—of in Rwanda’s own experience as chronicled people, firms, and institutions—to harness the in this report.1 Sustained high growth has opportunities on offer. another, higher purpose for Rwanda: to The implications of different growth escape from its tragic past. The country is pathways are staggering. At its current pace keen to put great distance between a bright of growth (4 percent per capita), Rwanda future and a painful past. will barely cross the threshold for lower- This is not to say that economic growth middle-income status by 2035. At growth of should be the only area of policy focus. 1 2 FUTURE DRIVERS OF GROWTH IN RWANDA The challenges of inequality, public service long-term consequences. Of all the coun- delivery, and environmental sustainabil- tries in the world, only Mozambique had ity, among others, are just as important for lower per capita income in 1994, and none development and need to be looked at in had lower life expectancy at birth than parallel, as is being done in Rwanda. Rwanda. Rwanda’s income ambitions are built Facing these grim conditions, a new on a strong and widely acknowledged Government of National Unity took over record of success. Emerging from the dev- in 1994. Its task was made more difficult by astation of the 1994 genocide against the the fact that its coffers were running empty, Tutsi, which itself followed three decades infrastructure (social and physical) had been of economic stagnation, the country has decimated, trade links were broken, businesses seen its average income rise three-and- and agricultural assets had collapsed, and a-half-fold. In 1994, only Mozambique institutions of governance needed to be rebuilt. was poorer than Rwanda, which today is Insecurity and instability loomed large, both ahead of 20 countries. This rapid progress internally and from threats within the region, was made possible by the second-fastest and social trust had fallen to worrying depths. growth of gross domestic product (GDP) Even so, relief, recovery, and reconstruc- per capita on the continent, sustained over tion efforts were swift and multipronged. two decades. National reconciliation and healing evolved Past successes justify bold ambitions but through homegrown initiatives, includ- do not guarantee them. The reform agenda ing the establishment of a National Unity for accelerating growth to even higher levels and Reconciliation Commission in 1999 and then sustaining it is complex and highly and, eventually, the Gacaca courts, mod- demanding, as described in this report. eled after a traditional approach to set- tling disputes. In December 1994, the Government of National Unity issued and A Strong Start despite Initial implemented (and ­ s ubsequently closely Conditions and Emerging followed) a Declaration of Principles that outlined its political, social, and economic Concerns agenda for a “New Rwanda.” The decla- The genocide against the Tutsi in 1994 left ration emphasized social stability, national in its wake a shattered and traumatized security, and a commitment to a market nation. It was, above all, a human trag- economy, backed by a capable state and a edy, with few precedents. More than one strong private sector (World Bank 2007). ­ m illion people perished, leaving behind In the absence of a viable private sector, s uffering in indescribable forms and on ­ the state led the economic recovery through immeasurable scales. About two m ­ illion large public investments that were financed Rwandans sought refuge in neighbor- through external assistance. At the same ing countries, and at least one million time, to strengthen the foundations for were internally displaced (Chukwuma growth, a series of reforms liberalized the 2003). The economic and social costs economy, 2 promoted private investment, were enormous. The country’s GDP col- and sought to privatize some of the state- lapsed that year, falling in half from an owned enterprises (SOEs) that dominated already low base. Inflation climbed to industry. A multiphase decentralization more than 60 percent. Four out of every program was launched in 2000. five persons were living below the official These early efforts proved effective. poverty line. Life expectancy at birth fell The poverty rate had declined to 60 per- to below 30 years, and hunger and food cent by the early 2000s, on the back of a insecurity were widespread, with severe solid economic recovery. GDP had recovered OVERVIEW 3 to pre-genocide levels by 2000. And, by FIGURE O.1  Rwanda’s global rankings on Ease of Doing Business 2005, per capita income had surpassed pre-­ indicators genocide peak levels, as had many health and education indicators such as infant and child mortality, life expectancy, and pri- Registering property 2 mary school enrollment rates.3 Government Getting credit 3 revenues (excluding grants) had picked up to 14 ­percent of GDP by the mid-2000s (from Protecting minority investors 14 4 percent in 1994), which together with the Overall 29 consolidation of expenditures led to higher government and national savings. The task Paying taxes 35 of accommodating and reintegrating the influx of returning refugees had also pro- Starting a business 51 ceeded well. Resolving insolvency 58 With the dual post-genocide objectives of economic recovery and social and politi- Getting electricity 68 cal stability largely met, a strong platform was laid for long-term growth and devel- Enforcing contracts 78 opment. This agenda was picked up early Trading across borders 88 under Vision 2020 (adopted in 2000), with Dealing with far-reaching development targets that were construction permits 106 later backed by ambitious reform programs 0 20 40 60 80 100 120 implemented under a series of poverty reduc- Global ranking tion strategies. Reform efforts were most noteworthy in strengthening business regula- Source: World Bank 2019. tions, as reflected in Rwanda’s strong stand- ing on the World Bank’s Doing Business supported by robust performance in each Indicators, where Rwanda is ranked 29th of the major economic sectors. Industry in the world (figure O.1)—above all other (propelled by construction) and services low-income countries and second only to (driven by information and communication Mauritius on the continent (World Bank technology [ICT] and trade and transport) 2019). Decentralization was deepened fur- each has grown at an annual rate between ther, aimed at strengthening quality and 9 and 10 percent since 2006, and agricul- accountability in service delivery. Large-scale ture (led by crops and livestock) has grown public investments continued to close the at 5.4 percent. infrastructure gaps, particularly in energy, Broad-based sectoral growth facilitated telecommunications, and road transport. rapid structural transformation, meaning the Trade integration was also accelerated with movement of labor out of low-­ productivity Rwanda’s accession to the East African agriculture and into industry and services Community (EAC) in 2009, which signifi- and from farms to cities. Close to two-thirds cantly brought down tariffs—from an aver- of GDP growth since 2000 has been on age of 16.5 to 11 percent. account of structural transformation, with Economic and social achievements in the rest coming from growth within sectors. the post-recovery phase (2006 onward) The percentage of the population in urban have continued to be impressive. Growth areas increased from 16 percent in 2002 to of GDP per capita has averaged 5 percent 27 ­percent in 2015, and the urban popula- a year since 2006, second only to Ethiopia tion more than doubled (from 1.5 ­ million to on the continent (figure O.2). GDP growth, million). Because the agriculture sector had 3.5 ­ for the most part, has been broad based, such low labor productivity at the beginning 4 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE O.2  Average growth of GDP per capita in of the process, the transformation brought Rwanda and select countries, 2006–16 significant gains. With close to 70 percent of agriculture—the sector the labor force still in ­ China even now with the lowest labor productivity Ethiopia figure O.3)—­ (­ significant potential remains India for realizing continued gains from structural Cambodia Bangladesh transformation, which is still in its early stages RWANDA Vietnam Sudan in Rwanda. Mauritius Indonesia Ghana Rwanda has also seen impressive improve- ments in social indicators, including gender Mozambique equality (table O.1). On many of the health Philippines Poland indicators, Rwanda is now closer to the aver- age of lower-middle-income countries, well Zambia Turkey ahead of its low-income peers. Maternal Tanzania and child mortality rates have fallen Colombia 80–90 ­ percent, and life expectancy at birth Uganda Malaysia has more than doubled (to 69 years) in the Thailand Korea, Rep. past 20 years or so. Basic ­ i nfrastructure— Congo, Dem. Rep. from roads to telecommunications and Namibia Botswana energy supply— has improved solidly. Chile Kenya A nationwide rollout of health care, includ- Singapore Nigeria ing health insurance coverage for more than Malawi Egypt, Arab Rep. 80 percent of the population, is also impres- Russian Federation sive. Rwanda has surpassed several financial inclusion targets, with almost 90 percent of Pakistan the adult population having some access to Niger financial institutions. Crucially, the institu- Togo ­ overnance have been strengthened tions of g Cameroon significantly, earning the government a repu- Brazil tation for efficiency and probity (Ggombe Senegal and Newfarmer 2017). In some areas, progress has been less South Africa satisfactory. Some of the more visible gaps Mexico are in educational outcomes. Rwanda’s cur- Grenada rent levels of human capital and its current trajectory of investment in human capi- tal are not consistent with its ambitions. Ukraine In some areas—for example, stunting and primary and lower-secondary school com- Chad pletion rates—Rwanda lags behind the aver- age of low-income countries (figure O.4). Burundi Madagascar Achieving rapid economic growth and job creation will require a large increase in the Jordan quality of human capital. A second area of policy concern is the Yemen, Rep. low savings rate. The domestic savings rate –6 –4 –2 0 2 4 6 8 10 is less than 10 percent, well short of the Percent investment rate of 26 percent and behind that of many of Rwanda’s regional peers Sources: World Development Indicators data (World Bank, various years); National Institute of Statistics of Rwanda data. (figure O.5). The domestic savings rate, OVERVIEW 5 which tends to move with income levels, FIGURE O.3  Sectoral labor productivity and annual change in has remained virtually unchanged since the share of employment in Rwanda, 2000–16 mid-2000s, despite a 70 percent increase in per capita income (figure O.6).4 The gap 2.5 Financial GDP per worker from economywide GDP per worker) Sectoral labor productivity (log deviation in sectoral services and has been made up by external assistance real estate and, increasingly, by inflows of foreign 2.0 direct investment (FDI). To meet its growth Administrative and support objectives, Rwanda will need to raise its 1.5 Utilities services Manufacturing domestic ­savings rate to at least 30 percent 1.0 Mining and quarrying and to attract more FDI in tradable sec- Hotels and restaurants tors. 5 The high-growth East Asian econo- Culture and other services 0.5 Human health and social work mies, whose record Rwanda would like to surpass, had savings rates that were at least Transport services 0 double Rwanda’s at similar income levels. Determining the causes of Rwanda’s low Agriculture Construction Wholesale and –0.5 retail trade savings rate and identifying specific reform options require more ­ i n-depth analysis. –1.0 Further, continued reliance on external –2.0 –1.5 –1.0 –0.5 0 0.5 assistance, in any case, carries risks. As Average annual change in employment share a share of GDP, this assistance has been (percentage points), 2000–16 declining since 2004, and, as Rwanda Source: Derived from Diao, Randriamamonjy, and Thurlow 2017. approaches middle-income status, it can Note: The bubble sizes indicate the share of total employment in 2016. be expected to continue declining. Even if Rwanda can fill the gap through inflows A third challenge is low productivity. of private capital, the associated macro- Despite high GDP growth, Rwanda lags economic challenges are substantial if the other countries on labor productivity (out- gap continues to be large for an extended put per worker) (figure O.7). Rwanda’s per- period. formance on this measure is explained to a TABLE O.1  Social and economic indicators in Rwanda compared with low- and middle-income-country averages, 1994 and 2015 Rwanda Low- and Initial Low-income- middle-income- condition Latest country country average Indicator (1994)a (2015)a average (2015)a (2015)a % of population below national poverty line 80 < 40 (2014) n.a. n.a. Gini indexb 0.49 (2000) 0.50 (2014) n.a. n.a. Immunization, measles (% of children ages 12–23 months) 76 (1996) 96 76 80 Improved sanitation facilities (% of population with access) 39 62 28 52 Improved water source (% of population with access) 62 76 66 90 Births attended by skilled health staff (% of total) 27 (2000) 91 49 (2012) 59 (2012) Maternal mortality ratio (modeled estimate, per 100,000 live births) 1,270 290 496 254 Mortality rate, under-five (per 1,000 live births) 284 41 76 53 Life expectancy at birth, total (years) 31 69 64 70 Secondary school enrollment (%, gross) 9 (1999) 37 n.a. n.a. Literacy rate, adult females (% of females ages 15 and above) 49 (1991) 68 49 (2010) 66 (2010) Literacy rate, adult males (% of males ages 15 and above) 68 (1991) 75 66 (2010) 80 (2010) Sources: World Development Indicators data (World Bank, various years); National Institute of Statistics of Rwanda data. Note: n.a. = not applicable. a. Unless specified otherwise. b. The Gini index is a measure of income inequality. Zero represents perfect equality, and 1 represents the most extreme inequality. 6 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE O.4  Human capital outcomes in Rwanda FIGURE O.6  Rwanda’s domestic savings and and other countries, by income level, 2015 investment rates, 1995–2015 30 Stunting rate 20 Primary % of GDP completion rate 10 Lower-secondary 0 completion rate –10 0 20 40 60 80 100 Percent 95 97 99 01 03 05 07 09 11 13 15 19 19 19 20 20 20 20 20 20 20 20 Rwanda Low income Net ODA received Net FDI in ows Middle income High income Domestic investment rate Gross domestic savings rate Source: World Development Indicators data (World Bank, various Sources: Calculations based on data from World Development years). Indicators (World Bank, various years) and National Institute of Statistics of Rwanda data. Note: FDI = foreign direct investment; ODA = official development assistance. FIGURE O.5  Average gross domestic savings rate FIGURE O.7  Labor productivity and GDP per for Rwanda and regional comparators, 2014–16 capita in Rwanda and other countries 100 25 90 Labor productivity (2010 US$, PPP) 80 20 70 60 50 Savings rate 15 40 30 10 20 10 5 0 400 4,000 40,000 GDP per capita (2010 US$, PPP) 0 All countries Rwanda Ethiopia Tanzania Uganda Kenya Rwanda Source: Calculations based on Penn World Tables 9.0 data (Feenstra, Inklaar, and Timmer 2015). Sources: Calculations based on data from World Development Note: PPP = purchasing power parity. Indicators (World Bank, various years) and National Institute of Statistics of Rwanda data. OVERVIEW 7 large extent by weak total factor productivity FIGURE O.8  Total factor productivity (TFP) and GDP per capita in (TFP) (­ figure O.8). TFP growth has fallen to Rwanda and other countries just 1 percent since 2008. Although resources have moved broadly from agriculture to 1.5 other sectors with higher labor productivity, 1.3 the allocative efficiency within sectors has been suboptimal in Rwanda. Within-sector 1.1 TFP (United States = 1) labor productivity either fell or rose only 0.9 slightly across most sectors from 2001 to 2016 (­ figure O.9). 0.7 Fourth, the private sector still maintains 0.5 a relatively limited presence, has shown low capacity for innovation, and lacks the scale 0.3 economies that are crucial for productiv- 0.1 ity. In large part, this reflects the historical legacy, whereby the formal private sector was –0.1 400 4,000 40,000 decimated after the genocide. Rwanda has GDP per capita (2011 US$, PPP) since been successful in streamlining business All countries Rwanda Log of all countries regulations, as seen, for example, in its high ratings on the Doing Business Indicators. But Source: Calculations based on Penn World Tables 9.0 data (Feenstra, Inklaar, and Timmer 2015). the work is by no means over. A key issue is Note: PPP = purchasing power parity. that costs faced by businesses are higher in Rwanda than in economies at similar stages of their development. For example, the high- FIGURE O.9  Average growth of labor productivity in Rwanda, growth East Asian economies had Rwanda’s by sector, 2001–16 cost levels only after they had crossed the upper-middle-income-country threshold Agriculture (­f igure O.10). 6 The high costs of energy, Transportation services finance, and trade logistics have been impor- Health and social services tant contributing factors. Private investors Hotels and restaurants will also look for stronger economic returns, Wholesale and retail trade for which a pickup in productivity growth Mining and quarrying would be essential. Fifth, in the midst of high growth and Culture and other services rapid structural transformation, the per- Financial services and real estate formance of the manufacturing sector has Administrative and support services been subpar. Growing at a pace just short Utilities of the overall economy, the sector’s share Manufacturing of total value added has dropped to just Construction 6 ­ percent—2 percentage points lower than at –3 –2 –1 0 1 2 3 4 5 the turn of the century. Only 4 percent of the Average growth rate in labor force is in manufacturing (two-thirds output per worker (%) in agro-processing), and even that share has been declining (World Bank 2015b). Source: Calculations based on Diao, Randriamamonjy, and Thurlow 2017. As a result, Rwanda has fallen behind in ­ generating manufacturing value addition (at about 20 percent) for a small economy (­figure O.11). seeking high growth. Goods exports have Finally, Rwanda’s outward orientation has edged up only slightly (from 4.8 percent of not kept pace with its economic objectives. GDP in 2005 to 8.4 percent in 2015), and The ratio of exports to GDP remains low much of the increase was accounted for by 8 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE O.10  Price levels and GDP per capita in Rwanda and (largely unprocessed) commodities such as comparator countries and economies, 2011 coffee, tea, and reexported minerals. Value addition from manufactured exports has 0.9 not made a major contribution to early- stage growth, even though the export base 0.8 has become more diversified. Export of ser- vices (dominated by tourism) has risen at an Index (level of U.S. GDP in 2011 = 1) 0.7 uneven pace, with their share of GDP rising 0.6 sharply from 2002 to 2007 and falling there- 0.5 after, finally settling in at about 6 ­percent. FDI in export-oriented activity also has been 0.4 relatively low, again despite Rwanda’s very strong performance on the Doing Business 0.3 Indicators and efforts to attract FDI. 0.2 0.1 Future Aspirations 0 Building on its strong record, Rwanda’s 500 5,000 50,000 f ut u re aspi rat ions a re a mbit ious , as GDP per capita (2011 prices, PPP) reflected in the country’s Vision 2020 and Vietnam Rwanda China Vision 2050 currently under preparation. India Korea, Rep. Taiwan, China These documents, products of extensive national consultations, are being opera- Source: Calculations based on Penn World Tables 9.0 data (Feenstra, Inklaar, and tionalized by a series of economic devel- Timmer 2015). Note: PPP = purchasing power parity. opment and poverty reduction strategies and seven-year government programs. For 2018 to 2024, they are being combined into a single process, the National Strategy for Transformation. FIGURE O.11  Manufacturing value added in the East African Community, 2006–16 Vision 2020 set the country on an ambi- tious course. Achieving middle-income 15 status and reducing the poverty rate to 20 percent by 2020 were among its key 14 objectives. The vision identified six pillars to Manufacturing value added (% GDP) 13 achieve its goals: (1) human resource devel- 12 opment and a knowledge-based economy, (2) regional and international economic 11 integration, (3) infrastructure development, 10 (4) a ­private sector–led economy, (5) produc- 9 tive and ­ m arket-oriented agriculture, and (6) good governance and a capable state. 8 Gender equality, environmental sustain- 7 ability, and long-term commitment to sci- 6 ence and technology were the cross-cutting themes to support the six pillars. 5 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Vision 2050 aspires to take Rwanda to upper-middle-income-country status by Kenya Sub-Saharan Africa Uganda Burundi Tanzania Rwanda 2035 and high-income status by 2050, with the intention of providing productive eco- Source: World Development Indicators data (World Bank, various years). nomic opportunities and higher-quality OVERVIEW 9 living standards to all Rwandan citizens FIGURE O.12  Projected GDP per capita for Rwanda under alternate (Government of Rwanda 2017). These aspi- growth scenarios, 2016–50 rations translate into double-digit average annual growth rates (more than 10 percent 14,000 in per capita terms), requiring Rwanda to 12,000 High-income cuto ($12,476) grow faster than China or the Republic of Korea at similar stages of their development 10,000 f igure O.12). If growth falls even slightly (­ US$, 2016 prices short of this, Rwanda’s upper-middle- 8,000 income-­ country ambitions would be pushed back by at least a decade. The aspirational 6,000 growth scenario has major underlying Upper-middle-income cuto ($4,035) 4,000 requirements, as described in box O.1. To achieve these ambitions, Vision 2050 2,000 adopts four broad priorities, which also Rwanda’s GDP per capita in 2016 underpin the design, policies, and actions 0 of the National Strategy for Transformation: 20 6 20 8 20 0 20 2 20 4 20 6 20 8 20 0 20 2 20 4 20 6 20 8 20 0 20 2 20 4 20 6 20 8 50 1 1 2 2 2 2 2 3 3 3 3 3 4 4 4 4 4 20 Rwanda, aspirational India 1. Higher quality of life and standard of China Thailand living. To move beyond meeting basic Korea, Rep. Botswana needs to ensuring a higher standard of liv- Taiwan, China ing for all people, Rwanda will focus on Sources: Calculations based on World Development Indicator data (World Bank, various (1) access to affordable high-quality edu- years); Penn World Tables 9.0 data (Feenstra, Inklaar, and Timmer 2015) cation and health care; (2) modern hous- Note: Alternate GDP per capita trajectories if Rwanda were to grow at the pace of other countries from a similar income level as Rwanda in 2016. ing and settlements with environmentally friendly and climate-resilient surround- ings; (3) comprehensive adequate social 4. Positive contributions to international security and safety nets; (4) universal peace and prosperity. Rwanda will forge access to daily amenities; and (5) compre- its own place in the world in the context hensive personal security and safety. of regional integration, multilateral and 2. Transformation for prosperity (develop- bilateral cooperation, freedom from aid ment of high-value and competitive jobs dependency, pan-Africanism, and South- and sectors). To improve productivity South cooperation. and competitiveness, Rwanda will tar- get diversified tourism, manufacturing These objectives will be underpinned by driven by competitive local industries, strong and sustainable macroeconomic fun- business and financial services, informa- damentals, effective institutions, and positive tion technology, logistics and aviation, Rwandan values. The values underpinning agroprocessing, science and technology economic and social progress are self-reliance innovation, construction, and extrac- and self-determination, dignity, unity and tive industries. All of these efforts will be Rwandan identity, integrity, equity (including underpinned by high-quality services in gender and youth), transparency and open- public and private sectors. ness, participation in the global community, 3. Development of modern infrastructure good governance and accountability, com- and productive livelihoods. This effort munity participation, local innovation, and would involve modernization with smart national stability. green cities, towns, and rural settlements; Centered on the Rwandan people, Vision well-designed transport facilities and 2050 considers Rwandans as the country’s services; and efficient public and private main resource, reflects the society’s core val- services. ues, recognizes a strong collaborative role 10 FUTURE DRIVERS OF GROWTH IN RWANDA BOX O.1  Major requirements of Rwanda’s income aspirations The aspirational income targets of Vision 2050 quality of education and worker skills. A much require substantial increases in savings and invest- stronger external orientation also would be nec- ment, as illustrated in the aspirational high-growth essary, with the ratio of exports to GDP more scenario summarized in table BO.1.1. The sce- than doubling to well over 40 percent by 2035. nario assumes that gross domestic product (GDP) The aspirational high-growth scenario would growth rises to 12.5 percent a year (10.2 percent further require significantly higher labor produc- in per capita terms) by 2022 and stabilizes at that tivity growth—aggregate and within each of the level for the foreseeable future (the slightly lower major sectors. The demands on industry would GDP growth rate in table BO.1.1 reflects slower be especially strong, with labor productivity in the growth from now until 2022). As a result, GDP sector having to switch from a declining trend in per capita rises to just over US$4,000 in 2035 and the past decade to solid 6 percent annual growth to more than US$12,000 in 2050, which is the in the coming period. Labor productivity growth in threshold currently used to determine high-income agriculture and services also would have to accel- status. The investment rate would have to rise to erate significantly. Even with strong within-sector about 40 percent, in a context of declining foreign improvements, almost 70 ­ p ercent of the overall assistance. The slack would have to be picked up growth of labor productivity in the next couple of by increased private investment, because public decades would be generated by structural transfor- investment is already at the limits set by financing mation, reflecting massive movement of labor out of options. Further, the savings rate would need to agriculture and into industry (mining, construction, increase four- to fivefold (from 8 percent in 2016) in manufacturing, utilities) and services. the next two decades to finance these much higher The requirements of less ambitious income levels of investment. growth trajectories (by matching, for example, the Total factor productivity growth also would growth rate of China or the Republic of Korea or need to accelerate rapidly (tables BO.1.2, BO.1.3, the slightly lower growth rate of Botswana, India, and BO.1.4)—to about 6 percent a year—and to or Vietnam at a similar stage) would be slightly less account for two-thirds of future growth. Another onerous, but not by much. Rwanda would require crucial requirement is greater human capital, significantly higher savings and private investment calling for strong improvements in the level and rates and productivity growth. TABLE BO.1.1  Aspirational high-growth scenario for Rwanda: Demand-side requirements, 2000–35 Share of GDP (%) Average annual growth (%) Indicator 2016 2025 2035 2000–16 2016–35 Gross domestic investment 28 39 40 12.9 13.6 Public 12 10 8 n.a. n.a. Private 16 29 32 n.a. n.a. Domestic consumption 92 77 62 7.3 9.2 Gross domestic savings 8 23 38 n.a. 21.0 External transfers (net) 4 3 2 n.a. 6.9 Gross national savings 12 26 40 n.a. 18.6 Exports of goods and services 17 27 46 12.4 17.7 Imports of goods and services 37 43 48 10.8 13.1 Trade balance −20 −16 −2 n.a. n.a. Current account balance −16 −13 0 n.a. n.a. Sources: Estimates based on World Development Indicators data (World Bank, various years); National Institute of Statistics of Rwanda data. Note: n.a. = not applicable. (Box continues next page) OVERVIEW 11 BOX O.1  (continued) TABLE BO.1.2  Aspirational high-growth scenario for Rwanda: Productivity requirements Average annual growth (%) Indicator 2000–14 1999–2008 2009–16 2015–35 2017–35 Labor productivity 5.3 — — 8.1 — Agriculture 3.1 — — 5.0 — Industry −2.8 — — 6.1 — Services 3.2 — — 6.1 — Total factor productivity — 4.3 1.1 — 6.0 Sources: Estimates based on World Development Indicators data (World Bank, various years); National Institute of Statistics of Rwanda data. Note: — = use of different time periods for labor productivity growth and total factor productivity growth. TABLE BO.1.3  Aspirational high-growth scenario for Rwanda: Sectoral share of employment, 2000–35 Share of total employment (%) Average annual growth (%) Indicator 2015 2025 2035 2000–14 2015–35 Total employment 100 100 100 2.5 2.8 Agriculture 66 49 36 0.6 −0.2 Industry 9 16 27 14.6 8.6 Services 24 36 37 9.4 4.8 Source: Estimates based on National Institute of Statistics of Rwanda data. Note: The assumed decline in average growth in agriculture from 2000–14 to 2015–35, despite higher labor productivity growth in the sector, in part reflects the massive movement of labor to other sectors. TABLE BO.1.4  Aspirational high-growth scenario for Rwanda: Sectoral share of GDP, 2000–35 Share of total GDP (%) Average annual growth (%) Indicator 2015 2025 2035 2000–14 2015–35 Total GDP 100 100 100 8.0 10.1 Agriculture 30 16 9 5.4 3.8 Industry 19 27 38 10.0 14.1 Services 51 57 53 9.6 10.1 Source: Estimates based on National Institute of Statistics of Rwanda data. Note: The assumed decline in average growth in agriculture from 2000–14 to 2015–35, despite higher labor productivity growth in the sector, in part reflects the massive movement of labor to other sectors. for both the state and the private sector, and is envisioned as the innovation, telecommu- signals an outward reach to make up for the nications, financial, and logistics hub for small domestic market and to access criti- East Africa. cal knowledge and ideas. Strong connective Rwanda in 2050 will be a vastly altered infrastructure and a focus on efficient urban place—an innovative, knowledge-based, and development are viewed as being crucial for globally integrated economy with a highly future growth. The vision further envisages a creative population and world-class enter- dynamic knowledge-based economy, with the prises. Not only will it be substantially richer, service sector playing a leading role: Rwanda but also its people will be universally endowed 12 FUTURE DRIVERS OF GROWTH IN RWANDA with much higher levels of human capital, middle-­i ncome countries. The innovations and its cities, hosting more than 60 percent are affecting what is produced, how it gets of the country’s population, will be brim- produced, and where it is produced—not just ming with economic vitality and serving as for advanced goods but also for more tradi- continental centers of specialized knowledge. tional manufactured goods. High levels of societal trust will be the glue New technologies associated with indus- holding the country firmly together. Rwanda try 4.0 —or the fourth industrial revolu- also will have comfortably transitioned to tion, including industrial automation and becoming an open economy (akin to today’s advanced robotics, digitalization and integra- Singapore)—with cross-­ border trade, con- tion of Internet-based systems (the Internet of nective infrastructure, investment flows, Things), and additive manufacturing (three- and people-to-people exchanges marking dimensional printing)—are part of a larger close economic ties across the continent and trend of automation and data exchange in beyond. Drawing on the vitality of its cities economic activity. For example, consumer and the highly developed capabilities of its goods are increasingly embedded with elec- people, Rwanda will transition to becoming a tronics, software, sensors, and network con- continental hub for higher-end services, hav- nectivity to enable these objects to connect ing maximized the industrial opportunities and exchange data, which is changing manu- along the way. The institutions of governance facturing. The service sector, traditionally a will evolve into even higher forms of capa- nontradable sector, is becoming a major part bility and accountability, with the country of the global trade story. ranked alongside high-income nations on key Technology and the information revolu- measures of governance. tion are also enabling disruptive innovations in business models. Apple is an early (and highly successful) example, with its disrup- Opportunities and Risks tion of a wide range of industries, including Global Megatrends personal computers, music, movies, media, and telecommunications. Other prominent Global megatrends of technological innova- examples include digital platform com- tions, shifting patterns of production, and panies like Airbnb and Uber, which have changing sources of cross-border investments ­ f undamentally altered the hospitality and have major implications for economic activ- taxi industries, respectively. Increased auto- ity in Rwanda and elsewhere. Global climate mation is further affecting production loca- change is another major trend, with significant tions, enabling, for example, some leading downside risks for all countries, but especially firms, albeit still in small measure, to return for those in the low- and middle-income world labor-intensive manufacturing back to high- that are reliant on agriculture and have lim- income economies and closer to consumers. ited resources to mitigate the impact. Rwanda Their disruptive nature notwithstand- can shape the net benefits of these forces in its ing, technological innovations, for the most favor with early and decisive action to develop part, bring upside opportunities for low- and the capabilities of its workers, firms, and pub- middle-income countries. There is evidence lic institutions, the right investment decisions, that more widespread use of scale-neutral and more effective business regulations. digital technologies, such as ICT, allows firms in emerging economies to access wider Technology Megatrends markets by cutting entry costs and by reduc- Global technological and business innova- ing the impact of distance.7 For example, tions, powered by the information revo- scale is expected to matter less with addi- lution, are increasingly disrupting the tive manufacturing technologies, such as patterns of production and trade across the three-dimensional printing. Combined with world, with major implications for low- and the demand for customized, fast-delivered OVERVIEW 13 goods, widespread use of digital technolo- Consequently, cheap labor as a source of gies could lead to geographically dispersed competitive advantage is increasingly giving production, potentially helping small econo- way to a more demanding ecosystem. It is mies like Rwanda to overcome some of their still not a zero-one choice for global produc- scale and locational constraints.8 ICT in the ers, which remain open to producing in both Internet of Things space, such as big data high-income and low- and middle-income and cloud computing, can similarly reduce countries. But their requirements are getting the impact of distance on firms’ competitive- more stringent with regard to infrastructure, ness. This, too, is positive news for Rwanda, available skills, logistics, other backbone given its landlocked geography and distance services, regulations, supplier base, and intel- from markets. The Rwandan government lectual property rights. There is also growing aims to capitalize on this emerging trend demand for low- and middle-income-country by continuing to invest in network connec- firms to adopt new technologies to remain tivity and sensor deployment in different competitive, but to do so they need to have applications. greater capacity to use the new technolo- In services, the information revolution has gies, raising the bar further in terms of the allowed low- and middle-income countries to mentioned prerequisites. go beyond traditional export sectors such as An early mover in some areas, Rwanda tourism and transport to more modern ones may be well placed to take advantage of these such as health care and business. These new trends. The cell phone revolution, for example, export products can be exported electroni- has lessened the need for landline connectivity, cally (taking distance out of the equation), reducing the potential cost of investing in such can achieve scale economies, and are an effec- legacy technologies.9 Similarly, drones can tive channel for technology diffusion. The potentially reduce the need for air t ­ ransport– spread of productivity-enhancing characteris- related infrastructure. The much-cited tics in services expands the range of activities example of Zipline, which uses unmanned available as pathways for development. drones to make 50–150 daily deliveries of Modern technological innovations also critical medical supplies to various locations pose major threats. To the extent that they are across Rwanda, is perhaps just the begin- labor efficient and skills intensive, the concern ning. Furthermore, commercial drone services is that they will narrow the paths for low- could transport items other than emergency and middle-income countries to benefit from supplies.10 The impact is already being felt in manufacturing. Some observers have raised agriculture. e-Soko—an electronic platform the specter of premature deindustrialization that gives farmers, consumers, and traders (Rodrik 2015). The implication is not just up-to-date market price information by short that machines will displace unskilled labor, message service—is widely used in Rwanda, but also that production will be retained in enabling farmers to market their agricultural higher-income economies. In addition to produce better and to get premium prices. reshoring of manufacturing activity from ICT-enabled modern services could enable low- and middle-income countries back to Rwanda to leapfrog—at least in part—the high-income countries, there is also a concern stage of manufacturing-led development in that the expected migration of labor-intensive the conventional process of structural change. activities from China to poorer economies Business process outsourcing services, for with lower labor costs might not happen or example, can be developed without a manu- might happen only in select locations that facturing core, with the potential to become a form part of China’s regional value chains. major export area. To derive the most benefits, Technological innovations also threaten to Rwanda will have to build its human capital exacerbate inequality, both within and across rapidly, enable the emergence of vibrant and countries, as countries in a position to lever- competitive domestic enterprises, and invest age technology gain higher incomes. further in building resilient market institutions. 14 FUTURE DRIVERS OF GROWTH IN RWANDA BOX O.2  The importance of regional value chains for growth and outward orientation in East Asia Light manufacturing has long been seen as a cata- into full package production. Some were eventually lyst for economic development. It is labor intensive, able to develop and commercialize brands of their is capital parsimonious, and offers viable develop- own (Gereffi 1999; Gereffi and Memedovic 2003). ment pathways for low- and middle-income coun- In the process, they offshored lower-value-adding tries. Various economies in East Asia, including activities to other countries in the region, which the Republic of Korea and Taiwan, China, began eventually upgraded into higher-value-adding func- their engagement with buyer-driven value chains in tions themselves, sending their previously held com- apparel and footwear, among others, by importing petencies elsewhere (Akamatsu 1962). The shift intermediate goods and assembling them into final of manufacturing activity from coastal China to products, often in export-processing zones. Over Vietnam is the latest step in this “flying geese” pat- time, firms in these countries were able to upgrade tern of industrial development. FIGURE O.13  Measures of global vertical integration in Rwanda, development (box O.2). These trends have 1995–2014 brought tremendous benefits to participating countries, seen in their growing heft in the 120 world economy: the share of low- and middle- income G-20 countries11 in world GDP rose 110 from 11 percent in 1990 to 28 ­ percent in 2016. China’s share alone increased from less than 100 2 percent to 15 percent, as its share of global manufacturing rose from less than 5 percent Percent 90 in 1990 to 10 percent in 2016. The world economy is now experienc- 80 ing structural shifts that could dramatically change the outlook for global value chains in 70 the coming years (OECD 2017). On the one hand, global value chains are maturing and 60 losing momentum because the “unbundling of production” has largely already happened, 19 5 19 6 19 7 19 8 20 9 20 0 20 1 20 2 20 3 20 4 20 5 20 6 20 7 20 8 20 9 20 0 20 1 20 2 20 3 14 9 9 9 9 9 0 0 0 0 0 0 0 0 0 0 1 1 1 1 19 as seen in the flattening of the share of foreign Foreign value added, share of exports per gross exports (WIOD 2013) value added in gross exports of goods and Foreign value added, share of exports per gross exports services since the onset of the 2008 global (WIOD 2016) financial crisis (figure O.13). On the other Manufacturing import content, share of manufacturing hand, new forces are reorienting produc- exports per gross exports (WIOD, WITS) tion within global value chains. Production costs have risen significantly in some low- Source: Hallward-Driemeier and Nayyar 2017. Note: WIOD = World Input-Output Database; WITS = World Integrated Trade Solution. and middle-income economies, eroding their competitiveness in labor-intensive manufac- Regional Value Chains turing activities and driving the transition Fragmentation of production into global from labor- to capital-intensive industries.12 value chains has created a critical oppor- Digitalization of production is also shaping tunity for countries to develop through global value chains by reorienting global pro- export-led ­ manufacturing. In East Asia, for duction and trade closer to demand.13 example, several countries have leveraged Production is becoming increasingly con- these opportunities to achieve export-led rapid centrated in regional and local hubs closer OVERVIEW 15 to end markets. This concentration has and exports and services trade. These oppor- implications for manufacturing-led growth tunities apply in particular to agribusiness and in Africa, where cross-border production food processing, subsectors in which Rwanda networks have yet to materialize because of currently has a revealed comparative advan- weak regional integration. Enormous oppor- tage (Hallward-Driemeier and Nayyar 2017). tunities for cross-border trade in food prod- These activities employ unskilled workers and ucts, basic manufactured goods, and services have relatively low intensity of physical capital therefore remain unexploited because of high and research and development. The products transport costs, nontariff barriers, and regu- typically are traded regionally (rather than latory constraints (Brenton and Isik 2012). internationally) because they are bulky to Enhanced regional integration is important transport (for example, wood products, bev- to derive the benefits on offer. The conti- erages) or require proximity to raw materials nent also will need to keep a close eye on the (for example, food production). The region rapid technological innovations and changing also offers major opportunities for trade in sources of FDI (discussed next). services, which already forms more than half T h e g r o w i n g r e g i o n a l i z a t i o n - i n -­ of Rwanda’s export earnings. There is sig- production trend can work in Rwanda’s nificant potential to boost services exports favor. The region around it abounds in further, while benefitting from the greater (largely untapped) opportunities (box O.3). competition, lower prices, and improved qual- Moreover, the ability of Rwandan firms to ity that come with services imports. compete globally depends on the “competi- tiveness” and “connectedness” of its neigh- Changing Sources of Outward FDI bors. Rwanda is therefore likely to seize The sources of FDI are changing, a trend that regional trade opportunities through intra- is likely to continue (figure O.14). In response regional trade and participation in regional to higher labor costs, labor-­ i ntensive man- value chains connected to the global market. ufacturing is moving out of China, and The strongest regional opportunities for with it new opportunities are emerging for Rwanda exist in commodity-based processing other low- and middle-income countries. BOX O.3  Opportunities in the region around Rwanda The regional economic blocks around Rwanda of Rwanda). Revived as a free trade area in 2000 abound in opportunities for trade, investment, and (following dissolution of a previous arrangement in transfer of ideas. They also offer scale economies 1977) and upgraded to a customs union in 2005, and specialization in production, which would be the community aims to create a common currency hard for Rwanda to achieve by relying solely on its and aspires to establish itself as a full political fed- own small, landlocked economy. eration. EAC members have duty-free preferential The most significant regional opportunity arises access to the United States through the African from Rwanda’s membership in the East African Growth and Opportunity Act and to the European Community (EAC), an ambitious platform for eco- Union through Everything But Arms. nomic, political, social, technological, and security The EAC Customs Union entails three key aspects: cooperation for six Great Lakes countries: Burundi, duty-free trade within the EAC; common external Kenya, Rwanda, South Sudan, Tanzania, and tariffs (CETs) on imports from a third country; a Uganda. The combined population of the EAC is and common customs procedures. The securities 170 million (about 15 times that of Rwanda), and exchanges of four countries (minus Burundi, which its combined gross domestic product (GDP) in 2016 does not have one) are members of the East African was US$163 billion (about 18 times larger than that Securities Exchanges Association. Other areas of (Box continues next page) 16 FUTURE DRIVERS OF GROWTH IN RWANDA BOX O.3  (continued) cooperation include a regional power pool, transport Free Trade Area (AfCFTA). COMESA, a free trade links, and trade facilitation. EAC-wide competition area with Rwanda as one of its members, has 40 times policy and law are in place. Rwanda’s population and 80 times its GDP. The Despite this good progress, several impedi- AfCFTA, an even bigger opportunity, is a pan-African ments prevent more meaningful economic integra- push to promote and harmonize trade liberalization tion. First, free trade is hampered by a long list of across the continent’s subregions. Once operational, duty exemptions and significant nontariff barriers. it will be one of the world’s largest free trade areas, Second, lack of policy harmonization holds back potentially involving 55 nations (only 44 have signed cross-border infrastructure development (Dihel, the agreement thus far) with a combined popula- Fernandes, and Mattoo 2012). Third, although tion of more than 1.2 billion and GDP of more member countries have committed themselves to than US$2 trillion (UNCTAD 2018). The United developing harmonized and complementary trade Nations Economic Commission for Africa estimates and transport policies, these commitments gen- that intra-African trade could increase more than erally remain on paper only.b Fourth, the region 50 percent under the AfCFTA if import duties were is still some way from enabling free cross-border eliminated, and more than 100 percent if nontariff mobility of labor and capital. Issues such as perma- barriers were also reduced (UNECA 2018). nent residency and the right of access to, and use Finally, there is significant untapped market of, land remain subject to national policies rather potential to the west in the Democratic Republic of than being part of the common market protocol. Congo. While conflict in the Democratic Republic Fifth, EAC member states still have to harmonize of Congo limited trade prior to 2007, exports have customs procedures in practice. Moreover, customs increased considerably in the last decade. By 2016, revenues are not pooled.c Finally, regional institu- Rwanda exported more goods to the Democratic tions are weak, and enforcement mechanisms are Republic of Congo than to the EAC. The main especially so. exports include livestock and crops, but there also Further opportunities lie in the Common Market is significant (informal) cross-border trade in ser- for Eastern and Southern Africa (COMESA) and vices such as finance, transportation, and wholesale the expected opening of the African Continental trading. a. Currently, there are three CET rates: 0 percent for raw materials, 10 percent for intermediate goods, and 25 percent for finished goods. b. For example, regulations on vehicle dimensions, axle-load limits, road transit charges, and highway codes have yet to be harmonized. Even common definitions of road classes and route numbers are missing. Similarly, rail connectivity is impeded by minimal integration of national technical standards, such as those for building and maintaining railway facilities. Shipping on inland waterways and lakes would benefit from common regulations on ship registration as well as safety standards, including those on periodic ship surveys, staffing requirements, and aids to navigation and radio communication (World Bank 2015a). c. Customs operations, including revenue collection, are managed by national authorities, creating delays and increasing transaction costs. FDI outflows from East Asia rose 7 percent 2000, as has the stock of Chinese FDI in in 2016 to US$363 billion, mainly because the continent. In 2015, 4 of the top 10 FDI of new outflows from China. On the back investors in Africa were Asian economies, of surging cross-border mergers and acqui- led by China.15 Notably, greenfield FDI sitions by Chinese firms, China has become into Africa is coming largely from low- and the second-largest Asian investor, after Japan ­ i ncome economies, whose compa- m iddle-​ (UNCTAD 2017). nies are also participating through mergers Investment abroad by Chinese firms tar- and acquisitions and purchases of assets held gets a wide range of manufacturing and ser- by high-income-country ­ m ultinationals. vices industries,14 with increasing attention In 2016, almost 80 p ­ ercent (US$73.6 billion to Africa as a destination. African countries of US$94.1 billion) of announced green- still account for a small share of net global field FDI projects in Africa were from low- FDI inflows (2.4 ­ p ercent in 2015), but the and middle-income countries, nearly half continent’s share has almost tripled since of it (US$36.1 ­ billion) from China alone OVERVIEW 17 (UNCTAD 2017). However, these projects FIGURE O.14  Share of global net outflows of foreign direct are mostly in real estate, natural gas, infra- investment, by region, 1994–2016 structure, renewable energy, chemicals, and automotive; FDI in manufacturing is still 90 relatively rare in Africa. 80 The top FDI investors in Rwanda are 70 already low- and middle-income nations. Faced with declining foreign assistance, 60 Rwanda surely needs more FDI to address 50 Percent its massive investment needs. However, it 40 also needs to make sure that FDI is targeted 30 increasingly to the tradable sector to support the massive export drive needed for high 20 growth. The regional trade agreements are 10 likely to enhance Rwanda’s (and the rest of 0 the continent’s) attractiveness to FDI, as they –10 promote a unified market structure and inte- 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 grated production networks. Europe and Central Asia Middle East and North Africa East Asia and Pacific Sub-Saharan Africa Global Climate Change Latin America and the Caribbean South Asia Climate change is potentially the most con- sequential global issue. Greenhouse gas emis- Source: World Development Indicators data (World Bank, various years). Note: Net outflows of investment from the reporting economy to the rest of the world. sions are on path to cause a 3.5°C to 4.0°C warmer planet by the end of the century. Climatic conditions, heat, and other weather Me d iu m - t er m cl i m at e proje c t ion s extremes considered highly unusual or for Rwanda indicate further increases in unprecedented today could become the new ­ temperature—between 1oC and 2.5oC—by normal. The impact of global climate change the middle of this century. Rainfall projec- is already being felt, with the number of cat- tions, though less certain, suggest increased egory 4 and 5 storms having risen sharply year-to-year variability in rainfall. The agri- over the past 35 years. The Arctic Ocean’s culture and livestock sectors will likely be ice has shrunk to its lowest area on record, affected. Rwanda will need to plan for these and global sea levels have risen about 10–20 vulnerabilities and seek to manage them with centimeters in the past century, increasing the climate adaptation policies and investments. risk of storm surges and fluctuations in pre- cipitation (World Bank 2016). Domestic Opportunities and Challenges The threat to Rwanda’s economy from climate change is already being felt. The aver- ­ Population Dynamics and the Promise of a age temperature in Rwanda has increased at Demographic Dividend a higher rate than the global average. Its rain- Rwanda’s population has grown about fall patterns are becoming more irregular and 2.5 percent a year since 1960, a pace that unpredictable, with shorter rainy seasons, has been sustained since 2000. This is a which has had a major impact on food pro- brisk but not exceptional pace in compari- duction. Rwandan agriculture is mostly rain- son with the rest of the low- and middle- fed (less than 10 p­ ercent of cultivable land is income world: 54 nations have seen a faster irrigated), which is why crop production is increase in population since 1960 (44 since highly vulnerable to c ­ limate- and weather- 2000). As a result of demographic shifts, related risks. These risks are compounded by influenced by relatively high but declining high levels of soil erosion and periodic floods fertility rates and sharp reductions in child and landslides. mortality, today’s Rwanda is a youthful 18 FUTURE DRIVERS OF GROWTH IN RWANDA nation, with a median age of just 19 years: growth remains high, Rwanda can expect a 40 percent of the population is under the age rapid decline in the poverty rate and a con- of 15, and almost 70 percent is under the age comitant rise in the share of its middle-class of 30 (UNFPA 2017). population. With 7.5 percent growth of Rwanda has the potential to harness its GDP per capita and with income distribu- demographic dividend. Yet risks associated tion assumed to remain unchanged, Rwanda with population increase are heightened by would virtually eliminate extreme poverty Rwanda’s high population density (the high- (those spending less than US$1.90 purchas- est in mainland Africa) and rising pressures ing power parity a day) somewhere between on its limited natural resources and the envi- 2035 and 2040 and more than halve the ronment. In addition, the fertility rate, one of share of the moderate poor (those spending the main drivers of population growth, has between US$1.90 and US$3.20 purchasing declined only gradually on average—from power ­ parity a day) from the 21 ­ percent rate an average of 5.6 births per woman in 2000 observed in 2014 (figure O.15). At the same to 4.2 births per woman in 2015, includ- time, more than 70 percent of its population ing a brief rise and fall during that period. would be middle class or higher (­ persons cat- At the current pace, Rwanda’s population, egorized as emerging consumers and global 11 ­m illion in 2016, is projected at least to middle class). Even if per capita growth comes double by 2050, which would more than dou- in at about 5.5 percent a year (still quite high ble the density of population to close to 1,000 by global standards), half of Rwanda’s popu- people per square kilometer of land area. lation would be middle class by 2040. Important gains from a demographic div- These projected shifts are likely to be rep- idend are possible if the decline in the fertil- licated at the regional level and will have ity rate—and thus in the population growth profound implications for the population’s rate—could be accelerated. Rwanda could purchasing power, preferences, aspirations, boost its GDP growth 1 percentage point and expectations of government. The thrust by targeting a lower dependency ratio—the of Rwanda’s growth strategy would still need number of people who are under and above to be oriented outward, producing tradable the working ages of 15–64 years for every goods and services to meet regional and 100 persons of working age—in parallel global demand rather than internal demand. with improved schooling (UNFPA 2017).16 This orientation is the only way for Rwanda Rwanda’s dependency ratio is relatively to overcome the constraints of being a small high at 75 percent. The United Nations economy and reach the scale economies and Population Fund estimates that the depen- specialization necessary for high growth. But dency ratio could fall to anywhere between rising consumer incomes will generate strong 53 and 64 percent by 2050, depending on demand for nontradable products, which will the pace of the decline in fertility. High- create an important secondary opportunity growth countries in East Asia bottomed out for local businesses. at dependency ratios of about 40 ­ p ercent, so perhaps there is scope to accelerate Strong Potential for Structural Rwanda’s progress. Policy options include Transformation better-targeted family planning measures With close to 70 percent of the labor force and continued emphasis on female educa- still in agriculture, which has much lower tion, health, and economic empowerment labor productivity than the major subsectors (UNFPA 2017). within industry and services ( ­figure O.16), a significant potential remains for realizing fur- Emerging Middle Class ther gains from structural ­transformation— High growth sustained over an extended the process by which labor and other period can profoundly affect the income resources shift from one economic sector to characteristics of the population. If GDP another. By this report’s estimate, more than OVERVIEW 19 FIGURE O.15  Growth scenarios for Rwanda and share of the population, by consumption category, by 2040 a. Aspirational high-growth scenario (7.5% per capita growth) b. High-growth scenario (5.5% per capita growth) Share of population, by consumption category (%) Share of population, by consumption category (%) 100 100 90 90 80 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 14 16 18 20 22 24 26 28 30 32 34 36 38 40 14 16 18 20 22 24 26 28 30 32 34 36 38 40 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Global middle class (> US$15.00 PPP/day) Moderate poor (US$1.90–US$3.20 PPP/day) Emerging consumer (US$5.51–US$15.00 PPP/day) Extreme poor (< US$1.90 PPP/day) Near poor (US$3.21–US$5.50 PPP/day) Source: Calculations based on National Institute of Statistics of Rwanda data. Note: PPP = purchasing power parity. 70 percent of Rwanda’s GDP growth in the FIGURE O.16  Labor productivity gaps in Rwanda, by sector, 2014 next two decades is likely to be accounted for by structural transformation, even under 1,000 970 a high-growth scenario. But gains from 900 structural transformation cannot be taken Sectoral productivity as a share 800 of average productivity (%) for granted. Only strong labor productivity 714 700 growth within sectors can assure continued 600 gains from structural transformation. For 500 that assurance, policies to boost competi- tiveness within sectors, as discussed in this 400 300 295 report, and well-functioning factor markets 250 251 180 207 (for land, labor, and capital) are essential. 200 141 100 43 0 Main Elements of Rwanda’s re n s m p ng m tio on s g es es e ce io rin ltu vic iti vic an ns Hu ica tati i Growth Strategy ct ur co Tra Min tu til u er er tru so un or ric U ac rs ls re ns uf Ag cia he m ns Co an Ot an As noted, aspirations have been set extremely M Fin high, targeting an upper-middle-income d an Rwanda by 2035 and a high-income econ- Sources: Estimates based on World Development Indicators data (World Bank, various omy by 2050. These aspirations translate years); National Institute of Statistics of Rwanda data. ­ ouble-digit average annual growth rates into d (more than 10 percent in per capita terms)— future investment and savings rates, human targets that will require Rwanda to grow faster capital development, export orientation, than any country (China and Korea included) technological innovation, and other forms of has in the past. The ­requirements—regarding productivity improvements—are demanding. 20 FUTURE DRIVERS OF GROWTH IN RWANDA Essential ingredients for s ­ uccess are strong For scale and specialization, Rwanda will leadership, social cohesion, and deep invest- need to make the most of external oppor- ments in core capabilities—of people, firms, tunities and enhance the benefits of urban and i ­nstitutions—to harness the global and agglomeration. To succeed in these areas, technology-related opportunities that are on Rwanda needs to have a competitive domes- offer. tic enterprise sector, both public and pri- The reform agenda is complex and highly vate, with a strong potential to do well in demanding: nothing short of an extraordinary competitive environments. Such enterprises effort will suffice, given the level of Rwanda’s themselves have three critical requirements: ambition. The hard work begins in class- a strong ecosystem for technological innova- rooms. The country needs a massive effort tion, world-class human capital, and robust focusing on human capital ­ development—its institutions of governance. This chain of pri- own education-focused Marshall Plan—if it orities forms Rwanda’s high-growth strategy. is to achieve its income ambitions. With all of Rwanda’s strategy for high grow th its achievements, Rwanda still lags the aver- thus has four essential and interdependent age of low-income countries in crucial aspects d rivers—innovation, integration, agglom- ­ of human capital—for example, in stunting eration, and competition (figure O.17). and in primary and lower-secondary school These future drivers of growth, in turn, completion. An important related issue is the would receive the necessary boost from high stunting in early years, with implications reforms in six high-­priority areas: (1) human for children’s future learning abilities and par- capital development; (2) export dynamism ticipation in the knowledge and services-led and regional integration; (3) well-managed economy that Rwanda envisages. urbanization; (4) competitive domestic The next requirement is to achieve higher enterprises; (5) agricultural modernization; investment rates. Rwanda already has a and (6) capable and accountable public insti- r elatively high investment rate of about ­ tutions. The six reform areas are discussed 26 percent. But double-digit growth rates in more detail in the report. would require investment to be significantly higher still—at least 35–40 percent—in an environment of declining external assistance. FIGURE O.17  Future drivers of Rwanda’s growth: Achieving this level of investment will require Innovation, integration, agglomeration, and competition a sharp increase in investment by the private sector, because public investment is already at the limit set by the available financing options; Highly a more than fourfold rise in the domestic sav- developed ings rate; and even higher FDI. human A higher-order challenge is to boost pro- capital Trade and Competitive ductivity growth, which also has a bearing regional domestic on Rwanda’s ability to raise the investment integration enterprises and savings rates. High productivity growth, DRIVERS OF GROWTH in turn, will require scale economies and eco- Innovation nomic specialization in areas of Rwanda’s Integration Agglomeration comparative advantage, with competition Modern Competition Well- and technology diffusion as essential comple- agriculture/ managed ments. Scale economies and economic spe- food sector urbanization cialization have proved essential for sustained Capable and high growth across the world (Commission accountable on Growth and Development 2008), but it is institutions even more important for a small, landlocked country like Rwanda. OVERVIEW 21 Doing well on each of these six neces- out (box O.4). The more challenging part is sary reform areas is what separated the to go beyond the necessary to the sufficient high-growth East Asian economies from conditions. Rwanda will be pushed to take countries that achieved rapid grow th high-risk strategic bets to gain high returns. for a decade or two, only to see it fizzle T hese efforts must be calibrated and BOX O.4  An international comparison of Rwanda’s long-term growth trajectory Only a handful of economies in East Asia have FIGURE BO.4.1  GDP per capita in Rwanda and select achieved an income trajectory even approaching the countries and economies one Rwanda aspires to achieve. A comparison with 35 GDP per capita (thousands of US$, 2010 PPP) those economies, therefore, can shed useful light on the fundamental areas that need to be at the center 30 of Rwanda’s policy focus. Rwanda’s long-term growth trajectory stacks 25 up well against the early growth records of global 20 economies that sustained rapid growth over several decades. Figure BO.4.1 compares Rwanda against 15 two Asian Tiger economies plus Brazil, China, Morocco, and Thailand. The figure is striking on 10 two counts. First, the long-term acceleration of growth in each of these economies started at similar 5 per capita income levels—between US$1,000 and 0 US$1,600 (in 2010 U.S. dollar purchasing power Base +2 +6 +10 +14 +18 +22 +26 +30 +34 +38 +42 +46 +50 parity). Second, by 2016, Rwanda (roughly a decade year into its postrecovery growth acceleration) had kept Number of years beyond base year pace with comparators (including China) at a simi- Thailand Korea, Rep. lar stage. Rwanda Morocco What happens from here on seems critical on Rwanda China (4% per capita growth) the basis of international comparisons. About a Brazil Rwanda decade into their respective growth accelerations— (7.8% per capita growth) Taiwan, China Rwanda’s current stage—successful economies Sources: Calculations based on Penn World Tables 9.0 data (Feenstra, such as China, the Republic of Korea, and Taiwan, Inklaar, and Timmer 2015); National Institute of Statistics of Rwanda data. China, started to pull ahead of the rest. They were Note: The base year is the year prior to the one in which long-term growth began to accelerate in each economy. It is 1951 for Taiwan, able in years 10–50 either to maintain or to surpass China, and Brazil; 1958 for Thailand; 1959 for Morocco; 1962 for the their growth records of the first 10 years. However, Republic of Korea; 1977 for China; and 2005 for Rwanda. The base year economies such as Brazil, Morocco, and Thailand for Rwanda is 2005 because that is when GDP per capita recovered to the pre-genocide peak levels. PPP = purchasing power parity. saw their growth plunge early in the latter period. The challenge for Rwanda is to follow the trajec- tory of the former group and to avoid that of the lat- critical period, the successful economies focused ter group. Doing so will mean raising its growth of on a core set of fundamentals. Common elements gross domestic ­product per capita to the 7–8 ­ percent included high rates of savings and investment; range. However, if recent-year growth rates of about strong agricultural productivity; early and sustained 4 percent per capita continue, then even after three emphasis on human capital; strong outward orien- decades or so, Rwanda could not expect to get past tation; robust urban agglomeration; a competitive where Brazil or Thailand is today. domestic enterprise sector; and strong, stable, and Rwanda therefore is at a seemingly important adaptable institutions of governance. Those condi- juncture. Decisions in the next 5–10 years will mat- tions are necessary for Rwanda to achieve its long- ter for meeting its long-term aspirations. In this term growth aspirations. 22 FUTURE DRIVERS OF GROWTH IN RWANDA managed carefully. Policy responses need on water and air pollution from early on, to first to address key cross-cutting constraints protect the quality of natural resources (air, (such as skills, finance, infrastructure, and land, water), to build climate resilience into business regulation), clarify the future role economic planning and infrastructure invest- of SOEs, and then selectively target areas for ments, and to meet international commit- direct support that are closely aligned with ments on cutting greenhouse gas emissions. Rwanda’s comparative advantages. Any The high-growth strategy thus enunci- such direct support will have to set clear ated, with emphasis on economic special- policy objectives and performance targets ization, scale economies, and trade, leaves for beneficiary firms, be coordinated closely open the question of food security at the across government entities, and include a national level. Rwanda will need to work rigorous system to monitor progress, enforce closely with its regional partners to ensure sanctions, and provide incentives to reward an uninterrupted flow of food staples—for success and punish failure—a model that example, through the provision of financial Korea followed closely in its early years of guarantees under a regional process. If the development. country decides to maintain self-sufficiency Any future growth strategy is unlikely in any particular staple, then it should do to succeed if it leads to further widening of so after a careful analysis of the full eco- income disparities or does not provide equality nomic implications of such a move. For of opportunity for all citizens. Growth accel- example, Japan and Korea still produce erations can widen inequalities in the initial their own rice and beef, but at significant stages as returns on skill levels rise and urban economic costs, which they fully recognize agglomeration leads to spatial concentration and willingly accept. of economic activity. Rwanda will need to be Finally, the reform agenda and invest- keenly mindful of its future trends. Building ment programs underlying the high growth a strong system of social protection to pro- strategy will be financially and institution- vide basic safety nets to the most vulnerable ally demanding. Given the limited fiscal and expanding access to quality public ser- space available and expected continued vices (education, health care, and basic infra- decline of external concessional financ- structure such as sanitation, potable water, ing, future interventions will require far electricity, and road and telecommunications greater involvement of the private sector, connectivity, for example) to give everyone a together with efforts to enhance access fair chance to enjoy the fruits of growth will to domestic and external capital markets be important efforts in this regard. A stronger and to strengthen the planning and imple- emphasis on educating girls and empowering mentation capabilities of the government. women can be especially fruitful in breaking The latter involves better prioritization the initial cycle of poverty and expanding the and greater efficiency of public spending, country’s base of human capital. increased flow of government revenues, and The growth strategy also will have to stronger and more transparent public finan- protect the environment: economic growth, cial management systems. Strong emphasis however fast-paced it may be, at the cost of on macroeconomic and social stability will the environment is not sustainable. Low- and remain essential to these efforts, even as middle-income countries often make the mis- Rwanda navigates a more uncertain global take of neglecting these aspects in the early economic and political landscape. stages of development, only to find the cost of reversing them later to be much higher—both Six Reform Areas of Importance financially and in terms of lost growth poten- tial. It is not necessary to adopt high-income- for Rwanda country environmental standards from the This section discusses in more detail the six beginning. But it is necessary to put a price reform areas of importance for Rwanda. OVERVIEW 23 Human Capital and Innovation is insufficient given the country’s ambi- tious aspirations. Moreover, investments Introduction in human capital, particularly efforts to Rwanda’s aspirations to reach upper-­ reduce stunting and improve basic educa- middle-income status by 2035 require tion, contribute to growth only after con- dramatic improvements in human capital. siderable time (the beneficiaries have to Economies that have grown rapidly over grow up). Only a small fraction of the pop- an extended period have made substantial ulation enters the labor force each year, so investments in the education and health even dramatic improvements in education of their citizens. Human capital includes will adjust the skills of the overall work- the knowledge and skills of the popula- force only slowly. The sooner Rwanda tion, and it results from investments in increases its human capital investments, education and health. Ultimately, human the sooner it will be on track to achieve capital investment and economic growth more rapid growth. form a virtuous cycle: greater human devel- opment increases economic growth, and Challenges and Opportunities greater economic growth finances further Human capital efforts are urgent for human development. Cross-country analy- Rwanda on a wide range of interventions, sis demonstrates that countries that invest including prevention of stunting, improve- in human capital early in the cycle enjoy ment of access to and quality of basic edu- the benefits of this virtuous cycle, whereas cation, increase in enrollment in higher countries that experience high initial eco- education, and building of innovation nomic growth without investing in human capacity. capital almost always fall into a vicious Despite the launch of several success- cycle of low human capital followed by a ful initiatives, stunting still represents a slowdown of economic growth (Ranis, major challenge in Rwanda. At last mea- Stewart, and Ramirez 2000). sure, 36.5 percent of children suffered from A broad set of policies is needed to support stunting. Furthermore, the elasticity of human capital formation. Building human stunting with respect to economic growth capital entails a wide array of interventions in Rwanda is less than half the elasticity in across the life cycle, starting with invest- the rest of the world, so future economic ments in early childhood, including prenatal growth cannot be counted on to eliminate and early child nutrition and cognitive stim- it. Stunting has significant short- and long- ulation, and continuing with high-quality term impacts on Rwanda’s human capital basic education, higher education oppor- and—ultimately—on economic growth. tunities, and skills that individuals gain as Adults who were stunted in childhood have adults, both through on-the-job training and poorer health and shorter height, are at risk through adult education. Interventions also of suffering noncommunicable diseases, include health investments beyond early child and have lower cognitive ability and fewer nutrition, from vaccinations for children to socioemotional skills. These effects trans- preventive and curative health care for chil- late into reduced productivity, lower wages, dren and adults. Finally, a strong economy and slower economic growth. By one esti- invests in innovation to continue creating mate, the per capita income of today’s new opportunities. workforce would be 10 percent higher if R e c o g n i z i ng t he c ou nt r y ’s s t rong adult Rwandans had not been stunted as future ambitions, progress on human children. These impacts spill over to the capital needs to be dramatic and swift. next generation, with stunted women more Under the current rate of improvement, likely to have babies who are underweight Rwanda will still have low-income-coun- and who have cognitive challenges (Galasso try rates of human capital in 2035, which et al. 2016). 24 FUTURE DRIVERS OF GROWTH IN RWANDA Rwanda faces the same two challenges completion and making improvements in in basic education as many other coun- quality difficult. tries: access and quality. Virtually all The concern about the quality of educa- ­ children in Rwanda begin primary school, tion is just as serious as the concern about but only about two-thirds of them com- access. If students are failing to learn, then plete it. One reason is out-of-pocket costs. economic growth will not follow from Another is high opportunity costs, which investments in education. Yet 85 percent are accounted for in part by high repetition of Rwandan students at the end of grade rates in the early grades. Nearly one-quarter 3 were rated “below comprehension” on a of first-grade students are made to repeat recent reading test, and one in six students that grade (USAID 2017). Repetition— in grade 3 could not answer a single read- along with other factors, such as limited ing comprehension question (EDC 2017). early childhood education coverage and late In addition, youth who are approaching the school ­starters—results in a large “bulge” job market also need better opportunities for of students in P1 (first grade) (figure O.18), skills training to help them to transition to lowering the likelihood of primary school employment. FIGURE O.18  Early-grade “bulge index” in Rwanda and select countries 3 Cabo Verde São Tomé and Príncipe 2 Mauritius Côte d'Ivoire Mali Tanzania Burkina Faso 1 Guinea South Africa Ghana Niger Senegal Gambia, The Zimbabwe Swaziland 0 Namibia Botswana Early-grade "bulge index" Comoros Equatorial Guinea Congo, Rep. –1 Lesotho Central African Republic Cameroon Congo, Dem. Rep. –2 Angola Benin Chad Togo Sierra Leone Ethiopia –3 Burundi Guinea-Bissau Uganda –4 Madagascar –5 Rwanda –6 Source: Bashir et al. 2018. Note: Low values in Rwanda reflect high enrollment in P1 (first grade) relative to the age group, low progression to P2 (second grade) (that is, many repeaters), and low preprimary coverage. OVERVIEW 25 Rwandan teachers attend and put forth FIGURE O.19  Projected share of Rwanda’s workforce with tertiary significant effort, yet this effort is under- education under two growth scenarios, 2010–50 mined by skills and pedagogy. Crucially, many teachers have limited command of the a. General education trend scenario (business as usual) language of instruction: a recent assessment 100 found that less than half of teachers are at the 80 “intermediate level” in English. A lack of pro- ficiency in the language of instruction inevi- 60 Percent tably affects student performance not only in English reading, writing, and speaking ability 40 but also in all other subjects that are taught in 20 English, including math, science, and social studies. The government has invested exten- 0 sively in teacher language training, but more 2010 2015 2020 2025 2030 2035 2040 2045 2050 intensive efforts are needed to bring teachers b. Fast-track scenario (at the pace of the the rest of the way to fluency in the language Republic of Korea and Singapore) of instruction. 100 Currently, relatively few Rwandans com- plete tertiary education, although enrollment 80 rates are rising rapidly. The latest numbers 60 Percent (2015) suggest that just 8 percent of tertiary- age youth are enrolled in tertiary education, 40 well below the level in middle-income coun- tries. Furthermore, even with rising tertiary 20 enrollments—and those enrollments have 0 doubled in the last 10 years—shifting the 2010 2015 2020 2025 2030 2035 2040 2045 2050 proportion of the population with tertiary No education Some or all secondary education in the workforce takes significant Some or all primary Some or all tertiary time, because only a small proportion of the workforce changes with each graduating Source: Construction based on data from Lutz, Butz, and KC 2014. Note: Panel a is based on the current rate of growth of enrollment in Rwanda. Panel b is cohort (figure O.19). Moreover, relatively few based on the rate of growth of enrollment in the Republic of Korea and Singapore. graduates are specializing in key job creation fields, such as science and engineering. Just Recommendations 6 ­percent of university students in Rwanda Rwanda faces an array of needs with regard are enrolled in engineering, manufactur- to human capital investment and needs to ing, and construction; and only 9 ­ p ercent expand investment at all levels of human capi- are studying sciences. If Rwanda intends tal formation. Each human capital investment to grow its manufacturing and technology builds on previous ones. Therefore, invest- sectors, then the number of students in sci- ments yielding returns that are realized more ences and engineering clearly also needs to quickly need to be balanced with investments grow. Even though historically Rwanda has in younger generations. For example, the invested more in higher education than other returns from investments in tertiary education countries in the region, public financing of and skills training are realized most imme- higher education has declined recently, even diately. However, their continued expansion as demand for higher education continues to is dependent on an increased flow of well-­ grow. Financing for innovation in Rwanda educated individuals from earlier levels, and also remains very low, most recently esti- their returns are complemented by a popula- mated at 0.4 percent of GDP (Lemarchand tion with broad literacy and numeracy. and Tash 2015), which is a tenth of the share Government should consider two prin- in countries like Korea and Singapore. ciples in setting investment priorities to 26 FUTURE DRIVERS OF GROWTH IN RWANDA maximize the impact on growth. First, it been demonstrated to increase parent and stu- is necessary to identify the “binding con- dent investment in education and to improve straints”: Which problems are so severe that, the quality of education. unless they are solved first, no amount of money or time spent solving the less severe Improve the Quality of Education.  Improving problems will help? Second, it is necessary the quality of teachers is essential. Teachers in to identify the investment required in activi- Rwanda generally attend school and teach, ties that yield economic growth both now which is itself a great challenge in many coun- and later. Complementary investments, such tries. But primary school teachers need to be as safety nets, can help households to make trained in improved pedagogies so that chil- investments that will contribute to inclusive dren in the first three years get a firm grasp on growth. literacy before transitioning to English. There A holistic approach is needed, with special is good evidence on effective programs to teach emphasis on the following areas. literacy in the early years, including partially scripted lessons and ongoing coaching of teach- Drastically Lower the High Rates of ers. Because these improvements take time, Stunting.  Reducing Rwanda’s stunting rate transitioning to English later (for example, requires multiple coordinated interventions, at the end of primary) may also be valuable. and incentives are key. Currently, many pro- Many teachers in the upper grades of primary grams to address these incentives in Rwanda and in secondary have a limited grasp of the suffer from limited coverage and inconsis- language of instruction: a recent assessment tent financing from a patchwork of donors. showed that only 43 percent of teachers have A big push is needed to improve food secu- the “intermediate level” in English. Rwanda rity and nutritional practices and to increase could consider leveraging its major investment access to clean water and good sanitation. in laptops in schools to provide teachers with Careful monitoring systems are required to regular opportunities to improve their English. evaluate which policies deliver the greatest Technology could also be applied to increase gains. The good news is that high rates of students’ learning performance.17 Rwanda also stunting are not inevitable. Many countries could consider recruiting English-speaking (including Peru and Senegal) have made dra- expatriates to remedy the immediate shortage matic gains, with political will, civil soci- of teachers and, for the longer run, to train a ety cooperation, media campaigns, robust core group of high-quality Rwandan teachers. monitoring, adequate resources, and a multi- pronged approach seen as key to their success. Strengthen the Provision of Technical and Vocational Training. Collection and Expand Access to Basic Education. In dissemination of information on the quality Rwanda, expanding access first means reduc- of skills providers and the returns to differ- ing costs and improving perceived benefits. ent skills would encourage youth to partici- Countries that have achieved widespread pate in sectors with high returns and help basic literacy have offered free primary to improve the quality of skills training pro- education. Achieving universal primary edu- grams. Many high-growth countries, includ- cation requires ensuring that children are not ing Korea, Malaysia, and Singapore, used an turned away for failure to pay incidental fees, activist approach to skills development by which may require increasing the per student setting a strategic direction, tone, and culture benefit paid to schools and reducing repetition for efforts to improve workforce skills; creat- in early grades so that children do not become ing an organizational infrastructure with the too old for their grade. It is also important to appropriate governance design; and ­ fostering provide information on the financial returns efficient and effective management of service to schooling because such information has delivery by providers. OVERVIEW 27 Develop a Quality Tertiary Education the amount of resources devoted to educa- System Focused on High-Return tion over time. Activities.  Rwanda has taken dramatic steps to improve the quality of tertiary edu- Provide Better Information on Out- cation in recent years, consolidating public comes.  Rwanda is an innovator in holding ser- universities into the University of Rwanda vice providers accountable, with performance for better governance. Increasing access to contracts for educators and health workers, financing, including loosening restrictions along with innovations in p ­ erformance-based on private financing, would help to expand pay (Basinga et al. 2011; Zeitlin et al. 2017). enrollment. Rwanda also needs to focus its In many countries, service provision has tertiary education system on key areas of improved because of the effective involve- investment: more science and engineering. ment of parents, pointing to the importance Strategies used in high-income countries to of dissemination of information on learn- encourage university students to enter high- ing outcomes and other elements of school return fields—including financing incentives performance. More regular measurement of (as in Argentina and Australia) and improv- learning outcomes and stunting reduction ing the quality of science and engineering would allow the government to evaluate the instruction in earlier grades (as in Norway effectiveness of different interventions and to and Poland)—could be considered. accelerate progress on both aims. Foster Innovation.  The tertiary education Transformation through Trade: Using sector is an ideal space to foster innovation in Exports and Regional Integration to Rwanda. Publications and patents in Rwanda Drive Future Growth have been rising, although from a very low base. Likewise, Rwanda has invested in a Introduction range of graduate and postgraduate centers To achieve its aspiration of becoming a high- for technical training, including Carnegie income country by 2050, Rwanda needs Mellon University and the various centers of to accelerate the growth of trade. Inflows excellence. Creating incentives for researchers of development assistance have financed to develop and adapt innovations that benefit a large share of investment and powered industries in Rwanda can help Rwanda to reap GDP growth in the past two decades, but the maximum returns to local innovation. they are likely to attenuate in the coming A practical way to do this follows the model two decades as Rwanda progresses toward common in high-income countries, where pri- middle-income status. Trade will become vate firms finance university research to solve an increasingly important driver of growth. production challenges. Given the nascent pri- Exports will provide foreign exchange to vate sector, the government will have to con- purchase ­much-needed investment in equip- tinue to play a supporting role. ment, ­ t echnology goods, intermedi- h igh-­ ate components, and product varieties and Allocate More Resources to Human Capital will foster productivity by allowing firms to Development.  The amount that success- exploit increasingly large economies of scale. ful countries spend on health and education Increased import capacity will facilitate varies, and financing is far from the only fac- access to high-­ technology goods and foster tor in successful human capital investment. competition that drives productivity. Trade Governance of the sectors arguably matters expansion also implies the need to attract far more and is a strength in Rwanda. It is FDI, because multinational companies bring impossible to imagine achieving universal managerial, technical, and design skills in ­ primary education and a significant increase while their networks facilitate access to new in secondary education without expanding export markets (Freund and Moran 2017). 28 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE O.20  Exports as a share of GDP and log GDP per capita in markets and to propel structural transfor- Rwanda, 2000–16 mation and growth (Hallward-Driemeier and Nayyar 2017). 125 Challenges and Opportunities Exports of goods and services (% of GDP) 100 On the basis of a comparison with countries experiencing high rates of economic growth, Rwanda’s exports as a share of GDP have to 75 increase significantly for Rwanda to achieve its income objectives for 2050. A review of global 50 Rwanda, 2007 Rwanda, 2016 experience shows a large spread of exports- Rwanda, 2000 to-GDP ratios as countries achieve upper- middle-income status, but that ratio tends to 25 Upper-middle-income countries be upward of 40 percent (figure O.20). Since 2000, Rwanda has consistently increased its 0 share of exports in national income, but the 4 6 8 10 12 share is still far below the average for upper- Log GDP per capita middle-income countries. Rwanda needs to have double-digit year-on-year export growth Source: Calculations based on Comtrade data (United Nations, various years). every year up to 2035 if it is to cross the 40 percent threshold. Meeting such a target is difficult but not impossible, given that the Trade expansion is central to creat- share of exports in national income is still low ing new, higher-productivity jobs that relative to that of other countries with com- faci l it ate g row t h t h roug h st r uc t u ra l parable levels of per capita income. Therefore, transformation. Moving labor from low- significant export opportunities are avail- productivity jobs mainly in agriculture able to Rwanda, and exports still have ample to higher-productivity jobs in a range of room to grow. mostly urban activities is imperative for Rwanda has begun laying the foundations growth. East Asia made this transition to for an export push. Over the last decade, high growth by relying on labor-intensive exports have grown about 20 percent annu- manufacturing for export. Rwanda, as ally, and the export portfolio has become with much of Africa, requires not only more diversified, with the share of coffee, labor-intensive manufacturing but also tea, and minerals falling from 41 percent in agribusiness, horticulture, and selected 2005 to 25 percent in 2015. New activities, services—what some have called “indus- particularly services exports, have surged in tries without smokestacks” (Newfarmer, importance. Tourism alone now accounts for Page, and Tarp, for thcom ing). T hese 23 percent of export earnings, and services activities, taken together, hold the prom- account for about half. New exports of goods ise of doing for Rwanda what manufactur- are also becoming important, especially ing did for East Asia in the 1990s. They exports to regional markets. In 2014, the are labor intensive and tradable, and they EAC accounted for 41 percent of manufac- have high value added per worker. As in turing exports, 66 percent of leather goods, traditional manufacturing, technologi- and 53 percent of horticultural products cal change is rapid and can spawn rapid sold abroad. Non-EAC neighboring mar- productivity growth. In a world of recent kets have also become important. Exports technological revolutions in ICT, manufac- to the Democratic Republic of Congo have turing techniques, and global value chains, increased considerably in the last decade: Rwanda has an opportunity to leverage by 2016, Rwanda exported more goods to greater integration into regional and global the Democratic Republic of Congo than to OVERVIEW 29 the EAC.18 Rwanda has also attracted more FIGURE O.21  Intra-bloc goods imports as a share of GDP before FDI of late, which is contributing to export and after joining the bloc performance. That said, Rwanda still exports below 18 the average of other countries at its per 16 15.8 capita income. Whereas exports in agro- 14 13.2 processing have been dynamic, exports in other manufacturing areas have been disap- 12 Share of GDP (%) pointing, declining from 8 percent of total 10 8.7 exports (2012–14) to 5 percent in 2016. 8 This decline is largely due to reductions 6.5 in some of the higher-skill export prod- 6 5.1 ucts such as machinery, mechanical appli- 4 3.0 ances, and electrical equipment. Further, 2.2 2.0 2.1 2 Rwanda’s exports remain concentrated in a 0 small number of firms. Over 2009–16, the EAC SADC ASEAN top 1 percent of exporters accounted for 2 years before 5 years after 10 years after more than 40 percent of the total value of exports (while the top 5 percent accounted Source: Shepherd, de Melo, and Sen 2017. for 80 percent). Exporter survival rates are Note: EAC = East African Community; SADC = Southern African Development Community; ASEAN = Association of Southeast Asian Nations. also low. Export performance continues to suffer from low labor productivity in key backbone services such as transport and getting goods to market and getting inputs ICT and low levels of private investment in to local producers. Rwanda is a land- tradable sectors. locked country, which means that trans- An important challenge is that the ben- port costs typically add some 50 ­ p ercent to efits to trade from regional integration in the cost of exporting products and import- the EAC have been limited so far. By remov- ing inputs. Because more than 90 percent ing internal tariff and nontariff barriers, of goods exports are transported by truck, regional integration is expected to result in Rwanda relies heavily on the land trans- trade creation. Yet a comparison of intra- port corridors of other countries for access regional imports as a share of GDP before to the sea. Almost all of its trade in global and after joining the EAC shows that this markets goes through two East African effect has been limited (figure O.21). After trade corridors: the Northern Corridor a small increase in the first five years, the (Mombasa) and the Central Corridor (Dar EAC’s intra-bloc imports of goods as a share es Salaam). The lack of facilitation at the of GDP fell to a level below initial integra- border, fragmentation of the supply chain, tion 10 years after membership. In contrast, and limited access to affordable air cargo other regional bodies experienced a consid- opportunities also contribute to the high erable increase in intra-bloc trade. After 10 cost of trading. years of regional integration, the Southern African Development Community (SADC) Recommendations saw a tripling of intra-bloc imports, while Meeting Rwanda’s export objectives requires the Association of Southeast Asian Nations a comprehensive trade policy that spans ser- doubled intra-bloc imports to almost 16 per- vices, industry, and agriculture. Analysis cent of GDP. from this report suggests that no one sector Trade connectivit y and high trans- can drive the necessary export and employ- port costs pose a constraint for Rwanda. ment growth on its own. Rwanda should The single most important determinant of look to produce high-quality products for the long-run trade growth, in fact, is the cost of region (especially in agribusiness and food 30 FUTURE DRIVERS OF GROWTH IN RWANDA processing) and to develop other sectors that Monitor the Exchange Rate Closely to are similarly tradable and productive, but less Maintain Export Competitiveness.  Export dependent on location (such as horticulture, drives in other countries have been sus- tourism, professional services, and ICT). tained by maintaining a competitive real The government thus has to build further on exchange rate over a long period. Similar the service sector (the largest current source benefits may be on offer for Rwanda, but of exports), strongly accelerate industrial greater flexibility and close monitoring of growth, and expand into other high-value exchange rates are needed to ensure that agricultural exports (such as horticulture). the real exchange rate remains in line with Six major policy priorities should figure fundamentals and that episodes of over- prominently in a comprehensive reform pro- valuation are avoided, especially in regional gram that uses trade to accelerate and sustain markets where manufactured goods and growth. new products predominate. A first step is to undertake more analytical work on the Harness the Regional Blocs as Platforms link between exchange rates and export for Transformation. The region offers growth,19 particularly as it affects poten- a crucial learning ground for exporting tially dynamic export sectors. Because higher-quality products, especially in man- imports consume capital inflows for for- ufacturing and agroprocessing. Entering eign exchange payments, it is advisable for regional supply chains can help to prepare the government to focus on export growth Rwandan firms to enter the global market. rather than on the trade deficit as a macro- The region can also be used to stimulate economic objective. within-sector productivity growth and to develop other tradable sectors such as tour- Invest in Economic Diplomacy.  The ­complex ism, transport, and professional services issues under consideration in trade negotia- through greater regional scale economies tions require more investment in staff knowl- and greater competition from leading firms. edge and expertise to usefully engage in and A key focus should be on revising and low- undertake negotiations, especially in highly ering the common external tariff within specialized sectors. the EAC to benefit Rwandan producers and exporters that use imported inputs and Improve Trade Connectivity by Lowering poor consumers who disproportionately Transport Costs.  Intensive collaboration consume heavily taxed imports. The inte- along Rwanda’s two trade corridors has grated market can also be extended by pro- already lowered costs and increased exports, moting harmonized standards in goods and but costs can decline even more. Lowering services and by reducing nontariff barriers. regional road tolls and ensuring their uniform For key sectors such as energy and finance, application across all EAC vehicles is one the EAC should develop regional supply avenue. Reviewing tax policy to ensure that chains that can result in economies of scale. truckers from all countries compete on a level Intra-industry regional trade competition playing field is another. Developing Rwanda can force firms to larger scale and drive as a regional logistics hub by attracting pio- out less productive firms. Finally, Rwanda neer foreign firms in logistics would further could advocate for a stronger EAC secre- lower transport costs and strengthen nascent tariat that can review and discuss potential services exports. RwandAir has provided the violations of common market protocols. basis for expanding air connectivity, and its The recently agreed African Continental efforts have reduced the cost of air cargo. Free Trade Area, a pan-African initiative More can be done here, in part by aggressively to liberalize continental trade, can also be pursuing open skies arrangements within the advantageous, once details are penned and EAC and SADC. Rail connectivity also could implementation modalities agreed to. become important eventually; a long-­ standing OVERVIEW 31 ambition to build a railway along the two upgrading and certification). The solid results trade corridors continues to hold promise for in the special economic zone offer a model reducing freight costs. that could be extended to the whole produc- tive sector. Upgrading the product quality Increase Service Sector Productivity.  Lower of domestic firms through a designated sup- policy barriers to services competition and plier development program would offer great stronger services trade within the region potential by enabling domestic firms to sup- would enhance competitiveness across the ply large international firms, which could EAC. Rwanda should look to broaden value facilitate integration within global value added tax exemptions on services exports chains. Such a program also would provide by aligning itself with leading African coun- an opportunity for strengthening regional tries such as Mauritius and South Africa. manufacturing exports. Rwanda would benefit from elevating tour- ism promotion in a separate, high-visibility Faster Urbanization, Greater strategy that seeks to expand leisure tourism Agglomeration beyond gorillas and to convince tourists to spend more time in Rwanda. In the decades Introduction ahead, Rwanda is strategically positioned to Few countries have attained upper-middle- take advantage of exports in mining, educa- income status without substantial urbaniza- tion, and business services. Realizing these tion. Urbanization is the spatial corollary opportunities requires seeking out foreign of structural transformation that involves firms and facilitating near-term services trade the movement of workers from less to more for EAC and SADC professionals, which productive sectors—typically from farm- can be strengthened by extending mutual ing to off-farm sectors in the early stages of recognition agreements beyond architec- development. Urbanization has the potential ture, accounting, and engineering to include, to generate enormous benefits by increas- for example, legal, finance, and consulting ing economic density, which, when man- services. Abolishing limits in work-permit aged well, facilitates the transmission of regimes for all eligible professionals is also knowledge and ideas, increases economies critical and would help to facilitate services of scale and opportunities for specialization, trade for short-term assignments. However, and improves firms’ access to both critical stimulating services exports requires con- services and a large pool of labor with a siderable investment in addressing Rwanda’s wide variety of skills. Cities are also instru- skills deficit. mental in matching skilled people with jobs that value those skills. The presence Attract Private (Foreign and Domestic) of these agglomeration economies means Investment into Tradable Sectors.  More pri- that the more workers and firms are added vate investment is needed in tradable sectors. to an urban location, the higher their indi- Building on its investor promotion activities, vidual productivity becomes. International the government could track firm performance ­ evidence suggests that a 1 percentage point and tailor specific incentives to priority increase in the urban population is associ- sectors. To improve Rwanda’s export compet- ated with a 3–8 percent increase in a coun- itiveness and compensate for relatively high try’s per capita income. transport costs, the government could offer The close association between urban- additional, integrated assistance across min- ization and economic development is seen istries (for example, by linking the Rwanda most vividly in the high-growth economies Development Board’s investment promotion of East Asia. These economies have had with the Rwanda Revenue Authority’s abil- intense spatial concentration of economic ity to exempt import tariffs and the Rwanda activity, typically in locations well con- Standards Board’s support for standards nected to domestic and external markets. 32 FUTURE DRIVERS OF GROWTH IN RWANDA Consider the “flat” development trajec- Rwanda, and urban areas have accounted tories of China, Korea, Thailand, and for 48 percent of national labor productiv- Vietnam in figure O.22, panel a. The flat ity growth over the past 15 years. However, slopes reflect the fact that their urban- urban areas in Rwanda—and Kigali in par- ization has been accompanied by sharp ticular—have not generated the kind of increases in per capita income. However, productivity gains and agglomeration econ- in an unsupportive policy environment, omies that rapidly growing cities in success- urbanization can just as easily be discon- ful East Asian economies have achieved. nected from economic transformation; The pace of rural-to-urban migration also many countries in Sub-Saharan Africa has been relatively muted, 20 which has lim- have urbanized even more rapidly than ited the potential benefits of urbanization. their East Asian counterparts, but with Institutional, regulatory, and infra- minimal or negative growth of GDP per st r uc t u ral refor ms a re needed along capita (figure O.22, panel b). three dimensions—spatial, sectoral, and Rwanda is on a path to rapid urban- i nstitutional—and should be supported ­ ization and needs to manage the process by even faster rural-urban migration and well to harness the urban agglomeration deeper economic specialization. The nature economies that are essential for future high of agglomeration economies, however, dif- growth. The country’s urban population fers by the function of different urban settle- has more than doubled since 2002, and ments. Rwanda’s small towns, secondary Kigali has been one of the fastest-­g rowing cities, and Kigali should perform comple- A frican ­ c ities. T his repor t’s analysis mentary functions, differentiated by the shows that urbanization has accounted for type of scale and agglomeration economy percent of national structural change in 37 ­ they can deliver. FIGURE O.22  Changes in urbanization and income in East Asia and Pacific and in Sub-Saharan Africa, 1985–2010 a. East Asia and Paci c b. Sub-Saharan Africa Korea, Rep. 60 80 Gambia, The Mauritania Malaysia 70 Mongolia Cameroon % of population, urban % of population, urban 60 Guinea-Bissau Fiji Nigeria Congo, Benin 50 Indonesia China Dem. Rep. 40 Togo Thailand 40 Lao PDR Madagascar 30 Vietnam 20 Zimbabwe 10 20 200 2,000 20,000 300 600 1,200 GDP per capita (constant 2005 international $) GDP per capita (constant 2005 international $) Sources: Based on World Development Indicators data (World Bank, various years); Economist 2012. Note: Bottom of arrow = 1985; top of arrow = 2010. OVERVIEW 33 Challenges and Opportunities within 20 kilometers of a secondary city, yet To achieve the rapid economic growth targeted the locational advantages are undetectable until 2050, Rwanda needs to have an efficient more than 5 kilometers from the city center. portfolio of urban locations. With its relatively Half of the rural population lacks access to small population, Rwanda requires one espe- a road network in good condition within a cially strong urban economic powerhouse—of 2-kilometer walking distance. Better rural- the sort that Seoul was for Korea or Tokyo for urban transport links around secondary cities Japan. Kigali has that potential. To fulfill its would improve both their access to inputs and urbanization potential, Rwanda also needs to their impact on surrounding areas. Stronger have a portfolio of complementary secondary rural-urban transport links also would cities and small towns. enhance the benefits of growing urbanization However, although Kigali has seen rapid along Rwanda’s major transport routes, most urbanization and strong urban infrastruc- notably along the Rubavu-Nyabihu-Musanze ture investments, its urban fabric remains corridor, but also along corridors running fragmented by small-scale, patchy land devel- south from Kigali to Burundi and the newly opment. Kigali needs to operate more like developed Kivu shoreline. an effective, integrated labor and product Outside of Kigali, own-source revenue market, which would increase its potential mobilization is low—estimated at 10 percent for realizing agglomeration economies and of district budgets in 2013–14. Districts rely reduce the costs of private inputs and public on central revenues and do not capture many infrastructure. As regards the secondary cit- of the gains from their investments. This ies and small towns, Rwanda should be ready reliance reduces both district resources for to support the emerging economic activities investment and the incentives to identify and in those locations. There is great uncertainty prioritize investments where they are most concerning what kinds of economic activities likely to unlock growth. will settle in which locations. All urban loca- tions are showing similar trends, ­ manifested Recommendations in fragmented spatial development, poor Kigali has a crucial role to play in Rwanda’s transport connectivity, and weak fiscal future development, because “urbaniza- handles. tion” economies—especially those involv- Urban fragmentation in Rwanda is driven ing knowledge spillovers—are generated by institutional and infrastructural con- mainly by large cities. Market forces— straints. Inefficiencies in land use management reflected by the location decisions of people are exacerbated by the ineffective imple- and businesses—have chosen Kigali. Durable mentation of master plans and urban design investments in infrastructure, housing, and documents with targets for densities, building other amenities are needed to ensure that structures, and land uses that do not match the city is productive and livable. However, current market demand. Rwanda essentially Rwanda’s small towns, secondary cities, and has been managing cities by regulating the Kigali should be performing complementary type and intensity of private development by functions, each differentiated by the type of location—epitomized by the zoning of struc- scale and agglomeration economy they can tural characteristics and building uses. Such deliver. These complementary urban locations mismatches, not uncommon in low- and do not have to be huge to generate agglom- m iddle-income countries, are particularly ­ eration benefits. The size of settlements mat- serious in Rwanda, because master plans are ters less than their function: with reasonable updated only irregularly. A lack of local plan- transport costs, towns can be large enough to ning capacity and inconsistent application of facilitate internal scale economies. Medium- regulations exacerbate the problem. size cities are often large enough for “local- Rural and urban connectivity is another ization” economies that come from thick key challenge. A third of Rwanda’s poor live input markets. The implication is that policy 34 FUTURE DRIVERS OF GROWTH IN RWANDA makers should focus on the functions of dif- Prioritize Investments.  Another priority is to ferent types of cities and support agglomera- provide efficient, affordable, and integrated tion though economic density. public transport and to boost investment in As a general principle, policy makers roads. Public transit routes to Kigali’s center should look to deliver a basic standard of ser- are congested and likely to become more so. vices to all people, while letting markets pick To limit sprawl, excellent central public trans- the pace and form of private sector devel- port should be combined with quality pub- opment by location. This principle is exem- lic transport to a small number of strategic plified by Korea, which allowed the dense peripheral locations. Grids should be provided concentration of economic activity in Seoul, (in Kigali and in secondary cities) to ensure while delivering basic services like education, that rapid peripheral development takes place health care, and clean water evenly across the in a planned way. Greater investment in roads population. Investments in human capital are also is necessary to improve the integration especially “safe” bets, because human capi- of Rwanda’s cities with the countryside and tal is a portable asset that can have powerful with neighboring countries and to strengthen effects on people’s welfare and mobility. connections between urban centers. Getting on the path to faster urbaniza- Kigali’s growth needs to be accompanied tion and greater agglomeration will involve by a more calibrated approach toward devel- reshaping regulatory policy, prioritizing oping a complementary set of secondary cit- investments, and strengthening institutions ies and small towns. Potential investments along three dimensions: spatial, sectoral, and in other settlements should be divided into institutional. On spatial reforms, a more bull- those that can be made at low risk before ish view is needed to unleash the economic market demand has emerged versus those potential of Kigali, which market forces have that are better suited after it has emerged. chosen as the main center of production and Here, choices need to be made between commerce. On sectoral reforms to enhance investments in place-specific durable assets the efficiency of capital investment in infra- (such as infrastructure, industrial complexes, structure, housing, and commercial struc- and housing) and people-focused portable tures, there is an urgent need to strengthen assets (such as health, education, and water land markets. On institutional reforms, and sanitation). International experience stronger intergovernmental coordination of shows that such investments should center economic and spatial planning processes is on increasing external connectivity. Stronger needed. connectivity, however, does not guarantee the greater development of secondary cities Reshape Regulatory Policy.  The main and other towns versus Kigali. Connectivity priority is to strengthen land markets so and specialization may well entail greater that private developers have the appropri- concentration of activities in Kigali. ate incentives to increase density. Rwanda The government has pursued a policy of should explore moving from the quantity- grouped villages (imidugudu). The program based regulation of land use, epitomized provides rural-urban links by preparing by highly prescriptive zoning requirements future urban dwellers for urban life while for structures, to price-based allocation. they are in rural areas, and aims to preserve Land prices, which reflect the demand for agricultural land. Improvement in the plan- and scarcity of locations, should become ning and establishment of settlements is the basis for allocating private investments needed to ensure that they serve the intended across space and for allowing structures to purpose while minimizing huge infrastruc- serve the evolving economy. Regulations on ture costs and potential future financial structures should be used to mitigate nega- losses. Prioritization of rural-urban connec- tive externalities like environmental degra- tivity and better anticipation of migration dation and structural safety. to larger urban areas are needed to address OVERVIEW 35 underlying concerns. This effort can be private and public firms—that responds complemented with “off-grid” rural infra- with agility to emerging domestic, regional, structure (for example, maintenance of rural and global opportunities is essential for roads and bridges) that delivers key services future growth. As Rwanda’s economy while placing smaller “bets” on the long-run develops and matures, the need for the pri- vocation of the location. vate sector to play a leading role in achiev- ing the rapid growth required to reach Strengthen Institutions.  Faster urbanization Rwanda’s ambitious income aspirations and greater agglomeration require stronger only increases. Private enterprises create institutions. Urban planning capabilities at jobs and generate income, drive economic the city and district levels need to be con- transformation, compete in global export siderably stronger. Plans should respond to markets, and, ultimately, drive aggre- markets and community needs, while safe- gate productivity growth and innovation. guarding against capture by special interests Moreover, the private sector has to provide through strong public oversight and engage- much of the huge investment needed to sup- ment. Further, broad institutional and gov- port future growth ambitions. ernance reforms such as clarification of land Rwanda needs to strengthen the com- and property rights set the foundations for petitiveness of the private sector as it under- urbanization and urban development. takes necessary infrastructural investments. Rationalization of planning activities To support the development of viable enter- is needed to foster institutional coordina- prises, the government has continued the tion. To accomplish this, economic planners strong track record of reforms initiated should start to think spatially, and spatial after the genocide against the Tutsi, as evi- planners should become more aligned with denced, for example, in Rwanda’s rapidly economic planning and goals. improved ranking in the Ease of Doing Further, economically connected districts Business Index, which now need to be fol- have to be encouraged and required to coor- lowed through with the reform measures dinate land use plans, transport and service described below. provision, major infrastructure like special The state has an equally important, but economic zones, tourism infrastructure, and complementary, role to play in correcting so on to exploit complementarities. market failures and defining its role in pri- Finally, stronger institutions are needed to vate investment—by implementing well- value land, to disseminate land values pub- targeted public investments, strengthening licly across uses, and to assign and protect bureaucratic capacity, withdrawing gradu- property rights. Rwanda has a strong basis ally from productive activity while enhancing for efficient and transparent valuation, with its regulatory and facilitation functions, and centralized and digitized landownership and developing capable and accountable market transaction records, a professional valuation institutions. Public investments will remain body, and credible institutions for oversight. important for several years to come, given Credible land valuation would enable urban the vast infrastructural and social needs of areas to fiscalize public investments in land the country; that said, with growing financ- through land value capture. ing needs under the National Strategy for Transformation and tightening limits on public borrowing, the relative share of pub- Competitiveness and Enterprise lic investment will decline over time as the Development for Innovation-Led private sector share increases. SOEs will also Growth remain crucial for several years to come, Introduction because the private sector needs time to build. Emergence of a vibrant, competitive, and The reform agenda to boost the competi- innovation-driven enterprise sector—both tiveness and innovativeness of the enterprise 36 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE O.23  Share of firms in Rwanda, by firm stages of development. The contrast is espe- size, 2011 and 2014 cially stark when compared with the relative price trajectories of the Asian high-growth 80 economies whose long-term performance 70 Rwanda aspires to surpass. It was only when 60 these economies reached income per capita of more than US$5,000 (2010 U.S. dollars, Share of rms (%) 50 purchasing power parity) that the ratio of 40 their price levels to those of the United States 30 exceeded Rwanda’s current ratio. 20 Low returns (real and perceived) also con- tribute to the slow uptake in private sector 10 activity. Rwanda’s labor productivity (output 0 per worker) is low by international standards, 1 2–4 5–9 10–19 20–99 100+ Firm size (number of employees) explained by negative or weak growth in within-sector labor productivity across most 2011 2014 sectors. Moreover, Rwanda’s TFP levels (out- Sources: Calculations based on 2011 and 2014 Rwanda Census of put generated by a given quantity of labor Business Establishments data (NISR 2011, 2015). and capital) are low for its income, and the rate of TFP growth has slowed significantly sector is comprehensive and demanding, in recent years, further widening the gap with reflecting Rwanda’s strong ambitions for other countries. future growth. W hat explains Rwanda’s weak T FP p erformance? A country’s TFP has two ­ Challenges and Opportunities key determinants. The first is the alloca- The enterprise sector in Rwanda, in its tive efficiency of its resources (land, labor, current form, remains relatively small, young, ­ c apital)—that is, the extent to which these ­ and concentrated in the nontradable sector. public and private resources get channeled Private sector firms are small (figure O.23); to their most productive use. The second is they lack the scale economies critical for the pace of technological innovation in the competitiveness and have limited export pres- ­ e conomy—that is, the pace at which the ­ ence. Moreover, as in other low-income coun- frontiers of technology and good business tries, informality is prevalent. practices expand. The significance of tech- The limited presence of formal enter- nological innovation will gather pace as the prises reflects high cross-cutting costs, lower Rwandan economy moves toward middle- productivity than what is needed to sustain income status and beyond, becoming an rapid economic growth, and low capacity for increasingly pressing subject of reform for innovation. future growth. Price levels and thus input costs are high Resource misallocation refers to ineffi- in Rwanda, which constrains the competi- cient enterprises commanding more resources tiveness of its enterprises. Low- and middle- (land, labor, capital) than warranted by their income economies tend to have lower price productivity levels. Analysis of Rwanda’s levels than high-income economies because of manufacturing sector suggests that TFP in the lower costs of nontradable goods and ser- Rwanda’s manufacturing sector could be dou- vices (the prices of tradable goods tend to be bled if resource misallocations were reduced similar across all countries). This divergence drastically (figure O.24). The scope for in prices tends to close as countries’ income improving productivity through this channel levels rise. Rwanda’s relative prices have fol- is likely even larger in other sectors.21 lowed such patterns, but at more elevated lev- Some of the underlying causes for mis- els compared with other economies at similar allocation are straightforward, rooted in OVERVIEW 37 FIGURE O.24  Resource misallocation in formal be strongly positioned on innovation capac- manufacturing in Rwanda and select countries ity. Currently, private enterprises do not serve as strong demanders of innovation. On top Kenya of that, managerial capabilities remain low, which undermines firms’ capacity and incen- Rwanda tives for upgrading technology and operating Ghana practices. Innovation is also hampered by the scarcity of links with foreign firms and Ethiopia export markets and by limited FDI in trad- India able sectors. The supply side of Rwanda’s innovation system is only starting to emerge. Côte d’Ivoire China Recommendations Rwanda’s demanding reform agenda to boost 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 the competitiveness and innovativeness of the Measure of misallocation enterprise sector reflects the government’s Source: Calculations based on NISR 2017. great ambitions for growth. Key areas include Note: Resource misallocation is measured by the dispersion the following. of marginal products of inputs across firms. Large dispersions suggest that frictions in input and output markets prevent the movement of productive resources across firms and that Reduce Business Costs by Tackling Cross- inefficient firms command more resources than warranted by their productivity. Cutting Constraints. Rwanda has made substantial reforms to improve the business environment and support firms. The first the way in which both the state and mar- imperative is to continue to improve afford- kets allocate resources. Decisions on public able access to finance. Over time, capital investment—accounting for more than half markets and nonbank financial institutions of total investment in the economy—need need to be developed. There also is consid- to be guided by clear analysis of economic erable potential for tapping into foreign returns. Rwanda has made good progress sources of funding, which would require in this regard, for example, in developing a actions at the regional level to harmonize robust public investment management sys- regulations and reduce barriers to services tem. Moreover, the government’s focus in its trade and investment. Next, to reduce the industrial policy needs to shift toward differ- cost and increase the reliability of electric- entiating firms and sectors by performance ity, Rwanda needs to forecast and manage (past or potential). domestic supply and demand of energy, Rwanda is in the early stages of building build capacity to procure and implement its innovation capacity, which is the ability power-purchasing agreements, and bet- to introduce new products, processes, ideas, ter integrate system planning and system technologies, and solutions. It needs to con- operations functions in the utility. Efforts tinue strengthening these capabilities for to reduce logistics and broadband costs are productivity-​ ­ led growth to meet its long-term also important for the competitiveness of income ­ ambitions. Although structural trans- enterprises in Rwanda. formation and improved allocation of resources can stimulate labor productivity growth for Improve Effectiveness of the Government’s ­onger another decade or so, sustaining it over a l Tax Incentives and Other Industrial Policy period will only be possible through the diffu- Interventions.  Rwanda has already estab- sion of innovation and technology. lished some of the key preconditions for an Innovation capacity of Rwandan firms has effective industrial policy. However, incen- been on an upward trend, but several chal- tives need to be linked more clearly to the lenges need to be addressed for Rwanda to productivity and export performance of 38 FUTURE DRIVERS OF GROWTH IN RWANDA beneficiary firms. Direct support should be competitive environment would encourage focused on maximizing foreign exchange firms to seek out the best available knowl- earnings (or savings) and targeted to edge and strengthen managerial capabilities strategic sectors well placed to generate to introduce new processes and technologies, exports, economic growth, and job cre- integrate them in the production system, allo- ation. Crucially, a strongly monitored cate skilled staff to use them, and make them ­ p erformance-based approach (focused on financially viable. Improvements in man- firm productivity and exports) should be agement practices can be achieved through mainstreamed in all interventions. various training and coaching services, Rwanda’s industrial incentives need to be improvements in graduate-level management better coordinated across government agen- courses, and efforts to encourage increased cies. Incentives also could do a better job of interaction between multinational firms and attracting FDI in tradable activities. local suppliers. Stronger coordination, moni- toring, and evaluation of funding programs Define the Future Role of SOEs and for science, technology, and innovation are Further Strengthen Their Corporate also needed. The agenda on the supply side of ­Gover­n ance.  SOEs played an important innovation is discussed in chapter 1. role in the early stages of Rwanda’s economic development and will continue to do so in the Transitioning Agriculture and Food to medium term. On the basis of sectoral assess- Be a Longer-Term Engine of Growth ment of the level of competition, competitive- ness of SOEs versus private firms, economic Introduction development goals, and social considerations, Agriculture has been a major source of a useful starting point would be to divide key national income and growth for Rwanda. It sectors into four groups: (1) sectors in which accounts for close to 70 percent of employ- SOEs will retain a monopoly; (2) sectors in ment, more than 30 percent of GDP, and which SOEs will compete with private firms; more than 50 percent of exports of goods. (3) sectors from which SOEs will withdraw Agricultural value added has risen more than when efforts to build up private sector capacity 5 percent a year over the past 15 years, which prove successful; and (4) sectors from which is a high rate of growth by global standards; SOEs will withdraw immediately because the and productivity in the sector has increased private sector is already capable and there strongly. Further, growth in agriculture has is no compelling social rationale for them. likely stimulated growth in other sectors Public-private boundaries can shift as the pri- through backward links (for example, by vate sector gains strength. Strong ­ government encouraging input industries such as fertil- support will still be needed, but its role will izers) and forward links (for example, by increasingly be to facilitate private investment increasing food processing). Widespread and build strong market institutions that sig- growth in the purchasing power of farm nal priority sectors. Enterprise development households in remote areas has also helped in Rwanda must somehow resolve the some- to mobilize unemployed local resources times-conflicting goals of alleviating market by increasing demand for local goods and failures and encouraging private sector devel- services. opment. The reluctance of private investors to Considerable government effort has been invest enough in public goods such as infra- devoted to developing the sector, especially structure is one market failure. Another may after the food price shocks of 2007– 08. be low investment capacity among Rwanda’s Major government efforts have included private sector. development of new irrigated land, large- scale land consolidation, and land regis- Build an Effective National Innovation tration. The development of cooperatives System.  On the demand side, a strong has been a major component of state-led OVERVIEW 39 collective action, although there is scope for implementation of those private sector deci- much more. Promotion of the use of fertil- sions. Like any rapidly growing economy, izers and improved seeds and effective inter- Rwanda needs to evaluate shifting patterns ventions in livestock production have also of demand in regional and global markets, contributed to growth in agriculture. discover its core advantages, and then assess With this strong start, the main question how they can best be deployed to specialize now is how long and to what extent agricul- in food and other agricultural products that ture will remain an engine of growth in the will maximize income gains. Rwandan economy. Does Rwanda’s highly Rapid increases in productivity also ambitious economic future continue to have require more mechanization, higher use of agriculture at its core, or will the sector inorganic fertilizer and improved seeds, and be useful mostly to launch high long-term improved access to financing. Despite sub- growth and then gradually take a back seat? stantial progress in land consolidation, most This report contends that the agriculture of Rwandan agriculture is carried out under sector will remain a major source of com- conditions of abundant labor on small plots, parative advantage for Rwanda, even as the mostly on hillsides. The use of labor-saving country climbs the income ladder over the machinery, such as tractors and combine coming decades. harvesters, is confined to a few small niches. However, mechanical solutions are needed Challenges and Opportunities to expand bench terraces and extend irriga- Periods of rapid growth in the past have tion at lower costs per unit, in both cases largely reflected either the expansion of agri- involving whole communities. Mechanical cultural land or the increased use of inputs solutions also are needed for postharvest (figure O.25). However, there is diminishing functions such as transport, processing, marginal scope for further expansion of cul- tivated land. There is significant potential for FIGURE O.25  Decomposition of sources of agricultural growth continuing to improve farmer skills, extend in Rwanda, 1961–2014 irrigated area, stabilize and expand terraces, boost the use of more adapted fertilizers, 6 5.55 4.01 5.15 and increase farmers’ use of better seeds. 5.04 5 Programs targeting these improvements Average annual net agricultural could boost agricultural growth over the next 4 output growth (%) decade or so. Maintaining rapid growth over 3 1.84 a longer period, however, requires greater 2 emphasis on TFP growth through more effi- cient allocation of resources and greater reli- 1 0.05 ance on technological innovation. 0 For stronger productivity growth, the sec- tor has to continue to modernize, become –1 more responsive to market signals, and –2 integrate more effectively with regional and 1961–70 1971–80 1981–90 1991–2000 2001–10 2005–14 global markets. Its contribution needs to TFP Area expansion Inputs intensification shift from supplying commodities primarily for domestic use to producing higher-value- Source: Compiled from U.S. Department of Agriculture 2017 data. Notes: The three sources of growth listed sum to output growth in the period in question. added goods as an integral part of food supply Output growth may be different than column height because of negative contributions to chains linked to regional and international growth in the period in question. Growth not explained by area expansion or increased use per hectare of inputs (including labor) is attributed to total factor productivity (TFP). TFP is markets. This effort requires an institutional, a combination of increased allocative and technical efficiency and technological change. infrastructural, and policy environment led Because the numerator of TFP (output growth) is highly affected by weather outcomes in any given year, it makes sense to consider TFP outcomes over at least a 10-year period by the market, with a commensurate gradual that includes both good and less good weather. This explains the strategy for handling shift in the public role to facilitating the fair differences between the early and later parts of the period 2001–14. 40 FUTURE DRIVERS OF GROWTH IN RWANDA and packaging. The declining use of inor- organic fruits and vegetables including jams ganic fertilizer also needs to be reversed, by and frozen dinners, packaged perishable ani- continuing to expand the coverage of loca- mal products, cut flowers, and gourmet teas tions and to improve the formulation of and coffees of specific geographic origins. fertilizers. The government is already mod- Even under current conditions, Rwanda’s ernizing its seeds regulations and is also tak- comparative advantage in these items rela- ing steps to legalize agricultural genetically tive to its neighbors is strong by conventional modified organisms. Finally, limited agricul- measures. tural finance remains a major constraint: the use of formal financial services in the agri- Recommendations culture sector is quite low. In the medium term, Rwanda will have an Environmental concerns arising from agriculture sector that is increasing in pro- global climate change and land degrada- ductivity and adapting to the future. State-led tion are a particular challenge that needs efforts to increase agricultural productivity to be managed proactively for improving and maximize the sector’s potential in Rwanda agricultural prospects in Rwanda. Rwanda remain relevant in the medium term. These has experienced a 1.4°C mean tempera- efforts involve extending the approach to con- ture increase since 1970 and is on track to solidation of production decisions and inten- experience another 1°C to 2.5°C increase sification of staple food crops. Public support over the next three decades. Higher tem- should continue to improve the efficiency of peratures have led to the spread of pests and small farms that are still not using fertilizer in diseases, impairing the health of livestock appreciable amounts, lack access to adequate and humans, lowering crop yields, harming agricultural water, are situated on unprotected food security, and decreasing export earn- hillside slopes, or use inferior seeds, so that ings. Rainfall patterns, already highly vari- such farms can achieve the technical efficiency able in Rwanda, have become even more of Rwanda’s better farmers. variable, with projections of a 20 percent Higher agricultural growth in the longer increase in variability by the 2050s. In the run can only be achieved by reaping the ben- east and parts of the south, this limits the efits of scale economies and specialization availability of water and feed for livestock through trade and the production of higher- and increases the vulnerability to diseases. value-added goods. Such efforts require At the same time, heavier rains, particu- a rapid response to market signals, ready larly in the north and west of the country, access to investment resources, technical increased floods and landslides, resulting expertise, and ability to organize production in crop losses, health risks, and damage to and provide appropriate incentives for work- infrastructure. Moreover, the combination ers, led by the private sector. The state needs of rain-fed, small-scale agriculture, high to continue playing a leading role as umpire, rainfall levels, and steep hillsides (where generator of public knowledge, provider of most of Rwanda’s agriculture is practiced) public goods such as infrastructure and basic leads to very high rates of soil erosion. research, and responder to other market fail- Rwanda can harness its comparative ures. Targeted high-return initiatives should advantage in high-value crops and livestock receive public financial support, accompa- products benefitting from a hillside envi- nied by a rigorous evaluation to determine ronment combined with extensive biomass, whether adequate returns are being achieved. intensive inputs of skilled labor, the ability to Public sector activities in agriculture will brand credibly (as in organic or Fair Trade), increasingly be targeted more effectively to pro- and the ability to support higher transport viding key public goods in seven key areas. costs to final markets. Potential products include branded mountain cheeses, essen- Adapt Rwanda’s Research and Regulatory tial oils, potato chips, baby food, certified Institutions to Evolving Opportunities and OVERVIEW 41 Threats.  A key element of Rwanda’s com- integrity of food security. Regional integra- parative advantage is the credibility of its cer- tion also should be pursued through continu- tifications, including ISO 22000 ­ certification ing efforts to adopt science-based standards, for food safety and Fair Trade standards product registrations, and certification of under the Rwanda Standards Board. Organic agricultural inputs. certification also may become advanta- geous. The high technical expertise of the Profit from the Big-Data Revolution in Rwanda Agriculture Board will be increas- Innovation, and Make the Benefits of Big ingly important as improving standards and Data Accessible at Reasonable Cost to Rwanda’s efforts to expand exports increase Smallholders.  This effort is critical both the frequency and importance of sanitary and to increasing innovation and private sector phytosanitary issues. Rwanda’s reputation for investment and to dealing with the smallest probity and accountability is one of its most farms and most densely settled rural areas precious assets and is especially important in in Africa. “Smart farming,” where just the agriculture, where quality and food safety right amounts of the right inputs are used issues are dominant concerns of consumers. for each parcel in response to information gathered by handheld devices with the right Improve the Effectiveness of Policies on sensors, offers one such opportunity. Big- Vertical Coordination. Improved policies data approaches such as blockchain technol- can be achieved by expanding participa- ogy could offer the potential for lowering tion in cooperatives and encouraging private the costs of small financial transactions that sector aggregators. More effective vertical require secure record keeping and decen- coordination in agriculture would be where tralized input, such as land registration and private sector partners provide skills, capi- mortgages. tal, and entrée into international markets and where farmers acquire higher incomes Link Public Investments in Infrastructure and new resources. Government can improve More Effectively to Agricultural Development general operating conditions for aggregators and Higher-Value-Added Products.  Irrigation and develop a knowledge platform regarding is clearly a constraint, especially in marsh- which forms of industrial organization work lands and in the drier eastern regions of the best for addressing specific industrial organi- country. High-value export crops (such as zation problems. The platform could provide horticulture for export) would likely cover the reliable data on vertical coordination, finance, cost levels much more easily if cost-effective prices, costs, and weather; this information is supply chains were in place that could realize critical to reducing uncertainties and lowering international prices. risks for private investors and should encour- age competition. Further Develop High-Level Human Capital Formation in Agriculture.  In particular, to Engage with Neighboring Countries on a achieve vertical integration, a more consis- Regional Division of Labor. A regional tent approach is needed to training managers division of labor would include both and technical experts working for aggrega- production based on comparative advantage ­ tors. Companies could be involved in this and knowledge generation and diffusion. effort through training programs to build up Rwanda has a major stake in increasing the nationally needed skill sets in agribusiness, importation of raw materials such as cere- high-value supply chains, and agricultural processing and animal pro- als for its food-­ technology. The best multinational com- tein industries, while increasing its exports panies have a track record of making these of processed foods and high-value calories, investments and provide examples of how for example, from milk and meat products. investments in human capital can be done to Ways should be explored to guarantee the mutual advantage. 42 FUTURE DRIVERS OF GROWTH IN RWANDA Address Land Degradation and Global Climate hold all government officials accountable for Change More Comprehensively. Although performance through the system of perfor- considerable progress has been made in mance contracts (imihigo); established the constructing robust “bench” (wide) ter- Joint Action Development Forum, a partici- races, much more needs to be done to secure patory planning mechanism to improve the Rwanda’s land asset. Fortunately, the soil and alignment of citizens’ and districts’ priorities; water management interventions c ­entral to and created an annual forum for national halting land degradation also would increase dialogue (Umushyikirano) to allow all citi- the ability of farmers to adapt to climate zens to have direct access to the most senior change through better water retention and leadership in government. improved soil quality. Landscape restora- These efforts have proved effective, well tion and conservation are a community-level reflected in Rwanda’s international rank- activity, not an individual one. Local govern- ings, with particularly strong performance ment, producer groups, and national techni- on indicators of government effectiveness, cal expertise and funding all have a role to control of corruption, rule of law, and reg- play in implementation. The development and ulatory quality (figure O.26). Reflecting use of climate-smart and big-data agricultural the government’s high-profile anticorrup- technologies can further help farmers to cope tion efforts, Rwanda was ranked 48 (out of with the challenges of climate change. 180 countries) for control of corruption on Transparency International’s Corruption Perceptions Index in 2018, marking a vast Capable and Accountable Institutions improvement over its 2006 ranking of 121. Introduction Emerging from very difficult initial condi- Challenges and Opportunities tions after the 1994 genocide against the Rwanda’s strong overall governance has been Tutsi, Rwanda’s leadership faced an enor- a driving force behind its rapid growth and mous challenge not only of rebuilding a dev- astated economy but also, more broadly, of FIGURE O.26  Benchmarking of Rwanda along the various dimensions of governance constructing a national identity and renewed sense of national purpose. Reestablishing peace and social stability and building the Control of decimated social capital of trust were viewed corruption as the underlying sociopolitical challenges Rule of law of governance (see Jones and Murray 2018). The government’s response was decisive in Regulatory quality several forms, establishing security and seek- ing to channel social energies toward the Government national purpose of raising incomes, improv- effectiveness ing opportunities, and providing health and Political stability education to all citizens. Along the way, the government created Voice and new homegrown solutions and programs to accountability create an integrated Rwandan community, 0 20 40 60 80 including revitalizing the community work Percentile rank, 2016 program to bring people together around a Mean of UMICs Mean of LICs common community purpose (“Umuganda”) Mean of LMICs Rwanda once a month; reestablished the traditional community courts (Gacaca) and created Source: 2016 Worldwide Governance Indicators from Kaufmann, Kraay, and Mastruzzi 2010. mediation committees (abunzi) to achieve Note: LIC = low-income country; LMIC = lower-middle-income reconciliation and mete out justice; sought to country; UMIC = upper-middle-income country. OVERVIEW 43 provides a key comparative advantage over specialization of justice sector employees, most peers in the region. But the require- including judges, prosecutors, and inves- ments of a middle-income Rwanda will be tigators; and improving case management different—more demanding—than those techniques would enhance the judiciary’s that sufficed in the low-income period. effectiveness in enforcing contracts. Further Building effective institutions is a long-term improvements in the efficiency of the land endeavor. The improvements are necessarily management system and provision of more incremental, building on global and local les- resources to implement the law on intellec- sons and factoring in the specific context of tual property would strengthen property the country. rights. Reviewing the performance of fiscal To strengthen institutional effectiveness incentives with a view to reducing expan- requires greater social capital of trust— sion in Rwanda’s tax expenditures, increas- between citizens, between civil society ing local government revenue collection, and government, and even within national ensuring that the state captures an appro- political leadership. In Rwanda, interper- priate share of the expected rise in mineral sonal trust was understandably very low in revenues, and taking further steps to raise the immediate aftermath of the genocide Rwanda’s strong public financial manage- against the Tutsi. Impressive progress has ment to international standards would been made since then on achieving reconcili- strengthen the state’s financial capacity. ation, promoting economic and social devel- A final set of governance challenges has to opment, and strengthening the structures of do with accountability mechanisms between governance. But more work is needed in this the government and citizens. Rwanda has in regard, given the strong positive relationship place homegrown initiatives, which aid in between trust and long-term development. creating demand for accountability by citi- Key challenges also relate to limited inno- zens through making them shareholders of vation for public civil servants and local development. governments. Imihigo has made a strong contribution to instilling a culture of results. Recommendations Building a more complex, richer society Strong growth over the next decades will requires an approach where civil servants, depend on specific institutional requirements, local officials, and civil society can serve as a including a market system that encourages source of innovation. It also requires the gov- entrepreneurs to take more risks in search of ernment to obtain and act on feedback from higher rewards by protecting private prop- citizens. Essentially, public institutions need erty and minimizing policy uncertainty. to become more innovative, better coordi- Strong growth also requires enhancing nated, and more adaptive. state effectiveness through three core insti- To achieve continued, rapid growth, an tutional pillars: (1) enabling innovative, effective market-based economic system is coordinated, capable bureaucracy and local needed to encourage private sector develop- governments; (2) using efficient market sig- ment, matched by financial resources of the nals to enable strong cooperation between the state. Despite Rwanda’s impressive improve- state and the business sector and maintaining ments in governance, private sector invest- fiscal prudence; and (3) strengthening public ment and the entry of new firms remain administration accountability. All three com- low, reflecting gaps in market institutions. ponents are essential for an effective state Creating special courts and fast-track pro- (figure O.27), which, in turn, is necessary for cedures to adjudicate small claims; further sustained high growth. promoting alternative means of resolving commercial disputes (such as arbitration, Pillar 1: Enable Innovative, Coordinated, mediation, and conciliation); boosting reli- Capable Bureaucracy and Local ance on technology; increasing training and Governments.  To achieve strong growth 44 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE O.27  A framework for assessing Pillar 2: Strengthen Market Institutions and effectiveness of state institutions in Rwanda Fiscal Prudence.  The evidence is compelling that functioning markets require well-defined State rules of the game, enforced transparently and effectiveness predictably. Advanced economies (almost as a rule) have a system of highly evolved eco- nomic institutions that convey prices, define Societal trust property rights, enforce contracts and com- petition policies, and close informational gaps Well-structured between buyers and sellers (Commission on Strong market State and innovative institutions accountability Growth and Development 2008). As part government of its long-term development, Rwanda has to develop these key market institutions (for competition, contract enforcement, and prop- over the next decades, well-structured, capa- erty rights), backed by strong financial capac- ble, and empowered civil service and local ity of the state. governments are needed, with the resources Establishing the Rwanda Inspectorate and and other incentives to be innovative (even Competition Authority and extending its entrepreneurial). Such innovation entails the mandate to monitor the impact of SOEs on incentives and resources to inform decision competition, promoting the principle of com- makers and challenge them with evidence and petitive neutrality to ensure equal treatment analysis when needed; to share information of all investors, and removing regulatory and coordinate actions within and outside the barriers to entry and rivalry in input sectors administration; and to take informed risks, would improve the competition framework. experiment, scale up successes, and learn Creating special courts and fast-track proce- from failures. Creating such institutions will dures to adjudicate small claims, promoting strengthen capacity and coordination within alternative means of resolving commercial the government and make progress on decen- disputes (such as arbitration, mediation, and tralization to encourage local initiative and conciliation); boosting reliance on technol- eliminate overlaps in responsibilities. ogy; increasing training and specialization Modifying the imihigo to adopt multiyear of judges, and improving case management targets with annual benchmarks (rather than techniques would enhance the effectiveness focusing only on short-term goals), making of the judiciary system for contract enforce- greater use of qualitative surveys, and placing ment. Further improvements in the efficiency a stronger focus on outcomes rather than pro- of the land management system, better public cesses could enhance innovation. Increasing information on market values of all catego- the importance of the joint imihigo and a ries of land, and provision of more resources strong delivery unit to coordinate policies on to implement the 2008 law on intellectual key cross-cutting issues would improve inter- property would strengthen property rights. agency coordination. Further increases in Increasing local government revenue col- training by the Rwanda Management Institute lection, ensuring that the state captures an should be supported by skills audits and func- appropriate share of the expected rise in tional reviews to identify the missing skills mineral revenues, and taking further steps to in the public sector. Regular compensation strengthen Rwanda’s public financial man- reviews, higher compensation for individu- agement will be crucial. A better balance has als with scarce skills, and the development of to be found between public investments that nonfinancial incentives could become increas- are expected to generate high returns over the ingly important to retain and attract highly long term versus investments in areas with skilled staff as competition for scarce skills potentially high short-term returns based on increases from private companies. market signals of scarcity. OVERVIEW 45 Pillar 3: Enhance the State’s Accountability and the free withdrawal from foreign to Citizens.  For institutions to remain legiti- exchange accounts in commercial banks mate, the state has to be “citizen-centered,” (Ggombe and Newfarmer 2017). supported by an effective feedback loop 3. This report considers the period 1995–2005 as the recovery period and the subsequent between civil society and local governments phase as the postrecovery period. and service providers. The capacity and moti- 4. This stagnation mostly reflects a decline vation of citizens to participate with state in private savings, as government savings actors in assessing service delivery and find- increased. The savings-investment gap has ing solutions are critical for Rwanda’s emer- been made up by external assistance and, gence as a strong, successful middle-income increasingly, by FDI inflows. economy. 5. Although future investment does not have to Reform measures can seek to strengthen be financed through domestic savings if the checks and balances further to improve country has ready access to external financ- accountability over the executive, local offi- ing, running large external imbalances over cials, and service providers. Strengthening long periods has proved unsustainable for countries. key watchdog agencies (the Public Accounts 6. Rwanda’s high costs stem from the small Committee, the Office of the Auditor market size, long distance to external mar- General, the Office of the Ombudsman, kets (by virtue of being landlocked), shallow and the Rwanda Governance Board) could and costly credit markets, high cost of energy improve government accountability. Greater generation, and scarcity of land that increases involvement of citizens in local decision mak- the rent on it. ing (for example, in setting imihigo objectives) 7. For example, Lendle and Olarreaga (2017) and reliance on more qualitative information find that the impact of distance on cross- in monitoring performance would strengthen border trade flows—across 61 countries and support for local government and improve the 40 product categories—is about 65 percent quality of services. smaller for eBay transactions relative to total international trade. More generally, Osnago and Tan (2016) find that a 10 percent increase in an exporter’s rate of Internet adoption leads Notes to a 1.9 percent increase in bilateral exports. 1. Worldwide, almost 80 percent of income 8. Initial signs are encouraging. FabLab, an growth of the poor is due to increases in innovation hub at the government-backed countries’ income levels (Dollar and Kraay ICT park, has started to work with three-­ 2016). Rwanda, too, has witnessed the close dimensional printers (BBC 2016). association between growth and general 9. Rwanda’s priorities in this regard are already improvements in welfare along several dimen- well reflected, for instance, in the soon-to- sions of welfare in the past two decades, as roll-out 4G LTE broadband network—a chronicled in this report. wireless broadband technology designed 2. Key l ibera l i zat ion measu res i ncluded to support roaming Internet access via c ­ ell (1) elimination of most price controls; phones and handheld devices. (2) liberalization of exchange rates in 10. Three drone ports, which will allow the 1995; (3) lowering of tariff rates from 35 to drone network to send supplies to 44 percent 18 ­p ercent; (4) liberalization of the mon- of the country, are expected to be completed etary and financial sectors, including adop- by 2020. tion of new currency exchange regulations, 11. These countries are Argentina, Brazil, China, creation of private commercial banks, and India, Indonesia, Korea, Mexico, the Russian privatization of state-owned banks; (5) lib- Federation, Saudi Arabia, South Africa, eralization of current account operations and Turkey. Korea is counted in this group (imports, exports, services); (6) elimination because it was not yet a high-income country or easing of certain restrictions on capital in 1990, the base year for the comparison. flows, including the transfer of capital and 12. For example, Chinese domestic companies revenues related to foreign direct investment have captured larger market shares and moved 46 FUTURE DRIVERS OF GROWTH IN RWANDA up the value chain in highly competitive manu- 20. Only 14 percent of Rwanda’s urban popula- facturing industries, such as high-end chemi- tion were recent migrants from rural areas cals, electronics (in particular, information in 2014, while only 20 percent of recent inter- technology), automotive, and aircraft. nal migrants moved from rural to urban areas 13. Robotics, automation, computerized manu- (NISR 2015). facturing, and artificial intelligence could 21. Efficiency of resource allocation across sec- reduce the advantages of production in low- tors can be just as important as allocation labor-cost emerging economies, discouraging within sectors. However, the available data offshoring from and favoring reshoring to do not permit a comparison of the magnitude these countries. of misallocation of resources across sectors in 14. Purchases of real estate properties by Chinese Rwanda and in other countries. individuals in high-income countries also contributed to the boom in FDI outflows. Outflows from Korea have also increased, References driven by rising greenfield investments. Akamatsu, K. 1962. “A Historical Pattern of 15. In 2015, the top 10 investors (by FDI stock) Economic Growth in Developing Countries.” in Africa were the United States, United Journal of Developing Economies 1 (s1): 3–25. Kingdom, France, China, South Africa, Italy, Bashir, S., M. Lockheed, E. Ninan, and J.- P. Tan. India, Singapore, Switzerland, and Malaysia. 2018. Facing Forward: Schooling for Learning 16. High dependency imposes costs at the house- in Sub-Saharan Africa. Washington, DC: hold and national levels, in the form of educa- World Bank. tion, health care, social protection, and other Basinga, P., P. J. Gertler, A. Binagwaho, A. L. B. essential needs of children and the elderly. Soucat, J. Sturdy, and M. J. C. Vermeersch. 2011. When dependency ratios are low, under con- “Effect on Maternal and Child Health Services ducive conditions, lower costs should support in Rwanda of Payment to Primary Health- higher growth through higher savings and Care Providers for Performance: An Impact investments, including in building human Evaluation.” Lancet 377 (9775): 1421–28. capital of children. BBC (British Broadcasting Corporation). 2016. 17. For example, Mindspark, an interactive “Rwanda’s Technology Revolution Helps ­ computer-assisted learning software, increased Country Forge New Path after Genocide.” math scores by 0.59 standard deviation and BBC , May13. http://www.bbc.com /news​ language scores by 0.36 standard deviation /­world-africa-36285889#. for a group of Indian students in 90 days Brenton, P., and G. Isik. 2012. De-Fragmenting (Muralidharan, Singh, and Ganimian 2017). Africa: Deepening Regional Trade Integration Similarly, the Adaptive Mathematics Platform in Goods and Services. Washington, DC: that has been used by half of students in grades World Bank. 3–6 in Uruguay improved math scores by 0.2 Chukwuma, O. 2003. “Rwanda—The Search standard deviation, equivalent to half a semes- for Post-Conflict Socio-Economic Change, ter of schooling (Perera and Aboal 2017). 1995 –2 0 01.” A f r ic a R e g ion Work i ng 18. The main exports to the Democratic Republic Paper 59, World Bank, Washington, DC. of Congo include livestock and crops, but the http://documents.worldbank.org /curated​ Democratic Republic of Congo also provides /e n /6 432714 68759893135/ Rwa nd a- t he​ an important source of (informal) cross-­ border -search-for-post-conflict-socio-economic​ trade in services such as finance, transporta- -change-1995-2001. tion, and wholesale trading (Lalui 2016). Commission on Growth and Development. 2008. 19. Nguyen (2017) shows that a competitive The Growth Report: Strategies for Sustained exchange rate is likely to boost Rwanda’s G ro w t h a n d I n c l u s i ve D e ve l o p m e n t . growth by boosting its exports and manufac- Washington, DC: World Bank. turing production but also cautions that the Diao, X., J. Randriamamonjy, and J. Thurlow. costs and institutional requirements of main- 2017. “Urbanization and the Future of taining an undervalued exchange rate have to E c onom i c D e velopm e nt i n Rwa nd a . ” be considered before embarking on any such International Food Policy Research Institute, policy. Washington, DC. Background paper for Future OVERVIEW 47 Drivers of Growth in Rwanda, World Bank, Hallward-Driemeier, M., and G. Nay yar. Washington, DC. 2017. Trouble in the Making: The Future Dihel, N., A. M. Fernandes, and A. Mattoo. of M a n u fa c t u r i ng - L e d D e ve l o p m e n t . 2012. “Developing Professional Services in Washington, DC: World Bank. Africa through Regional Integration.” In Inglehart, R., C. Haerpfer, A. Moreno, C. Welzel, De-Fragmenting Africa: Deepening Regional K. Kizilova, J. Diez-Medrano, M. Lagos, Trade Integration in Goods and Services , P. Norris, E. Ponarin, and B. Puranen, eds. edited by P. Brenton and G. Isik. Washington, 2014. World Values Survey: Round Six— DC: World Bank. Country-Pooled Datafile 2010–2014. Madrid: Dollar, D., and T. Kraay. 2016. “Growth Is Still JD Systems Institute. http://www.worldvalues​ Good for the Poor.” European Economic survey.org/WVSDocumentationWV6.jsp. Review 81 (January): 68–85. Jones, W., and S. Murray. 2018. “Consolidating Economist. 2012. “Daily Graph: The Urbanisation Peace and Legitimacy in Rwanda.” Country Trap.” Economist (online), October 2. Study, London School of Economics, Oxford EDC (Education Development Center). 2017. Commission on Fragility, Grow th, and Literacy, Language, and Learning Initiative Development, London. (L3): National Fluency and Mathematics Kaufmann, D., A. Kraay, and M. Mastruzzi. Assessment of Rwandan Schools. Report 2010. The Worldwide Governance Indicators: for t he U. S . Agenc y for I nter national M e t ho d o l og y a n d A n a l y t i c a l I s s u e s . Development. Waltham, MA: EDC. Washington, DC: Brookings Institution Press. Feenstra, R. C., R. Inklaar, and M. P. Timmer. Lalui, A. 2016. “Analysis of Cross-Border Trade 2015. “The Next Generation of the Penn in Services: Issues and Trends with a Focus on World Table.” American Economic Review Rwanda and DRC.” TradeMark East Africa, 105 (10): 3150–82. www.ggdc.net/pwt. Kigali. Freund, C., and T. Moran. 2017. “Multinational Lemarchand, G. A., and A. Tash, eds. 2015. I nve stor s as E x p or t Sup erst a rs: How Mapping Research and Innovation in the Emerging Market Governments Can Reshape Republic of Rwanda. GO-SPIN Country Comparative Advantage.” Working Paper 17-1, P rof i le s i n S c ie nc e , Te ch nolo g y, a nd Peterson Institute for International Economics, Innovation Policy, vol. 4. Paris: United Nations Washington, DC. E duc at ion a l , S c ient i f ic , a nd Cu lt u ra l Galasso, E., A. Wagstaff, S. Naudeau, and Organization (UNESCO). M. Shekar. 2016. “The Economic Costs of Lendle, A., and M. Olarreaga. 2017. “Can Online Stunting and How to Reduce Them.” Policy Markets Make Trade More Inclusive?” ADBI Research Note, World Bank, Washington, DC. Working Paper 742, Asian Development Bank Gereffi, G. 1999. “International Trade and Institute, Tokyo. I ndu s t r i a l Up g rad i n g i n t he App a rel Lutz, W., W. P. Butz, and S. KC. 2014. World Commodity Chain.” Journal of International Population and Human C apital in the Economics 48 (1): 37–70. Twenty-First Century. Oxford, U.K.: Oxford Gereffi, G., and O. Memedovic. 2003. “The University Press. Global Apparel Value Chain: What Prospects Muralidharan, K., A. Singh, and A. J. Ganimian. for Upgrading by Developing Countries?” 2017. “Disrupting Education? Experimental United Nations Industrial Development, Evidence on Technology-Aided Instruction in New York. https://ssrn.com/abstract=413820. India.” NBER Working Paper 22923, National G gomb e , K ., a nd R . N e w fa r mer. 2 017. Bureau of Economic Research, Cambridge, MA. “Rwanda: From Devastation to Services— Newfarmer, R., J. Page, and F. Tarp, eds. First Transformation.” W IDER Working Forthcoming. Industries without Smokestacks: Paper 2017/84, United Nations University African Industrialization Revisited. Oxford, World Institute for Development Economics U.K.: Oxford University Press. Research, Helsinki. Ng uyen , H . 2017. “Rea l E xcha nge R ate G over n ment of Rwa nda. 2017. “7 Yea rs Misalignment and Its Relationship to Output Government Programme: National Strategy Growth in Rwanda.” Background paper for for Transformation (NST1): 2017–24.” Draft. Future Drivers of Growth in Rwanda, World Government of Rwanda, Kigali. Bank, Washington, DC. 48 FUTURE DRIVERS OF GROWTH IN RWANDA NISR (National Institute of Statistics of Rwanda). Forward for the Development of Africa: ECA’s 2011. Establishment Census 2011. Kigali: Vera Songwe.” UNECA (news), March 21. NISR, September. https://www.uneca.org/stories/signing​-afcfta​ ——— . 2 015. R w a n d a Po v e r t y P r o f i l e -giant-stride-forward-development​ - africa​ Report 2013–2014: Results of Integrated -eca%E2%80%99s-vera-songwe. Living Conditions Survey (EICV). Kigali: UNFPA (United Nations Population Fund). 2017. Government of Rwanda. September. “Unlocking Rwanda’s Potential to Reap the ———. 2017. “Seasonal Agricultural Surveys— Demographic Dividend.” Policy Brief, UNFPA, Season A (May) and Season B (September).” New York, October. NISR, Kigali. United Nations. Various years. Comtrade OECD (Organisation for Economic Co-operation database. New York: United Nations. and Development). 2017. “The Future of U S A I D ( U . S . A g e n c y fo r I nt e r n at io n a l Global Value Chains: ‘Business as Usual’ or Development). 2017. Literacy, Language, ‘A New Normal’?” Directorate for Science, and Learning Initiative: National Fluency Technology, and Innovation Policy Note, and Mathematics Assessment of Rwandan OECD, Paris. Schools. Washington, DC: USAID. Osnago, A., and S. W. Tan. 2016. “Disaggregating U . S . D e p a r t m e nt of A g r i c u lt u r e . 2 017. the Impact of the Internet on International I nternational Agricultural Productivit y Trade.” Policy Research Working Paper dat abase. Wash i ng ton , DC: E conom ic 7 785, World B a n k , Wash i ng ton , DC . Research  Service. https://www​.ers​.usda.gov​ ht t p s: //op e n k nowl e d g e .worldb a n k .or g​ /data​ - prod​ u cts ​ / ­i nternational-agricultural​ /­handle/10986/24866. -productivity/. Perera, M., and D. Aboal. 2017. “Evaluación Wo r l d B a n k . 2 0 0 7. R w a n d a : To w a r d del impacto de la plataforma adaptativa Sustained Growth and Competitiveness. de matemática en los resultados de los 2 vols. Washing ton, DC: World Bank. aprendizajes.” Centro de Investigaciones ht t p s: //op e n k nowle d g e .worldb a n k .or g​ Económicas, Montevideo. / ­handle/10986/7700. Ranis, G., F. Stewart, and A. Ramirez. 2000. ———. 2015a. “Building a Reform Consensus “Economic Growth and Human Development.” for Integrated Corridor Development in the World Development 28 (2): 197–219. East African Community.” World Bank, Rodrik, D. 2015. “Premature Deindustrialization.” Washington, DC. NBER Working Paper 20935, National Bureau ———. 2015b. Rwanda: Employment and Jobs of Economic Research, Cambridge, MA. Study. Washington, DC: World Bank. Shepherd, B., J. de Melo, and R. Sen. 2017. ———. 2016. Vietnam 2035: Toward Prosperity, “Reform of the EAC Common External Tariff: C re a t i v i t y, E q u i t y, a n d D e m o c ra c y . Evidence from Trade Costs.” IGC Policy Paper, Washington, DC: World Bank. International Growth Centre, Kampala. ———. 2019. Doing Business 2019: Training for UNCTAD (United Nations Conference on Trade Reform. Washington, DC: World Bank. and Development). 2017. Science, Technology, ———. Various years. World Development and Innovation Policy Review. Geneva: I nd icators dat abase. Wash i ng ton , DC: UNCTAD. Wo rl d B a n k . h t t p s : //d a t a ​ / wo rl d b a n k​ ———. 2018. “The Task Ahead Is Great Say . o r g ​/­d a t a - c a t a l o g / wo r l d - d e v e l o p m e n t­​ African Leaders as Continent-wide Free Trade -indicators. Deal Is Signed.” UNCTAD (news), March Zeitlin, A., C. Leaver, O. Ozier, and P. Serneels. 21. http://unctad.org/en/pages/newsdetails​ 2017. “Recruitment, Motivation, and Retention .aspx?OriginalVersionID=1695. Effects of Teacher Performance Pay.” Project UNECA (United Nations Economic Commission Report S-38408-RWA-1, International Growth for Africa). 2018. “Signing AfCFTA Giant Stride Centre, London School of Economics, London. 1 Human Capital and Innovation Introduction test, and one in six students in grade 3 could not answer a single reading comprehension The government’s aspirations to reach upper- question (EDC 2017). Holding the amount middle-income status by 2035 will require of knowledge constant, there is no clear link dramatic improvements in both the amount between education and growth: if students and the quality of human capital. Economies are failing to learn, then investments in edu- that have grown rapidly over an extended cation will not stimulate economic growth. period have made substantial investments Human capital investments begin in early in the health and education of their citi- childhood, including prenatal and early child zens. These efforts are urgent: Rwanda’s nutrition and cognitive stimulation; they investment in human capital lags behind the continue with high-quality basic education, ­ average of low-income countries—for exam- higher education opportunities, and on-the- ple, in stunting and in primary and lower- job training and adult education. They also secondary school completion (­ f igure 1.1). include investments in health, including Moreover, investments in human capital early child nutrition but also vaccinations contribute to growth only after beneficiaries for children and preventative and curative have grown up. Only a small fraction of the health care for children and adults. Finally, a population enters the labor force each year, strong economy invests in innovation to cre- so even dramatic improvements in educa- ate new opportunities. To achieve long-term, tion will improve the skills of the overall rapid growth, dramatic progress is needed in workforce only slowly. Under the current a wide range of policies supporting human rate of improvement, Rwanda will still have capital formation. low-income-country rates of human capital Government can invest in a range of in 2035. The sooner Rwanda increases its interventions to address stunting, includ- human capital investments, the sooner it will ing investments to improve food security, be on a track to achieve more rapid growth. access to clean water and good sanitation, Rwanda also has to improve the quality access to nutritional supplementation, and of human capital: 85 percent of Rwandan awareness of good nutritional practice. The students at the end of grade 3 were rated National Early Childhood Development “below comprehension” on a recent reading Program (NECDP) has been established to 49 50 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE 1.1  Human capital outcomes in Rwanda and other Many teachers in the upper grades of primary countries, by income level, 2015 and in secondary have a limited grasp of the language of instruction: at last measure, only 43 percent of teachers were assessed at the “intermediate level” in English. Although a Stunting rate great improvement over a decade ago, this level of competency falls far short of what is needed for effective teaching. Rwanda may Primary wish to leverage its major investment in lap- completion rate tops in schools to provide teachers with regu- lar opportunities to improve their English. Technical and vocational training is hard Lower-secondary to get right, so Rwanda needs to ensure completion rate quality and encourage youths to participate in high-return programs. Many high-growth 0 20 40 60 80 100 countries (the Republic of Korea, Malaysia, Percent and Singapore) have supported skills devel- Rwanda Low income Middle income High income opment by setting a strategic direction, tone, and culture for efforts to improve workforce Source: World Development Indicators data (World Bank, various years). skills, creating an organizational infrastruc- ture with appropriate governance, and fos- address issues of coordination and inconsis- tering efficient and effective management tent financing. Rwanda will need a big push of service delivery by providers. Collection across areas, with careful monitoring to eval- and dissemination of information on the uate which policies deliver the greatest gains. quality of skills providers and the returns Expanding access to basic education means to different skills would encourage youth reducing the costs for families. Countries that to participate in high-return sectors and have achieved widespread basic literacy have help to improve the quality of skills training offered free primary education. Although programs. Rwanda provides primary education without Rwanda is building its tertiary sector, charge, families are often obliged to pay inci- but more focus is needed on high-return dental fees. Achieving universal primary edu- activities. University enrollment has dou- cation will require ensuring that children are bled from a decade ago, and Rwanda has not turned away for failure to pay incidental taken dramatic steps to improve the quality fees and providing families with information of tertiary education, consolidating public so that they understand the financial returns universities into the University of Rwanda to schooling. for better governance. Increasing access Improving the quality of education in to financing, including loosening restric- Rwanda means improving the quality of tions on private financing, would help to teachers. Rwanda’s teachers generally attend expand enrollment. At the same time, it is school and teach, but teachers need better important to encourage university students training so that children in the first three to enter high-return fields; high-income years get a firm grasp on literacy before tran- countries do so by offering financing incen- sitioning to English. Because these improve- tives (as in Argentina and Australia) and by ments take time, transitioning to English improving the quality of science and engi- later (for example, at the end of primary) neering instruction in earlier grades (as in may also be valuable, so that children do not Norway and Poland). transition to a new language until they have The tertiary sector is an ideal space to mastered basic skills in the first language. foster innovation. Publications and patents H u m an C a p ita l and I nnovation 51 in Rwanda have been rising, and Rwanda Why Is Human Capital Important has invested in a range of graduate and for Growth? postgraduate centers for technical train- ing. Creating incentives to develop and Investments in human capital have been adapt innovations that benefit industries linked closely to growth. A decade ago, a in Rwanda can help to reap the maximum high-level commission identified 13 econo- returns to local innovation. In high-income mies that had grown at an average rate of countries, private firms finance university 7 percent a year or more for 25 years or research to solve production challenges. In longer since 1950. They concluded, “Every Rwanda, given the small private sector, gov- country that sustained high growth for long ernment may need to continue to play a sup- periods put substantial effort into schooling porting role. its citizens and deepening its human capital” It will be impossible to expand human (Commission on Growth and Development capital dramatically without spending more. 2008). Human capital includes the knowl- Each human capital investment builds on edge and skills of the population and results previous investments. The returns from from investments in education and health. investments in tertiary education and skills Ultimately, human capital investment and training are realized almost immediately. economic growth form a virtuous cycle: However, for returns to expand, Rwanda greater human development increases eco- needs to increase the flow of well-educated nomic growth, and greater economic growth individuals from earlier levels and, ultimately, finances further human development. Cross- to produce a population with broad literacy country analysis demonstrates that coun- and numeracy. tries that invest in human capital early in Better information can improve the qual- the cycle enjoy the benefits of this virtuous ity of human capital investments. Rwanda cycle, whereas countries that experience high is an innovator in holding service providers initial economic growth without investing accountable, with performance contracts in human capital virtually always fall into a for educators and health workers, along vicious cycle of low human capital followed with innovations in performance-based pay by a slowdown of economic growth (Ranis, (Basinga et al. 2011; Zeitlin et al. 2017). Stewart, and Ramirez 2000). In many countries, service provision has Human capital is essential because it influ- improved because of the effective involve- ences growth directly and increases total fac- ment of parents, pointing to the importance tor productivity. In the long-run history of of dissemination of information on learn- global economic growth, education leads to ing outcomes and other elements of school greater technological development, and tech- performance. nology tends to complement skills, thus rais- This chapter analyzes each element ing the returns to education (Commission on of human capital in turn—health (with Growth and Development 2008). Current a focus on stunting and fertility), basic cross-country estimates of the drivers of eco- education, skills training, and tertiary nomic growth demonstrate that human capi- e ducation—with an analysis of the cur- ­ tal leads to higher total factor productivity, rent state, key challenges, and potential which in turn boosts economic growth (Ranis, investments to help Rwanda to meet its Stewart, and Ramirez 2000). Regardless of ambitions. This discussion is informed by what other natural resources Rwanda may lessons from the experiences of countries identify, human capital investments are crucial that have achieved rapid gains across these to a sustained, high economic growth strategy. areas and begins by proposing a frame- To achieve growth via services or manu- work for thinking about human capital facturing or innovation, Rwanda needs investments in Rwanda. human capital investments. Economic growth 52 FUTURE DRIVERS OF GROWTH IN RWANDA through services comes not from low-skill Korea’s population was at that level. A decade services, such as petty trading, but from ser- into its high growth, 85 percent of Korea’s vices that are higher on the supply chain, such working-age population had at least a pri- as processing goods for sale in the market. mary education and 4 percent had completed To organize and support such an industry at tertiary. Current projections for Rwanda scale requires workers with skill. Economic 10 years from now are 72 ­ percent with at least growth through manufacturing can employ primary and just 1.2 percent with completed low-skill workers, certainly, but a sizable tertiary. Furthermore, Korea experienced cohort of high-skill managers and technicians growth in an era when average human capi- is essential. Higher-skill workers are capable tal was much lower globally, so the returns to of carrying out higher-skill tasks, which add even low levels of human capital were higher. greater economic value. Economic growth In recent years, human capital investments through innovation requires a significant have expanded globally, which means that cohort of workers with strong training in higher and better investments will be needed their field, usually with tertiary or postgradu- to sustain high rates of economic growth. ate education. For a highly skilled cohort of tertiary graduates, Rwanda needs to have A Framework for Thinking about a broad base of secondary school graduates with good skills. Human Capital Korea offers an example of the link Investments in human capital have a strong between human capital and high, sustained link to economic growth. Investments in economic growth. When Korea was at the human capital continue throughout the life same level of per capita income as Rwanda cycle and include investments in early child- today, in 1967, it had much higher rates hood development, basic education, higher of skill in the workforce than Rwanda education, skills formation as adults, health (­ f igure 1.2). For ­ example, whereas about services, and innovation (figure 1.3). All of 59 percent of Rwanda’s population has at these investments are tied to growth over the least a primary education, 70 percent of long term. For example, Rwanda had a stunt- ing rate of 36.5 percent at last measure, and FIGURE 1.2  Rates of human capital in Rwanda and the Republic of economic estimates suggest that per capita Korea at similar income levels income for today’s workforce is 10 percent lower than it would be if no one in the work- 100 force had been stunted as a child (Galasso et al. 2016; World Food Programme 2016). The quality of learning has a clear, positive 80 association with economic growth, much more so than the enrollment rate (Hanushek 60 et al. 2008). While the relationship between Percent higher education and economic growth 40 depends on the quality of higher educa- tion, “One does not get electrical engineers 20 and computer scientists without investing in higher education,” as Hanushek (2016) puts it. On-the-job training likewise shows clear 0 1965 1970 1975 1980 2015 2020 2025 2030 economic returns (Saraf 2017). Health ser- Korea, Rep. Rwanda vices are essential to a productive workforce, No education or incomplete primary Completed secondary and innovation is key to job creation. Completed primary Completed tertiary Rwanda—along with other countries— Sources: Data from Barro and Lee 2015; Lee and Lee 2016. faces an array of needs with regard to human Note: In 1967, the Republic of Korea was at a similar level of per capita income as Rwanda today. capital investment. Investments tend to favor H u m an C a p ita l and I nnovation 53 FIGURE 1.3  The array of human capital investments Early childhood Basic education Higher education Other skills Nutrition Primary University On-the-job training Education Secondary Technical/polytechnic Adult education Health services Innovation today’s needs (persons in the workforce or FIGURE 1.4  The trade-off between lower returns now and higher returns later about to enter it) or tomorrow’s needs (per- sons early in their education). Investments in favor of today’s needs—such as invest- Highest Rate of return to investment returns ments in vocational training, innovation, and higher education—tend to deliver the quickest returns because the beneficiaries are about to enter the workforce. Investments in favor of tomorrow’s needs—reducing stunt- ing, improving early child development, and Quickest returns increasing the quality of primary education— tend to deliver the highest returns because Early years Basic education Higher education early investments increase the productivity Time to realize investment of later investments (Carneiro and Heckman 2003; figure 1.4). Children with good nutri- tion in the early years go on to learn more Second, it is necessary to identify the invest- effectively throughout their school career, ment that will be required in activities that and children who master reading in the early yield economic growth both now and later. grades are most equipped to learn profes- Considering these two criteria together, sional or technical skills later. Of course, Rwanda will continue to make investments Rwanda will continue to invest in each stage. in both the early and the later years (for eco- But investments yielding returns that are nomic growth both now and sustained over realized more quickly need to be balanced time) and to identify the binding constraints with investments in younger generations; in each of those areas. Complementary otherwise, higher education will forever be investments, such as safety nets, can help playing catch-up. households to make investments that will Government should consider two prin- contribute to inclusive growth (box 1.1). ciples in setting investment priorities to maximize the impact on growth. First, it is necessary to identify the “binding con- Stunting straints.” Which problems are so severe that, Current Status unless they are solved first, no amount of money or time spent solving the less severe Despite the launch of several successful ini- problems will help (van der Berg et al. 2016)? tiatives, representing strong political will 54 FUTURE DRIVERS OF GROWTH IN RWANDA BOX 1.1  What is the role of social safety nets for economic growth? Rwanda has invested heavily in social safety nets, have invested in productive assets such as livestock through the Vision 2020 Umurenge Program and (Ralston, Andrews, and Hsiao 2017). others. Social safety nets play two potential roles Second, social safety nets have the poten- in sustained economic growth. First, they increase tial to increase the rate of economic growth. In the inclusivity of growth. Not all economic growth Zambia, households that had received three years is equally effective at reducing poverty. Indeed, of transfers were spending 67 percent more than economic growth has reduced poverty less in Sub- the size of the transfer. This suggests a multiplier Saharan Africa than in other regions (Chuhan-Pole effect. Households are able to convert every dollar et al. 2013). They also protect the most vulnerable received in transfers into much more than a dol- households from extreme poverty and give them lar’s worth of income through both agricultural opportunities to make productive investments. production and off-farm activity (Handa et al. Cash transfers in Brazil and Mexico dramatically 2018). Furthermore, safety nets may contribute to reduced inequality, improving the quality of life for growth by increasing human capital investments the poorest households (Soares, Ribas, and Osório and helping households to manage risk (Alderman 2007). Across Africa, households receiving transfers and Yemtsov 2013). and civil society cooperation, stunting still The poorest 20 percent of the population has represents a major challenge in Rwanda.1 a stunting rate of 49 percent, whereas the At last measure, 36.5 percent of children richest 20 ­percent has a rate of 21 percent suffered from stunting. Although this is a figure 1.7). Between 2000 and 2015, stunt- (­ major improvement since 2005, when half ing rates for children in richer households of all children were stunted, it exceeds global declined 3.2 percent a year, whereas rates and regional rates of stunting. The global rate is 22.9 percent, while East Asia and FIGURE 1.5  Stunting rate in Rwanda versus Pacific, Latin America and the Caribbean, the world, 1995–2015 and Sub-Saharan Africa have achieved lower levels of stunting (figure 1.5). Rates are also 60 higher in Rwanda than in many countries with similar levels of income (figure 1.6). 50 Senegal has achieved significantly lower lev- els of stunting—18.7 percent—because of 40 major programs to tackle child undernutri- Percent tion. Furthermore, the elasticity of stunting 30 with respect to economic growth in Rwanda and elsewhere in Africa is less than half that 20 in the rest of the world, so future economic growth cannot be counted on to elimi- 10 nate stunting. Although Rwanda has made impressive progress on acute malnutrition— 0 wasting—chronic malnutrition, represented 1995 2000 2005 2010 2015 by stunting, has been slower to improve, and Rwanda stunting shows strong links with household Sub-Saharan Africa poverty and national economic growth. East Asia and Pacific Latin America and the Caribbean The average level of stunting masks variations related to socioeconomic and Source: World Development Indicators 2017 data (World Bank, ­ geographic characteristics across Rwanda. various years). H u m an C a p ita l and I nnovation 55 FIGURE 1.6  Prevalence of stunting and GDP per capita, 2017 FIGURE 1.7  Prevalence of stunting in Rwanda, by wealth quintile, 2000–15 70 70 60 60 Stunting prevalence among Prevalence of stunting (%) 50 children under ve (%) 50 40 Rwanda 40 30 30 20 20 10 10 0 2000 2005 2010 2014–15 0 10 20 30 40 50 Poorest quintile GDP per capita (US$, thousands) National average Richest quintile Source: World Development Indicators 2017 data (World Bank, various years). Source: Eozenou and Shekar 2017. for children in poorer households declined MAP 1.1  Prevalence of stunting in Rwanda, by district, 2014–15 only 1 percent a year (Eozenou and Shekar 2017). The western districts of Rwanda have a higher ­concentration of stunting (map 1.1), MUSANZE NYAGATARE and stunting levels are much higher among NYABIHU BURERA rural residents (40.6 ­ p ercent) than among GICUMBI GATSIBO urban residents (24 percent).2 Among children RUBAVU GAKENKE RULINDO GASABO under five, boys have a higher rate of stunting NGORORERO KAYONZA percent) than girls (33 percent). Children (41 ­ MUHANGA RUTSIRO RWAMAGANA KICUKIRO whose mothers have attained secondary KAMONYI KARONGI NYARUGENGE education or higher have stunting ­ levels of RUHANGO NGOMA BUGESERA KIREHE 20 percent, compared with 47 percent for NYAMASHEKE NYANZA children whose mothers have no education. NYAMAGABE Households struggle to satisfy the nutritional RUSIZI HUYE needs of their children as family size increases. GISAGARA NYARUGURU Among poorer populations, there is a posi- tive relationship between stunting rates and ≥40%: Very high 20–29%: Medium Stunting prevalence increasing birth order—approximately 40 per- by district (%) 30–39%: High <20%: Low cent of children in the first two birth orders District boundaries are stunted, and this rises to 47 percent of IBRD 42943 | MAY 2017 children in the third and 50 percent of those ­ Source: Eozenou and Shekar 2017. in the fourth birth order. Poorly nourished mothers have children with higher levels of health and shorter height, are at risk of stunting than well-nourished mothers. suffering noncommunicable diseases like This stunting has significant short- and hypertension, and have lower cognitive abil- long-term impacts on human capital and— ity and fewer socioemotional skills. This ultimately—on economic growth. Adults relationship translates into reduced produc- who were stunted in childhood have poorer tivity and lower wages, which ultimately 56 FUTURE DRIVERS OF GROWTH IN RWANDA affects economic growth. By one estimate, households drink untreated water (NISR, the per capita income for today’s workforce various years). Only one-third of two-year- could be 10 percent higher in Rwanda if olds have access to both improved water the adults had not been stunted as children sources and sanitation facilities (Eozenou and (Galasso et al. 2016). By another estimate, Shekar 2017). The lack of access to WASH stunting currently leads to a loss of US$820 facilities induces diarrheal diseases, inhib- million—11.5 ­ p ercent of Rwanda’s gross its children’s ability to absorb nutrients, and domestic product (USAID 2014). These leads to malnourishment and stunting. impacts spill over to the next generation, with Despite improvements over the past stunted women more likely to have babies 15 years, care practices for mothers and who are underweight and who have cognitive children remain inadequate. The first 1,000 challenges (Galasso et al. 2016). days of a child’s life are a critical period of intervention: inadequate care during this period leads to irreversible damage. Rwanda Proximate Causes has made great strides in increasing exclusive Lack of diversity in food intake and inad- breastfeeding rates until the six-month mark; equate frequency of meals before the age of in 2015, 80 percent of children were breast- two increase the incidence of malnutrition fed (USAID 2014). However, suboptimal and stunting. Inadequate diets are a result complementary feeding practices between 6 of limited knowledge about feeding prac- and 24 months of age hinder early childhood tices and food insecurity. Approximately development—between 6 and 9 months of 40 ­ p ercent of households in Rwanda are age, one in three infants is not fed the rec- food insecure. These households typically ommended mix of breast milk and solid food are rural, depend on low-income agricul- (USAID 2014). With respect to care prac- ture, and have few adult household mem- tices for mothers, only 42 percent of women bers. They often have high micronutrient receive the recommended four antenatal vis- ­ deficiencies: “45% of food insecure house- its (World Bank 2017a). Other care practices, holds had not consumed any protein-rich including facility delivery and postnatal care, food in the week prior to the Comprehensive also have increased, but coverage in rural Food Security and Vulnerability Analysis Rwanda remains low. Survey” (Hjelm 2016). As a result, only Across individual interventions (adequate 20 ­ p ercent of Rwandan ­ c hildren con- access to food, care, and WASH practices), sume iron-​ rich food and only 18 percent of improved care practices (specifically ante- children between 6 and 24 months meet the ­ natal visits and facility delivery) are most requirements for a ­minimum acceptable diet significant in explaining changes in ­ ­ stunting. (Hjelm 2016). Within these early months, the An intervention that addresses only care first 1,000 days are particularly crucial, with practices reduces the probability of being longer-term impacts (Shekar et al. 2016). stunted by 8 percent. Given that stunting There is a high incidence of deficient is a result of multiple interrelated factors, WASH (water supply, sanitation, and hygiene) these interventions are most effective when practices and poor environmental health, combined. For example, interventions that which influences stunting levels. Although improve both environmental health and care access to improved water and sanitation practices reduce the probability of stunting facilities has increased significantly since by 12.5 percent. 2005, children from poor and vulnerable households do not have sufficient ­ coverage. Investments for Transformation According to the 2014/15 Demographic and Health Survey, more than 91 percent of rural High rates of stunting are not inevitable, households do not have a handwashing sta- and many countries have made dramatic tion and approximately 60 percent of rural gains over time. Many countries that already H u m an C a p ita l and I nnovation 57 had stunting rates below 50 ­ p ercent have FIGURE 1.8  Reduction in stunting in countries with less than achieved significant reductions in stunt- 50 percent stunting ing (figure 1.8), including Peru, Senegal, and others. If Rwanda were to reduce Mauritania (1974–97) stunting in the coming years at the same Bangladesh (1984–2014) rates as quick-reducers have done, stunt- Cambodia (1987–2010) ing could be significantly lower in coming Bolivia (1996–2012) years (­figure 1.9). Peru reduced its stunting Guinea-Bissau (1996–2012) by 50 percent over the course of 10 years Honduras (1986–2010) through a program endorsed publicly by Peru (1986–2010) high-level political officials, including Egypt, Arab Rep. (1978–2003) micronutrient supplements (which Rwanda Vietnam (1994–2012) already includes in its program), a mas- Tajikistan (1998–2014) sive media campaign, a careful monitoring Equatorial Guinea (1999–2013) system, and a ­ performance-based incentive Tunisia (2000–10) program. Senegal likewise made impressive Swaziland (1981–2008) gains in a short time period, despite weak Mexico (2000–10) Haiti (1992–2013) economic growth, through high-level politi- Albania (1997–2014) cal commitment, nutrition financing, and a Uzbekistan (1996–2006) multisectoral strategy. China (1997–2009) How did Peru reduce its stunting rates? Senegal (2000–14) Peru’s strategy for reducing stunting had four São Tomé and Príncipe (2001–12) key components: Kyrgyz Republic (1997–2010) El Salvador (2000–14) 1. Performance-based incentive program. Mongolia (2008–14) The Ministry of Economy and Finance Colombia (2004–14) coordinated efforts by providing a mon- 50 45 40 35 30 25 20 15 10 5 0 etary incentive through its results-based budgeting program. Every agency’s budget Reduction in stunting by initial rate was dependent on improving stunting- Source: Based on data from World Development Indicators (World Bank, various years). related indicators, such as the number Note: The time period next to the country name indicates the period over which the of child-growth checkups and nutrition stunting reduction was calculated. The bars indicate the level of stunting at the beginning (left) and end (right) of the period. counseling sessions. 2. Media campaigns. Across Peru, there was How did Senegal reduce stunting rates? a drive to deepen parents’ understanding of malnutrition through various outlets, • Nutrition financing. The government tied including radio and television. The dis- financing to nutrition-based outcomes to semination of information made the prob- improve the demand for and supply of lem more visible, which increased the health and nutrition services. demand for health and nutrition services • Community-based multisectoral strategy. among vulnerable populations. The Prime Minister’s Office coordinated 3. Monitoring systems. The government efforts between local governments and implemented regular data collection and civil society organizations to deliver health monitoring systems in order to focus on and nutrition services, such as health edu- results and to ensure accountability. cation and breastfeeding promotion, to 4. Multisectoral effort. The government inte- communities. grated nutrition into social protection pol- icies. The supply of health and nutritional Reducing Rwanda’s stunting rate will services was tied to Juntos, a nationwide require multiple interventions, but incen- conditional cash transfer program. tives are key. Stunting results from poor child 58 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE 1.9  Projected reduction in stunting in Rwanda, 2000–35 The Economic and Poverty Reduction Strategy for 2013 –18 and the National Projected prevalence of stunting, height for age Strategy for Transformation (NST) 2017– 60 2024 prioritized nutrition, food security, and early childhood development. There are dis- 50 trict plans to eliminate malnutrition for all 30 districts, including several initiatives at the % of children under 5 40 local and central levels. Monitoring and eval- uation systems measure progress on nutri- 30 tion-specific performance indicators. These monitoring systems include baseline and end- 20 line surveys to understand important nutri- Upper-middle tion behaviors, practices, and coverage levels 10 income (2015) and to assess the quality of service delivery. High income (2015) In addition, information systems are used to 0 monitor the health of vulnerable women and 2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050 children. However, key barriers are hinder- Rwanda (2015): 38% Peru (1986–2010): 44% to 15% ing the impact of these initiatives. Despite the Uzbekistan (1996–2006): 39% to 20% development of an overarching national strat- China (1997–2009): 38% to 9% egy, a lack of consistency in communicating Mexico (2000–10): 40% to 14% messages around childhood stunting has led Mongolia (2008–14): 30% to 11% to inefficiencies. Often, stunting prevention is not a priority in donor-funded activities Source: Based on data from World Development Indicators (World Bank, various years). and is not aligned with WASH intervention Note: Projections are based on actual rates of stunting reduction in other countries. The legend shows the period over which the reduction in stunting was calculated, as well as strategies. Hence a key action is to fast-track the level of stunting at the beginning and end of the period. All comparator countries implementation of the National Social and began their period of rapid reduction at a level comparable to that of Rwanda. Behavior Change Communication Strategy for Integrated Early Childhood Development, nutrition, but solving it requires a range of Nutrition, and WASH (2018–2024). investments: food security, access to clean More human and financial resources water and good sanitation, access to micro- should be devoted to preventing stunting. nutrient supplementation, and awareness of Several existing interventions, such as ante- good nutritional practices. Many programs natal care, micronutrient supplementation to address these issues are in place, and the for pregnant women, malaria treatment in NECDP has been established to address issues pregnancy, zinc supplementation, and the of coordination and inconsistent financing. public provision of complementary food, To reap the gains in long-term growth from could be scaled up across Rwanda in order having well-nourished children, Rwanda will to improve coverage. Sustainability and need to make a big push across areas and to accountability mechanisms are key aspects of monitor carefully which policies deliver the effectively scaling up an intervention. Service greatest gains, Without regular measurement, delivery should build the capacity of local it is impossible to gauge progress. One key governments and civil society. Nutrition- factor in Kirehe District, which reduced stunt- specific indicators tied to performance con- ing in the youngest children to 7 ­ percent in tracts at all levels of implementation can serve 2015, was incorporating nutrition indicators in as an effective accountability mechanism. performance contracts at all levels, including Finally, interventions should target the the household (World Bank 2017a). first 1,000 days of a child’s life. Because Rwanda already has a strategy for reduc- the first 1,000 days are crucial in deter- ing stunting, but greater effort is required. mining a child’s developmental trajectory, H u m an C a p ita l and I nnovation 59 interventions should target the demand for birth rates will allow for a favorable demo- and supply of services during this critical graphic transition, with more working-age period. Specifically, interventions should tar- adults and less youth dependency. Concurrent get complementary feeding practices between investments in economic and social spheres 6 and 24 months of age. Food fortification are crucial, because a large working-age pop- and the promotion of diversified, nutrient- ulation with no access to employment could rich food for both mothers and children dras- become a source of social risk. tically reduce the incidence of stunting. Investments for Transformation Fertility Several policy options are available for accel- erating a decline in fertility rates. Current Status Between 2005 and 2010, Rwanda achieved • Improve adoption of, access to, and effec- one of the fastest reductions in fertility tiveness of family planning programs. rates—a drop from 6.1 births per woman This could include increasing the use of to 4.6—in Sub-Saharan Africa. 3 This nota- contraceptives among underserved popu- ble decline was a result of several popula- lations (teenagers and women in rural tion reforms, particularly the new National areas), scaling up community-based dis- Population Policy of 2003, which addressed tribution of family planning services, and high fertility and mortality rates as well as retraining health care providers on perma- maternal and early childhood morbidity. The nent methods of contraception (intrauter- provision of family planning services is likely ine devices, implants, and sterilization). to have had the most significant impact on Examples include Bangladesh, India, and lowering fertility rates, primarily by increas- Sri Lanka, where female and male ster- ing the use of contraceptives among mar- ilization was incentivized to increase the ried women, from 17 percent in 2005 to adoption of family planning services in 52 ­percent in 2010 (Westoff 2013). In addi- the 1970s (Silva de Tiloka 2017). tion, influential mass media campaigns by • Education programs. The Comprehensive the government increased contraceptive use Sexuality Education Program could be and decreased family sizes. A substantial implemented in schools, and students downward trend in under-five child mortal- could be given improved access to youth- ity, from 196 per 1,000 births in 2000 to friendly facilities that specialize in sexual 38.5 in 2015, may have influenced fertility and reproductive health services. Effective as parents saw a rise in their children’s sur- education campaigns that reduce teenage vival rates.4 These reforms, coupled with an pregnancies and early marriage could improving socioeconomic environment, led be scaled up. Rwanda could continue its to changes in reproductive behavior and a efforts to encourage small family norms. subsequent decline in birth rates. Brazil took a unique approach to educat- However, fertility rates are now declining ing its citizens about reproductive health slowly, similar to pre-2005 levels. Between and family planning by including relevant 2010 and 2015, fertility rates fell from 4.6 messages in soap operas (Silva de Tiloka to 4.2 births per woman. Rwanda needs to 2017). accelerate this decline in order to decrease the • Address neonatal mortality rates (death pressures of a high youth dependency ratio in the first month of life). Health care on the country. As of 2015, 40 percent of the facilities could be better equipped to population was under the age of 15, limiting address postpartum health complica- the ability to save and invest for the future tions. In addition, women could have (UNFPA 2017). Achieving a steep decline in greater access to antenatal, postnatal, and 60 FUTURE DRIVERS OF GROWTH IN RWANDA delivery care. In order to address high secondary school age—is slightly lower, at neonatal mortality rates in Vietnam, all about 27 percent, because it does not capture children under the age of six from poor overage students due to grade repetition in households were eligible for free health earlier years. care through the Health Care Fund Although access has improved over time, for the Poor (WHO, World Bank, and Rwanda still lags behind the average of coun- Ministry of Health, Vietnam 2015). tries of similar income. One measure of access is the primary or secondary completion rate. An accelerated decline in fertility can This measures not only whether youths are positively affect economic growth if Rwanda enrolling in school but also whether they are harnesses the demographic dividend. As completing levels of schooling. Currently, the working-age population grows and Rwanda’s primary completion rate is about youth dependency ratios fall, incomes rise. 61 percent, as opposed to about 66 percent in The decline in youth dependency ratios can low-income countries as a whole. Likewise, increase investments in the health and edu- completion rates for lower-secondary school cation of every child, creating a more pro- are at 34 percent in Rwanda versus 37 per- ductive workforce in the future. Evidence cent in low-income countries as a whole. shows that lower fertility can improve labor Rwanda will not achieve upper-middle- force participation rates and education lev- income status without a dramatic increase els, particularly for women. It can also affect in school completion. Even the bottom 25th national savings rates and investment (Karra, percentile of upper-middle-income countries Canning, and Wilde 2017). In order to har- have primary completion rates of 94 percent, ness the benefits of this favorable age struc- about 50 percent higher than Rwanda’s cur- ture, Rwanda will simultaneously have to rent rate. The median primary completion develop a healthy and skilled workforce, cre- rate in upper-middle-income countries is ate quality jobs in sectors other than agricul- nearly 100 percent (figure 1.10). Likewise, the ture, and strengthen governance systems by median lower-secondary completion rate for improving efficiency and accountability. upper-middle-income countries is 87 ­ percent, more than 2.5 times Rwanda’s current rate. The disparity is even greater for upper-­ Basic Education secondary completion. Expanding basic education, together with ensuring quality, is Current Status essential for Rwanda’s sustained growth. Rwanda faces the same two challenges as Even among students in school, the qual- many other countries: access and quality of ity of education in Rwanda needs to be education. Rwanda has made great strides raised. The first recommendation coming out in ensuring access to education. Net pri- of the 2017 Joint Review of the Education mary school enrollment—the proportion Sector was to emphasize “quality of educa- of children of primary school age who are tion at all levels, particularly in lower grades actually in primary school—exceeds 95 per- for Kinyarwanda, English, and mathemat- cent. Secondary school enrollment, although ics” (Ministry of Education, Rwanda 2017). much lower, has also been rising. Gross sec- Why so? Recent measures indicate that stu- ondary ­enrollment—the number of children dents in P1 (first grade) can read 7.7 words in secondary school divided by the number per minute in Kinyarwanda, and students in of children of secondary school age—is 37 P2 can read 25 words per minute. Although percent, more than three times as high as the number of words that children need to be it was in 2000. This is great progress. Net able to read in order to understand what they secondary ­ enrollment—the number of chil- are reading—a reasonable definition of liter- dren of secondary school age in secondary acy—varies from language to language, this school divided by the number of children of level is well below literacy in any measured H u m an C a p ita l and I nnovation 61 language. In P1, 60 percent of students could FIGURE 1.10  School completion rates in primary and secondary not answer a single reading comprehension school in Rwanda compared with countries at different levels of income, 2017 question, and more than 85 percent were graded as reading “below comprehension,” meaning that they could not answer four out a. Primary completion rate of five comprehension questions correctly.5 At 100 the end of P3 (third grade), the final year of 90 instruction in Kinyarwanda, the percentage 80 of students scoring zero on the comprehen- Percent sion test is better: just 17 percent or one in six 70 students. However, 85 percent were still read- Rwanda 60 ing “below comprehension,” and 50 percent 50 were still reading between 1 and 32 words per minute. In P4 (fourth grade), the official 40 Low income Lower-middle Upper-middle High income language of instruction switches to English. income income At the end of P4, reading comprehension in Kinyarwanda (which is still taught as a school b. Lower-secondary completion rate 100 subject) rises to 40 percent—still well under 90 half—but nearly 50 percent of learners can- 80 not answer a single comprehension question 70 in English, the language of instruction for all Percent 60 of their subjects over the course of the year.6 50 40 75th 50th Rwanda Proximate Causes 30 25th 20 Virtually all children in Rwanda begin pri- Low income Lower-middle Upper-middle High income mary school, but fewer than two-thirds of income income them complete it. Some of the reasons include c. Upper-secondary completion rate out-of-pocket costs and paying for uniforms 100 and some materials. The elimination of all 90 school fees is key to increasing school par- 80 ticipation and advancement. This elimination 70 60 comes at a cost, but in its absence, universal Percent 50 primary completion will remain elusive, and 40 the poorest children will remain those least 30 likely to complete. 20 Rwanda Children also face opportunity costs: by 10 attending school, they sacrifice whatever else 0 Low income Lower-middle Upper-middle High income they might be doing with their time. Part of income income the failure to advance in school can come 75th percentile 50th percentile 25th percentile from mistaken perceptions about the returns to schooling. In the Dominican Republic and Source: Based on World Development Indicators 2017 data (World Bank, various years). Madagascar (although not in rural China), communicating accurate data on the future access to data on wages may be poorly situated wages of students who complete various levels to evaluate the returns to schooling, in which of schooling significantly increased school par- case they may underestimate the benefits rela- ticipation (Jensen 2010; Loyalka et al. 2013; tive to the costs. Rwandan data from 2010 Nguyen 2008). In Mexico, a similar program suggest that, for wage employment, the return even increased student test scores (Avitabile to an additional year of schooling is equal to and de Hoyos 2015). Parents with limited 22 percent in Rwanda; in fact, Rwanda has 62 FUTURE DRIVERS OF GROWTH IN RWANDA the highest measured returns to education of figure 1.11). The bulge results in exception- (­ any country in the world (Montenegro and ally large class sizes: classes in P1–P3 have Patrinos 2014). Likewise, Rwandans with more than 50 pupils each, with 64 pupils in more education, regardless of whether or not P1 classes. One estimate suggests that elimi- they have wage work, have more assets.7 nating the bulge in P1 would “save about High repetition rates in the early grades 5 percent of the primary education budget,” increase the opportunity cost of complet- which would allow for more investments ing primary and secondary education. In in other parts of the system (Crouch and a recent study, 14 percent of learners in Merseth 2017). Although children in P1 may P1–P4 were repeaters. In P1, that share not have much economic value to the house- was 23 ­ p ercent, so nearly one-quarter of hold, as children grow older without pro- first-grade students were repeating (USAID gressing or completing primary school, their 2017). This—along with other factors, such economic value from participating in child- as limited early childhood education cov- care or agricultural activities increases. erage and late school starters—results in The quality of instruction in these first a large “bulge” of students in P1, lowering years of school has major implications for the likelihood of primary school completion all future educational investments. If stu- and making improvements in quality difficult dents do not learn to read in the first few FIGURE 1.11  Early-grade “bulge index” in Rwanda and select countries 3 Cabo Verde São Tomé and Príncipe 2 Mauritius Côte d'Ivoire Mali Tanzania Burkina Faso 1 Guinea South Africa Ghana Niger Senegal Gambia, The Zimbabwe Swaziland 0 Namibia Botswana Early-grade "bulge index" Comoros Equatorial Guinea Congo, Rep. –1 Lesotho Central African Republic Cameroon Congo, Dem. Rep. –2 Angola Benin Chad Togo Sierra Leone Ethiopia –3 Burundi Guinea-Bissau Uganda –4 Madagascar –5 Rwanda –6 Source: Bashir et al. 2018. Note: Low values in Rwanda reflect high enrollment in P1 (first grade) relative to the age group, low progression to P2 (second grade) (that is, many repeaters), and low preprimary coverage. H u m an C a p ita l and I nnovation 63 years of school, they cannot effectively pupils meet quality standards and advance learn mathematics or science or social sci- appropriately. ence. Furthermore, the use of mother-tongue Once in school, why do children fail to instruction in the first three years of primary learn? Not all schooling is created equal. school heightens the importance of mak- Across school systems, students learn dramat- ing sure that children master reading in the ically different amounts of knowledge. For early years of primary school. Students who example, a recent study compared the aver- are taught in their mother tongue in the age years of schooling and level of knowledge early years learn to read faster than those of young adults (figure 1.12). (Unfortunately, who are taught in an unfamiliar language. because Rwanda has not participated in Learning to read becomes more difficult for international exams, it is not included in this children who enter school after the language analysis.) Young adults between the ages of of instruction has transitioned to English, 25 and 29 in Singapore, for example, have which is one reason research has suggested completed nearly 15 years of schooling. that a later transition to English would allow A young adult of the same age in the United students to first master literacy in a familiar States would have completed a similar num- language (Bashir et al. 2018). ber of years of schooling but would have Some have argued that eliminating auto- the cognitive ability of a Singaporean young matic promotion will improve quality. adult who had completed less than 12 years However, repetition is exactly what drives the of schooling. In other words, the quantity bulge in early grades. Rather than repealing a of education is similar across the two coun- policy of automatic promotion, the solution is tries, but the quality of education differs dra- to improve the quality of instruction so that matically. In South Africa, young adults have FIGURE 1.12  Gap between actual and learning-adjusted years of schooling in select countries and economies Number of years of schooling (population ages 25–29) 15 12 9 6 3 0 Isr d Bo ahr ia Ka Jor an Ku ep. Ja ly tsw ain Ca ana Ko t, A C a SA b R e n, H Chin . Isl un a Ire ep. Ko hs n a, n hu it Ze co Sa er tar ic ry Ita l ala a M ia Ne Mo alta No land Fe Qa ay Sin Ara n ga bia So ov ore h A ia Sw frica d A haila n Un rab Tur nd d ir y Un Kin ates d S om s R, ep ae te ite Em ke d M ani ng ra hil lan za da re ta i io T ede Lit wa B ral ys ut en am ga rw w roc p na ta R R ud at p ite gd st a k Au l d S ian ng gyp Ira ss Ho E ite Ru Un Actual years Learning-adjusted years Sources: World Bank 2018. Note: Years of schooling in Singapore are the same as learning-adjusted years because Singapore, which scored highest on the Trends in International Mathematics and Science Study (TIMSS) mathematics assessment in 2015, serves as the basis for comparison in this illustration. For the purposes of this illustration, data for years of education in the United Kingdom are adjusted using the TIMSS score for England. Note that for all countries and economies, the size of the adjustment will reflect the scale of the metric used to make it. 64 FUTURE DRIVERS OF GROWTH IN RWANDA completed roughly 11 years of schooling, review of interventions intended to improve but their level of knowledge is equivalent to learning outcomes found the highest, most just 6 years of Singaporean education (World consistently large effects for interventions Bank 2018). Furthermore, economic growth that sought to improve the quality of teach- analysis demonstrates that what matters for ing, much more than efforts to improve economic growth is not the years of school- school infrastructure, materials, or manage- ing completed, but rather the amount of ment (Snilstveit et al. 2016). Likewise, stud- learning (Hanushek and Woessmann 2012). ies from countries as diverse as Ecuador and The World Bank’s World Development Pakistan have demonstrated that the differ- Report 2018 on education highlights the ence between a highly effective teacher and proximate determinants of learning (World a less effective teacher tends to equal at least Bank 2018). These determinants include the amount of learning that students nor- having prepared and motivated learners mally gain over the course of a regular school (or students), skilled and motivated teach- year and sometimes much more (Popova ers, school inputs (such as infrastructure and et al. 2017). materials, including technology) that comple- The World Bank’s regional study of basic ment the interactions between teachers and education, Facing Forward: Schooling with learners, and effective school management Learning in Africa, identifies Rwanda- (figure 1.13). A significant element of hav- specific priorities for policy makers to ing prepared, motivated learners is ensur- consider (Bashir et al. 2018). Rwanda has ing nutrition at early ages and eliminating invested heavily in basic education, prepared stunting, as discussed earlier. After the stu- good policies and plans—often informed dents themselves, teachers are central to the by evidence and experience from other learning experience. A recent, comprehensive c ountries—and aligned donor financing ­ accordingly. Yet lagging completion rates FIGURE 1.13  Key proximate determinants of learning and poor learning outcomes suggest that, although the country’s overarching policies as presented in sector strategies and plans are generally sound, more attention needs to be focused on bridging the gap between policy and implementation (box 1.2). A focus on rs Le implementation means paying more atten- he ar tion to schools and to teachers and admin- c ne a Te istrators, the main personnel involved in the rs delivery of education services. Systematic reinforcement and oversight of these frontline agents and the provision of adequate learn- LEARNING ing conditions in schools are thus critical. Rwanda can learn from other countries in this regard as well as from its own experience Sch with policy implementation. ts Rwandan teachers seem to put forth oo pu lm relatively high effort. Systematic data on in na ol a ge o teacher absenteeism and punctuality are me S ch nt not available, but certain statistics suggest that Rwanda is ahead of other countries in the region in this regard. In a survey from 2008, teacher absenteeism was reported Source: World Bank 2018. at under 10 percent in government and H u m an C a p ita l and I nnovation 65 BOX 1.2  Confronting a learning crisis in basic education: Key areas for action Rwanda is not alone among African countries in con- basic education with quality; (2) ensuring effective fronting a learning crisis in basic education. Bashir management and support of teachers; (3) increasing et al. (2018) identify four top areas for action by financing of education and focusing spending and ­ countries—like Rwanda—that share this challenge: budget processes on improving quality; and (4) clos- (1) completing the unfinished agenda of universalizing ing the institutional capacity gap. government-subsidized primary schools the government of Rwanda has invested and under 14 percent in secondary schools extensively in teacher language training, (Bennell and Ntagaramba 2008). 8 More with initial intensive efforts and subse- than 90 percent of respondents say that quently a shift to more in-school mentoring. teachers come to school on time. By contrast, No doubt these efforts are largely respon- recent surveys across seven African coun- sible for the fact that many teachers have tries found average teacher absenteeism rates reached an intermediate level. However, of 23 percent, reaching as high as 45 percent continued intensive efforts will be necessary in Mozambique (Bold et al. 2017). In those to bring teachers the rest of the way to flu- same countries, teacher absenteeism from ency in the language of instruction. class was even higher: on average, teachers Is teacher pay a binding constraint? were not in class teaching during 44 percent Teacher pay in Rwanda falls in varying places of class time, reaching as high as 57 percent across the spectrum of low-income African in Uganda. Although comparable numbers countries (figure 1.14). Evidence from are not available for Rwanda, the fact that Indonesia suggests that average teacher sala- tardiness is not regarded as a broad chal- ries do not have a major impact on day-to-day lenge suggests that teacher effort in class teacher effort (De Ree et al. 2018). However, may be higher. This relatively high effort low salaries may have an impact on the qual- means that, rather than needing to solve a ity of individuals who apply to become teach- wide array of challenges regarding teacher ers, and major inequality between primary effort, as many countries must, Rwanda can and secondary teachers can create strange focus on improving teacher skills. incentives, such as encouraging all of the Teachers do not have the skills they need. more qualified candidates to become second- Many teachers have limited command of ary school teachers. Recent moves to incor- the language of instruction. In one recent porate bonuses based on performance may be assessment (in 2014), only 43 percent of useful, although current teacher performance teachers were assessed at the “intermedi- contracts (imihigo) are not well defined with ate level” in English (Upper Quartile and regard to learning outcomes. Institute of Policy Analysis and Research Rwanda 2015). A lack of proficiency in the Investments for Transformation language of instruction will inevitably affect student performance not only in English Improving basic education is fundamental reading, writing, and speaking ability but to achieving sustained economic growth, also in all other subjects that are taught and large gains for new cohorts are pos- in English, including math, science, and sible. President Paul Kagame recently high- social studies. With the shift of language of lighted the need for a “revolution” in the instruction from French to English in 2008, quality of Rwandan education “at all levels” 66 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE 1.14  Lag in teacher salaries in Rwanda, by primary and secondary levels Central African Republic Average Tanzania Upper secondary Rwanda Sudan Burundi Uganda Sudan Average Central African Republic Rwanda Tanzania Lower secondary Burundi Uganda Central African Republic Uganda Tanzania Average Primary Rwanda Sudan Burundi 1,500 2,500 3,500 4,500 5,500 6,500 7,500 8,500 9,500 Average teacher salary (US$ PPP) Source: Ministry of Education and Sports, Uganda 2014. Note: PPP = purchasing power parity. (Tashobya 2017). Korea and Singapore Achieving universal access to basic educa- prioritized investments in basic education tion will require resources, but it can pay as the foundation of long-term economic off, as was the experience in the Republic of growth. In Korea, “Considerable capital Korea (box 1.3). The budgetary implications accumulation and investment in primary can be reduced, if necessary, by targeting education during [the 1950s] allowed a this policy to areas with the lowest comple- gradual shift up the value-added chain tion rates. toward more sophisticated commodities.” It is impossible to know whether per- Korea staged its educational investments: formance is improving without regularly “The focus of the government’s educational measuring it. Regularly assessing students plan has moved from primary to secondary and then sharing that information can play education and finally to the tertiary level, a transformative role. First, having more according to the nation’s economic advance- information about school performance can ment” (Pillay 2010). Likewise, in the 1960s, create incentives for all participants in the Singapore expanded education; but, in education system to focus on learning. If the late 1970s and 1980s, it made major citizens have access to regular data on stu- investments in quality, permitting a shift dent learning outcomes, they are more likely into a broad-based knowledge economy to demand improved learning outcomes (Rodríguez, Dahlman, and Salmi 2008). from schools. In many communities, parents If Rwanda desires long-term growth, then themselves have limited literacy, so if they quality education in the early school years is do not receive credible information, they crucial. Furthermore, many countries have may find it difficult to evaluate the quality of shown that large gains in primary school education. In Brazil, the government sets tar- completion are possible (figure 1.15). gets that combine ­ student passing rates with The way to expand access is to reduce costs student test scores. Each year, school and and improve perceived benefits. Countries state performance on these targets is widely that have achieved widespread basic lit- publicized, and communities hold their local eracy have offered free primary education. government accountable. Second, regular H u m an C a p ita l and I nnovation 67 FIGURE 1.15  Current and projected primary and lower-secondary completion in Rwanda, 2000–35 a. Projected primary completion rate b. Projected lower-secondary completion rate High income (2015) High income 100 100 (2015) 90 Upper-middle 90 % of relevant age group income (2015) 80 % of relevant age group Upper-middle 80 income (2015) 70 70 60 60 50 40 50 30 40 20 30 2000 2010 2020 2030 2040 2050 2000 2010 2020 2030 2040 2050 Rwanda (2015): 61% Rwanda (2015): 34% Cambodia (2002–15): 60% to 95% Turkey (1990–2010): 35% to 102% India (1999–2014): 69% to 98% Botswana (1986–2006): 36% to 88% Ghana (1992–2012): 68% to 98% Nepal (1992–2015): 37% to 84% Colombia (1982–2002): 69% to 94% El Salvador (1991–2011): 45% to 85% Peru (1973–93): 65% to 87% Indonesia (1984–2004): 36% to 74% Source: Based on World Development Indicators 2017 data (World Bank, various years). Note: Projections are based on the rate of increase in other countries. The legend shows the period over which the primary or lower-secondary completion expansion was calculated, as well as the level of primary or lower-secondary completion at the beginning and end of the period. All comparator countries began their period of rapid expansion at a level comparable to that of Rwanda. BOX 1.3  Increasing access to education and training in the Republic of Korea Since the 1950s in Korea, gross enrollment rates •• Engaging the private sector. The private sector have risen rapidly across elementary, secondary, and played an active role in secondary and higher edu- tertiary schools. Elementary school enrollment rates cation by setting up educational institutions and have been high (approximately 100 percent) since donating resources, allowing the government to the early 1950s. In 2010, approximately 80 percent focus on strengthening elementary education. of the workforce had a high school or college degree. •• Providing effective financing. Under the Edu- Since 2006, 15-year-old students in Korea have per- cational Grant Act, the government linked a formed consistently better than the Organisation for percentage of the internal tax revenue to the Economic Co-operation and Development average education budget and channeled these resources in mathematics, science, and reading. Their learn- toward elementary and secondary schooling. As ing outcomes are comparable with levels in Finland; of 2010, 20.27 percent of the internal taxes went Hong Kong SAR, China; and Japan. to the education budget. The government’s invest- How has Korea achieved high gross enrollment ment in education, as a percentage of GDP, has rates? remained above 4 percent since 2005. •• Providing vocational training. The government •• Ensuring access to education. The Korean govern- addressed the skills gap in the workforce by ment enforced a Six-Year Compulsory Education investing in vocational education. The c ­ountry Expansion Plan from 1954 to 1959 and allocated 80 made a concerted effort to build respect for percent of its education budget toward ensuring com- technical professionals. In 1973, it implemented pulsory education for students in elementary school. the ­ Initiatives for Technical High Specialization ­ (Box continues next page) 68 FUTURE DRIVERS OF GROWTH IN RWANDA BOX 1.3  (continued) Schools, which specialize in one of four themes: •• Addressing brain drain. The government built mechanical, experimental, specialized, and gener- research institutions to address the problem of alized. The government provided scholarships and brain drain and to foster talent within Korea. stipends to encourage enrollment in these schools. For example, American professors served as con- Currently, under the Vocational Training Special sultants and resided in Korea, in order to help Measures Act, private companies must provide to transform Seoul National University into a mandatory training for their ­ employees. The pre- renowned institution. cursor to this act were government ­ subsidies to companies for employee training. Source: Hong and Lee 2016. information on performance can help A key example of high-quality teacher schools themselves to improve learning per- training is ongoing coaching rather than formance and the effectiveness of interven- one-time training opportunities. “Teacher tions. In Mexico, providing information coaching has emerged as a promising alter- on school performance to both schools and native to traditional models of professional communities helped head teachers and class development” (Kraft, Blazar, and Hogan teachers to understand whether innova- 2017). These programs have been imple- tions in the education system were having mented in high-income countries like the the desired effect. Regular ­i nformation on United States to good effect, with large learning helped a scholarship program in impacts on the quality of instruction and Cambodia to reach—over time—the most sizable impacts on student achievement. vulnerable students. In Singapore, one of the pillars of high Upgrading the skills of primary teach- school performance is sustained teacher ers is one crucial step (Bashir et al. 2018). professional development (box 1.4). More From the more than 100 recent studies on recently, these coaching approaches have improving learning in high-, middle-, and been implemented successfully in South low-income countries, one key finding is Africa, with large impacts on student learn- that certain types of teacher training can ing, more than double the size associated have a strong impact on the quality of edu- with more traditional teacher professional cation (Evans and Popova 2016). But this development (Cilliers and Taylor 2017). In teacher training is not about general teacher Kenya, the government used existing school training, which commonly takes place in inspectors for this purpose (Gove, Poole, one-day or one-week seminars. Rather, and Pipe 2017). The way to improve teach- teacher training tends to be effective when ers’ classroom practice is through improved it is sustained over time and reaches the supervision in their schools. Another way is teacher at her or his level of need. Good through improved training of school prin- teacher development has significant con- cipals. In the United States, a recent inter- tact hours and focuses on concrete tasks, vention trained school principals to observe including mentoring for beginners and peer and provide constructive feedback to teach- learning (Popova, Evans, and Arancibia ers, resulting in improved student learning 2016). In countries like Singapore, high-­ (Fryer 2017). quality teachers are part of the recipe for It may be tempting to think that Rwanda success in quality education (box 1.4). can import high-quality teachers from H u m an C a p ita l and I nnovation 69 BOX 1.4  Achieving world-class education in Singapore According to the Global Competitiveness Report, •• Collaborative system. Key stakeholders in the edu- 2007–08, Singapore ranks first when it comes to qual- cation system—policy implementers, researchers, ity education (World Economic Forum 2008). The and educators—work closely together. The National country has consistently performed well in tests that Institute of Education, which conducts research and evaluate education systems around the world such as teacher training, works closely with the Ministry the Trends in Mathematics and Science Study (TIMSS), of Education. Officials from the ministry also meet the Progress in International Reading Literacy Study regularly with school principals in order to ensure (PIRLS), and the Programme for International Student that their policies are rooted in context. Assessment (PISA). Since 1997, through its vision •• Links to economic development. As a nation of “Thinking Schools, Learning Nation,” Singapore without natural resources, Singapore consistently has designed its education system to prepare students focuses on leveraging its human capital. Over the for the opportunities of the 21st century. last 40 years, it developed an education system How has Singapore improved its quality of that adapted to the changing needs of its economy. education? ­ Currently, the focus is on innovation, creativity, and research to contribute to the global knowl- •• Teacher recruitment, development, and ­ retention. edge economy. In 2004, “Teach Less, Learn More” Teaching is a highly respected profession in was introduced to shift students from rote learning Singapore—only the top third of high school ­ toward learning through experimentation and dis- graduates are recruited as teachers. Singapore covery, connecting topics to real life and character promotes teacher retention through a variety building. Singapore also has a unique system for of monetary incentives—for example, a gener- identifying labor force needs and projecting future ous bonus scheme based on an appraisal system demand. These skills are then developed in students increases salaries over the course of a teacher’s through higher and postsecondary education. career. There is also a strong focus on professional •• Priority given to mathematics, science, and technical development among teachers. In addition to mak- skills. Mathematics and science are core subjects at ing 100 hours of professional development avail- the primary and secondary levels, often taught by able to every teacher, the government reimburses specialist teachers. Although children are taught far teachers up to a certain amount to improve their less material than students in other countries, teach- skills and knowledge. This reimbursement allows ers cover these subjects in greater depth. The “Model teachers to join courses, subscribe to magazines, Method,” which involves visual aids to teach math- or purchase software, among other things. ematics, is used across schools in Singapore. Sources: Lee et al. 2008; NCEE n.d.; OECD 2010. outside, but surrounding countries also university students and members of the busi- struggle with quality teaching, and their ness community to contribute to the quality teachers are unlikely to speak Kinyarwanda. of schools. The answer lies within Rwanda’s own teach- Recruiting expatriates can help to rem- ers, but they need to receive a different kind edy the shortage of teachers, especially of of support than they are used to receiving. teachers who are proficient in English. The This support also involves providing bet- current ratio of pupils to teacher in primary ter management to teachers, as education schools in Rwanda is 58, among the highest policy makers have done in Shanghai (box in the world; the ratio of pupils to teacher 1.5). Rwanda may be able to adapt existing proficient in English would be much higher. social structures, such as the monthly day of Countries like Botswana and Oman car- service—called ­ u muganda —to encourage ried out overseas recruitment in the 1970s well-educated members of society such as and 1980s to close the teacher gap created 70 FUTURE DRIVERS OF GROWTH IN RWANDA BOX 1.5  Lessons from China: Teacher management in Shanghai In 2009 and again in 2012 , when Shanghai’s school an “empowered management” contract, with 15-year-olds topped the scores on the global PISA funding attached, to improve a lower-­ p erforming tests, the world wondered how this feat could have school, in which the principals and teachers involved been achieved in an education system that only would spend time in each other’s school to work on three decades earlier had been left in shambles by improving management and instruction. China’s Cultural Revolution (1966–76), a decade Teachers handle relatively large classes, but they when teachers were sent to the countryside to per- are equipped to teach well through a system of con- form manual labor. tinuing professional development. Mentor teachers Veteran foreign observers and Chinese educators gather weekly with teachers from the same subject consider Shanghai’s method for teacher management or grade level to discuss some aspect of instruc- to be one of the system’s most valuable and unique tion that needs improvement. Classroom observa- assets, one that enabled the city to achieve its remark- tion occurs frequently: teachers observing each able results (Tucker 2014). The method has little to other, trainee teachers observing master teachers, do with the workshop model of professional develop- and teachers visiting from other schools circulating ment for teachers and everything to do with enabling among classrooms. The public nature of teaching effective teaching and rewarding continuous, disci- practice exerts a subtle influence on teachers who plined improvement of teacher performance. Three perform less well, putting social pressure on them to ingredients stand out: systemwide structural reforms, improve and allowing them to see how other teach- active experimentation in schools, and purposeful ers get better results. management of teachers’ work. The system incentivizes all teachers to progress In the 1990s, when the focus shifted from up the career ladder, which requires honing not just expanding access to improving quality, Shanghai one’s own teaching skills but also getting involved piloted a new national curriculum, began encourag- in the professional development of colleagues and ing more active kinds of pedagogy, and abolished improving the overall instructional quality of the primary school examinations—the latter to encour- school. As teachers move up the career ladder they age schools to focus on deeper learning rather than may lead action research projects, often in collabo- on teaching to the test. The city also moved to nar- ration with others. For this research, they focus on row the gap between top-performing “key” schools an education issue of interest, evaluate the literature, and those at the bottom that served large num- experiment with different approaches in the school, bers of children from poor households, including and report on the results. Teachers who wish to migrant families from rural areas. It experimented reach the top of 13 levels (master teacher level) are with many approaches: pairing high-performing expected to spend time teaching in a lower-income schools with struggling ones to bring up the qual- area. Being recognized as a master teacher confers ity of leadership and teachers in the latter; asking great honor on the small percentage of teachers principals from high-performing schools to manage who qualify; such teachers not only work in their more than one school; forming clusters of schools own school but also share their expertise with other to share resources; and giving a high-performing schools as well. by rapid educational expansion (Al Maskari would impair the quality of education. Also et al. 2015; Kesamang et al. 2012). Rwanda essential would be recruiting expatriate might consider hiring more teachers from teachers from countries that have a curricu- neighboring English-speaking countries, lum consistent with Rwanda’s. given the limited capacity of Rwanda’s Expert teachers from abroad can be teacher training instit utions and the used to train a core group of high-quality acute demand for teachers in the system.9 teachers. During Korea’s development era However, dependency on expatriate teach- (1954–61), the country retrained its medi- ers may lead to high turnover rates, which cal faculty at U.S. universities, leading to H u m an C a p ita l and I nnovation 71 improved teaching methods and a new computers and software. Most technologi- organizational culture in Korean medical cal interventions are expensive and chal- education (Lee and Taejong 2015). Indeed, lenging to implement, especially in areas Rwanda has some history of this approach, with limited technical and human capacity. requesting teachers from Kenya to pro- Indeed, the Ministry of Education encoun- vide in-service training to Rwandan teach- tered similar difficulties when implement- ers in 2011 and establishing agreements ing the One Laptop per Child program with Ugandan teachers in 2009 to provide and building “smart classrooms” in pri- English instruction (Ochs and Yonemura mary and secondary schools in Rwanda 2013). Short-term teacher trainings tend (Ministry of Education, Rwanda 2016). to have little impact, so an effort along Moreover, technology delivers learning these lines would need to bring teachers gains only when it affects pedagogical from high-quality education systems for practices. For example, the One Laptop sufficient time to change practice among per Child program alone could not have Rwandan teachers. been effective without being accompanied Te c h n o l o g y c a n b e a p p l i e d t o by adaptive learning software (Perera and increase students’ learning performance. Aboal 2017). M i ndspa rk, an interac tive computer- Rwanda has made major investments in assisted learning sof t ware, increased education technology, deploying laptops math scores by 0.59 standard deviation to many schools across the country. As of and language scores by 0.36 standard 2013/14, there were nearly 270,000 lap- deviation for a group of Indian students tops in primary schools across Rwanda, in 90 days (Muralidharan, Singh, and although few schools (less than 1 in 10) Ganimian 2017). Students signed up to had I nternet access (L emarchand and take six 90-minute sessions after school Tash 2015). The number of users per com- per week; actual attendance was about puter is 13 in primary schools and 27 in 58 percent. The software provides stu- secondary schools (NISR 2016). Before dents with extensive, high-quality learn- investing further in information and com- ing materials as well as individualized munication technology infrastructure and activities based on each student’s learning related interventions, it is crucial to focus progress, which is difficult to achieve in a on technologies that are compatible and classroom with high heterogeneity in stu- potentially yield the largest returns to the dent performance. More promising, a sim- existing education system. For example, ilar intervention implemented at scale has although coaching provided under the also shown positive impact on learning. School-Based Mentoring Program has The Adaptive Mathematics Platform that significantly improved teachers’ English has been used by half of students in grades skills, the program has had difficulty 3–6 in Uruguay improved math scores by recruiting and deploying mentors in every 0.2 standard deviation, equivalent to half school across Rwanda (R EB 2016). In a semester of schooling (Perera and Aboal other countries, technology has provided 2017). The online platform is accessible at innovative solutions to improve adults’ no cost through the laptops distributed to language skills (box 1.6). Regular prac- all students in Uruguayan public schools tice with high-quality ­ l anguage-learning as part of a national education program, software has the potential to improve Plan Ceibal. On average, 68 percent of teacher skills more than occasional prac- students completed at least one exercise on tice with an in-school mentor alone. the platform during the three-year evalua- Linking high English proficiency with tion period. job certification and remuneration can However, to be effective, technology increase the incentives to practice and has to be about much more than installing improve. 72 FUTURE DRIVERS OF GROWTH IN RWANDA BOX 1.6  Technology for improving teachers’ English skills Technology has been an effective way to improve volunteers over Skype. Learning outcomes were adult ability in a new language in other countries. comparable to those in community college classes, The LaGuardia Community College in New York but at a faster pace (Wilson 2014). City used a Blackboard website and podcasts to Technological innovations hold promise to help adult immigrants with basic English profi- enhance teachers’ language skills in a shorter time ciency to develop language skills that are expected than traditional approaches. First, learning soft- in the lodging industry. The curriculum was ware can reach teachers in rural and remote areas designed specifically for the hospitality industry, through computer or mobile devices when a school and podcast lessons were tailored to the needs of mentor is not available. Second, adaptive learning individual students. Students in the program made software enables teachers to learn at their own level substantial gains in both English and computer and pace, better addressing the needs of a teach- skills (Warschauer and Liaw 2010). Another pro- ing force with various levels of English proficiency. gram, OneAmerica in Washington State, piloted Third, language-learning software can combine an English teaching program using a combination with pedagogical training as part of teacher pro- of technology and classroom instruction. Adult fessional development. Current school mentors can students took self-paced English courses using travel across schools and provide on-site support LiveMocha software through computer or mobile following the most effective practices (Evans and devices and participated in English discussion with Popova 2016). Skills Training among the population 25–30 years old (NISR 2017). Current Status Increasing numbers of youth are enroll- In order to reap more immediate economic ing in technical and vocational education returns from human capital investment, and training (TVET), up nearly 40 percent youth who are approaching the job mar- between 2011 and 2015. Relative to the ket need better training to help them to rest of the region, Rwanda has high TVET transition. Census projections suggest that enrollment (figure 1.16). As more students ­ the population of working-age Rwandans complete primary and lower-secondary will increase by roughly 240,000 per year school, these numbers will likely continue between 2016 and 2025 (World Bank to rise. This population has the potential to 2017c). At the same time, business people meet significant needs for technicians as well rated “inadequately educated workforce” as for professional graduates who come from as the second largest constraint to doing traditional higher education. business in Rwanda, after “access to financ- ing” (World Economic Forum 2017). Audits Proximate Causes from across sectors (agriculture, energy, manufacturing, mining, urbanization, and Rwanda is already innovating in its voca- meetings, incentives, conferences, and exhi- tional training system. The five-year plan bitions) have indicated the need for major (2013–18) for priority skills development skills upgrading.10 The 2017 Labor Force highlighted seven fields of special focus— Survey puts the share of youth who are infrastructure (including energy), agricul- not in education, employment, or training ture, natural resources, investment, trade (NEET) at 25 percent among the popula- and industry, information and communica- tion 16–19 years old, 35 percent among the education— tion technology, and health and ­ population 20–24 years old, and 40 p ­ ercent and set targets based on needs assessments H u m an C a p ita l and I nnovation 73 FIGURE 1.16  Enrollment in secondary TVET, by region and within Africa, circa 2000 and 2014 b. By country Angola Cameroon Congo, Dem. Rep. Rwanda Mali a. By region Tanzania Mauritius Congo, Rep. Europe and Central Asia Gambia, The Niger Ethiopia Latin America and the Caribbean São Tomé and Príncipe Liberia Burundi Equatorial Guinea East Asia and Paci c Botswana South Africa Togo Middle East and North Africa Mozambique Senegal Uganda Seychelles Sub-Saharan Africa Guinea Lesotho Burkina Faso South Asia Central African Republic Cabo Verde Benin 0 5 10 15 20 25 Madagascar Ghana % enrollment in TVET secondary Guinea-Bissau Chad Circa 2000 Circa 2014 Sudan Mauritania Eritrea Comoros Kenya Swaziland 0 5 10 15 20 25 30 35 40 45 50 % enrollment in TVET secondary Source: Calculations based on data from United Nations Educational, Scientific, and Cultural Organization, Institute for Statistics (Arias, Santos, and Evans 2018). Note: On panel b, Angola = 45.2. TVET = technical and vocational education and training. in each area. A midterm review of the plan Kenya, income for those who had vocational revealed that many Rwandans are gain- training in hairdressing was more than six ing training and skills, although not nec- times higher than for those who had training essarily in the priority subfields: In 2016, in tailoring (Hicks et al. 2015). Getting indi- Rwanda formed the Capacity Development viduals into high-quality programs with high and Employment Services Board to bring returns is key. skills development and employment creation Two key issues stand out in Rwanda’s together under the Ministry of Public Service skills development. On the one hand, the and Labor (World Bank 2017c). unskilled nature of the bulk of the labor force Technical and vocational training tends requires further increases in agricultural to have highly variable returns. Training productivity and job creation in industries programs in some fields bring high returns, related to agriculture (for example, agribusi- whereas others provide little benefit above ness and agroindustry) to improve employ- simply obtaining work in the field. In the ment outcomes for the most vulnerable group United States, technical education in of workers. On the other hand, the rapid health delivers returns more than twice as expansion of education means that new labor high as returns in all other fields (Stevens, force entrants will be increasingly educated, Kurlaender, and Grosz 2015). For women in which requires the creation of more skilled 74 FUTURE DRIVERS OF GROWTH IN RWANDA and semiskilled jobs, preferably through the skills. Second, they created an organizational entrance of larger formal firms. Expansion infrastructure with the appropriate gover- of economic activity in the secondary cities nance design. Third, they fostered efficient will be important in this regard, as will be and effective management of service delivery the ongoing efforts to expand access to elec- at the level of providers. tricity and improve the domestic and external Five principles emerge from the analysis connectivity of the economy. of international skills programs (Roseth and Valerio 2016): Investments for Transformation 1. Align resources to the strategy. Ensure What can Rwanda learn from international that there is political and technical will as examples in skills training? Building work- well as funding to back up the national force skills for employment and economic workforce development strategy. growth is the aspiration of most low- and 2. Bring employers into the process. Engage m iddle-income countries. The task poses ­ employers not just for providing intern- many difficulties for which solutions often ships and apprenticeships but also for con- involve government action to varying degrees tributing to policy making and providing of intensity. In high-income economies, the training. This engagement requires clear government’s role is often muted or focused on mechanisms and incentives. Malaysia has areas of specific needs (for example, appren- had success engaging firms in the voca- ticeships in the United States in recent years tional training space, largely through leg- and training for mid-level jobs through com- islation (box 1.8). munity colleges rather than through four-year 3. Improve the standing of TVET. TVET colleges). In low-income economies or devel- is often seen as the last choice of stu- opmental states that are driven by aspira- dents who do not succeed in university. tions for rapid economic transformation, the Improving the quality of systems and dem- government’s role tends to be more active and onstrating the returns to students who broader in scope, if only because the institu- are well suited to technical training can tions for skills provision are nascent or poorly improve the attractiveness of these pro- developed and information gaps impede the grams. Korea, Singapore, and Taiwan, smooth functioning of the market for skills. China, all sought to make technical and But effective skills training builds on a foun- vocational training a more attractive dation of basic skills, which is why Singapore’s option for students (box 1.9). excellent training system builds on a strong 4. Focus on what matters. If the objective foundation of basic education (box 1.7). is job placement, then programs should The development experience of econo- be measured on job placement. If the mies such as Chile, Ireland, Korea, Malaysia, objective is competency, then programs Singapore, and Taiwan, China, exemplify should be measured on an objective set of this activist role of government in facilitat- competencies. Performance information, ing the building of skills for employment and made publicly available, will let students, growth. Although their approaches respond their families, and employers make opti- to c­ ountry-specific conditions and there- mal choices. fore vary in design, these countries all had 5. Gather and use data. Without systematic to solve challenges in at least three areas to data collection, many of these practices build a system capable of achieving sustain- are impossible. able delivery of quality services: strategy, Getting more people into priority pro- organization, and implementation. Three les- grams requires information. Information sons stem from the experiences of these coun- can be a powerful tool to help individuals tries. First, they set a strategic direction, tone, to access high-return programs. Campaigns and culture for efforts to improve workforce that share information on the returns to H u m an C a p ita l and I nnovation 75 BOX 1.7  A strong foundation in basic education for transforming TVET in Singapore I n 20 07, Singapore’s I nstit ute of Tech nical advance to the next level of education or training and Education (ITE) beat more than 100 competitors (2) reposition TVET as part of the postsecondary from 30 countries to win the inaugural Global system (Government of Singapore 1991). The second Award in Transforming Government given by the recommendation led to ITE’s creation in 1992; the Ash Institute for Democratic Governance and first contained details on giving all students a firm Innovation at Harvard University’s John F. Kennedy grounding in English, mathematics, and proficiency School of Government (Law 2015). The award in the mother tongue. All primary pupils in the first brought attention to one of the world’s most suc- four grades would follow the same curriculum, with cessful approaches to providing technical educa- the following time allocation: 33 percent for English, tion. Instead of being a last-resort option for low 27 percent for the mother tongue, and 20 percent ­ a cademic achievers, by 2007 ITE was enrolling for mathematics. At the end of fourth grade, pupils 25 percent of each cohort of tenth-grade completers. would separate into classes of English, mathemat- For those with the lowest academic results, ITE ics, science, and mother tongue of varying difficulty, used new pedagogies particularly suited for expe- with the most academically and linguistically able riential learners, equipping them for success in the pupils learning English and the mother tongue at a labor market and, for some “late bloomers,” for first-language level (10 percent); the majority learning further education. This single shift in perspective English language as a first language and the mother combined with a massive restructuring of the sys- tongue as a second language (70–75 percent); and tem to make Singapore a global leader in technical weaker pupils learning English as a first language and vocational education and training (TVET). and the mother tongue at oral proficiency (15–20 per- Few are aware, however, of the critical impor- cent). For the last group, the reduced stress on mother tance of a firm foundation in basic education for tongue freed up time to concentrate on English and ITE’s success. A landmark review of the school mathematics. Secondary education would last four to system, including vocational training, revealed five years, with students following either an academi- deep problems among the lowest academic achiev- cally or a technically biased course. The latter course ers, those exiting the system with just eight years of would cater to the 15–20 percent least academically ­ primary schooling (Government of Singapore 1991). inclined students for whom English, mathemat- Some 85 percent of each cohort of these students ics, and computer applications would be compul- took up TVET, but as many as 60 percent could sory; strengthening their proficiency in English and not complete the lowest-level courses for semiskilled mathematics would continue to be emphasized. For workers, the main reason being their weak academic students in the academically biased course, some of preparation in mathematics and English (Varaprasad whom may follow the TVET after completing their 2016). Moreover, 75 percent of those who did gradu- course, compulsory subjects would include English, ate were not getting jobs in their area of training. mathematics, and the mother tongue. The poor results came to be recognized as a serious The committee’s two recommendations presented impediment to the country’s transition to its next very significant challenges. Their successful imple- phase of industrial development in the early 1990s. mentation created a strong foundation for TVET The 12-member Review Committee recom- and set the stage for changing the face of TVET, mended two remedies: (1) provide all children with in the process producing what Varaprasad (2016) at least 10 years of general schooling before they describes as “a modern miracle of transformation.” education have been successful at improving distributed a simple flyer presenting the learning in Mexico and reducing dropouts returns across different vocational train- in the Dominican Republic and Madagascar ing programs had a sizable but short-term (World Bank 2018). Recent campaigns have impact—it was effective at getting youth into sought to inform youth of the returns to higher-return programs, but not necessar- different sectors. A program in Kenya that ily at getting them to complete them (Hicks 76 FUTURE DRIVERS OF GROWTH IN RWANDA BOX 1.8  Partnering with firms to invest in workforce skills in Malaysia Malaysia’s levy-based system has evolved over the Firms could make claims for allowable train- past 30 years and now offers the government a ing expenses, up to the limit of their firm-specific p owerful tool for collaborating with employers ­ levy contributions. In 2016, the PSMB collected a on the skills agenda. The Eleventh Malaysia Plan significant amount in grants to fund nearly 90,000 (2016–20) envisions using the system to raise the training places through various schemes, the most share of Malaysians in skilled jobs, from 28 percent popular being training planned by firms them- in 2015 to 35 percent by 2020. selves. The establishment of the National Human The first piece of legislation, an act of Parliament Resources Center in 2011 intensified support for enacted in 1993, created the levy system to replace small and medium enterprises to invest in skills. The an ineffective tax incentive scheme in force creation of the HRDF Pool Fund in 2016—financed since 1987. It established the Human Resources by a 30 percent claim on levies held in firm-specific Development Fund (HRDF) to hold funds compris- accounts and government budget allocations— ing periodic government grants and monthly levies sought to increase the focus on skills for priority paid by firms into firm-specific accounts. It also economic sectors authenticated by internationally created the Human Resource Development Council recognized credentials, not just certificates of atten- (HRDC) and a secretariat to administer levy collec- dance at training courses. Evaluations suggest that tion and grant disbursements for approved training HRDF grants have helped to raise firm and worker programs. In 2001, a new act replaced the HRDC productivity. with the Pembangunan Sumber Manusia Berhad Malaysia’s levy-based system complements par- (PSMB)—Registration of Employers and Payment allel efforts to provide effective training. Legislation of Levy—an agency under the Ministry of Human in a separate act in 1993 created the National Resources. Structured as a corporation, PSMB is Occupational Skills Standards system—supervised governed by representatives of businesses and gov- by the National Vocational Training Council—to ernment entities and is always chaired by a busi- align skills certification with standards established ness leader—practices already in place under the by industry, particularly those defined by levy-­ HRDC. paying firms. Legislation in subsequent years cre- In 1993, the system required only ­ manufacturing ated various accreditation bodies for public and firms with 50 or more employees to pay a levy; just private providers of education and training, cul- three years later, it also required smaller manu- minating in the 2007 creation of the Malaysian facturing firms with 10 more employees to do so, Qualification Agency, to bring standards for pro- along with some firms in services. By 2017, its cov- vision and certification within a single framework. erage had spread to 63 subsectors in three indus- Under the Eleventh Malaysia Plan, further efforts to tries: services, manufacturing, and mining and enhance the coherence and responsiveness of provi- quarrying, with nearly 18,000 firms registered. As sion to emerging needs include (1) establishing the of April 2017, firms with 10 or more workers were National Skills Development Council to consolidate required to pay a levy of 1 percent of payroll, while TVET and academic programs under a single gov- those with 5–9 workers were required to pay a levy erning body and (2) creating industry skills coun- of 0.5 percent. The current plan envisions register- cils and critical skills committees to inform the ing all firms. council’s work. Sources: Gobilee 2016; PSMB n.d., 2016; World Bank 2013; Yaacob 2016. et al. 2015). As in basic education, regular high-return fields, they credited the influence monitoring and publication of information of a role model (Buehren and Van Salisbury can be crucial to increase the returns of skills 2017; Campos et al. 2015). A program that programs. gives youth the chance to interact with a role Mentors also play a key role in skills devel- model in target, high-return fields may have a opment. When young women in Ethiopia and significant impact on moving youth into stra- Uganda were asked why they had entered tegic sectors. H u m an C a p ita l and I nnovation 77 BOX 1.9  Developing skills for economic growth and transformation in the “Asian Tiger” economies Five to six decades ago, the Republic of Korea, relevant targets through additional structures. Singapore, and Taiwan, China—the original Asian In Singapore, this was accomplished through the Tiger ­ economies—faced existential threats, rampant Economic Development Board and the Council for joblessness and poverty, and overwhelming odds in Professional and Technical Education; in Korea, their pursuit of industrialization. Yet, in the space through Presidential Commissions on Education of a generation, all three have lifted living standards Reform and other advisory bodies; and in Taiwan, to levels that took high-income countries three or China, through the Council for Economic Planning more generations to reach. Proactive policies in edu- and Development’s Manpower Development cation and training were a critical enabler. What Commission and the National Youth Commission approaches did the three economies share? Are they (Ashton et al. 2002). still relevant today? They all affirmed the value of technical and The Asian Tigers each followed their own path to vocational education and training (TVET). While develop their economies. Singapore opted to attract emphasizing a solid foundation in basic education, multinational corporations to spearhead its drive up all three economies used their limited resources the value chain; Korea developed its own national at the postsecondary level, especially in the early conglomerates (the chaebols) modeled after Japanese decades, to expand TVET rather than academic counterparts; and Taiwan, China, relied on indig- higher education. The employment of TVET gradu- enous small and medium family-owned businesses ates was a key metric used to reflect the value of and on some state-owned enterprises (subsequently T V ET. In Singapore, substantial investments, privatized), which adopted, where appropriate, beginning in the early 1990s, created the Institute leading-edge technology (Ashton et al. 2002). These of Technical Education as an explicit part of the economies shared three features in their approaches postsecondary system (no longer part of general to education and training: (1) mechanisms to link education) that today enrolls 25 percent of tenth- decisions about the output of graduates to priority grade completers (Law 2015). Singapore’s Economic skills needed by the new industries; (2) regulation of Development Board also partnered with foreign access to the various types of education and train- firms to establish numerous stand-alone training ing; and (3) continuous upskilling of the workforce institutes, which, after 20 years, were consolidated aligned to employer demand. Mistakes were made in 1993 to create a new polytechnic, mainstreaming and corrected, in the process building domestic best practices in industry-relevant training (Tan and capacity to sustain reforms over decades and achiev- Nam 2012). In Korea, as the chaebols were being ing tangible economic transformation. established and demand for semiskilled labor was They all aligned education and training to the rising, the government expanded vocational second- evolving needs of their new industries. All three ary programs, which at the peak in 1973 graduated Asian Tigers created what might be described as 60 percent of all high school graduates (Lee and “super-ministries” as the institutional mechanism Hong 2014). Parental pressure proved too power- for this purpose. These entities had the power, ful to sustain these high shares, however (Yoon over many decades, to steer, coordinate, and, if and Lee 2009). In Taiwan, China, heavy govern- necessary, override the priorities of other min- ment investment boosted TVET students’ share in istries and interest groups, including parents, so secondary schools from 40 ­ p ercent in 1960 to 72 as to keep the needs of the economy paramount. percent by 1990. In line with its development strat- In Singapore, it was the Ministry of Trade and egy, in 1970 the government created the National Industry; in Korea, the Economic Planning Board; Taiwan Institute of Technology, which was autho- and in Taiwan, China, the Council for Economic rized to offer degree-level training; it also set and Planning and Development. These high-level entities achieved high targets for enrollment in science and set overall economic strategy, oversaw the distilla- engineering in tertiary education (for example, in tion of its implications for education and training, 1984 these fields enrolled 48 percent of undergradu- and formalized the setting and enforcement of ates) (Ashton et al. 2002). (Box continues next page) 78 FUTURE DRIVERS OF GROWTH IN RWANDA BOX 1.9  (continued) They all offered and funded continuous upskill- on equipping workers to improve their productivity ing of workers. In the 1980s, Singapore’s National (Tan and Nam 2012). Finally, in Taiwan, China, the Productivity Board sponsored modular training to levy scheme met with limited success partly because equip workers for jobs higher up the value chain in the of the predominance of small and medium enter- country’s target industries (for example, pharmaceuti- prises. Budget allocations thus remain a major source cals, petrochemicals, and electronics). Of the 170,000 of funding for upskilling the workforce. In all three participants during 1983–86, 75 percent were in the countries, employers are involved in training their Basic Education for Skills Training course, which was workers, by practicing or hosting workplace learning, designed explicitly for workers with an incomplete providing inputs for curriculum design, or contribut- primary education to improve their English and math- ing faculty for course delivery. ematics skills (Law 2015). For employers, Singapore Rwanda is already learning from the Asian previously relied on wage policies (found ineffective Tiger economies. The political leadership is strongly and thus dropped) and levies to stimulate employ- committed to investing in workforce skills across the ­ ers’ investment in worker training. In 2001 the Life full spectrum of the education and training system. Long Learning Endowment Fund, seeded by a large The country faces different conditions, however. allocation from the government’s budget surpluses, The headwinds of today’s integrated global econ- replaced the levy scheme. Interest income from the omy are possibly stronger, and the workforce needs endowment now funds incentives for both workers rising levels of general education to harness tech- and firms to invest in workforce skills. Such spend- nology’s full promise. The Asian economies were ing also creates a countercyclical macroeconomic relentless in building systemic capacity to manage dynamic, protecting jobs and increasing training education and training systems, treating skills their ­ during an economic downturn, as it did during the for transformation as teamwork across multiple 2008–09 global financial crisis. Korea’s levy scheme entities, working collaboratively and pragmatically was initially ­ successful—for about a decade—and to overcome implementation difficulties, relying on was later replaced by the Employment and Insurance market forces to shape incentives, but also not shy- Scheme. Firms use their claims to subsidize approved ing away from government action, and using simple in-house reskilling and retraining (Yoon and Lee performance metrics to correct mistakes, test new 2009). In recent years, a tripartite arrangement ideas, and check for progress. A similar mind-set (called BRIDGE) involving the chaebols, universities, will help Rwanda to overcome its own enormous and small and medium enterprises has emerged to challenges of system construction in the coming ensure that skills upgrading remains solidly focused years. Sources: Almeida, Behrman, and Robalino 2012a, 2012b; Ashton et al. 2002; Green et al. 1999; Law 2015; O’Hare 2008; Rodrik 1995; Sung 2006; Tan and Nam 2012; Tan et al. 2016; Yoon and Lee 2009. For the significant portion of the popula- programs that provide physical capital— tion who will continue creating their own whether livestock in India or cash grants in jobs, capital-centric programs may be more Ghana or Uganda targeted toward growing effective and cheaper to implement than businesses—have had strong positive impacts simple training programs. Many Rwandans (Blattman and Ralston 2015). Some of these remain self-employed, often in the informal programs combine skills ­ training with cash sector. To allow their enterprises to grow, one grants. Physical capital may be more of a approach is to seek to increase their human constraint to entrepreneurs than human capital (through training) or their ­ physical capital, or physical capital may allow them capital (through cash transfers, in-kind to multiply whatever gains they receive from capital, or subsidized credit). Skills training skills training. Regardless, providing physical for self-employed workers tends to be very capital for youth to enable them to become expensive, and the impacts on wages are entrepreneurs is a bold, promising approach often low or nonexistent. At the same time, supported by increasing evidence. H u m an C a p ita l and I nnovation 79 Tertiary Education and FIGURE 1.17  Conceptual framework outlining the links between higher education and economic growth Innovation Current Status Entrepreneurship Specialization Productivity Jobs Higher education is important to Rwandan Private economic growth. Higher education can lead to economic growth through both pri- Economic vate and public channels, with higher indi- Increased spending growth Higher Sustained vidual incomes and job creation as well as education income investments in research and development Tax revenues growth (­ figure 1.17). Recent empirical analysis shows Poverty reduction that investment in education has a significant impact on growth (Bloom et al. 2014). Rwanda has low, but rapidly rising R&D Public FDI Social enrollments in tertiary education. The lat- ­ Governance Safety development est numbers—for 2015—suggest that just 8­ percent of tertiary-age youth are enrolled Source: Bloom et al. 2014. in tertiary education, well below the level Note: FDI = foreign direct investment; R&D = research and development. in today’s upper-­ m iddle-income coun- tries. Furthermore, even with rising tertiary FIGURE 1.18  Projected share of Rwanda’s workforce with tertiary enrollments—and those enrollments have ­ education under two growth scenarios, 2010–50 doubled in the last 10 years—shifting the pro- portion of the population with tertiary educa- a. General education trend scenario (business as usual) 100 tion in the workforce takes significant time, because only a small proportion of the work- 80 force changes with each graduating cohort (­ figure 1.18). However, with rapid expansion 60 Percent of new cohorts, significant improvements are possible in the coming years. 40 There is historical precedent for a rapid 20 rise in tertiary graduates. Countries such as Korea and Turkey achieved dramatic 0 increases in tertiary education. If Rwanda 2010 2015 2020 2025 2030 2035 2040 2045 2050 were to grow its tertiary enrollment at b. Fast-track scenario (at the pace of the the same rate as these countries did in Republic of Korea and Singapore) their fastest periods, it could reach a high 100 rate of ­t ertiary enrollment (figure 1.19). 80 Furthermore, universal tertiary enrollment is not the objective. Whereas tertiary enroll- 60 Percent ment is very high in some high-income coun- tries (93 percent in Korea and 86 p ­ ercent 40 in the United States), other high-income 20 countries have much lower rates, with just 57 ­ p ercent enrollment in the United 0 Kingdom and 74 percent in high-income 2010 2015 2020 2025 2030 2035 2040 2045 2050 countries as a whole. In middle-income No education Some or all secondary countries, the average higher education Some or all primary Some or all tertiary enrollment rate is just 47 percent. Source: Construction based on data from Lutz, Butz, and KC 2014. One key role of the tertiary sector is gen- Note: Panel a is based on the current rate of growth of enrollment in Rwanda. Panel b is erating and adapting innovations for use in based on the rate of growth of enrollment in the Republic of Korea and Singapore. 80 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE 1.19  Current and projected tertiary enrollment in Rwanda, 2000–35 Projected gross enrollment rate in tertiary education 100 90 80 High income (2015) 70 60 Percent 50 40 Upper-middle income (2015) 30 20 10 0 2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050 Rwanda (2015): 8% Korea, Rep. (1981–2001): 8% to 82% Turkey (1995–2015): 20% to 95% Colombia (1995–2015): 16% to 56% Mongolia (1995–2015): 14% to 69% Finland (1971–91): 13% to 50% Source: Based on data from World Development Indicators (World Bank, various years). Note: The projected enrollment rate in Rwanda is drawn from the rate of growth in enrollment of other countries. The legend shows the period over which the tertiary enrollment expansion was calculated as well as the level of tertiary enrollment at the beginning and end of the period. All comparator countries began their period of rapid expansion at a level comparable with that of Rwanda. FIGURE 1.20  Number of Rwandan scientific publications available applications has been quite low, with some on the Web of Science, 1970–2015 indication of a recent increase (Lemarchand and Tash 2015). 200 Rwanda is taking action to foster inno- Publications listed on the Web of Science 175 vation. Over the last decade, the govern- ment has taken steps to increase innovation 150 capacity. One of these steps was to estab- 125 lish the research-focused University of 100 Rwanda, formerly 10 other public univer- sities. There are a total of 29 institutes of 75 higher learning across the country. Another 50 step was to establish various institutes that house research and innovation activities, 25 including the National Industrial Research 0 and Development Agency, the Rwanda 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 Agriculture Board, the Rwanda Biomedical Center, Carnegie Mellon University-Africa, Source: Lemarchand and Tash 2015. the African Institute for Mathematical Sciences, and kLab, an open space for infor- Rwandan industry. Rwanda is making prog- mation technology entrepreneurs to col- ress in that area. In scientific publications, laborate. In addition, the government has Rwanda most recently ranked 25 out of entered strategic partnerships to establish 53 countries in the African region and 125 in a range of “centers of excellence” in fields the world. The number of scientific publica- as diverse as mathematics, biomedical engi- tions per year has grown dramatically, as neering, data science, and innovative math- shown in figure 1.20. The number of patent ematical pedagogy.11 In its most recent H u m an C a p ita l and I nnovation 81 National Science, Technology, Innovation, But expanding tertiary education—and and Research Policy, Rwanda seeks to estab- improving the capacity to innovate therein— lish a clear set of goals to strengthen innova- depends crucially on expanding primary and tion and to integrate science and technology secondary education. With upper-secondary education system. at every level of the ­ completion at only 18 percent and lower-­ Nearly a century of economic theory sug- secondary completion at only one-third, the gests that a country like Rwanda can reap university can draw on only a limited num- high returns from adoption and adaptation in ber of graduates, even fewer if the university innovation, which partnerships with higher selects only those most prepared for a tradi- education can help to develop. The economist tional tertiary education as opposed to tech- Joseph Schumpeter coined the term “creative nical or vocational training. As ­ figure 1.21 destruction” 75 years ago, referring to a pro- shows, almost every country with higher than cess of “industrial mutation,” in which new 20 ­percent tertiary enrollment has at least 60 processes push out (or “destroy”) the old ones percent of its youth graduating from lower- (Cox and Alm 2008). Innovation drives that secondary school. Likewise, expansions in process: it underlies economic growth and is tertiary education require expansions in upper- a crucial element of “how countries achieve secondary education, at least to a degree. prosperity” (Cicera and Maloney 2017). Countries that are farther from the devel- Proximate Causes opment frontier—low- and middle-income countries—have a unique opportunity to Relatively few graduates are specializing in benefit from innovation. Firms in Bolivia and key job-creation fields, such as science and Burundi—and Rwanda—should be able to engineering. According to last years’ statis- take advantage of innovation in Singapore tical yearbook (NISR 2017), nearly half of and Sweden and to experience extraordinary university students in Rwanda were studying growth while investing far less, adapting social science, business, and law (table 1.1). innovations rather than coming up with them The next biggest group was studying educa- (Cicera and Maloney 2017). tion. Just 6 percent were studying engineering, FIGURE 1.21  Tertiary enrollment rate as a function of secondary completion rates a. Lower secondary b. Upper secondary 100 100 90 90 80 80 Tertiary enrollment rate (%) Tertiary enrollment rate (%) 70 70 60 60 50 50 40 40 30 30 20 20 Rwanda 10 10 Rwanda 0 0 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 Completion rate (% of relevant age group) Completion rate (% of relevant age group) Sources: Based on data from World Development Indicators (World Bank, various years) and from United Nations Educational, Scientific, and Cultural Organization, Institute for Statistics. Note: Data, which were accessed in 2017, are for the most recent year for the country. 82 FUTURE DRIVERS OF GROWTH IN RWANDA manufacturing, and construction, and 9 per- figure 1.22). Indeed, in recent years, region (­ cent were studying sciences. This chapter has public financing of higher education has discussed the importance of improving the declined, even while the demand for tertiary quality of education in Rwanda, so having ­ education continues to grow (University of tertiary students in education may be worth- Rwanda 2016). At the same time, financ- while. But if Rwanda intends to grow its man- ing for innovation in Rwanda remains very ufacturing and technology sectors, then the low, most recently estimated at 0.4 percent number of students in science and engineering of gross domestic product (Lemarchand clearly will need to grow. and Tash 2015). In countries like Korea and Historically, Rwanda has invested more Singapore, that share is closer to 4 percent. in higher education than other countries in the region. In 2000, it spent a larger propor- Investments for Transformation tion of its education budget on higher edu- cation than almost any other country in the Rwanda is already making dramatic reforms to its tertiary education system, both refocusing its public resources on ­ ­ science and TABLE 1.1  Distribution of tertiary students in Rwanda, by field of study engineering training and seeking to improve % of quality. In 2017, the government suspended Field students some courses in 10 universities because they Education 14 failed to meet various quality standards. Humanities and arts 3 In 2013, those public universities were con- Social science, business, and law 45 verted into the single University of Rwanda Sciences 9 Engineering, manufacturing, and construction 6 in order to improve quality and efficiency. Agriculture 5 Rwanda needs to focus its tertiary edu- Health and welfare 9 cation system on key areas of investment: Service 10 more science and engineering. The cen- Source: NISR 2017. ters of excellence and Carnegie Mellon FIGURE 1.22  Spending on tertiary education in Rwanda versus other countries, circa 2000 and 2015 a. Circa 2000 b. Circa 2015 Rwanda Rwanda Maximum Maximum 75th percentile 75th percentile Sub-Saharan Sub-Saharan 50th percentile 50th percentile Africa Africa 25th percentile 25th percentile Minimum Minimum Low income Low income Lower-middle income Lower-middle income Upper-middle income Upper-middle income High income High income 0 10 20 30 40 0 10 20 30 40 50 Spending (% of total budgetary expenditure) Spending (% of total budgetary expenditure) Source: Based on World Development Indicators data (World Bank, various years). H u m an C a p ita l and I nnovation 83 BOX 1.10  Achieving high-quality tertiary education in the Republic of Korea Over the past 60 years, Korea has prioritized Korea also implemented several reforms to expanding access to education; as a consequence, its address its quality of tertiary education. Two-year tertiary system has grown rapidly. In order to meet teacher training programs were extended to four the increasing demand for education, the Korean years in order to improve teacher practices and government encouraged philanthropists to establish knowledge. Reforms included the expansion and private universities. These reforms resulted in one modernization of university libraries and research of the highest gross enrollment rates in the world. systems. Resources were funneled toward collabo- In 2015, Korea had a gross enrollment rate of 93 rating with foreign scholars, publishing in leading percent at the tertiary level, which far exceeded the international journals, and modernizing equipment. world average of 36 percent. Korea also lowered student-teacher ratios to 15 in order to improve the learning experience. Sources: Jones 2013; Kim 2008. University-Africa, which offers master’s recalls again that effective ­ tertiary reform degrees in engineering and information tech- relies on effective reform of basic education. nology, are good starts. This means shifting Encouraging innovation is part of trans- students into strategic areas in the University forming Rwanda’s universities. In Korea, of Rwanda as well. The extreme strategy for encouraging researchers to publish ­ innovative this shift is exemplified by Japan. In 2015, research was a key part of the strategy for Japan’s Ministry of Education ordered top developing world-class tertiary education (box universities to reduce their humanities and 1.10). The University of Rwanda currently has social science programs dramatically. Of incentives in place for publication, provid- 60 national universities, 26 will either close ing researchers who publish with additional or reduce these faculties and the relevant resources for research. To maximize the pro- student admissions (Grove 2015). Notably, ductivity of this research, Rwanda will want Japan’s top two universities—the two that to encourage researchers to adapt existing are among the world’s top 100 universi- innovations to the Rwandan context, thereby ties, University of Tokyo and University reaping the highest gains at the lowest cost. of Kyoto—are not making adjustments, so Japan will retain key programs in these Financing areas. But it will shift students into stra- tegic growth areas of study. Another, less These interventions require funding, but draconian, option includes providing finan- Rwanda is currently underspending on edu- cial incentives for students, such as offering cation (figure 1.23). Although infrastruc- scholarships focused on science, technology, ture needs (such as roads) often seem more engineering, and mathematics (STEM), as pressing, data from spending in 83 countries Argentina, Australia, and Denmark have suggest that reallocating resources from done (OECD 2012). The University of infrastructure spending to social spending Rwanda is already focusing scholarships on (that is, from roads to schools) can promote STEM. But a third key element of this strat- growth (Acosta-Ormaechea and Morozumi egy involves improving the quality of STEM 2017). Other analysis suggests that invest- instruction in basic education, so that more ments in schools are more likely to drive students are ready to study it in higher edu- growth than many infrastructure invest- cation, as Australia, Norway, and Poland ments (Bose, Haque, and Osborn 2007). each has done in different ways. This option In the short run, infrastructure may deliver 84 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE 1.23  Current spending on education and health in Rwanda relative to other countries in the world a. Expenditure on education b. Expenditure on health 9 14 8 13 7 12 Spending (% of GDP) Spending (% of GDP) 6 11 5 10 4 Rwanda 9 3 8 2 Rwanda 1 7 0 6 6 7 8 9 10 11 12 0 2 4 6 8 10 12 14 16 18 Log GDP per capita, PPP (current international $) Log GDP per capita, PPP (current international $) Source: Based on World Development Indicators 2017 data (World Bank, various years). Note: PPP = purchasing power parity. quick gains (Atolia et al. 2017), but for sus- not be the same for Kinyarwanda, but it is tained growth, economies need to invest unlikely to be significantly lower. in broad-based literacy. Regardless, major 6. All statistics in this paragraph are drawn from investments in human capital are an essen- EDC (2017). 7. Based on analysis of the 2015 Demographic tial part of Rwanda’s growth strategy. and Health Survey for Rwanda (NISR, vari- ous years). 8. At least two studies report that significant Notes proportions of schools are concerned with 1. Stunting refers to children who are dramati- absenteeism, but they do not indicate whether cally shorter than they should be. Specifically, the absolute level of absenteeism is high or low it refers to children who are 2 standard devia- (Bennell and Ntagaramba 2008; Valentine tions shorter than the median height for that and Ongus 2017). age in a reference population. 9. Rwanda has been hiring more teachers to 2. All statistics in this paragraph are drawn from respond to the increased number of stu- World Bank (2017b). dents; however, the pupil-teacher ratio in 3. The fertility rate decline is based on Rwanda’s primary schools remains high. The average Demographic and Health Surveys from 2005 pupil-teacher ratio in primary schools was and 2010 (NISR, various years). This sharp 41 for low-income countries in 2015 (World decline varies from the change presented by Bank, various years). To reach that average, the World Development Indicators (4.52 total even assuming no growth in the number of births per woman in 2005 to 3.92 in 2010), new students, Rwanda needs to hire more which uses five-year period, interpolated data than 20,000 additional teachers (based on from the United Nations Population Division’s NISR 2016). Currently, the system hires about World Population Prospects annual data series 2,500 to 3,000 teachers per year, which is (UN Population Division, various years). close to the training capacity of its teacher 4. Data for 2015 are from the World Development training institutions. Indicators (World Bank, various years). 10. Owen and Associates Limited (2017). Similar 5. One benchmark is that learners should be reports exist for the other sectors listed. reading 45–60 words per minute by the end 11. Data in this paragraph come from Republic of of second grade (Abadzi 2008). The rate will Rwanda (2016). H u m an C a p ita l and I nnovation 85 References Bashir, S., M. Lockheed, E. Ninan, and J.- P. Tan. 2018. Facing Forward: Schooling for Learning Abadzi, H. 2008. “Efficient Learning for the in Sub-Saharan Africa. Washington, DC: Poor: New Insights into Literacy Acquisition World Bank. for Children.” International Review of Basinga, P., P. J. Gertler, A. Binagwaho, A. L. B. Education 54 (5-6): 581–604. Soucat, J. Sturdy, and M. J. C. Vermeersch. Acosta-Ormaechea, S., and A. Morozumi. 2017. 2011. “Effect on Maternal and Child Health “Public Spending Reallocations and Economic Services in Rwanda of Payment to Primary Growth across Different Income Levels.” Health-Care Providers for Performance: An Economic Inquiry 55 (1): 98–114. Impact Evaluation.” The Lancet 377 (9775): Alderman, H., and R. Yemtsov. 2013. “How Can 1421–28. Safety Nets Contribute to Economic Growth?” Bennell, P., and J. Ntagaramba. 2008. “Teacher Policy Research Working Paper 6437, World Motivation and Incentives in Rwanda: A Bank, Washington, DC. Situational Analysis and Recommended Al Maskari, Z. F. Al Mawali, A. Al Harthi, Priority Actions.” Rwanda Education NGO and A. Al Rasbi. 2015. “Oman.” In TIMSS Cooperation Platform, Kigali. 2015 Encyclopedia: Educational Policy and Blattman, C., and L. Ralston. 2015. “Generating Curriculum in Mathematics and Science, Employment in Poor and Fragile States: Evidence edited by I. V. S. Mullis, M. O. Martin, from Labor Market and Entrepreneurship S. Goh, and K. Cotter. Boston, MA: TIMSS Programs.” Working Paper, University of and PIRLS International Study Center, Boston Chicago; World Bank, Washington, DC. College. Bloom, D. E., D. Canning, K. Chan, and D. L. Almeida, R., J. Behrman, and D. Robalino. Luca. 2014. “Higher Education and Economic 2012a. “Overview.” In The Right Skills for the Growth in Africa.” International Journal of Job? Rethinking Effective Training Policies for African Higher Education 1 (1): 22–57. https:// Workers, edited by R. Almeida, J. Behrman, doi.org/10.6017/ijahe.v1i1.5643. and D. Robalino. Washington, DC: World Bold, T., D. Filmer, G. Martin, E. Molina, B. Bank, Social Protection, Human Development Stacy, C. Rockmore, J. Svensson, and W. Network. Wane. 2017. “Enrollment without Learning: ———, eds. 2012b. The Right Skills for the Job? Teacher Effort, Knowledge, and Skill in Rethinking Effective Training Policies for Primary Schools in Africa.” Journal of Workers. Washington, DC: World Bank, Social Economic Perspectives 31 (4): 185–204. Protection, Human Development Network. Bose, N., M. E. Haque, and D. R. Osborn. 2007. Arias, O., I. Santos, and D. K. Evans. 2018. The “Public Expenditure and Economic Growth: Skills Balancing Act in Sub-Saharan Africa. A Disaggregated Analysis for Developing Washington, DC: World Bank. Countries.” The Manchester School 75 (5): Ashton, D., F. Green, J. Sung, and D. James. 2002. 533–56. “The Evolution of Education and Training Buehren, N., and T. Van Salisbury. 2017. “Female Strategies in Singapore, Taiwan, and S. Korea: Enrollment in Male-Dominated Vocational A Development Model of Skill Formation.” Training Courses: Preferences and Prospects.” Journal of Education and Work 15 (1): 5–30. Africa Gender Policy Brief, World Bank, Atolia, M., B. G. Li, R. Marto, and G. Melina. Washington, DC. 2017. “Investing in Public Infrastructure: Campos, F., M. Goldstein, L . McGorman, Roads versus Schools.” Vox, August 9. A. M. Munoz Boudet, and O. Pimhidzai. Avitabile, C., and R. de Hoyos. 2015. “The 2015. “Breaking the Metal Ceiling: Female Heterogeneous Effect of Information on E nt repreneu rs W ho Suc c e ed i n M a le - Student Performance: Evidence from a D om i n ate d S e c tor s .” Pol ic y R e s e a rch Randomized Control Trial in Mexico.” Policy Wo r k i n g P a p e r 75 0 3 , Wo r l d B a n k , Research Working Paper 7422, World Bank, Washington, DC. Washington, DC. Carneiro, P., and J. J. Heckman. 2003. “Human Barro, R., and J.-W. Lee. 2015. Education Capital Policy.” In Inequality in America: Matters: Global Schooling Gains from the What Role for Human Capital Policies? by 19th to the 21st Century. Oxford: Oxford J. J. Heckman and A. Krueger. Cambridge, University Press. MA: MIT Press. 86 FUTURE DRIVERS OF GROWTH IN RWANDA Chuhan-Pole, P., L. Christiaensen, A. Dennis, Gobilee, R. A. 2016. “The Role of Training Levies G. Kambou, M. Angwafo, M. Buitano, in Financing the Expansion of Technical and V. Korman, C . Galinda Pardo, and A. Vocational Education and Training (TVET) in Sanoh. 2013. Africa’s Pulse: October 2014. South East Asia: The Experience of Malaysia Washington, DC: World Bank, Office of the through the Human Resource Development Chief Economist, Africa Region. Fund (HRDF).” PowerPoint slides, Planning, Cicera, X., and W. F. Maloney. 2017. The Innovation Strategy, and Development Division, HRDF, Paradox: Developing-Country Capabilities Kuala Lumpur. and the Unrealized Promise of Technological Gove, A., M. K. Poole, and B. Pipe. 2017. Catch-up. Washington, DC: World Bank. “Designing for Scale: Reflections on Rolling Cilliers, J., and S. Taylor. 2017. “Monitoring Out Reading Improvement in Kenya and Teachers and Changing Teaching Practice: Liberia.” New Directions for Child and Ev id e n c e f ro m a Fi eld E x p e r i m e nt .” Adolescent Development 155 (Spring): 77–95. Georgetown University, Washington, DC. Government of Rwanda. 2016. “Mid-Term Commission on Growth and Development. 2008. Review Report of the Five-Year Priority Skills The Growth Report: Strategies for Sustained Development Programme.” Government of G ro w t h a n d I n c l u s i ve D e ve l o p m e n t . Rwanda, Kigali. Washington, DC: World Bank. Government of Singapore. 1991. “Improving Cox, W. M., and R. Alm. 2008. “Creative Primary School Education.” Report of the Destruction.” In The Concise Encyclopedia of Review Committee, Government of Singapore. Economics, 2d ed., edited by D. R. Henderson. Green, F., D. Ashton, D. James, and J. Sung. 1999. Indianapolis, IN: Liberty Fund. “The Role of the State in Skill Formation: Crouch, L., and K. A. Merseth. 2017. “Stumbling Evidence from the Republic of Korea, at the First Step: Efficiency Implications of Singapore, and Taiwan.” Oxford Review of Poor Performance in the Foundational First Economic Policy 15 (1): 82–96. Five Years.” Prospects 47 (3): 175–96. Grove, J. 2015. “Social Sciences and Humanities De Ree, J., K. Muralidharan, M. Pradhan, and Faculties ‘to Close’ in Japan after Ministerial H. Rogers. 2018. “Double for Nothing? Intervention.” Times of Higher Education , Experimental Evidence on an Unconditional September 14. Teacher Salary Increase in Indonesia.” Quarterly Handa, S., L. Natali, D. Seidenfeld, G. Tembo, Journal of Economics 133 (2): 993–1039. and B. Davis. 2018. “Can Unconditional EDC (Education Development Center). 2017. Cash Transfers Raise Long-Term Living Literacy, Language, and Learning Initiative Standards? Evidence from Zambia.” Journal (L3): National Fluency and Mathematics of Development Economics 133 ( July): Assessment of Rwandan Schools. Report for 42–65. U.S. Agency for International Development. Hanushek, E. A. 2016. “Will More Higher Waltam, MA: EDC. Education Improve Economic Growth?” Eozenou, P. H.-V., and M. Shekar. 2017. Stunting Oxford Review of Economic Policy 32 (4): Reduction in Sub-Saharan Africa. Washington, 538–52. DC: World Bank. http://documents.​­ worldbank​ Hanushek, E. A., D. T. Jamison, E. A. Jamison, .org/curated/en/126621505397202676​/­Stunti and L. Woessmann. 2008. “Education and ng-reduction-in-Sub-Saharan-Africa. Economic Growth.” Education Next 8 (2): n.p. Evans, D. K., and A. Popova. 2016. “What Really Hanushek, E ., and L . Woessmann. 2012 . Works to Improve Learning in Developing “Do Better Schools Lead to More Growth? Countries? An Analysis of Divergent Findings Cognitive Skills, Economic Outcomes, and in Systematic Reviews.” The World Bank Causation.” Journal of Economic Growth 17 Research Observer 31 (2): 242–70. (4): 267–321. Fryer, R. 2017. “Management and Student Hicks, J. H. H., M. Kremer, I. Mbiti, and E. Achievement: Evidence from a Randomized Miguel. 2015. Vocational Education in Kenya: Field Experiment.” Working Paper, Harvard A Randomized Evaluation; 3ie Grantee Final University, Cambridge, MA. Report. New Delhi: International Initiative for Galasso, E., A. Wagstaff, S. Naudeau, and Impact Evaluation (3ie). M. Shekar. 2016. “The Economic Costs of Hjelm, L. 2016. “Comprehensive Food Security Stunting and How to Reduce Them.” Policy and Vulnerability Analysis.” World Food Research Note, World Bank, Washington, DC. Programme, Rome. H u m an C a p ita l and I nnovation 87 Hong, S.-C., and J.-H. Lee. 2016. “Accumulating Lemarchand, G. A., and A. Tash, eds. 2015. Human Capital for Sustainable Development Mapping Research and Innovation in the in Korea.” Korea Development Institute, Republic of Rwanda. GO-SPIN Country Seoul. P rof i le s i n S c ie nc e , Te ch nolo g y, a nd Jensen, R. 2010. “The (Perceived) Returns to Innovation Policy, vol. 4. Paris: United Nations Education and the Demand for Schooling.” E duc at ion a l , S c ient i f ic , a nd Cu lt u ra l Quarterly Journal of Economics 125 (2): Organization (UNESCO). 515–48. Loyalka, P., C. Liu, Y. Song, H. Yi, X. Huang, J. Jones, S. R. 2013. “Education Reform in Korea.” Wei, L. Zhang, Y. Shi, J. Chu, and S. Rozelle. OECD Economics Department Working Paper 2013. “Can Information and Counseling Help 1067, OECD Publishing, Paris. Students from Poor Rural Areas Go to High Karra, M., D. Canning, and J. Wilde. 2017. School? Evidence from China.” Journal of “The Effect of Fertility Decline on Economic Comparative Economics 41 (4): 1012–25. Growth in Africa: A Macrosimulation Model.” Lutz, W., W. P. Butz, and S. KC. 2014. World Population and Development Review 43 (51): Population and Human Capital in the Twenty- 237–63. First Century. Oxford, U.K.: Oxford University Kesamang, M., C. Mudongo, O. Keatimilwe, Press. and T. Bolaane. 2012. “Botswana.” In TIMSS Ministry of Education, Rwanda. 2016. “ICT in 2011 Encyclopedia: Educational Policy and Education Policy.” Ministry of Education, Curriculum in Mathematics and Science , Kigali. edited by I. V. S. Mullis, M. O. Martin, C. ———. 2017. 2017/18 Forward-Looking Joint A. Minnich, G. M. Stanco, A. Arora, V. A. S. Review of the Education Sector Summary Centurino, and C. E. Castle, 177–90. Boston, Re por t . K igali: M inistr y of E ducation, MA: TIMSS and PIRLS International Study June. Center, Boston College. Ministry of Education and Sports, Uganda. 2014. Kim, Y.- C. 2008. “Understanding Korean “Teacher Issues in Uganda: A Shared Vision Educational Policy: Universalization of for an Effective Teachers Policy.” UNESCO, Tertiary Education.” Korea Education Policy I nternational I nstitute for E ducational Research Institute, Seoul. Planning (IIEP), Pôle de Dakar, Paris. K raf t, M. A ., D. Blazar, and D. Hogan. Monteneg ro, C . E ., and H . A . Patrinos. 2017. “The Effect of Teacher Coaching on 2014. “Comparable Estimates of Returns Instruction and Achievement: A Meta-Analysis to Schooling around the World.” Policy of the Causal Evidence.” Harvard University, Research Working Paper 7020, World Bank, Cambridge, MA. Washington, DC. Law, S. S. 2015. A Breakthrough in Vocational Muralidharan, K., A. Singh, and A. J. Ganimian. and Technical Education: The Singapore 2017. “Disrupting Education? Experimental S t o r y . S i n g a p o r e : Wo r l d S c i e n t i f i c Evidence on Technology-Aided Instruction in Publishing. India.” NBER Working Paper 22923, National L e e , J.- H ., a nd S . C . Hong. 2014. T he Bureau of Economic Research, Cambridge, Development of Vocational High Schools in MA. Korea during the Industrialization Period . NCEE (National Center on Education and the Seoul: Ministry of Strategy and Finance. Economy). n.d. “Singapore: Teacher and Lee, J.-W., and H. Lee. 2016. “Human Capital Principal Quality.” NCEE, Washington, DC. in the Long Run.” Journal of Development Nguyen, T. 2008. “Information, Role Models, Economics 122 (C): 147–69. a nd Perc eive d Re t u r n s to E duc at ion : Lee, K. W., and K. Taejong. 2015. “Real-Time Experimental Evidence from Madagascar.” Impact Evaluation of a Capacity-Building Working Paper, Abdul Latif Jameel Poverty Health Project in Lao PDR.” KDI Journal of Action Lab, Cambridge, MA. Economic Policy, November 30. NISR (National Institute of Statistics of Rwanda). Lee, S. K., C. B. Goh, B. Fredriksen, and J.-P. 2016. Statistical Yearbook 2016 . Kigali: Tan. 2008. Toward a Better Future: Education Government of Rwanda. and Training for Economic Development in ———. 2017. Labor Force Survey 2017. Kigali: Singapore since 1965. Development Practice Government of Rwanda. in Education Series. Washington, DC: World ———. Various years. Rwanda: Demographic and Bank. Health Survey. Kigali: Government of Rwanda. 88 FUTURE DRIVERS OF GROWTH IN RWANDA Ochs, K., and A. Yonemura. 2013. “Managing Ralston, L., C. Andrews, and A. Hsiao. 2017. Teacher Migration: Implications for Policy and “The Impacts of Safety Nets in Africa: What Practice in Africa.” Commonwealth Education Are We Learning?” Policy Research Working Partnerships 2012/2013 (online). http://www​ Paper 8255, World Bank, Washington, DC. .cedol.org/wp-content/uploads/2013/09​/ ­M an Ranis, G., F. Stewart, and A. Ramirez. 2000. aging-teacher-migration-Ochs-and-Yonemura​ “Economic Growth and Human Development.” .pdf. World Development 28 (2): 197–219. OECD (Organisation for Economic Co-operation REB (Rwanda Education Board). 2016. “School- and Development). 2010. “Singapore: Rapid Based Mentor Program Framework, 2017– Improvement Followed by Strong Performance.” 2022.” REB, Kigali, July 12. http://www​.reb​ In Strong Performers and Successful Reformers .rw/fileadmin/default/templates/2017_doc​ in Education: Lessons from PISA for the /SBMP/SBMP_Framework.pdf. United States, ch. 7. Paris: OECD. Republic of Rwanda. 2016. “National Science, ———. 2012. Science, Technology, and Industry Technology, Innovation, and Research Policy.” Outlook 2012. Paris: OECD. Draft, Republic of Rwanda, Kigali, July. O’Hare, D. 20 08. “Education in I reland: Rodríguez, A., C. Dahlman, and J. Salmi. Evolution of E conom ic and E ducation 20 08. K n owl edge an d Inn ovation for Policies since the Early 1990s.” In An African Competitiveness in Brazil. Washington, DC: Exploration of the East Asian Education World Bank. Experience, edited by B. Fredriksen and Rodrik, D. 1995. “Getting Interventions Right: J.-P. Tan. Development Practice in Education How South Korea and Taiwan Grew Rich.” Series. Washington, DC: World Bank. Economic Policy 20 (4): 55–107. Owen and Associates Limited. 2017. “Skills Audit Roseth, V., and A. Valerio. 2016. “Improve in Agriculture Sector in Rwanda 2017.” Report Workforce Development Systems in 5 (Not So submitted to Capacity Development and Simple) Steps.” World Bank, Washington, DC. Employment Services Board (CESB), Kigali, Saraf, P. 2017. “On-the-Job Training: Returns, May. Barriers to Provision, and Policy Implications.” Perera, M., and D. Aboal. 2017. “Evaluación Policy Research Working Paper 8090, World del impacto de la plataforma adaptativa de Bank, Washington, DC. matemática en los resultados de los apren- Shekar, M., J. Kakietek, J. D. Eberwein, and D. dizajes.” Centro de Investigaciones Económicas, Walters. 2016. “An Investment Framework Montevideo. for Nutrition: Reaching the Global Targets Pillay, P. 2010. Linking Higher Education for Stunting, Anemia, Breastfeeding, and and Economic Development: Implications Wasting.” World Bank, Washington, DC. for Africa from Three Successful Systems. Silva de Tiloka, T. S. 2017. “Population Control Cape Town: Centre for Higher Education Policies and Fertility Convergence.” Journal of Transformation (CHET). Economic Perspectives 31 (4): 205–28. Popova, A., D. K. Evans, and V. Arancibia. Snilstveit, B., J. Stevenson, R. Menon, D. 2016. “Training Teachers on the Job: What Phillips, E. Gallagher, M. Geleen, H. Jobse, Works and How to Measure It.” Policy T. Schmidt, and E. Jimenez. 2016. The Impact Research Working Paper 7834, World Bank, of Education Programmes on Learning and Washington, DC. School Participation in Low- and Middle- Popova, A., D. K. Evans, M. Breeding, and V. Income Countries: A Systematic Review Arancibia. 2017. “The Global Landscape of Summary Report. 3ie Systematic Review In-Service Teacher Professional Development Summary 7. London: International Initiative Programs: The Gap between Evidence and for Impact Evaluation (3ie). Practice.” World Bank, Washington, DC. Soares, F. V., R. P. Ribas, and R. G. Osório. 2007. PSMB (Pembangunan Sumber Manusia Berhad). “Evaluating the Impact of Brazil’s Bolsa Família: n.d. “Human Resources Development Fund Cash Transfer Programmes in Comparative Malaysia: Moving Forward.” Ministry of Perspec t ive.” I P C Eva luat ion Note 1, Human Resources, Kuala Lumpur. United Nations Development Programme, ———. 2016. Annual Report: “Think Different.” International Poverty Centre, Brasilia. Kuala Lumpur: Human Resource Development Stevens, A. H., M. Kurlaender, and M. Grosz. Fund (HRDF). https://www.hrdf.com.my. 2015. “Career Technical Education and Labor H u m an C a p ita l and I nnovation 89 Market Outcomes: Evidence from California Valentine, J., and R. W. Ongus. 2017. “Extrinsic Communit y Colleges.” N BER Working Motivation and Its Effects on the Productivity Paper 21137, National Bureau of Economic of Teachers of English Language Subject in Research, Cambridge, MA. Primary Schools in Gasabo District of Kigali Su ng, J. 20 06. E xpl aining the E conomic City, Rwanda.” Paper prepared for the “English Success of Singapore: The Developmental Language in Education and Research” confer- Worker as the Missing Link . Cheltenham: ence, Mount Kenya University, Kigali, March 7–8. Edward Elgar. van der Berg, S., N. Spaull, G. Wills, M. Tan, J.-P., K. H. Lee, R. Flynn, V. Roseth, and Gustafsson, and J. Kotzé. 2016. Identifying Y.-J. J. Nam. 2016. Workforce Development in Binding Constraints in Education: Synthesis Emerging Economies: Comparative Perspectives Report for the Programme to Support Pro- on Institutions, Praxis, and Policies. Directions Poor Policy Development . University of in Development Series: Human Development. Stellenbosch, Department of Economics, Washington, DC: World Bank. May 24. http://resep.sun.ac.za/wp-content​ Tan, J.-P., and Y.-J. J. Nam. 2012 . “Pre- /uploads/2016/05/PSPPD_BICiE-email.pdf. Employment Technical and Vocational Varaprasad, N. 2016. 50 Years of Technical Education and Training: Fostering Relevance, Education in Singapore: How to Build a Effectiveness, and Efficiency.” In The Right World Class TVET System. Singapore: World Skills for the Job? Rethinking Effective Scientific Publishing. Training Policies for Workers , edited by Wa rschauer, M ., and M .-L . Liaw. 2010. R. Almeida, J. Behrman, and D. Robalino, “Emerging Technologies in Adult Literacy 67–103. Washington, DC: World Bank, Social and Language Education.” National Institute Protection, Human Development Network. for Literacy, Emerging Technologies in Tashobya, A. 2017. “Quality Education Must Adult Literacy and Language Education, Be Our Priority—Kagame.” New Times, Washington, DC, June. December 19. Westoff, C. F. 2013. “The Recent Fertility Tucker, M. ed. 2014. Chinese Lessons: Shanghai’s Transition in Rwanda.” Population and Rise to the Top of the PISA League Tables. Development Review 38 (s1): 169–78. Wash i ng ton , DC: Nat iona l C enter on W HO (World Health Organization), World Education and the Economy, Center on Bank, and Ministry of Health, Vietnam. 2015. International Education Benchmarking. Success Factors for Women’s and Children’s UNFPA (United Nations Population Fund). 2017. Health: Viet Nam. Geneva: WHO. “Unlocking Rwanda’s Potential to Reap the Wilson, J. 2014. “Investing in English Skills: The Demographic Dividend.” Policy Brief, UNFPA, Limited English Proficient Workforce in U.S. New York, October. Metropolitan Areas.” Metropolitan Policy UN Population Division. Various years. World Program, Brookings Institution, Washington, Population Prospects. New York: United D C . h t t p s : / / w w w. b r o o k i n g s . e d u / w p​ Nations. https://esa.un.org/unpd/wpp/. -­content ​ /uploads/2014/09/metro_20140924​ Un iversit y of Rwa nda. 2016. “Resou rce _investing​_in_english_skils_report.pdf. Mobilisation Strategy and Implementation World Bank. 2013. “Malaysia Workforce Plan.” Directorate of Resource Mobilisation and Development, Country Report, Systems Approach Investment, University of Rwanda, November. for Better Education Results.” Education Global Upper Quartile and Institute of Policy Analysis Practice, World Bank, Washington, DC. www​ and Research Rwanda. 2015. Final Evaluation .worldbank.org​/education/saber. Report: Evaluation of Results-Based Aid ———. 2017a. “Project Appraisal Document: in Rwandan Education. Edinburgh: Upper Stunting Prevention and Reduction Project, Quartile. Rwanda.” World Bank, Washington, DC. U S A I D ( U . S . A g e n c y f o r I n t e r n at i o n a l ———. 2017b. “Rwanda Nutrition Situation Development). 2014. “Rwanda Nutrition Analysis and Policy Options.” World Bank Profile.” USAID, Washington, DC. Group, Washington, DC. ———. 2017. Literacy, Language, and Learning ———. 2017c. “Rwanda Priority Skills for Initiative: National Fluency and Mathematics Growth (PSG) Program-for-Results.” Program Assessment of Rwandan Schools. Washington, Appraisal Document Report 116204-RW, DC: USAID. World Bank, Washington, DC. 90 FUTURE DRIVERS OF GROWTH IN RWANDA ———. 2018. World Development Report 2018: Yaacob, M. 2016. “Private Sector Engagement in Learning to Realize Education’s Promise. TVET—Malaysia’s Experience.” Presentation Washington, DC: World Bank. at the “8th Annual Expert Meeting of the ———. Various years. World Development Initiative on Employment and Skills Strategies Indicators database. Washington, DC: World in Southeast Asia (ESSSA),” Ministry of Bank. http://databank.worldbank.org/data​ Department of Skills Development, Philippines /home.aspx. and Ministry of Human Resources, Malaysia, World E c o nom i c For u m . 2 0 0 8 . G l o b a l Cebu City, October 11–12. Competitiveness Report, 2007– 08. Geneva: Yoon, J. H., and B.-H. L ee. 20 09. “T he World Economic Forum. Transformation of the Government-Led ———. 2017. “Rwanda 52nd.” I n Global Vocational Training System in Korea.” In Competitiveness Index, 2016–2017 (online edi- Vocational Training. International Perspectives, tion). Geneva: World Economic Forum. http:// edited by G. Bosch and J. Charest, 163–86. reports.weforum.org/pdf/gci-2016-2017/WEF​ London: Routledge. _GCI_2016_2017_Profile_RWA.pdf. Zeitlin, A., C. Leaver, O. Ozier, and P. Serneels. World Food Programme. 2016. “Report Finds 2017. “Re cr u it ment , Mot ivat ion , a nd Malnutrition Rates in Rwanda Have Improved, Retention Effects of Teacher Performance but More Work Is Still Needed.” World Food Pay.” S-38408-RWA-1, International Growth Programme News, April 5. Centre, London. 2 Transformation through Trade: Using Exports and Regional Integration to Drive Future Growth Introduction growth through structural transformation. Moving labor from low-productivity jobs, To achieve its aspiration of becoming a high- mainly in agriculture, to higher-productivity income country by 2050, Rwanda will have jobs in a range of mostly urban activities to accelerate the growth of trade. Inflows is imperative for growth. East Asia made of development assistance have financed a this transition to high growth by relying on large share of investment and powered gross labor-intensive manufacturing for export. domestic product (GDP) growth in the past Rwanda, as with much of Africa, will require two decades, but they are likely to attenu- not only labor-intensive manufacturing but ate in the coming two decades as Rwanda also agribusiness, horticulture, and selected progresses toward middle-income status. services—what some have called “industries Trade will become an increasingly impor- without smokestacks” (Newfarmer, Page, tant driver of growth. Exports will provide and Tarp, forthcoming). These activities, foreign exchange to purchase much-needed taken together, hold the promise of doing for investment in equipment, high-technology Rwanda what manufacturing did for East goods, intermediate components, and prod- Asia in the 1990s. They are labor intensive uct varieties and will foster productivity by and tradable, and they have high value added allowing firms to exploit increasingly large per worker. As in traditional manufacturing, economies of scale. Increased import capac- technological change is rapid and can spawn ity will facilitate access to high-technology rapid productivity growth; some exhibit scale goods and foster competition that drives and agglomeration economies (Ebling and productivity. Trade expansion also implies Janz 1999; Ghani and Kharas 2010). In a the need to attract foreign direct investment world of recent technological revolutions in (FDI), since multinational companies bring information and communication technol- in managerial, technical, and design skills at ogy (ICT), manufacturing techniques, and the same time that their networks facilitate global value chains (Hallward-Driemeier and access to new export markets (Freund and Nayyar 2017), Rwanda has an opportunity Moran 2017). to leverage greater integration into regional Trade expansion is central to creating and global markets and to propel structural new, higher-productivity jobs that facilitate transformation and growth (Nayyar 2017). 91 92 FUTURE DRIVERS OF GROWTH IN RWANDA On the basis of a comparison with coun- capabilities, trading with the region can tries experiencing rates of economic growth serve as a vital training ground for learning similar to those required for Rwanda to to export and produce higher-quality goods. achieve upper-middle-income status by It also can stimulate productivity in priority 2035, exports as a share of GDP will have sectors (notably services) through competitive to increase significantly. Rwanda will need pressures and takeovers from leading firms in double-digit year-on-year export growth for the region. every year up to 2035. Meeting such a target To assist the government in meeting its will be difficult but not impossible. Efforts medium-term trade objectives, this chapter begin with recognition that the share of highlights five major priorities, together with exports in national income is still low relative a subset of policy options, that should figure to that of other countries with comparable prominently in a comprehensive program per capita income levels; it is low even when using trade to accelerate and sustain growth: compared with that of other landlocked countries. Rwanda cannot achieve its growth 1. Harness the EAC and SADC as a plat- aspirations without a major export push. form for transformation by aligning Meeting Rwanda’s export objectives regional incentives, harmonizing stan- requires a comprehensive trade policy dards, and exploiting economies of scale. that spans services, industry, and agricul- Accomplishing this will require assertive ture. Whereas other fast-growing countries negotiations to reform the common exter- (notably in East Asia) relied primarily on nal tariff (CET) and the disadvantages it export-oriented manufacturing to drive up imposes on Rwandan producers and con- productivity and absorb low-skill work- sumers, to remove nontariff barriers that ers, Rwanda still faces significant barriers restrict Rwanda’s exports, and to harmo- to becoming globally competitive in manu- nize competing tax incentives that erode facturing because of its remote location the tax base while providing minimal eco- and detachment from global supply chains. nomic benefit. Maintaining a competitive Analysis from this chapter suggests that no value for the Rwandan franc, especially one sector will be able to drive the necessary relative to neighboring currencies, is criti- export and employment growth on its own. cal to export competitiveness. Rwanda will instead have to produce high- 2. Improve trade connectivity by lower- quality products for the region (especially in ing transport costs within and across the manufacturing and agroprocessing) and to region. Recent progress in lowering trans- develop other sectors that are similarly trad- port costs that put Rwanda at a significant able and productive, but less dependent on disadvantage relative to competitors in the location (such as tourism, professional ser- region has already stimulated exports, but vices, and ICT). The government of Rwanda more can be done to reduce delays at bor- will thus have to build further on the service ders and in transit and to form a logistics sector (the largest current source of exports), hub in Kigali. No less important is aggres- strongly accelerate industrial growth, and sive pursuit of open skies agreements that expand into other high-value agricultural would pave the way for RwandAir to export goods (such as horticulture). expand into new markets. Regional integration offers a key oppor- 3. Increase service sector productivity, both tunity to stimulate transformation. Trading as a critical input to other priority sec- with neighboring countries, particularly the tors and as a source of exports. Rwanda East African Community (EAC) and the already is a services-exporting economy— Democratic Republic of Congo, offers sig- some 50 percent of foreign exchange earn- nificant opportunities for Rwanda. Although ings are from tourism and other services the current volume of direct trade is still exports. Accelerating their development limited because of comparable production requires promoting regional competition TR ANSFORMATION THROUGH TR ADE 93 by reducing obstacles to investments in in the investment climate, and establishing services, harmonizing regulations and tax an integrated set of strategies to address both policies that otherwise hamper regional general and sector-specific challenges. competition, and seizing opportunities to Joining the EAC helped to bring down develop new services exports. Permitting tariffs, while also spurring substantial easy access to imported skills through reductions in transport costs and time spent liberalization of professional services can at border crossings. The accession into the remedy near-term skills deficits and pro- EAC’s CET reduced average tariff rates mote skills upgrading and learning for from 16.5 to 11 percent, which strongly Rwandans. benefitted intraregional trade, especially 4. Stimulate foreign and domestic invest- with Tanzania and Uganda.1 Regional inte- ment into tradable sectors by using selec- gration also enabled greater cooperation tive time-bound and performance-driven on trade facilitation along East Africa’s incentives. High priorities include map- two trade corridors: the Northern Corridor ping tax incentives to ensure that they (Mombasa) and the Central Corridor (Dar conform to overall development objec- es Salaam), which has significantly reduced tives and then monitoring performance of the overall cost and time to move goods to beneficiaries to ensure that commitments port (see priority 2). are fulfilled. They also include target- The investment climate has improved ing investment promotion efforts more considerably, largely through establishment effectively and improving investor after- of the Rwanda Development Board (RDB). care once foreign companies set up in The government has focused extensively on Rwanda. improving and streamlining the country’s 5. Accelerate industrialization through business environment. The extent of this diversification, value addition, and qual- progress is notable, as is evident in the World ity upgrading. Ongoing efforts to add Bank’s Doing Business rankings, where value to coffee production could be com- Rwanda moved from 158 in the world in plemented through more aggressive efforts 2007 to 29 in 2019 (figure 2.1), considerably in mining (see chapter 4 of this report). higher than the regional average (table 2.1). A supplier development program could increase local value added by linking large FIGURE 2.1  Rwanda’s Ease of Doing Business exporters to local sources. Ranking, 2007–19 With the adoption of these five priorities, 180 the government will be in a very strong posi- tion to raise exports and use trade to cre- 150 ate productive jobs and foster the structural transformation necessary to drive growth. 120 Ranking 90 Setting the Stage: Policy, Performance, and Challenges 60 Recent Developments in Trade Policy 30 Rwanda has progressively improved its policy framework governing trade. Since 0 2000, the government has adopted a wide 07 08 09 10 11 12 13 14 15 16 17 18 19 range of reforms to stimulate export growth. 20 20 20 20 20 20 20 20 20 20 20 20 20 These reforms include joining the EAC and Source: Doing Business database (World Bank, various years). COMESA, making significant improvements Note: Out of 190 countries. 94 FUTURE DRIVERS OF GROWTH IN RWANDA TABLE 2.1  Ease of Doing Business 2019 ranking Recent Export Performance for Rwanda and average for the East African Community Rwanda’s policy framework has enabled it EAC to diversify exports while raising the value Indicator Rwanda average of traditional commodity exports. Rwanda Ease of doing business rank 29 119 succeeded in growing its exports by about Starting a business 51 116 20 percent annually from 2000 to 2014, Dealing with construction permits 106 143 albeit from a low base. By 2016, total exports Getting electricity 68 129 Registering property 2 112 of all goods and services reached 20.1 per- Getting credit 3 83 cent of GDP. A comparison between 2005 Protecting minority investors 14 96 and 2016 (table 2.2) shows that the country’s Paying taxes 35 97 exports have become less dependent on the Trading across borders 88 142 three traditional exports: tea, coffee, and Enforcing contracts 78 91 Resolving insolvency 58 110 minerals. Although the share of these tradi- tional exports dropped from 41 to 25 per- Source: Doing Business (World Bank, 2019). Note: Out of 190 countries. Data cover the period June 2, 2017–May 1, cent, their overall value rose from US$126 2018. EAC = East African Community. million to US$415 million. Trade policy that focuses on adding value has limited the impact of international com- Improvements have been particularly evident modity price fluctuations on exports. This since establishment of the RDB. Registration connection is exemplified by the government’s and aftercare have improved significantly, coffee strategy. Through active engagement, company registration can now be completed the average farm gate price of coffee cherries within a few hours, and all permits and doc- grew from FRW 50 per kilogram in 2002 to uments can be obtained at RDB’s one-stop FRW 250 per kilogram in 2015 (box 2.1). shop. Investment promotion activities have These efforts also helped to offset drops in also become more plentiful and targeted to international prices, which fell between 2011 priority sectors, while private sector con- and 2015 and only recently stabilized. By cerns are better addressed across the wider continuing to raise value added, Rwanda is government. As a result, Rwanda now ranks laying a foundation for resurgent earnings second in investment ­ climate in Sub-Saharan once commodity prices rebound. Africa, after Mauritius. Rwanda is rapidly becoming a service- The government has adopted a broad exporting economy. Today, half of Rwanda’s trade policy framework of general and total export earnings come from services sector-specific strategies that directs atten- ­ (table 2.2). In 2016, for example, tourism tion to high-priority areas. Recent years have constituted 23 percent of exports, while seen the development of a large set of policy transport, ICT, construction, and finance documents, offering both general and sector-­ jointly accounted for another 11 percent. specific plans. These policies tend to align This does not count services associated closely with regard to both recommended with the large volume of reexports. For ser- policy instruments and priority sectors. vices to become a driver of growth in the Through extensive private sector consulta- future requires high-level attention to effi- tions, some of these sectoral policies have ciency (a point to which the chapter returns been particularly effective (illustrated later below). in box 2.1 for coffee exports). The integrated Rwanda’s (nonmineral) industrial sec- nature of this policy environment also means tor has also started exporting, particu- that it is increasingly providing guidance for larly for agroprocessing. Since 2005, light various ministries (see, for example, box 2.4 manufactu ring has remained roughly on the cross-governmental strategy for tex- constant in export share (7 percent), but tiles, apparel, and leather). export value has increased from US$23 TR ANSFORMATION THROUGH TR ADE 95 TABLE 2.2  Rwandan exports, by sector, 2005 and 2016 2005 2016 Sector and subsector US$ (millions) Share (%) US$ (millions) Share (%) Services Tourism 49 16 390 23 Government goods and services 38 12 254 15 Transport 33 11 91 5 Information and communication technology 0 0 41 2 Construction 0 0 37 2 Financial services 0 0 28 2 Maintenance and repairs 0 0 7 0 Total services 120 39 848 50 Industry Minerals 65 21 282 17 Light manufacturing 23 7 94 6 Agroprocessing 1 0 72 4 Total industry 88 28 447 27 Agriculture Coffee and tea 61 20 133 8 Other agriculture 0 0 32 2 Horticulture 1 0 9 1 Total agriculture 62 20 174 10 Reexports 41 13 216 13 Total 311 100 1,685 100 Source: Calculations based on Comtrade data (United Nations, various years). BOX 2.1  Transformation of the coffee sector In the late 1990s, Rwanda’s government closely beans from volume to quality incentivized farmers to controlled the coffee sector, dictating a single coffee raise the value of their coffee (processing it in a wash- price for the entire season and providing no incen- ing station). The sector was also liberalized to attract tive for farmers to upgrade to higher-value coffee new washing stations and farmer associations. processing. With little investments, the existing The effect of these policies was dramatic. The stock of coffee trees aged, soil fertility declined, and number of washing stations increased from 2 in 2002 crops were weakened by insects and fungal diseases. to 245 in 2015. This expansion increased exports of By 2000, coffee production had declined in volume, fully washed coffee from less than 1 percent in 2002 to and 90 percent of Rwanda’s crop was classified as 21 percent in 2010 and up to 50 percent in 2015. The low-quality, “ordinary” coffee. rise in quality increased the average farm gate price To break out of this “low-quantity, low-quality” of coffee cherries from about FRW 50 per ­ kilogram trap, the National Coffee Strategy was adopted in in 2002, to FRW 200 per kilogram in 2010, and to 2002. The strategy fundamentally restructured the FRW 250 per kilogram in 2015 (figure B2.1.1). This coffee sector by drastically increasing the share of focus on value addition was the main reason why high-quality (“fully washed”) coffee from 1 percent coffee exports rose from US$14.6 million in 2002 to in 2002 to 60 percent in 2006. Shifting the pricing of more than US$62 million in 2015. (Box continues next page) 96 FUTURE DRIVERS OF GROWTH IN RWANDA BOX 2.1  (continued) FIGURE B2.1.1  Number of coffee-washing stations and average price of coffee cherries in Rwanda, 2002–15 300 300 250 National average farm gate price of Number of co ee-washing stations 250 co ee cherries (FRW/kg) 200 200 150 150 100 100 50 0 50 02 11 05 08 14 20 20 20 20 20 Number of CWFs Average farm gate price of cherries, FRW/kg Note: CWF = coffee-washing station; FRW/kg = Rwanda francs per kilogram. Source: Morjaria and Steenbergen 2017. FIGURE 2.2  Value of Rwandan exports, by destination, 2001–16 million to US$94 million (table 2.2). This reflects a rise in exports for some new 700 industries, including apparel and leather products, mechanical appliances, and bev- 600 Rwanda enters EAC erages. Rwanda has also started exporting agroprocessing goods, where key new sec- 500 tors such as grain milling and processed Millions of US$ 400 animal products now account for 4 percent of overall exports. 300 Regional markets have expanded consid- erably. After joining the EAC customs union 200 in 2009, Rwanda experienced a consider- 100 able rise in intraregional exports of goods (figure 2.2). Better access to the ports of 0 Mombasa and Dar es Salaam also contrib- uted to large increases in exports. Intra-EAC 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 East African Community Congo, Dem. Rep. Rest of the world trade offers a particularly important mar- ket for nontraditional exports. In 2014, the Source: Calculations based on Comtrade data (United Nations, various years). EAC accounted for 41 percent of manufac- Note: EAC = East African Community. turing goods, 66 percent of leather goods, TR ANSFORMATION THROUGH TR ADE 97 and 53 percent of horticultural products sold FIGURE 2.3  Exports of Rwandan goods, by destination, 2016 abroad (MINICOM 2015). Non-EAC neigh- boring markets have also become important. 35 32 Although conflict in the Democratic Republic 30 of Congo limited trade prior to 2007, exports have grown considerably in the last decade. 25 25 By 2016, Rwanda exported more goods to the Democratic Republic of Congo than to 20 Percent the EAC (figure 2.3). The main exports to the 16 15 14 Democratic Republic of Congo include live- stock and crops, but cross-border (informal) 10 8 trade in services such as finance, transporta- tion, and wholesale trading is also significant 5 3 2 (Lalui 2016). 0 . ity n t ia ica ca ep as io As fri un er eE .R Un Remaining Challenges rA Am m m dl an he m De id rth pe Co Ot M o, ro No an Export-oriented industrialization is grow- ng Eu ric Co ing slowly given Rwanda’s growth aspira- Af st Ea tions. The share of Rwanda’s goods and services exports coming from nonmineral Source: Calculations based on Comtrade data (United Nations, various years). industrial products has fluctuated in recent years, moving up from 5 percent in 2006 to FIGURE 2.4  Agroprocessing and other manufacturing as a share of 11 percent in 2012 and down again to 9 per- total exports in Rwanda, 2005–16 cent in 2016 (figure 2.4). Whereas agropro- cessing exports have been dynamic (rising 20 from almost nothing in 2006 to 4 percent of all exports in 2016), export performance in other manufacturing areas has been dis- 15 appointing. These exports recently declined from 8 ­percent of total exports (2012–14) to 8 8 8 11 6 Percent 5 percent in 2016. 10 5 Direct export opportunities within the 7 EAC are constrained by comparable produc- 6 6 6 tion capabilities. Although further regional 4 5 5 4 4 4 4 3 integration brings considerable trade ben- efits, these countries all produce similar 2 0 1 1 0 0 1 goods, as illustrated in figure 2.5, which 0 ranks each country’s revealed comparative 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 advantage in 2013 and displays Rwanda’s Agroprocessing exports share top six products. All EAC countries have Other manufacturing exports share vegetables as either first- or second-ranked Source: Calculations based on Comtrade data (United Nations, various years). revealed comparative advantage, all have food products in their top six products, and all besides Tanzania specialize in hides economies) and by pushing for greater spe- and skin products (such as leather). This cialization (particularly with regard to prod- duplication limits the potential of direct uct quality). intraregional trade. Trade opportunities Export-oriented industrialization is grow- can be created for Rwanda by embracing ing slowly, while direct export opportu- such similarities (developing regional scale nities within the EAC are constrained by 98 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE 2.5  Ranking of revealed comparative advantage for East African Community countries, 2013 14 12 12 11 11 10 8 8 Ranking 6 6 6 6 6 5 5 4 4 4 4 4 3 3 3 3 2 2 2 2 1 1 1 0 Minerals Vegetables Hides and skins Raw materials Consumer goods Food products Rwanda Uganda Kenya Tanzania Source: Calculations based on World Integrated Trade Solution data (WTO 2018). comparable production capacity. Commodity of some higher-profit markets. It placed exports have grown considerably, but growth additional pressure on the development of has been undermined by recent global prices. nontraditional exports to improve the trade The export sector is dominated by a small balance, and these products have grown rap- number of firms, low productivity plagues idly in recent years. The share of traditional sectors where many people work, and more commodities fell from about 40 percent of private investment is needed in tradable all exports (including services) in 2005 to sectors like tourism and manufacturing. only 25 percent by 2016. Rwanda, despite recent export growth, still Over the period 20 09 –16, the top faces large structural challenges. 5  ­p e r c e n t of e x p o r t e r s i n Rw a n d a International commodity prices have accounted for more than 80 percent of the undermined export performance, but also total value of Rwandan exports, and the increased the pace of diversification. The top 1 percent of exporters accounted for Rwandan export sector historically has more than 40 percent of total export value been concentrated in a few commodities, (figure 2.7). Although such a highly skewed including coffee, tea, and minerals. This distribution is not unusual for countries concentration was beneficial when interna- at similar income levels, it has remained tional prices were high and rising, as was relatively constant over the last decade. the case up to 2011. However, since 2011, On average, these Rwandan firms export export volumes have continued to rise at only two products to 1.5 destinations. a steady pace, but commodity prices have Similarly, about 59 percent of all exporting stagnated or fallen, particularly for min- firms in 2016 sold only a single product to erals (figure 2.6). Mineral goods are still a single destination (accounting for about exported mainly without processing, which one-third of total export value). Rwanda’s has made Rwanda particularly vulnerable to export sector is much less diversified than price drops. This vulnerability was due to those of its regional peers, with exporting the global slowdown in demand from China firms in Uganda (3.6 products), Tanzania as well as restrictive legislation passed in (4.2 products), and Kenya (7.2 products) all the United States that pushed Rwanda out exporting considerably more products on TR ANSFORMATION THROUGH TR ADE 99 average (World Bank 2017). Concentration FIGURE 2.6  Value and volume index of Rwandan commodity of the type and destination of products may exports, 2000–15 also be contributing to Rwanda’s low rate of export firm survival (Jaud and Freund 700 2015). Rwanda’s exporter survival rates are low. 600 Only 30 percent of new exporters in Rwanda Index (year 2000 = 100) 500 are expected to be exporting products in the following year. This is comparable to the 400 level in Tanzania (31 percent), but lower than in Kenya (34 percent), Uganda (36 percent), 300 and especially Ethiopia (51 percent). Many 200 Rwandan exporters are small traders mov- ing goods across the border with neighboring 100 countries, particularly Democratic Republic of Congo, rather than producing goods them- 0 selves. Larger exporters have a much higher 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 survival rate. Moreover, firm survival is also Value index of Volume index of higher when a larger number of firms export commodity exports commodity exports the same product. Other, cross-firm export Source: Calculations based on Comtrade data (United Nations, various years). synergies may further explain this dynamic (World Bank 2017). The diversification of export goods has been driven by the sale of Rwanda’s cur- FIGURE 2.7  Share of export value in Rwanda, 2009–16 rent products in more national markets. Decomposing export grow th over the 100 period 2008–13 shows that expanding the 90 86 89 89 87 85 85 81 sale of existing products into new national 80 78 markets accounted for fully 61 percent of % of total export value 70 export growth (table 2.3). Selling existing 60 54 57 products to markets where Rwanda had a 50 44 46 45 45 presence prior to 2008 accounted for some 41 40 36 20 percent of export growth in this period. 30 However, only 19 percent of export growth 20 was due to the introduction of new prod- 10 ucts, mainly in markets where Rwanda 0 had a presence—notably regional mar- 2009 2010 2011 2012 2013 2014 2015 2016 kets. This underscores the importance of Share of top 1% exporters Share of top 5% exporters regional markets as a platform for learning to export. Source: Adjusted from World Bank 2017. Although labor productivity has improved, it has been particularly limited in the service sector. Average output per worker in Rwanda TABLE 2.3  Decomposition of merchandise increased by about 6 percentage points export growth, 2008–13 between 2005 and 2014. About two-thirds Products of this increase came from shifts in labor Market Existing New from low- to high-productivity areas, while the other one-third came from increases Existing 20 19 in within-sector productivity. Sectors with New 61 <1 the highest growth in labor productivity Source: MINICOM 2015. 100 FUTURE DRIVERS OF GROWTH IN RWANDA include finance and agriculture (figure 2.8). Low productivity in services undermines productivity declined in service sectors Yet, ­ a vital input for exporters. Services play an such as hotels and restaurants, construction, important role as intermediate inputs for and utilities. Other sectors like transport other sectors in the economy. Firms need a and ICT improved, but still have relatively variety of backbone services (such as logis- low productivity compared with the region tics, finance, and telecommunications), and (World Bank 2017). these needs are often more pronounced for higher-quality exporting firms. For that rea- FIGURE 2.8  Labor productivity in Rwanda, 2005 and 2014 son, Hoekman and Shepherd (2015) found that a 10 percent improvement in services 12 productivity increases merchandise exports of EAC countries by almost 0.5 percent. This 10 9.7 low productivity is driven in part by a policy (constant 2014 prices, millions) Average output per worker 8.3 environment that restricts trade in services. 8 7.1 If the EAC were to lower the restrictions on 6.1 trade in services to the level in Ghana (the 6 5.2 5.2 African country with the lowest trade barri- 4 ers for services, with an index of 18), exports 2.9 2.4 2.5 2.5 2.1 of EAC countries could increase substan- 1.7 1.8 1.7 2 1.3 1.4 tially: by 13 percent for Rwanda and some 0.3 0.4 20 percent for Kenya, Tanzania, and Uganda 0 (Hoekman and Shepherd 2015). es ies g T ts g es n re More private investment is needed in the IC rin in io an tu vic vic ilit ct in t, tu ul ur er er or tru M Ut tradable sectors. Private finance has expe- ric ac sta ls rs sp ns uf Ag cia he re an Co an rienced strong growth over the last decade, Ot an ls, Tr M te Fin Ho with private credit nearly tripling between 2005 2014 2000 (10 percent of GDP) and 2016 (28 per- cent of GDP). However, most of this private Source: Adjusted from World Bank 2017. Note: ICT = information and communication technology. finance has gone to sectors with little to no trade, such as construction and real estate, and to households (microfinance). By con- FIGURE 2.9  Rwanda’s financing stock, by sector, 2006–15 trast, in 2015 only 12 percent of the stock of private finance was in manufacturing and 3,500 18 percent was in tourism (figure 2.9). Given the broader aspirations of the government for 3,000 private-led transformation, more needs to be done to attract additional FDI and domes- Rwanda francs (billions) 2,500 tic private capital into tradable sectors. This 2,000 issue is discussed further later in the chapter. 1,500 Envisioning the Future 1,000 To meet its aspirations to become an 500 upper-middle-income economy by 2035, Rwanda will have to accelerate export 0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 growth significantly. On the basis of a comparison of Rwanda with countries Tourism Infrastructure Financial services Manufacturing Household Construction and real estate experiencing similar rates of economic growth, exports as a share of GDP will Source: Calabrese, Papadavid, and Tyson 2017. have to rise considerably for Rwanda to TR ANSFORMATION THROUGH TR ADE 101 meet upper-middle-income-­ country status. FIGURE 2.10  Exports as a share of GDP and log GDP per capita in Compared with upper-middle-income coun- Rwanda, 2000–16 tries (­f igure 2.10), the spread of exports- to-GDP ratios is large, but ratios tend 125 to be upward of 40 ­ p ercent. Since 2000, Exports of goods and services (% of GDP) Rwanda has consistently increased its share 100 of exports in national income, but the share is still far below the average for upper-­ middle-income countries. Therefore, signifi- 75 cant export opportunities are available to Rwanda, and exports still have ample room 50 Rwanda, 2007 Rwanda, 2016 to grow. Rwanda, 2000 To meet these objectives, Rwanda will 25 Upper-middle-income have to achieve double-digit rates of year- countries on-year export growth up to 2035. Although not impossible and in line with the growth 0 rates achieved between 2005 and 2016, 4 6 8 10 12 such growth implies sustaining consistently Log GDP per capita high export growth for nearly two decades. Historically, only a few countries have Source: Calculations based on Comtrade data (United Nations, various years). achieved this feat. In developing an export-oriented growth income rises around the world, the share of strategy, Rwanda will have to make strate- agricultural products in total expenditure gic decisions regarding which sectors have will decline, whereas global consumption the greatest potential and merit additional of other products (manufactured products, investment. These decisions include consid- services) will rise (Szirmai 2012). This sec- ering not only the type of explicit incentives tion focuses mainly on manufacturing and provided to sectors but also the wider sup- services for exploring medium-term export port structure, including the regulations, opportunities. skills training, and investment priorities that the government of Rwanda will have to set out in the medium term. The focus here is The Potential of Manufacturing Exports not to highlight specific products or types Most countries that recently achieved high of firms to support but rather to identify growth did so through export-oriented the activities that are likely to contribute to manufacturing. In East Asian countries, growth and job creation as part of Rwanda’s export-oriented manufacturing played a development strategy. critical role in boosting productivity growth Although significant opportunities exist (Newman et al. 2016). It did so in part by for increasing agricultural exports (as noted being export oriented, which strongly helped in chapter 5), transforming m ­ anufacturing to increase the within-sector productivity of and services will be even more critical in the manufacturing firms. Having to compete in medium term. Countries s ­ pecializing in agri- the international arena meant that firms had culture are often more vulnerable to price vol- to accelerate their learning process to remain atility in international markets. Agriculture competitive: firms that did not adjust went also tends to have lower ­ average labor under, freeing up labor and resources for productivity than other sectors, whereas ­ other companies. Greater interaction with productivity improvements are often closely other international actors also helped to linked with labor-saving technologies (lim- absorb foreign technology, while larger mar- iting the potential for long-term growth of ket size allowed firms to exploit economies labor productivity). In addition, as per capita of scale, further improving labor productivity 102 FUTURE DRIVERS OF GROWTH IN RWANDA (Herrendorf, Rogerson, and Valentinyi 2013). countries,3 Hallward-Driemeier and Nayyar Manufacturing also stimulated across-sector (2017) identified indicators for each of the productivity improvements by absorbing four pro-development characteristics and high levels of labor from low-productivity then classified sectors as “high,” “medium,” agriculture into higher-productivity modern or “low” in each category. To relate these industries (structural transformation). This to Rwanda’s current resource endowments, was made possible because manufacturing table 2.4 color-codes them for their potential was labor intensive (with manageable invest- contribution to growth in Rwandan exports ments in physical capital), while offering sig- and development (green as greatest potential, nificant opportunities for low-skill workers, blue as medium potential, and orange as cur- who could easily transition from agriculture rent potential). into factory employment without much skills The manufacturing sectors can be separated training (World Bank 1993). Recent evidence into different clusters (Hallward-Driemeier thus suggests that high productivity growth and Nayyar 2017), each with different poten- in the manufacturing sector explains about tial for Rwanda’s long-term growth. 50 percent of the catch-up in relative aggre- Commodity-based regional processing gate productivity across countries (Duarte sectors are the most immediate manufactur- and Restuccia 2010). ing priority for Rwanda. These sectors are Four components historically made man- linked closely with the use of agricultural ufacturing a major driver of development raw materials (for example, food processing) (Hallward-Driemeier and Nayyar 2017): and so can stimulate backward links with other sectors (for example, agriculture). Their 1. Tradedness. Manufacturing’s export ori- potential is derived from their high scope to entation helped to improve the potential employ unskilled workers and relatively low size of the sector, while also stimulating intensity of physical capital and R&D. These within-sector productivity (for example, goods typically are traded regionally (rather through competition, technology absorp- than internationally) because they are bulky tion, and scale economies). to transport (for example, wood products, 2. Labor productivity. The sector offered beverages) or they require proximity to raw opportunities for relatively high and materials (for example, food products). These growing output per worker over time. sectors likely have the greatest long-term 3. Scope to employ unskilled workers. The potential for Rwanda and are the only sub- sector’s ability to absorb a large number sector in which it has a revealed comparative of low-skill workers, at substantial pro- advantage (Hallward-Driemeier and Nayyar ductivity premiums, helped to promote 2017). Their prospects depend in part on an structural transformation. integrated and competitive regional market 4. Capital and research and development (see priority 1) and a focus on quality upgrad- (R&D) intensity. Modest requirements ing for Rwanda’s producers (see priority 5). for initial physical capital and R&D Low-skill labor-intensive tradables have meant that manufacturing had relatively strong pro-development characteristics, but it low barriers to entry. A gradual increase will take time for Rwanda to attain a com- in capital requirements, in line with more petitive edge in these areas. Sectors such as advanced production of manufactured textiles, garments, and leather are the only goods, helped to increase labor productiv- type of manufacturing that combines a high ity over time. degree of international trade with a high share of low-skill employment (Hallward- Yet not all manufacturing sectors exhibit Driemeier and Nayyar 2017). For this reason, all of these components. By combining an these sectors are often the highest priorities international data set 2 with firm-level data for low- and middle-income countries seek- from World Bank Enterprise Surveys for six ing to stimulate structural transformation. TR ANSFORMATION THROUGH TR ADE 103 TABLE 2.4  Manufacturing subsectors, grouped by pro-development characteristics Labor Scope to employ Capital and R&D Manufacturing sector Tradedness productivity unskilled workers intensity Commodity-based regional processing       Food, beverages, and tobacco products Medium (mainly Medium Medium Low Wood and wood products regional) High Other nonmetallic mineral products Fabricated metal Paper and paper products; printing and publishing Medium Rubber and plastics products Basic metals Low-skill labor-intensive tradables Textiles, wearing apparel, and leather products High Medium High Low Furniture; manufacturing n.e.c. (not specified) Capital-intensive processing Coke and refined petroleum products Medium (mainly High Low High Chemicals and chemical products regional) Medium-skill global innovators         Machinery and equipment n.e.c. (not specified) High Medium Medium High Transport equipment Electrical machinery and equipment High High-skill global innovators Computer, electronics, and optical equipment High High Low High Pharmaceutical products Source: Based on Hallward-Driemeier and Nayyar 2017. Note: R&D = research and development. Color-coding is as follows: green = greatest potential, blue = medium potential, and orange = current potential. However, because of their global export ori- TABLE 2.5  Labor costs in manufacturing in entation, efficiency-seeking investors mainly select countries select a location on the basis of the lowest Labor cost per worker in Country Years covered manufacturing (US$ 2010) possible input, production, and transport costs (rather than proximity to markets or Bangladesh 2007–13 835 Ethiopia 2011–15 909 natural resources) (Andersen, Kett, and von Rwanda 2011–15 1,505 Uexkull 2017). So far, this tendency has lim- Tanzania 2006–13 1,777 ited the potential of these sectors in Rwanda. Kenya 2007–13 2,118 Its landlocked location results in high trade Sources: Gelb et al. 2017. For Rwanda, calculations based on Rwanda costs. Even though Rwanda appears to have Revenue Authority pay-as-you-earn tax data. a cost advantage within the EAC, its labor costs are high compared with those of global The opportunities from other manu- leaders such as Bangladesh and, increasingly, facturing sectors are less straightforward, Ethiopia (table 2.5). Rwanda can only offset held back by higher skill, capital, and R&D these disadvantages in the short term through requirements. Mineral processing reflects higher productivity. In the medium term, one opportunity for exporting to the global lower transport costs, programs to improve market. However, as shown in table 2.6, this worker productivity, and access to growing sector will require considerable initial capi- neighboring markets could open the door for tal investment and tends to employ relatively more efficiency-seeking investment. few (low-skill) workers. Key for structural 104 FUTURE DRIVERS OF GROWTH IN RWANDA TABLE 2.6  Service subsectors in Rwanda, grouped by pro-development characteristics Labor Scope to employ Capital and R&D Service subsector Tradedness productivity unskilled workers intensity Low-skill services         Construction Low Low High Low Hotels and restaurants Wholesale and retail Medium Medium Medium Transport services (regional) Tourism High High-skill services         Professional Medium High Low Medium Financial High Information and telecommunication technology Source: Based on Hallward-Driemeier and Nayyar 2017. Note: Color-coding is as follows: green = greatest potential, blue = medium potential, and orange = current potential. transformation in the medium term will be variety will drive an increasingly diverse the extent to which the mineral sector creates availability of product lines on store shelves. positive spillovers into other manufacturing Much as it has in the case of the European industries.4 Opportunities in other sectors integration, this process of country special- are more remote, depending on sizable capi- ization in production and diversification in tal investment (for example, automation and consumption can drive productivity gains robotics for transport manufacturing) or on throughout the region. cutting-edge R&D (for example, for com- puter equipment). Entry into these sectors is The Potential of Services Exports likely to be shaped by unforeseen advances in production technology interacting with the Services are increasingly embedded in other pace of skills development in Rwanda. products (industry and agriculture) and In several of these activities, intraregional constitute a growing share of value added trade can play a pivotal role. Over time, in production processes. Many services are regional trade offers two avenues for growth. “embodied” in both manufacturing and The first is participation in regional value agricultural production, either as inputs (for chains, often connected to global markets. example, design, marketing, or distribution) Coffee is one example: Rwanda exports cof- or as enablers for trade to take place (such as fee cherries to roasters in the global markets. logistics services or e-commerce platforms). Similarly, Rwanda already is part of regional For that reason, more than one-third of the value chains linking manufacturing centers global value of gross manufactures exports in Kenya and Tanzania with Burundi and comes from the value added of embodied ser- the Democratic Republic of Congo. The sec- vices, which has risen considerably in recent ond avenue is through intraindustry trade in decades (Baines, Lightfoot, and Smart 2011; manufactures, including, for example, dif- Bamber et al. 2017). Services play a similarly ferentiated consumer products. This avenue important role in manufacturing in the EAC relies on unfettered access to the relatively (Hoekman and Shepherd 2015). large EAC market and on the development The increased “servicification” of manu- of firms that can operate productively at facturing reflects the growing interdepen- scale. Competition will drive the allocation dence of the two sectors and may limit the of production across the region and national extent to which either sector can grow specializations, while consumer desires for independently of the other. Rather than TR ANSFORMATION THROUGH TR ADE 105 prioritizing either the manufacturing or the with substantial investments in the Kigali service sector over the other, recent examples Convention Center, RwandAir, and tour- suggest that they may be interdependent. For ism assets, such as Akagera and Nyungwe example, in China from 2000 to 2014, ser- National Parks. These laudable efforts vices input into manufacturing accounted for warrant even more attention. However, 38 percent of annual growth in services value tourism alone will not be enough to meet added, while manufacturing input into ser- the export targets. Even under optimistic vices accounted for 30 percent. Stimulating growth assumptions, tourism can create service sector productivity is thus a critical only a portion of the higher-productiv- short-term priority that will assist in driv- ity jobs needed; additional efforts are ing labor productivity and export growth in required to expand services exports other other sectors (see priority 3). than tourism services as well. Selected services also share the pro-­ Opportunities for other low-skill services development characteristics traditionally are more limited, but they exist in retail and associated with the manufacturing sector transport services. The degree of traded- and should be given similar high priority. ness is critical for low-skill sectors, which The analysis of pro-development charac- often provide relatively little productivity teristics can also be extended to the service growth, as illustrated by Baumol’s (1967) sector (table 2.6). This reflects the increased “cost disease.”5 To improve their labor pro- tradability of services that has been made ductivity requires capital accumulation, possible in recent decades. Because of the rev- innovation, or economies of scale, all of olution in ICT, rapidly declining telecommu- which are highly related to ability to trade. nication costs, and expansion of broadband For that reason, some export opportunities Internet services, many services that origi- could be derived from wholesale and retail nally required a direct interaction between trade (especially cross-border trade with the customer and producer are now “unbundled” Democratic Republic of Congo) and from and can be delivered at a distance—within transport services (particularly within the and across borders. This offers another impor- EAC), but other low-skill service sectors tant channel of growth for low- and middle- offer a less direct priority for an export-­ income economies (Loungani and Mishra oriented medium-term strategy. 2014). McMillan, Rodrik, and Sepulveda Developing a high-skill service sector will (2017) found that during the 2000s the major- be difficult, but it is critical to Rwanda’s ity of growth-enhancing structural change in long-term growth. Many of the “criti- Sub-Saharan African was derived from the cal services inputs” for manufacturing fall shift of labor from agriculture into services. under high-skill service sectors (professional However, there are important differences to services, finance, and ICT); globally they consider across service sectors, particularly in also ­constitute a large share of all services their degree of tradedness and their scope to exports. Such services are also particularly employ unskilled workers. critical for Rwanda, because they are less The biggest immediate opportunities for constrained by the tyranny of a landlocked services trade lie with tourism. Table 2.6 geography. Rwanda’s superior regulatory and illustrates why tourism growth has been policy environment may facilitate the devel- so important for Rwanda in recent years— opment of new goods and services exports, it is the clearest example of a service sec- such as exporting standards for certifica- tor that is highly tradable but also offers tion and certification-based products, such significant scope to employ unskilled as fertilizers, seeds, and pharmaceuticals. workers in higher-productivity activi- Similarly, if Rwanda is able to continue its ties. The country has already launched development of higher education, particu- new efforts to accelerate tourism growth larly in specialties close to Rwanda’s natural 106 FUTURE DRIVERS OF GROWTH IN RWANDA comparative advantages, these specialties too The region may also be a springboard to could become an eventual services export. future growth, particularly for fostering the Already Rwanda has attracted top-flight uni- development of new services and product versity subsidiaries, such as Carnegie-Mellon. exports. Already, the great bulk of nontra- A recent grant from the Buffet Foundation ditional goods exports is sold either to the will allow it to build a first-class agriculture Democratic Republic of Congo or to neigh- school. Another logical services export, given boring countries in the EAC. Transport its location, could be mining services. To costs to reach global markets, though still develop mining services will require invest- high, are falling. With a collective GDP ment from foreign companies. some 25 times the size of the Rwandan econ- omy, the region (EAC plus the Democratic Overall Priorities for Export Promotion Republic of Congo) offers the potential for substantial long-term growth of both ser- Achieving Rwanda’s export objectives will vices exports (for example, logistics manage- require a multisectoral approach to stimu- ment) and products exports. Increasingly, late services trade, manufacturing, and agri- new products are likely to be associated cultural exports. No single industry can with regional value chains and exploita- provide the necessary export growth on its tion of economies of scale in sectors such as own. Similarly, important interdependencies agroprocessing. between sectors (most notably in services and Achieving Rwanda’s ambitious growth manufacturing) will prevent any sector from targets will require a governmentwide growing too large without sufficient inputs push. Policies determining export growth from others. For that reason, the most likely extend far beyond the Ministry of Trade scenario for high growth relates to a situation and Industry and the RDB to include sev- where services, industry, and agriculture all eral other ministries. For example, the poli- experience high year-on-year export growth cies of the Ministry of Infrastructure affect up to 2035 (figure 2.11). road infrastructure and the cost of electric- ity, vital for manufacturing exports. The FIGURE 2.11  Cumulative exports as a share of GDP in Rwanda, Ministry of Environment will oversee poli- 2005–16, and potential trajectory to upper-middle-income cies toward the parks, vital to tourism. The status by 2035 Ministry of Finance and Economic Planning oversees tax policy and tax incentives affect- 45 ing the decisions of private companies, 40 including FDI. Moreover, Rwanda’s success 35 in exports is contingent on its negotiations with regional and foreign partners. The need 30 for concerted efforts of the entire economic 25 cabinet becomes clear when considering the Percent following policy priorities for developing 20 exports. 15 Five policy clusters are high priority. The 10 next sections elaborate on the rationale for each priority and offer policy recommenda- 5 tions to assist the government in achieving its 0 medium-term objectives: 07 11 13 15 17 19 21 23 25 27 29 33 05 09 31 35 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Services exports Industry exports Agriculture exports 1. Harness the EAC as platform for trans- Services (target) Industry (target) Agriculture (target) formation, by aligning regional incentives, harmonizing standards, and exploiting Source: Calculations based on Comtrade data (United Nations, various years). economies of scale. TR ANSFORMATION THROUGH TR ADE 107 2. Improve trade connectivity by lower- FIGURE 2.12  Intra-bloc goods imports as a share of GDP before ing transport costs within and across the and after joining the bloc region. 3. Increase service sector productivity, both 18 as a critical input to other priority sectors 15.8 16 and as a source of exports, capitalizing on 14 13.2 regional competition and opportunities. 4. Stimulate foreign and domestic investment 12 Share of GDP (%) into tradable sectors by using selective and 10 8.7 performance-driven incentives. 8 5. Accelerate industrialization through diver- 6.5 sification, value addition, and quality 6 5.1 upgrading. 4 3.0 2.2 2.0 2.1 2 0 Priority 1: Harness the EAC as a EAC SADC ASEAN Platform for Transformation 2 years before 5 years after 10 years after The benefits to trade from regional integra- Source: Shepherd, de Melo, and Sen 2017. tion in the EAC have been limited so far, espe- Note: EAC = East African Community; SADC = Southern African Development Community; ASEAN = Association of Southeast Asian Nations. cially compared with other regional ­ bodies. By removing internal tariff and nontariff barriers, regional integration is expected to has been limited partly because of compara- result in trade creation. Yet a comparison ble production capabilities across EAC part- of intraregional imports as a share of GDP figure 2.5), which has restricted ner states (­ before and after joining the EAC (figure 2.12) the potential of intraregional trade. shows that this effect has been limited for the Rwanda can use the EAC as a platform EAC. After a small increase in the first five for transformation both by specializing years, the EAC’s intra-bloc imports of goods production (particularly in terms of qual- as share of GDP fell to a level below initial ity) and by embracing comparable produc- integration 10 years after membership.6 In tion capabilities (regional scale economies). contrast, other regional bodies experienced The EAC offers a crucial learning ground a considerable increase in intra-bloc trade. for exporting higher-quality products, After 10 years of regional integration, the especially in manufacturing and agropro- SADC saw a tripling of intra-bloc imports, cessing. Entering regional supply chains while the Association of Southeast Asian can help to prepare Rwandan firms to enter Nations doubled intra-bloc imports to almost the global market. The EAC can also be 16 percent of GDP. The overall trade effect used to stimulate within-sector produc- of EAC membership has therefore been some- tivity growth and develop other tradable what disappointing (Shepherd, de Melo, and sectors such as tourism, transport, and pro- Sen 2017). fessional services through greater regional Rwanda has been using the EAC mainly scale economies and greater competition as a destination for conventional exports, from leading firms. limiting its potential benefits. When Rwanda The EAC offers important untapped joined the EAC in 2009, the main impact opportunities for learning to export. The on trade was from the elimination of tariffs EAC is an important market for Rwanda’s from some countries and the adoption of a nontraditional exports, especially in higher- CET. Although the EAC has had some trade-­ skill production. In 2014, for example, it generating effects (figure 2.2), the overall accounted for 41 percent of manufacturing impact on developing Rwanda’s export sector exports, 66 percent of leather goods, and 108 FUTURE DRIVERS OF GROWTH IN RWANDA 53 percent of horticultural products sold a Rwandan firm becomes an exporter, 8 its abroad (MINICOM 2015). However, few average output increases 41 percent and out- Rwandan firms make use of these oppor- put per worker increases 17 percent. This also tunities. Spray and Wolf (2016) found that results in critical spillover for a firm’s domes- Rwanda’s sectors have widely varying partici- tic suppliers, which are found to increase pation rates in the external sector (table 2.7). their output and their output per worker by The mining sector is best integrated into 4 percent simply by being connected to a global supply chains, with 21 percent of new exporter. Increasing the share of export- firms exporting, mostly to Organisation for ers and improving supplier connectivity are Economic Co-operation and Development important mechanisms to increase a sector’s (OECD) countries. Manufacturing shows the average labor productivity. reverse, with 15 percent of firms exporting, Better access to imports will further assist primarily to EAC countries. Agriculture and firm exports and productivity. Imports are agroprocessing as well as services7 have lower typically of higher quality than is available export rates, spread between the EAC and domestically and offer a wider variety of OECD countries. inputs. Faced with high international qual- Comparisons with Uganda suggest that ity requirements, exporting firms tend to rely Rwanda can significantly increase partici- more strongly on imported inputs. This is pation in the export sector. Export partici- particularly pronounced in Rwanda, where pation in Uganda lies at about 20 percent more than 85 percent of all imports flow for both manufacturing and agroprocess- to exporting firms (figure 2.13). Improving ing firms, which is higher than in Rwanda access to imports (through trade facilita- (Spray and Wolf 2016). Crucially for man- tion and tariff exemptions) was an impor- ufacturing, this participation rate was tant channel through which the Kigali achieved by first increasing exports to the Special Economic Zone (KSEZ) helped EAC, which then slowly increased exports firms to improve their average productivity to the OECD. (Steenbergen and Javorcik 2017). This, in Expanding the share of exporting firms turn, helped nascent firms in priority sectors will help to increase labor productivity. Exporting firms face greater competition from other international firms and interact FIGURE 2.13  Total imports of exporting versus with more demanding customers that require nonexporting firms in Rwanda, 2008–14 higher quality. For that reason, they also tend to have higher average labor productiv- 3 ity than firms operating only in the domestic sectors (Melitz 2003; Tybout, de Melo, and Total imports (FRW, billions) Corbo 2015). Spray (2017a) found that, when 2 TABLE 2.7  Rwanda’s labor productivity and participation in the external sector, 2015 % unless otherwise noted Output Exporters 1 per worker Supplying Sector (US$) Total To EAC To OECD exporters Agriculture and agroprocessing 8,166 12 6 8 2 Manufacturing 5,729 15 10 2 14 0 Mining 16,080 21 2 13 7 2008 2009 2010 2011 2012 2013 2014 Services 9,855 5 2 1 9 Exporter Nonexporter Source: Spray and Wolf 2016. Note: EAC = East African Community; OECD = Organisation for Economic Co-operation and Source: Spray 2017a. Development. Note: FRW = Rwanda francs. TR ANSFORMATION THROUGH TR ADE 109 (such as light manufacturing and agropro- relative to Kenya and Tanzania. In addition cessing) to expand their sales, increase to the direct tariffs that their producers and their value added, and increase their over- consumers must pay, importing firms in these all employment. Ensuring access to cheap countries face implicit “extra tariffs” related imports within and outside the region will to the high costs of inland transport. Argent thus be critical to strengthen the export sec- (2014) shows that, accounting for Rwanda’s tor and improve labor productivity further. inland transport costs of US$3,000 per con- To become a platform for Rwanda’s trans- tainer, a kilo of rice faces an ad valorem formation, the EAC needs to move toward equivalent (AVE) rate of natural protection an enabling environment that better serves of 53 percent, whereas cement faces an AVE Rwanda. Better aligning regional incentives to rate of natural protection from outside of Rwanda’s productive sector will be an impor- the EAC of 136 percent. Similarly, because tant first step to improving its overall compet- of their higher weight-to-value ratio, natu- itiveness, most notably by adapting the CET ral protection is significantly higher on raw and by considering the use of a competitive and intermediate inputs (AVE estimated at exchange rate. Extending the integrated mar- 37 percent on average) than on consumption ket through harmonized s ­ tandards in goods goods (22 percent) or capital goods (2 per- and services is another priority. Effectively cent) (Argent 2014). A large portion of the exploiting regional economies of scale is tariff protection offered by the CET in effect also key, which can be done by adopting and taxes both inputs into Rwandan industry and extending proactive regional initiatives. Rwandan consumers to create jobs in the coastal countries. Since its inception, the CET has held a Reform the CET to Benefit Rwanda’s three-band tariff structure (0 percent for raw Producers materials, 10 percent for semifinished prod- Lowering tariffs will help to increase the ucts, and 25 percent for finished products), productivity of exporting firms. Tariffs are but in practice there are significant deviations a key factor influencing the price of inputs and challenges to implementing the CET. for exporters, with two-thirds of Rwanda’s The most important is the sensitive items (SI) imports bought by exporting firms (Spray list, a set of exceptions to the three-band rule 2017a). Lowering tariffs should be Rwanda’s for goods that are considered to have high main objective in the upcoming EAC nego- potential for import substitution and thus tiations to reform the CET. Some interest enjoy extreme levels of tariff protection (rates groups are advocating increasing the num- between 30 and 100 percent). However, few ber of bands from three to five (see Kenya Rwandan firms benefit from the SI list, and Association of Manufacturers 2017), with Rwanda is a net importer of almost all of the the likely consequence that Rwandan pro- items on it. In 2013, for example, Rwanda ducers and consumers will have to pay higher only exported US$1.2 million of SI list goods prices and off-setting job creation will be to the EAC but imported US$148 million from minimal (Frazer 2017). Others have advo- EAC partner countries, more than 122 times cated ending derogations that have permit- the amount of SI list exports (Argent 2014). ted some importers in some countries to have Although some Rwandan firms p ­ roduce SI access to inputs at international prices. The list products (for example, sugar, maize flour, outcome of these discussions will be impor- wheat flour, rice, and cement), most of these tant for Rwanda because tariffs are a power- products are exported to the Democratic ful force affecting the competitiveness of the Republic of Congo. At the same time, the SI economy and prices people have to pay for list may be hurting export competitiveness, products. because high input tariffs can impair firm Rwanda, along with Burundi and Uganda, productivity (Amiti and Konings 2007; Frazer operates at a significant cost disadvantage 2016b). 110 FUTURE DRIVERS OF GROWTH IN RWANDA The SI list is particularly damaging to the exports—that is, significant increases in the poor. Many of these items are food staples, growth of exports over a seven-year period and the increased prices from high tariffs or longer—occurred in 97 episodes over the disproportionately affect poor households. period 1980–2006. Surges were more likely Frazer (2012) estimated that the SI list to occur in open economies and typically ­ lowered the average poor person’s income by were preceded by a currency depreciation of 3.8 percent (two weeks’ worth of wages). 20 percent or more, which was subsequently To shield itself from the damaging impact maintained with low volatility. Tradables suf- of the SI list, Rwanda has made extensive use fer disproportionately from government or of “stays of application” (deviations from market failures that keep poor countries from the EAC-agreed tariffs). Rwanda has tar- converging toward countries with higher geted goods that are either important con- incomes (Rodrik 2008). The effect of the real sumer goods (for example, sugar and rice) exchange rate is even stronger for exports of or manufacturing inputs. However, this services than for exports of goods; it is espe- pragmatic solution has resulted in inconsis- cially strong for exports of modern services tencies in EAC tariffs and attracted criticism (Eichengreen and Gupta 2012). As low- and from neighboring countries (Argent 2014). middle-income countries shift from export- The system is also biased against small and ing primarily commodities and merchandise medium enterprises (SMEs) because only to exporting traditional and modern services large firms can readily circumvent customs in the course of their development, appropri- payments for the CET. For that reason, ate policies toward the real exchange rate Rwanda should push for formal rules to become even more important. phase out the SI list and cap tariff rates at Targeting the real exchange rate is difficult 25 percent (Frazer 2017). in Rwanda, in part because it is a small econ- Another challenge lies in the misclassifica- omy, inflation is often volatile and imported tion of tariffs. Goods used as raw or inter- through food prices, and capital inflows mediate inputs in the manufacturing sector can be large relative to exports. Moreover, (which should have a tariff of either 0 or Rwanda does not have the financial instru- 10 percent) continue to be applied high tariffs ments of conventional sterilization. Finally, of 25 percent. Analysis by Frazer (2017) com- discussion of exchange rates raises important paring the CET rate with the sectoral break- concerns about managing inflation. Because down of imports suggests that Rwanda has imports consume foreign capital inflows that about 200 product lines, imported primarily might otherwise contribute to an apprecia- by the manufacturing sector, that are charged tion of the Rwandan franc, the government tariff rates of 25 percent or higher. All such would be well advised to focus on export lines may require reclassification. Goods growth rather than on the trade deficit as a should be allocated to tariff lines on the basis macroeconomic objective. of their main use. Monitoring this allocation Because of its centrality to export growth, for all countries and adopting a common list continual monitoring of exchange rates, espe- at the EAC level would constitute an impor- cially relative to neighbors, is vital. Further tant step toward improved transparency and study is recommended of the growth in consistency.9 nontraditional exports in response to the movement of the value of the Rwanda franc against other currencies, as well as of setting Monitor Maintenance of a Competitive up a monitoring mechanism to ensure that Real Exchange Rate, Especially with central bank policy is adequately informed Respect to Other EAC Countries of the implications for export growth. This Arguably the most important price in the mechanism is especially important for cur- economy is the real exchange rate. Freund and rency movements within the EAC because the Pierola (2012) found that sustained surges in region is the destination for a large portion of TR ANSFORMATION THROUGH TR ADE 111 Rwanda’s manufacturing and nontraditional and deep integration (see Durevall 2011; exports. Because the government aspires to Newfarmer and Soderbom 2012). set up a monetary union with a common Another reason to work on the deep inte- currency, this issue is all the more important gration of the region now but to defer estab- over the long run. lishing a monetary union is that the gains B e c au s e e s t abl i s h i n g a mon e t a r y from deep regional integration are much union requires convergence in trade and higher than the benefits from a single cur- ­ m acroeconomic policies among all mem- rency. A common currency is estimated to lead bers, the government should focus on fully to an increase of GDP in the EAC of only one- realizing the goals of the common market tenth of 1 percentage point (0.01 percent). as a near-term priority and work toward Mayer and Thoenig (2016) note: the convergence of macroeconomic policies. Monetary unions entail surrendering mon- The implementation of a common cur- etary and exchange rate policies to a central rency, by alleviating exchange rate risk, authority and require policies, particularly would boost intraregional trade. However, fiscal policies, to be sustainable over the the quantitative impact is small, around medium run. The experience of the European four times smaller than the impact of a common market. Our interpretation is Union, particularly Greece, has illuminated that those expected economic gains are the problems associated with inconsistent too low for justifying paying the political, fiscal policies and subsequent adjustments. institutional, and legal costs attached to These problems are heightened in the pres- the implementation of a common currency. ence of possible asymmetric shocks often This is naturally reinforced by the fact that associated with volatile commodity prices. our evaluation of economic gains does not With the new discoveries of oil in some account for the costs of abandoning dis- EAC countries, the possibilities of external cretionary monetary policy as a tool for shocks that affect countries differently are achieving macro-stabilization. elevated. That said, deepening collaboration within the EAC on elements of trade and Adopting a currency union requires a full improving macroeconomic coordination, common market as well as common fiscal particularly in monetary and exchange rate policies to underpin its success. The steps of policies, can pay dividends even if a mone- liberalizing trade in goods and services as tary union is not established. Deepening col- well as harmonizing tax and subsidy policies laboration on reducing restrictions on trade are worthy in of themselves, even if subse- in goods and services—by facilitating trade, quent steps toward monetary integration are reducing nontariff barriers, and adopting left to the distant future. common regulations and standards—would promote regional integration and contribute Use Standards to Foster Regional Exports to regional price stability. If greater commu- nication among monetary authorities were to Harmonizing standards across the EAC facilitate common monetary decision rules offers both an opportunity and a poten- (for example, on inflation targeting) and tial risk. The EAC has adopted more than these efforts were to reduce price-level vola- 1,500 product standards, with plans for tility within the EAC, trade would expand 5,000 more. Although uniformity can by allowing firms to set up distribution net- improve quality, international product stan- works that would otherwise be disrupted dards also can be prohibitively difficult and by today’s volatility in real exchange rates. expensive for smaller firms, limiting compe- Finally, adopting procedures to allow for tition. The Rwanda Standards Board (RSB) greater labor mobility—particularly in pro- therefore distinguishes between two stan- fessional services—can help to put in place dards. Inspection standards are mandatory the fundamentals of regional price stability (with penalties for noncompliance) so as to 112 FUTURE DRIVERS OF GROWTH IN RWANDA protect consumer welfare. Certification stan- Better enforcement of regional standards dards are voluntary and cover higher, inter- will help to strengthen Rwanda’s exports. national standards (such as the International Recent examples of neighboring countries Organization for Standardization’s ISO 9001 certifying domestic products that fail to meet for quality management) that firms use for minimum standards undermines the reliabil- marketing purposes. Currently, some EAC ity of regional “S-mark” quality brand. The regional standards go far beyond interna- RSB rigorously enforces its standards, which tional standards in setting very demanding can lead to unfair competition for Rwanda’s requirements that have little or no connec- products within the EAC. A similar challenge tion with actual food risks or buyer needs. lies with “counterfeit” versions of Rwanda’s Such requirements can result in additional products abroad, which are often indistin- trade costs and at worst can disconnect poor guishable to foreign consumers but may be farmers from poor consumers in regional and unsafe and hurt the reputation of the “Made domestic markets (Keyser 2014). Formalizing in Rwanda” brand. Current “peer assess- a dual set of standards, tailored to differing ment” processes to deal with such violations needs of firms, will be a key for harmonizing are slow and ineffective and do not address standards in East Africa. the underlying difference in national enforce- Rwanda has a comparative advantage in ment standards. Strengthening the EAC the region with regard to product quality. Secretariat to enforce minimum regional The proliferation of counterfeit products in assessment procedures would strongly benefit neighboring countries—in seeds, fertilizer, Rwanda. and pharmaceuticals, to name a few critical Rwanda also should use its policy and products—constitutes an opportunity for the regulatory environment to build an export “Made in Rwanda” brand (box 2.2 describes capacity in standards applications. Rwanda a program to stimulate product upgrading in could choose to become a regional hub for Chile). Given its comparative advantage in product testing and certification by consoli- governance, Rwanda is uniquely placed to dating all testing units among different gov- become a regional leader for product quality ernment bodies into one agency that would and reliability. This potential is undermined allow testing to be offered faster and at more by poor regional assessment processes. competitive rates to producers. BOX 2.2  Stimulating product upgrading through a supplier development program An important obstacle to raising product quality is A supplier development program can help to that the average low-income consumer is sensitive to develop these domestic supply chains. Domestic prices and may not be willing to pay the higher cost supply chains are developed by bringing together of quality products (Bastos, Silva, and Verhoogen large anchor firms and potential local suppliers and 2016). Although the export market may provide then providing additional support and incentives to the demand, firms often do not know exactly what ensure that suppliers receive the appropriate train- is required of them or how best to compete in such ing so that their products meet the specific qual- a market. Yet countries often have a small number ity and specification standards of the anchor firm. of highly demanding domestic firms, generally large Arraiz, Henriquez, and Stucchi (2013) evaluated multinational anchor firms. Becoming a supplier for this program in Chile and found that participating multinational firms thus offers an important training suppliers saw a large increase in their sales, employ- ground for domestic firms, enabling them to learn to ment, and survival rates. Rwanda would benefit meet specific quality and specification requirements. from a similar program. Source: Steenbergen and Sutton 2017. TR ANSFORMATION THROUGH TR ADE 113 Establish New Initiatives to Develop area in terms of costs, network coverage, Regional Scale Economies, Especially in and reliability. Energy and Finance • Airspace management. To improve air connectivity and increase competition, Rwanda should build on recent initiatives to this project aimed to establish “open develop regional market opportunities. When skies” within the EAC. So far, open skies Rwanda joined the EAC in 2009, regional have been achieved for only two routes: integration was shallow and limited mainly Entebbe to Juba (launched in August to reciprocal market access for partner states. 2014) and Entebbe to Nairobi (launched In recent years, this integration has expanded in January 2015), but with considerable to include several projects to develop regional impact (see priority 2). scale economies. One important avenue has been the Northern Corridor Integrated The Northern Corridor initiatives have Projects (NCIP) Initiative between Kenya, been highly effective. This is best illustrated Rwanda, and Uganda. The NCIP operates through the One Network Area project, at the highest political level (often driven which offered a one-time regulatory interven- directly by presidential decrees), together tion that strongly improved competition and with strong technical and administrative helped to establish a regional telecommuni- support to align domestic regulatory frame- cations market. Within the first 10 days of works. Several important recent projects have adoption, phone traffic between Kenya and come out of the NCIP, including the follow- Rwanda increased by almost 200 ­ p ercent ing (Vanguard Economics 2017): (figure 2.14). Within four months, daily min- utes of roaming traffic increased by more • A single customs territory. To reduce trans- than 950 percent. Regulatory reductions port time and cost along the Northern encouraged suppliers to compete through Corridor transport network, this project lower prices, and volumes expanded suffi- started regional collection and monitor- ciently to allow for greater profitability for ing of data. This effort was followed by telecommunications firms and lower prices the adoption of a wide number of reforms for consumers. related to customs control and trade facili- tation, elimination of nontariff barriers FIGURE 2.14  Impact of One Network Area on mobile phone traffic (such as weighbridges), and adoption of between Rwanda and Kenya technology upgrades to track and moni- tor container movements. These efforts 40 Daily minutes of roaming tra c between strongly reduced trade costs (see priority 2). Implementation of One Network Area 35 • East Africa tourist visa. This project enabled international visitors to obtain 30 Rwanda and Kenya a single tourist visa covering Kenya, 25 Rwanda, and Uganda, encouraging visi- 20 tors to spend time in multiple countries 15 during a single trip and to increase the overall length of their stay in the region. 10 • A single network area. High regional 5 roaming charges led to high cross- 0 border interference in communication. 14 Oc 1, 2 4 4 t. 3 14 4 t. 5 14 t. 6 14 t. 8 14 Oc . 9, 2 14 Oc 0, 2 4 Oc 2, 2 4 Oc 13, 2 4 Oc 4, 2 4 Oc 15, 2 4 6, 4 14 t. 1 01 t. 2 01 t. 4 01 t. 1 01 t. 1 01 t. 01 t. 1 01 t. 01 t. 1 01 Creation of the One Network Area, cov- Oc , 20 Oc , 20 Oc , 20 Oc , 20 Oc 7, 20 t 0 20 Oc , 2 Oc , 2 Oc , 2 t. 1 ering Kenya, Rwanda, South Sudan, and t. Oc Uganda, aimed to increase telecommuni- From Rwanda to Kenya From Kenya to Rwanda cations connectivity along the corridor by moving gradually toward a single network Source: Kelly and Kemei 2016. 114 FUTURE DRIVERS OF GROWTH IN RWANDA This coordinated, competition-inducing TABLE 2.8  Cost of electricity in 2016 in select approach is worth emulating in other areas, East African Community countries especially for electricity. Unlike other African Cost (U.S. cents per Country kilowatt-hour) regional organizations like the Economic Community of West African States and Rwanda 17.9 Kenya 17.6 SADC, the EAC does not have specific pro- Uganda 22.7 tocols aimed at promoting regional electric- Tanzania 15.1 ity investments. The region made an initial Source: Calabrese, Papadavid, and Tyson 2017. attempt in 2005 at cooperation through the Eastern Africa Power Pool, which aimed in establishing regional energy markets or to pool energy resources, promote power adopting cross-border electrification projects. exchanges between utility companies, and Recent attempts to establish a 200-megawatt reduce the costs of power supply. Rwanda is interconnected transmission line between a member of the Nile Basin Initiative and the Uganda (Mbarara) and Rwanda (Birembo) Eastern Africa Power Pool. However, apart have been halted (Nalule 2016). A presi- from using its share of two regional hydro dential initiative would help to provide new plants,10 Rwanda is not connected to any momentum and even speed up some of the regional transmission ­ network (AfDB 2013). regional reforms. Rwanda would strongly benefit from Additional efforts also should be made a competitive regional energy market. to develop a regional financial market in the Interconnection can provide important ben- EAC. The financial sectors of EAC partner efits to sharing a reserve margin with other states are small and underdeveloped, which countries. Rwanda’s current 30-megawatt can limit investment, especially for large, interconnection capacity with Burundi and lumpy investment projects. Regional finan- the Democratic Republic of Congo has served cial integration can help to achieve economies it in critical periods of shortages in recent of scale, which would deepen domestic finan- years (AfDB 2013). It also offers an oppor- cial opportunities, expand the opportunities tunity for more regular electricity transfer for financial intermediation, and reduce the to take advantage of short-term diversity of overhead costs of maintaining financial infra- demand between the two power systems. structure. However, significant restrictions Rwanda’s current energy use and produc- are preventing the flow of capital across bor- tion are limited, with only about 16 percent ders. At the same time, there is also a need to of Rwanda’s households connected to the remain aware of the risks of increasing cross- grid and a total generation of 210 megawatts border financial links. The path forward of installed capacity. However, Rwanda has should thus combine facilitating capital flows signed a wide range of new electricity con- between EAC members with providing more tracts, which are expected to come on board careful oversight of the volumes and conse- in 2020 and could more than double capac- quences of these flows (Wagh, Lovegrove, ity (512 megawatts). This capacity will likely and Kashangaki 2012). exceed overall demand in Rwanda. Sharing this energy with neighboring Tanzania Strengthen the EAC Secretariat to and Uganda will provide a way to limit the Shield the Region from National potential risk of excess energy (EAPP 2014). Beggar-Thy-Neighbor Initiatives Establishing regional energy markets may further help to lower costs and drive electric- The EAC risks moving backward in several ity production costs down to Tanzania’s level areas, because the common market proto- (table 2.8). col is under threat from various national A new presidential initiative could help to b eggar-thy-neighbor initiatives tabled by ­ establish a regional energy market. Despite partner states. Several recent issues have initial efforts, little progress has been made slowed EAC regional integration and are TR ANSFORMATION THROUGH TR ADE 115 placing it under threat of moving backward. the cost of exporting products and import- The Burundi crisis resulted in high-level fric- ing inputs. Because more than 90 percent tions that ultimately led Burundi to block all of goods exports are transported by truck trade with Rwanda in 2015. More countries (Cameron and Viviers 2017), Rwanda relies faced with trade deficits have responded heavily on the land transport corridors of by adopting national initiatives aimed other countries for access to the sea. Almost at import substitution (such as Uganda’s all of Rwanda’s trade in global markets goes recent “Build Uganda, Buy Uganda” strat- through two East African trade corridors: egy). Diverging positions on regional trade the Northern Corridor (Mombasa) and the (related to s­ econdhand clothing, for exam- Central Corridor (Dar es Salaam). Distance ple) and claims of unfair competition in remains the single most important source of agricultural input subsidies (such as for trade costs worldwide (Rastogi and Arvis Tanzanian rice) have exacerbated tensions 2014).11 Yet the lack of facilitation at the bor- in the region. Amid all of these challenges, der, fragmentation of the supply chain, and the EAC Secretariat is facing increasing limited access to affordable air cargo oppor- financial hardship given cutbacks in donor tunities are equally important factors that contributions. contribute to the high cost of trading. Rwanda benefits most from a strong, rule- oriented EAC and should thus advocate for a Lower Transport Costs Further through more powerful secretariat. As a smaller mem- Intensive Collaborations on Rwanda’s ber state, Rwanda would significantly benefit Two Trade Corridors and Potentially from having a stronger EAC Secretariat with through a Regional Rail Network the ability to prevent dominance of any indi- vidual country. The EAC Secretariat could Although the Central Corridor has histori- also improve collective decision making by cally had lower costs for Rwandan trade, sev- offering member states a transparent over- eral initiatives along the Northern Corridor view of regional economic dynamics and pro- have sharply reduced transport times. These viding lessons from regional best practices initiatives include the introduction of a single (for example, in tax administration, indus- customs declaration (reducing the number trial policy, and regulatory harmonization). of declarations required from three to one), Encouraging new high-level initiatives that a regional customs transit guarantee (a sin- deepen the responsibilities of the EAC while gle regional customs bond recognized in all advocating for granting additional powers to three territories), and a joint electronic cargo the EAC Secretariat to enforce such tasks will tracking system (allowing containers to be be pertinent. Increasing the financial contri- tracked in real-time) (Vanguard Economics butions to the EAC Secretariat by all member 2017). Transport times have fallen 70 percent states is another prerequisite for deepening (from 21 to 6 days; figure 2.15) and could be regional integration. reduced even further to as little as 1.4 days (NCTTCA 2017). The increased competi- Priority 2: Improve Trade tion between corridors has placed downward Connectivity by Lowering pressure on the transport price for a 40-foot Transport Costs within the container along the Central Corridor (which fell from US$4,200 in 2013 to US$3,700 in Region and with the World 2016), so transport remains marginally less The single most important determinant expensive through the Central Corridor than of long-run trade growth is reducing the through the Northern Corridor (figure 2.16). costs of getting goods to market and get- Rwanda should continue to identify barriers ting inputs to local producers. Rwanda is a to intra-EAC transport of goods, agree on landlocked country, which means that trans- detailed reforms, and monitor the outcomes port costs typically add some 50 percent to using the two corridor observatories. 116 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE 2.15  Average transport time, by route lowering the rate at which it is applied appear to be the best options for Rwanda.13 25 As with road tolls, tax policy toward truck- ing within the EAC also will influence com- 21.1 petitiveness. The value added tax (VAT) 20 threshold of FRW 12 million reduces the 14.0 incentives for creating larger-scale transport 15 firms (which are typically more competitive), Days 10.3 because in-house transport services from 10 businesses that pay VAT are better able to 6.3 recover investment costs. Reviewing tax pol- icy to ensure that truckers from all countries 5 compete on a level playing field is a high 3.1 3.6 3.8 3.8 priority. In addition, the corporate and per- 0 sonal income tax burden on domestic truck- 2013 2014 2015 2016 ing firms is high because they face informal Mombasa to Kigali costs that are not deductible for tax pur- Dar es Salaam to Kigali poses. Because of this fact and the strategic Sources: Calculations based on CCTTFA 2013, 2014, 2015, 2016; nature of the industry, Rwanda might con- NCTTCA 2014, 2015, 2016a, 2016b. sider adjusting the accepted deductibles for Note: CCTTFA = Central Corridor Transit Transport Facilitation Agency; NCTTCA = Northern Corridor Transit Transport Coordination Authority. Rwandan truckers. A long-standing ambition to build a rail- way along both transport routes (Northern FIGURE 2.16  Transport rates for a 40-foot Corridor and Central Corridor) holds con- container, by route siderable promise. The Kenyan govern- ment recently completed the first leg of the 5,000 Northern Corridor, developing a parallel 4,800 4,700 standard-gauge railway between Mombasa and Nairobi along an old meter-gauge colo- 4,500 nial rail line, with plans to extend it all the 4,200 way to Kigali. Railways can reduce freight US$ costs considerably compared with roads, in 4,200 part because of the lack of weighbridge stops 4,000 and police checkpoints, and can bring goods 4,000 to port more quickly than trucks. Railways 3,800 3,700 also help to reduce the prevalence of truck 3,625 transport, limiting traffic congestion while 3,500 2013 2014 2015 2016 lowering the costs to maintain roads and Mombasa to Kigali operate weighbridges (Frazer 2016a). Dar es Salaam to Kigali However, rail construction is extremely Source: Calculations based on CCTTFA 2013, 2014, 2015, 2016; costly and should only be pursued when NCTTCA 2014, 2015, 2016a, 2016b. financed through special, long-term bonds Note: CCTTFA = Central Corridor Transit Transport Facilitation Agency; NCTTCA = Northern Corridor Transit Transport that offer highly concessionary rates. Most Coordination Authority. rail construction in Africa is financed through foreign assistance (including the For Rwanda, road tolls that discriminate recent Kenyan expansion, which relied on against its truckers dampen services income Chinese financing). Faced with a strict debt and reduce competition. Negotiating within ceiling, Rwanda should avoid financing a new the EAC framework for applying the road rail project through direct borrowing, which toll equally across all EAC vehicles12 and otherwise would displace public investments TR ANSFORMATION THROUGH TR ADE 117 with higher short-run returns in favor of DP World to develop and operate the Kigali investments that take decades to produce Logistics Center. Five potential components positive net social benefits. Instead, Rwanda typically make up a logistics hub. The first should use other financial instruments, such is a major distribution center for imported a new “big bond” suggested by Birdsall as ­ consumer goods, machinery and parts, and and Okonjo-Iweala (2017), where donors other goods requiring short delivery times would offer long-term bonds at subsidized to importers in neighboring countries, rates (borrowing in the market against future most prominently the Democratic Republic aid flows). Such a long-term, low-interest of Congo. The second is a group of one or bond could offer a good source of finance for more centers for consolidating agricultural railways because of its synchronicity with the products for export by land and air. The payoff horizon of a regional railway project. third is one or more distribution centers that supply imported and domestic products to retail outlets throughout the country. The Develop Rwanda as a Regional Logistics fourth, which may be co-located with the Hub by Attracting Pioneer Foreign third, is a location near the country’s large Firms in Logistics industrial zones where the service providers Attracting international logistics services for these industries cluster. The fifth is an may further improve the efficiency of cross- air transport hub for both passengers and border supply chains and reduce trade costs. freight linking countries in West Africa with Logistics performance depends not only on Asia and Europe. Although these five activi- infrastructure but also on the institutions ties have some synergies, they are separate and processes of trade—such as process- businesses, compete in different markets, ing by customs—and the quality of services use different networks, and offer different available for trade, which itself depends on value propositions in terms of cost, time, regulation and competition (Rastogi and and reliability. The government has already Arvis 2014). According to the logistics per- made progress on developing these com- formance indicator, Rwanda and other EAC ponents, so creating a regional hub now countries are perceived to outperform their requires finding synergies and ways to mar- income-group peers (World Bank, various ket new advantages. years). A new international airport at Bugesera Despite these significant improvements, is planned in the near future, equipped logistics services are provided mostly by the with modern cargo facilities. These facili- owners of the goods (first-party logistics pro- ties include a cargo village with a pack- viders) or the assets (second-party logistics ing area, cold chain facilities, and multiple providers) used to provide these services.14 ground-handling companies. A neighbor- Other than freight forwarders, third-party ing special economic zone that will be service providers such as Kumwe are limited. closely integrated with the airport’s facili- The international third-party logistics pro- ties will also be operational and is expected viders providing services in Rwanda, mostly to increase air freight demand. The airport Bollore (transport and warehousing services) will further support Rwanda’s tourism by and DHL (courier services or airfreight) oper- offering self-check-in, different (closed) ate mainly out of Kenya and, to some extent, boarding gates, multiple duty-free shops, Tanzania. As trade between Rwanda and its several restaurants, and a business lounge. neighbors and the world expands, the gov- Exploring ways to extend the capacity of ernment should actively encourage them to the current airport in the short term will establish operations in Rwanda. enable the government to push back the Several projects are ongoing, including necessary timeline for delivery, which may the recent award of a 25-year concession limit cost overruns to this high-profile pub- by Dubai-based marine terminal operator lic investment project. 118 FUTURE DRIVERS OF GROWTH IN RWANDA Aggressively Pursue Open Skies (2016) estimated that liberalization between Arrangements in Air Transport the five EAC countries could result in 9 per- cent lower average fares and 41 percent more The best way to expand air transport is frequent flights. These changes, in turn, to reduce restrictions on competition. would be expected to increase passenger Governments everywhere often support percent for the five countries and traffic 29 ­ their own national carriers by making it dif- 42 ­percent for Rwanda. Important indirect ficult for foreign airlines to access certain effects of open skies would be an additional routes. Yet allowing nondomestic airlines 45,800 tourism visits for Rwanda and a to land and take passengers to a third coun- 2.5 percent increase in Rwandan air cargo try (“5th Freedoms”)15 can significantly trade. Jointly, in the short term, these effects reduce air fares and increase passenger traf- are expected to provide an additional 30,700 fic. Examples of the effectiveness of such jobs (7,300 for Rwanda) and to increase GDP open skies agreements can be found in the by US$124 million a year (US$24 million for Nairobi–Johannesburg route, which saw a Rwanda). In the medium term, these benefits 69-fold increase in passenger volume when could grow much larger, because they would liberalized in 2003. Similarly, when routes for provide an important strategic means for the SADC opened, fares dropped 18 ­ percent Rwanda to lower the cost of its exports in (Bofinger 2017). goods and services. Air traffic has increased considerably for EAC routes that have been liberal- ized. Discussing open skies for the EAC started through a Northern Corridors Priority 3: Increase Service Initiative, first introduced by the heads of Sector Productivity, Both as a state in October 2013. By December 2014, Critical Input to Other Priority negotiations were finalized for bilateral Sectors and as a Source of air service agreements along two routes: Exports Entebbe to Juba and Entebbe to Nairobi. RwandAir has pro­ a ctively targeted new Trading in services provides a double value: routes for which ­ bilateral air service agree- exports of services earn foreign exchange and ments have been granted. The total direct assist diversification, while imports of ser- benefit to the airline industry on the two vices can lead to greater competition, lower routes is estimated at US$40 million prices, and better quality, thus enhancing effi- (Vanguard Economics 2017). Other flights ciency gains and competitiveness. Backbone within Northern Corridor a ­ irspace have service sectors (such as transport, telecom- also grown steadily, in anticipation of new munications, and professional ­ services) serve EAC open skies agreements. Between 2013 as critical inputs into manufacturing global and 2016, the annual number of passengers value chains, which can help to raise exports on RwandAir routes in this ­ a irspace grew and improve sectoral productivity (see pri- by 65 percent, while ticket prices on those ority 1). To realize these benefits, there is a routes fell 14 percent. need to lower policy barriers to services com- The potential benefits from airspace lib- petition and to strengthen services trade for eralization are large and should be pur- EAC professionals. Services trade also offers sued with vigor. Many intra-EAC routes prominent opportunities for direct exports. are still restricted for third-party operators. To realize these benefits, there is a need to This is driven partly by resistance from elevate tourism promotion and continue to Kenya Airways, as it attempts to reposi- address persisting skills shortages for profes- tion itself in the EAC market (Vanguard sionals by leveraging private sector knowl- Economics 2017). InterVistas Consulting edge and financing. TR ANSFORMATION THROUGH TR ADE 119 Lower Policy Barriers to Services exports are not exempted from VAT (with Competition and Services Exports the exception of tourism, transport, and finance) because services exports often Fostering greater competition among sup- have no physical border control, making pliers across borders would lower prices to verification more difficult than for goods. consumers and producers. While other fac- As a result, most of Rwanda’s services now tors (for example, small scale, high operat- face a type of export tariff that puts export- ing costs, inadequate regulation) contribute ers at a competitive disadvantage (because to inefficient services, the payoff to increased their products are now 18 percent more competition and efficiency could be large. expensive). Recent improvements in OECD Hoekman and Shepherd (2015) argue that countries show that services exports can be competition improves services efficiency and tracked reliably by requiring the exporting thus benefits downstream users, particularly firm to present evidence that the customer manufacturing firms. Using cross-country resides outside the country and has paid data for the world and East Africa, they for the services (OECD and WTO 2017). found that a 10 percent reduction in a coun- Rwanda should follow the OECD guidelines try’s Services Trade Restrictiveness Index that allow for exemption of VAT for any is associated with a 4.4 percent increase in goods, services, and intangibles (such as dig- manufactured exports from a country like ital products) with similar evidence require- Rwanda. ments. Following these guidelines would Competition can occur through either also help Rwanda to collect more reliable cross-border provision of services by foreign data on services exports (which are gathered companies or the establishment of a commer- through national surveys, limiting under- cial presence in the country.16 To foster com- standing of firm-level dynamics). Broad petition, EAC member states could negotiate exemptions on services exports would move to permit greater cross-border services trade Rwanda ahead of the EAC partner states, within the EAC itself. At present within the but in line with leading African countries EAC, some key markets are tightly knit oli- such as Mauritius and South Africa. gopolies, typically with a dominant supplier. The absence of competition is one reason why some industries experience high costs. Elevate Tourism Promotion For example, despite the advent of quality Internet service and the spread of mobile Earnings from tourism are the single largest telephony, telecommunications in Rwanda source of exports in Rwanda, almost equal- still suffer from higher-than-average con- ing the value of mineral, tea, and coffee sumer prices and costly restrictions.17 Given exports combined in 2016. Tourism receipts the government’s desire to promote the ICT almost doubled from US$200 million in sector, it may be useful to reexamine why 2010 to US$390 million by 2016. About half telecommunications prices are high. Other of this revenue comes from a small number examples relate to energy and finance, the of leisure tourists (about 100,000 between use of airspace, and services trade for EAC 2012 and 2014), 30 percent comes from professionals. business tourists (about 400,000 between The imposition of VAT is constraining 2012 and 2014), and 20 percent comes from Rwanda’s service exports. VAT is commonly other sources (for example, visiting family charged only in the destination where the and friends). transaction takes place. To avoid double The government’s M ICE (meetings, taxation, exporters are exempted from pay- incentives, conferences, and exhibitions) ing VAT after they declare the products to strategy aims to grow tourism receipts, customs. However, in Rwanda most services as gorilla tourism approaches saturation. 120 FUTURE DRIVERS OF GROWTH IN RWANDA Through its investments in the Kigali share of export earnings and has some of Convention Center, RwandAir, and hotel the best potential for future growth, it also construction, the government has put in merits greater institutional support. place a set of facilities that can drive tour- Working with tour operators to give ism growth for years to come. Moreover, it Rwanda more visibility in regional tours is has established promising contractual rela- another promising avenue. Most tour opera- tionships with blue-ribbon foreign compa- tors market Rwanda as part of a regional nies, such as Marriott and Radisson, to aid package, allowing only a few days in-­ country in bringing tourists and large conferences for gorilla trekking. Although regional to Rwanda. Conference tourism (MICE) is packaging has enabled Rwanda to piggy- doing well, increasing by 24 percent in 2014 back on the established East African safari and offsetting some of the decline in other market, this may be an opportune time to ­ business travel. distinguish Rwanda as an alternative, offer- The biggest rewards will be obtained ing more ­variety in a more secure environ- if the country can expand leisure tour- ment with less travel. The government might ism beyond gorillas and convince tour- consider giving priority—or discounts—for ists—both leisure and MICE —to spend gorilla permits to tourists who plan to stay more time in Rwanda. The average visitor in Rwanda longer than the few days typically spends only about 3.2 days in the country. necessary to visit the gorillas, at least during If visitors were enticed to double the aver- the peak season. age length of their stays, the sector could Several other initiatives—some easy and generate another US$60 million without some more difficult—might be integrated increasing the number of tourists (English into a higher-visibility strategy: 2017). Rwanda has two other national parks that offer complementary attrac- • The regional tourist visa, a worthwhile tions. Indeed, Akagera now receives more initiative, should be expanded to include visitors than Volcano National Park and Tanzania. its gorillas because of greater interest from • Training in tourism merits more invest- nationals. ment, both to satisfy Rwanda’s inter- To scale up tourism, both the Ministry nal demand but also to position any of Trade and Industry and the RDB should local training institution as a center improve their capacities to champion the for excellence that could become part tourism sector and to develop midrange of Rwanda’s educational exports. It tourism opportunities (for example, bird- might be possible to finance a project watching and adventure tourism). Rwanda that builds the human and institutional can also improve its targeted “destination capacity to manage the tourism sector, management authorities” to help develop fills infrastructure and skill gaps, and specific locations (for example, Lake Kivu) attracts private investors. by providing public goods such as infra- • In the long run, if the government were structure and sign-posting and by coor- able to establish working tourism agree- dinating private sector actors. Offering a ments with the Democratic Republic of website with comprehensive, timely infor- Congo, Rwanda could enhance its own mation, which enables visitors to book already high-growth tourist industry by g overnment permits online and brings ­ becoming a gateway to several remarkable together the vacation packages of differ- tourist attractions in eastern Democratic ent private tour operators, is another area Republic of Congo, including, for exam- for improvement. A well-staffed authority ple, the Mount Nyiragongo lava lake. would have greater capacity to engage in A regional accord could help to protect all of these planning and support activities. Lake Kivu’s development from threats of Because the sector accounts for the largest pollution. TR ANSFORMATION THROUGH TR ADE 121 • Finally, the government could develop a domestic law) (Dihel and Goswami 2016). program to enhance the links of the tour- Similar restrictions are also considered ist sector with the rest of the economy important nontariff barriers in the move- through increases in local purchases ment of goods within the EAC (EABC by hotels and by tourists themselves. 2016). Country-specific costs of regula- Developing the local supply chains to tion can have a serious impact on the entry hotels is a priority. decisions of SMEs (especially if firms do not expect large sales in the foreign mar- ket), which further increases sectoral costs Further Liberalize Services Trade for and impairs productivity. EAC Professionals Services trade for professionals would Professional services are critical for improv- benefit from recognizing additional licenses ing the productivity of other economic and standards obtained in other EAC coun- activities, especially for SMEs. Although tries and from adopting common qualifi- professional services account for only a small cations criteria. The EAC has introduced fraction of the service sector (and services mutual recognition agreements (MRAs) in exports), they play an important role in shap- three professional services: accounting and ing economic development. Business services architectural services have been covered in such as accounting, legal support, and engi- Rwanda since 2013, and engineering ser- neering offer important intermediate inputs vices have been covered since 2016. So far, for other agricultural, industrial, and service intra-EAC movement of these professionals sectors. Dihel, Fernandes, and Mattoo (2012) remains limited, with about 130 persons show that, for firms in East Africa, there is registering in another country to work as a strong positive correlation between the use accountants (1 percent of the total EAC pro- of professional accounting and legal services fession), 22 to work as architects, and 9 to and average labor productivity. This effect work as engineers (Hook 2013). Yet these is particularly pronounced for SMEs, which agreements have already offered important have more than 40 percent higher labor pro- benefits for certifications and learning insti- ductivity than similar enterprises that do not tutions (box 2.3). use professional services.18 Improving access The MRAs should be given formal treaty to affordable, high-quality professional ser- status as soon as possible, be broadened in vices can thus help to boost productivity for their purpose, and be expanded to other the wider economy. professions. The three concluded MRAs Harmonizing regional standards for were negotiated by professional bodies and EAC professionals offers an important subsequently presented to the EAC Council opportunity to foster professional services. of Ministers. Although this approach con- Differences in regulation provide criti- tributed to fast recognition, it also gave the cal nontariff barriers that limit integra- MRAs an uncertain legal status that weak- tion in the East African Common Market ened their implementation and created a and prevent the exploitation of econo- situation in which conflicts with domestic mies of scale. An important example for provisions (such as nationality requirements services lies with professional accredita- in competition policy, public procurement, tion requirements. Previously, each coun- or freedom of establishment) would not hold try in the EAC had its own criteria that up in a court of law. The MRAs should thus either would require extensive retraining be given treaty status that formally estab- to receive the necessary licenses to oper- lishes their legal precedence over such domes- ate (for example, in accounting) or would tic policy issues. They should also address prevent some professionals from operat- unintended consequences such as when EAC ing altogether (for example, in Kenya and nationals with foreign qualifications are Tanzania, only nationals can practice ineligible for MRA benefits (Hook 2013). 122 FUTURE DRIVERS OF GROWTH IN RWANDA BOX 2.3  The benefits of a mutual recognition agreement for Rwanda’s engineering professional body The Institute of Engineers of Rwanda (IER) was estab- rules and regulations. Such an “audit” helped the IER lished in 2013 through a parliamentary act. Its main to improve its standards and to establish the required responsibility is to regulate, promote, and raise train- procedures and documentation in a very short period ing standards of the engineering profession. It was also of time. This effort was accelerated by extensive sup- Rwanda’s key representative to negotiate the mutual port from other EAC countries, including the Ugan- recognition agreement (MRA) for engineering with dan Institute of Engineering, and quickly turned the EAC, which it signed in March 2016. To date, the the IER into one of the best-organized professional MRA has had limited impact on intraregional trade bodies in Rwanda. This process also improved the in engineers (which has been low). Instead, the main quality of accredited engineers. Following the EAC- benefit has been to formalize Rwanda’s engineering prescribed process, Rwanda now requires engineers profession and professional body. to submit and defend a “technical report” in a forum Before the MRA could be signed, the EAC came of professional peers, further raising the reputation to inspect whether the IER followed the appropriate for quality of Rwanda’s engineers within the EAC. To extend services trade further, MRAs relationship between per capita services should be drafted for other professional exports and tertiary gross enrollment rates sectors, including legal, finance, and consult- ­ (figure 2.17) as well as secondary comple- ing professionals. tion rates (figure 2.18). Taking the predicted To facilitate short-term assignments, the level of exports to GDP required to achieve EAC should further abolish work-permit upper-middle-income status and assuming regimes for all eligible professionals. The that half of it derives from services exports MRA still requires eligible professionals (the current share of services exports in total from EAC partner states to obtain permits exports) yields the average level of skills asso- prior to being employed, which is a burden- ciated with services exports per capita. This some and slow process (Basnett 2013). It also calculation suggests that Rwanda trails sig- requires registration for professional bodies nificantly behind on both tertiary enrollment in both the “home” and “visiting” country, and secondary c ­ ompletion. Although these each requiring an initial certification process figures are only indicative (and some coun- and an annual financial contribution. These tries have managed to achieve such services administrative burdens prevent profession- exports with less human capital), they illus- als from accepting short-term assignments. trate the significant challenge facing Rwanda Instead, any eligible professional certified in the next two decades with regard to from any EAC professional body should be increasing its human capital. Part of this con- automatically exempt from work-permit straint may be relieved by greater inflows of regimes. professionals from the EAC (see priority 4), but for the most part human capital will have to be increased through significant additional Address Persisting Skills Shortages for Professionals by Leveraging Private investments in education (discussed in chap- Sector Knowledge and Financing ter 1 of this report). Expansion of the private sector in East Rwanda is facing a skills deficit that, if not Africa is opening up opportunities that remedied, will constrain potential growth for require specialists with a new set of skills. high-skill services exports. There is a strong The growth of the financial sector, for TR ANSFORMATION THROUGH TR ADE 123 example, has drastically increased the FIGURE 2.17  Tertiary gross enrollment rate and log services demand for accountancy and legal services. exports per capita, 2000–16 A skills assessment estimated that Rwanda needs 5,000 accountants in the public sec- 100 tor and another 2,325 accountants in the financial sector to meet demand, yet it has 80 Tertiary gross enrollment rate only 6 percent of that number (ICPAR 2017). Rwanda is lagging behind the rest 60 of the EAC in the number of professionals and is far behind African leaders in ser- vices exports such as Mauritius and South 40 Rwanda, 2015 Africa. This lag is particularly noticeable for Upper-middle-income Rwanda, 2001 accountants and lawyers (figure 2.19). 20 countries Skills shortages provide both an oppor- tunity and a risk for trade in services within 0 the EAC. The MRAs provide a key means to 0 2 4 6 8 10 fill high-paying vacancies in the private sec- Log services exports per capita tor using regional experts. Yet, if this is seen to be a one-sided movement (for example, Source: Calculations based on World Development Indicators data (World Bank, various years). from Kenya to Rwanda), the use of regional experts may lead to resentment and result in FIGURE 2.18  Upper-secondary completion rate and log services a backlash to EAC regional integration. For exports per capita, 2000–16 that reason, it is important to increase the number of professionals in Rwanda through 100 a combination of student loan programs, private sponsorships, and in-company Upper-secondary completion rate mentoring. 80 Skills mismatches are another seri- ous concern for professionals. The EAC 60 is faced with jobless accountants despite hig h demand (Dihel, Fernandes, and 40 Mattoo 2012). Similar patterns are found Rwanda, 2015 in Rwanda, where low-quality training has Rwanda, 2002 20 resulted in an important distinction between accountancy graduates (often struggling to find employment) and accountants reg- 0 istered with the Institute of Chartered 0 2 4 6 8 10 Accountants of Rwanda (having received Log services exports per capita additional training). Potential explanations Source: Calculations based on World Development Indicators data (World Bank, various years). for this pattern include poor links between the education system, labor market, and professional associations. has started supporting polytechnics to offer Integrating professional skills within poly- accounting modules that can also fast-track a technics can help to address skills shortages. student’s accreditation. Such initiatives have In previous years, formal classes in account- increased the total number of students from ing often covered separate material and did 165 in 2012 to 1,771 in 2016 (ICPAR 2017). not contribute to a student’s certification as Extending them to other professional skills a public accountant. Now, the Institute of could help to increase the overall number of Certified Public Accountants of Rwanda professionals in Rwanda quickly. 124 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE 2.19  Number of professionals per 100,000 inhabitants in select African countries a. Accountants b. Lawyers Rwanda 4 Tanzania 2 Uganda 6 Uganda 4 Tanzania 15 Rwanda 5 Botswana 38 Botswana 12 Kenya 48 Kenya 19 South Africa 79 Mauritius 42 Mauritius 110 South Africa 43 0 20 40 60 80 100 120 0 10 20 30 40 50 Number of professionals Number of professionals Sources: For accountants, ICPAR 2017; for lawyers, Dihel, Fernandes, and Mattoo 2012. Note: ICPAR = Institute of Certified Public Accountants of Rwanda. Setting up private sector–sponsored stu- Priority 4: Stimulate Foreign dent loan programs offers another oppor- and Domestic Investment into tunity. At the moment, all student loan programs are on a needs basis. However, Tradable Sectors by Using for many service sector jobs, students are Selective and Performance- expected to earn large returns on this invest- Driven Incentives ment, so offering commercial loans may M o r e p r ivat e i nve s t m e nt i s n e e d e d make sense. There is considerable interest i n t radable sec tors l i ke tou rism a nd from the private sector in helping to develop manufacturing. Private finance has experi- ­ special-purpose financial instruments that enced strong growth over the last decade, can be used for such purposes, but seed fund- with private credit nearly tripling between ing from government and some broad sup- 2000 (10 percent of GDP) and 2016 (28 port are needed to limit the risk in case of percent of GDP). However, most of this defaults. private finance has gone to sectors with Government could work more closely with little to no trade, such as construction, real industry to ensure relevant skills training. To estate, and microfinance (households). By avoid skills mismatches, the curricula for pro- contrast, in 2015 only 12 percent of the fessionals in both polytechnics and universi- stock of private finance was in manufactur- ties should be aligned more closely with the ing, and 18 percent was in tourism. Given needs of particular industries. The industrial the broader aspirations of the government bodies offer a particularly helpful platform for private sector–led transformation, more that can negotiate on behalf of dominant needs to be done to attract additional FDI firms and the curriculum review board. To and domestic private capital into tradable enable students to signal their completion of sectors. an industry-certified course, industrial bodies Rwanda recently has experienced a should introduce and enforce accreditation remarkable increase in FDI as a share of GDP. schemes (where absent). Whereas FDI inward stock constituted only TR ANSFORMATION THROUGH TR ADE 125 about 3 percent of GDP in 2000, by 2016 FIGURE 2.20  FDI stock as a share of GDP in select East African its share had increased to almost 27 ­ percent countries, 2000–16 (figure 2.20). This has been made possible ­ partly through establishment of a strong 45 international reputation and brand, exem- 40 plified by Rwanda’s rank (second-highest in 35 Africa) on the World Bank Doing Business Indicators, a broad range of special tax incen- 30 tives to attract investors, and a highly pro­ 25 Percent active investment promotion agency (Rwanda 20 Development Board) that also offers exten- sive investor aftercare. 15 Rwanda’s main FDI stock centers on 10 telecommunications and financial ser- 5 vice sectors, but it has recently expanded 0 to include more manufacturing, tourism, 2005 2006 2007 2009 2010 2012 2013 2014 2016 and mining. Manufacturing now amounts Tanzania Uganda Ethiopia Kenya Rwanda to 15 percent of all FDI (table 2.9), partly due to proactive government partnerships Source: United Nations Conference on Trade and Development (UNCTAD) data. in the garment sector including a desig- Note: FDI = foreign direct investment. nated action plan, targeted investment mis- sions, and designated government support TABLE 2.9  Stock of foreign direct investment in Rwanda, by sector, 2015 for eligible investors. FDI to other sectors has been more sporadic, driven to a large Sector US$ (millions) Share (%) extent by large individual projects: the top Information and communication technology 592 28 four projects constitute more than 26 per- Financial and insurance activities 477 22 cent of all registered investment. Although Manufacturing 329 15 such large projects can quickly raise over- Tourism 219 10 all investment, the number of investments Agriculture 104 5 Mining 90 4 needs to increase to sustain this level and Wholesale and retail trade 89 4 strengthen resilience. Electricity, gas, steam 70 3 Other 61 3 Transport and storage 46 2 Increase Efforts to Attract Additional Total 2,139 100 “Priority” FDI Source: BNR 2017. Comparisons with other Eastern African countries suggest that more can be done to of a new anchor firm could help to allevi- attract FDI. Rwanda’s inward stock of FDI ate specific sectoral challenges (for example, as a share of GDP exceeded that of Ethiopia finance, skills, logistics) and generate impor- and Kenya for the first time in 2016, yet still tant spillovers for local buyers and suppliers trails behind that of Tanzania and Uganda (Farole and Winkler 2014). Making use of (figure 2.20). Additional efforts in Rwanda such opportunities, the RDB could pack- might well attract more FDI. age and present high-potential opportuni- Although Rwanda welcomes all invest- ties to investors in strategic sectors through ment, prioritizing sectoral investment targeted investment missions in key markets would help to assist strategic sectors. FDI (MINEACOM 2017). This effort would has the greatest effectiveness for economic mimic the recent successes in garments growth when it strengthens the wider sec- (box 2.4). Evidence suggests that sectors tor. Firms often complement each other in explicitly targeted by investment promotion a broader value chain, so the introduction agencies (IPAs) are associated with more than 126 FUTURE DRIVERS OF GROWTH IN RWANDA BOX 2.4  The strategy to transform the textiles, apparel, and leather sector In 2015, the Ministry of Trade and Industry clas- Economic Planning, through the Rwanda Revenue sified the textiles, apparel, and leather sector a key Authority, exempted key garment inputs from import priority industry, with high potential for manufac- tariffs. Finally, all new investors received support turing growth, employment, and exports. However, for training, technology upgrading, and certifica- the domestic apparel sector was facing a range of tion through a coalition of the Rwanda Standards challenges, including lack of investment, limited Board, the Workforce Development Authority, and skills, and fragmented value chains, all of which the National Integrated Rural Development Agency. limited Rwanda’s competitiveness and restricted Initial results of this initiative are impressive, potential exports. mobilizing significant private investment and for- In response, a cross-government strategy was eign direct investment. Sectoral growth accelerated drafted in 2016 to transform the textile, apparel, and from 3 percent in 2015 to 10 percent in 2016, while leather sector. The Rwanda Development Board took exports increased from US$16 million (1.1 percent the lead in promoting private direct investment from of all exports) to US$20 million (1.5 percent of all both foreign and domestic sources through targeted exports). More than 5,000 jobs are expected to be investment missions. The Ministry of Trade and created by 2020. This approach now stands as a Industry ensured prioritized access to land in a new wider model for industrial policy and for attracting special economic zone. The Ministry of Finance and foreign direct investment in Rwanda. Source: MINICOM 2017. twice the FDI inflows as nontargeted sectors Make Integrated Government Support (Javorcik 2004). Selective and Performance Driven Investor promotion can be strengthened further through a designated tracking Offering integrated government support can system. Establishing an integrated sys- stimulate production for priority sectors. The tem that compiles all challenges reported Rwandan government has been experiment- by investors will strongly assist the RDB ing with different approaches to industrial in identif ying which issues have been policy. One key means has been the KSEZ, addressed and in providing better investor which has attracted both foreign and domes- aftercare, which can also help to attract tic investors in nontraditional, tradable sec- additional F DI inf lows (Harding and tors. Between 2013 and 2016, the KSEZ Javorcik 2013).19 A tracking system also accounted for between 4.5 and 10 percent of can bring together data from other gov- all national goods exports and provided an ernment sources (for example, tax and important source of new manufacturing and customs) to track a firm’s key performance agroprocessing exports. The KSEZ also has indicators, such as total investments, prof- three important, nonfiscal benefits over other its, exports, and employment. This data locations: more reliable infrastructure and could then be compared with a firm’s over- utilities, better trade facilitation, and closer all amount of government support, which ties to government agencies to make firms includes tax incentives, other financial better aware of relevant incentives, laws, and incentives (for example, training support), regulations. This type of integrated govern- and nonfinancial benefits received. Such ment support was found to contribute to large a monitoring system could help to ensure firm-level increases in sales, value added, and that government support is conditioned on employment (Steenbergen and Javorcik 2017). performance and to assess whether a sec- The aspiration of Rwanda’s new Private tor’s overall incentives are in accordance Sector Development Strategy is to main- with its overall potential. stream integrated government support across TR ANSFORMATION THROUGH TR ADE 127 all priority value chains. Through the new government expenditure. Three deductions Growth Anchor Firm Initiative, the Private for corporate income tax constitute between Sector Development Strategy aims to extend 75 and 90 percent of all tax incentive costs: the benefits of the KSEZ to all anchor firms nonoperational expenses, investment allow- in priority value chains (including light man- ances, and investment expenditures. With ufacturing and ­ agroprocessing). Given the some adjustments and better targeting, there limited resources available, it may be more may be scope for increasing the return to effective to address specific challenges com- tax incentives, while at the same time rais- prehensively (for example, access to export ing more revenue (discussed in chapter 4 of finance, trade infrastructure) for a select this report). These incentives merit review number of priority anchor firms than to to ensure that they are having their desired offer a cross-cutting focus across the whole effect in attracting foreign and domestic private sector. This support is further tied to investment and that efforts are paying divi- broad export performance targets between dends in the form of new exports. the government and the private company (MINICOM 2017). This system closely Priority 5: Accelerate mimics the Republic of Korea’s support- ive approach to large holding companies Industrialization through (­chaebols), where the government provides Diversification, Value Addition, broad support and then rewards or penal- and Quality Upgrading izes enterprise managers on the basis of their Increasing the size of the industrial sector is performance outcomes (Khan 2007). ­ a critical component of Rwanda’s structural To assist industrialization, an integrated transformation, because of its large employ- government support project needs to be ment and export potentials. Increasing min- selective, performance driven, and time eral exports through value addition will be bound. Admission needs to be based on a critical to raising exports (chapter 4 of this demonstrated high potential for exports, report). Developing nascent agroprocess- employment, and domestic productivity ing and other manufacturing sectors will enhancement. To avoid firm abuse of govern- be another important avenue. Rwanda has ment benefits, the state should be able to dis- expanded nonmining industrial sectors by continue support for inefficient firms. Firm exporting new products such as garments, participation also should be time bound (for mechanical appliances, and grain milling. example, five years in length) to ensure that Expanding their production further will participants make optimal use of integrated be key to continuing Rwanda’s structural government support. transformation. Revise Tax Incentives to Align with Create a Supplier Development Export Promotion Objectives Program to Promote Domestic Supply Chains Tax incentives are another key policy that influences price incentives to producers. Another way to increase exports is to deepen Under the new 2015 Investment Code, the domestic supply chain of value added. To the government provides substantial tax the extent that domestic suppliers can sub- incentives to new investment in selected stitute for imports now going into exports, activities. Although the new code targets increasing domestic value added contributes incentives ­b etter than before, these incen- to embodied exports—a key objective of the tives are expensive. According to Bode, government’s “Made in Rwanda” program Steenbergen, and Lohmann (2017), they (MINEACOM 2017). This program requires amounted to about 3.3 percent of GDP in upgrading quality to improve links between 2016, equivalent to almost 10 percent of domestic suppliers and large international 128 FUTURE DRIVERS OF GROWTH IN RWANDA “anchor” firms. As Sutton (2014) argues, harmonizing regional standards are criti- “The most powerful engine of capability cal priorities and opportunities to promote building lies in firm-to-firm ­ interactions in Rwandan products. Preparing companies to supply chains.” Firm-to-firm interactions are meet product standards for certification also particularly important because such anchor has long-term export benefits. In Rwanda, firms often have detailed technical and quality TradeMark East Africa and the RSB have requirements for their supplies that may differ been providing hazard analysis and critical from generic quality certification. Globally control point (HACCP) training since 2011 recognizable producers with “brands” to to help firms to prevent food contamination uphold thus often require suppliers to com- and prepare them to attain further certifi- ply with additional “‘private standards” for cations. HACCP certification is a difficult social, environmental, and safety norms that process, requiring considerable time and may go beyond national legal requirements investment; with 24 applicants, only 10 enter- (UNIDO 2010). prises were ultimately expected to be certified Chile’s Supplier Development Program to HACCP and one to ISO 22000 (the global offers a model of ways to foster domestic food safety management standard). This pro- supply chains in a market-supportive way. cess has contributed to higher-value domestic It eschews the mandated rules for domes- contracts and increased exports in the EAC tic value addition that led to failure in some (TradeMark East Africa 2015). However, Latin America countries and eventually led steering this process through external actors the World Trade Organization to enact strict is costly and difficult to scale up. limits on performance requirements. The In-country processes are needed to sup- program aimed to improve commercial links port standards upgrading and offer cost- between SMEs and their large foreign custom- effective monitoring. Professional bodies ers. It helped buyers to identify potential sup- (for exa mple , Rwa nda’s Cha mber of pliers and married this help with subsidized Agriculture and its product-specific mem- credit and market support to create links. bers) could play a much larger role in these Arraiz, Henriquez, and Stucchi (2013) found processes. These bodies could be the first that this program was significantly effective point of contact for producers to learn in increasing sales, employment, and the sus- about product standards and to receive sup- tainability of SMEs. Similar programs have port for quality upgrading. They also could been launched in Colombia and El Salvador. absorb the initial cost for small-scale, rou- Rwanda is considering the creation of a tine quality-monitoring exercises in return supplier development unit in the RDB that ­ for producers’ membership contributions. could facilitate links between anchor firms Initial support from government (under and local suppliers and identify the train- leadership of the RSB) and development ing needed for suppliers to meet the tech- partners to train these private sector bodies nical ­ specifications and quality standards. could yield long-term rewards. Steenbergen and Sutton (2017) illuminate how this could be implemented effectively in Rwanda. A supplier development pro- Conclusion gram could also be linked to the Export Rwanda has set for itself a huge but wor- Development Facility at the Development thy challenge: to become an upper-middle- Bank of Rwanda. income country by 2035. Achieving the growth rates necessary to make this goal a reality would be almost unprecedented in the Use Standards to Foster Regional history of development. The difficulties can- Manufacturing Exports not be underestimated. Achieving this goal Upgrading product quality, establishing will require harnessing the full potential of and satisfying standards for products, and regional and global markets—not only to TR ANSFORMATION THROUGH TR ADE 129 expand exports at a rapid pace but also to   7. Services exports are underreported because tap into new sources of FDI, opportunities of the use of customs data, which do not regis- for technological progress, and knowledge, ter most services trade. with the objective of speeding structural  8. Spray (2017a) found that a recent large change through dynamic employment cre- (broadly exogenous) reduction in transport costs significantly increased the likelihood of ation. The five priorities laid out in this firms exporting. He exploits this fact using chapter provide the outlines of a road map— an instrumental variables approach to iden- abstract, but indicative of ways in which tify the impact on productivity for firms that Rwandans themselves might fill in the details export for the first time as well as for their and achieve trade-led growth. domestic suppliers.   9. The main way in which Rwanda deals with the misclassification is to make use of the Notes Duty Remission Scheme, which leaves SMEs 1. Rwanda’s accession to the EAC involved three at a competitive disadvantage. This allows shifts: a new tariff structure, free trade with manufacturing firms to avoid tariff duties Tanzania and Uganda (as Rwanda already had when they can prove that a good is used as a free trade with Burundi and Kenya under the basic input, thus improving the competitive- Common Market for Eastern and Southern ness of their finished products. However, the Africa Free Trade Agreement), and removal of scheme has also created inconsistencies within transport costs from tariff calculations, which the EAC and presents serious challenges for further reduced Rwanda’s effective trade tax customs administration. Larger firms have rates (Argent 2014). also been more successful in obtaining import 2. UNIDO Industrial Statistics (INDSTAT) tariff exemptions on their inputs, whereas database; U N Comtrade database; and the effective tax rates facing SMEs are much University of Minnesota’s Integrated Public closer to the official CET rate. The current Use Microdata Series (IPUMS) International CET is biased against SMEs. database. 10. These plants are owned by a private actor and 3. Brazil, China, the Arab Republic of Egypt, have equal joint participation from Burundi, India, Nigeria, and the Russian Federation. the Democratic Republic of Congo, and 4. Copper production offers an important inter- Rwanda. Rwanda purchases 3.5 megawatts of national example. Whereas the United States electricity from Rusizi I plant and 12 mega- in the early 20th century used its copper min- watts from Rusizi II. ing sector to develop expertise in chemistry 11. Trade costs are the price equivalent of the and metallurgy that laid the foundations for reduction in international trade as compared subsequent diversification and industrializa- with the potential implied by domestic pro- tion, Chile focused on mining alone and failed duction and consumption in the origin and to develop similar industries (Maloney and destination markets. Higher bilateral trade Valencia Caicedo 2016). costs result in smaller bilateral trade flows 5. Baumol’s example is that, to play a string (Arvis et al. 2013) quartet, the same number of musicians is 12. With reference to the transit toll, the Kenya needed today as was needed in the 19th cen- Public Road Tolls Act states, “Vehicles bear- tury, so the productivity of classical music ing Kenyan registration are for the time being performance has not increased. However, the exempted,” suggesting that lawmakers recog- real wages of musicians (as in all other pro- nized that the discriminatory practice could fessions) have increased greatly since the 19th require revision. century. 13. Removing the road toll for vehicles regis- 6. EAC membership expanded gradually over tered in any EAC country altogether would time, with Kenya, Tanzania, and Uganda join- align with the stated long-term vision for ing in 2005, and Burundi and Rwanda joining the EAC to become a fully functional cus- in 2009. For Burundi and Rwanda, “5 years toms union, with free flow of goods across after” captures their import share 1 year after borders, but it also would have a significant joining, and “10 years after” reflects 6 years impact on Rwanda’s own road toll income of membership in the EAC. (which totaled just under FRW 514 billion 130 FUTURE DRIVERS OF GROWTH IN RWANDA (approximately US$8 million) in fiscal year Andersen, M., B. Kett, and E. von Uexkull. 2011–12, making up 27 percent of the cur- 2017. “C or porate Ta x I ncent ive s a nd rent road fund revenues and probably close FDI in Developing Countries.” In Global to the current total raised by Kenya and Investment Competitiveness Report Tanzania collectively). 2017/2018: Foreign Investor Perspectives 14. A first-party logistics provider is a firm trans- and Policy Implications, ch. 3, 73 –99. porting products from a sender to a receiver. Washington, DC: World Bank. https://doi​ A second-party logistics provider is a carrier .org/10.1596/978-1-4648-1175-3_ch3. that owns the means of transport, such as a Argent, J. 2014. “Review of EAC Common External shipping line, airline, or trucking company Tariff.” TradeMark East Africa, Kigali. that owns, leases, or charters its vessels. Arraiz, I., F. Henriquez, and R. Stucchi. 2013. A third-party logistics provider offers out- “Supplier Development Programs and Firm sourced or “third-party” logistics services Performance: Evidence from Chile.” Small to companies for parts of their supply chain Business Economics 41 (1): 277–93. management function. A r vis, J.-F., B . Shepherd, Y. Duval, and 15. Schlumberger (2010) describes the nine “free- C . Utoktham. 2013. “Trade Costs and doms” regulating air transport through the Development: A New Data Set.” Economic lens of African air service. Premise 104, World Bank, Washington, DC. 16. In World Trade Organization parlance, these Baines, T., H. Lightfoot, and P. Smart. 2011. reflect mode 1 and mode 3 types of services “S er v it i z at ion w it h i n M a nu fac t u r i ng: trade, respectively. Additional modes include Exploring the Provision of Advanced Services consumption abroad (mode 2, for example, and Their Impact on Vertical Integration.” tourism) and temporary presence of persons Jour n al of M anufac turing Technolog y (mode 4, for example, servicing short-term Management 22 (7): 947–54. construction contracts). Bamber, P., O. Cattaneo, K. Fernandez-Stark, 17. Argent (2014) found that, to protect its mar- G. Gereffi, E. van der Marel, and B. Shepherd. ket share, the dominant mobile phone supplier 2017. “Diversification through Servicification.” charged consumers more to call an out-of-­ Unpublished paper, World Bank, Washington, network phone in competing networks than DC. to call London. Basnett, Y. 2013. “Labour Mobility in East 18. One example is the use of accounting tech- Africa: An Analysis of the East African nicians to provide basic record keeping ser- Community’s Common Market and the Free vices, which can help firms to meet economic Movement of Workers.” ODI Development gains by shifting to a more organized and Policy Review 31 (2): 131–48. targeted management style (Bloom et al. Bastos, P., J. Silva, and E. Verhoogen. 2016. 2016). “Export Destinations and Input Prices.” Policy 19. Harding and Javorcik (2013) found that Research Working Paper 6914, World Bank, an IPA’s ability to attract FDI depends on Washington, DC. strongly on quality, which also includes the Baumol, W. J. 1967. “Macroeconomics of ability of staff to handle investor inquiries. Unbalanced Growth: The Anatomy of Urban A 1–10 quality rating of IPAs found that a Crisis.” American Economic Review 57 (3): 1-point rating increase was associated with a 415–26. 15 percent increase in FDI inflows. Birdsall, N., and N. Okonjo-Iweala. 2017. “A Big Bond for Africa.” Project Syndicate, Center for Global Development, Washington, DC, References April 17. Af DB (African Development Bank). 2013. Bloom, N., J. Van Reenen, A. Choudhary, and “Rwanda Transport Sector Review and Action R. Lemos. 2016. “Management in Pakistan: Plan.” African Development Bank Group, First Evidence from Punjab.” International Tunis. Growth Centre, Islamabad. A miti, M., and J. Konings. 2007. “Trade BN R (National Ban k of Rwanda). 2017. Liberalization, Intermediate Inputs, and “Monetary Policy and Financial Stability Productivity: Evidence from Indonesia.” Statement—22nd February 2017.” Government American Economic Review 97 (5): 1611–38. of Rwanda, Kigali. TR ANSFORMATION THROUGH TR ADE 131 Bode, M., V. Steenbergen, and J. Lohmann. 2017. Duarte, M., and D. Restuccia. 2010. “The “Attracting Investments While Raising Tax Role of the Structural Transformation in Revenue: Recommendations for Reviewing Aggregate Productivity.” Quarterly Journal of Rwanda’s Tax Incentives.” IGC Policy Note, Economics 125 (1): 129–73. International Growth Centre, Kigali. Durevall, D. 2011. “East African Community: Bofinger, H. C. 2017. “Air Transport in Africa: Pre-conditions for an Effective Monetary Oppor t u n it ies a nd I ssues.” Dra f t pre - Union.” I nternational Grow th C entre, pared for Industries without Smokestacks: London. African Industrialization Revisited, EABC (East African Business Community). 2016. ed ited by R . New fa rmer, J. Page, a nd “East African Common Market Scorecard.” F. Tarp. Oxford: Oxford University Press, EABC, Nairobi. forthcoming. EAPP (Eastern Africa Power Pool). 2014. EAPP Calabrese, L., and A. Eberhard-Ruiz. 2016. Regional Power System Master Plan 2014. “What Types of Non-Tariff Barriers Affect Addis Ababa: EAPP, February. the East African Community?” Overseas Ebl i ng, G ., a nd N. Ja n z . 1999. “E x por t Development Institute, London. and Innovation Activities in the German Calabrese, L., P. Papadavid, and J. Tyson. 2017. Service Sector: Empirical Evidence at the “Rwanda: Financing for Manufacturing.” SET Firm Level.” ZEW Discussion Paper 99-53, Report, June, Overseas Development Institute, Center for European Economic Research, London. Mannheim. Cameron, M., and W. Viviers. 2017. “Identifying Eichengreen, B., and P. Gupta. 2012. “Exports Realistic Export Opportunities for Rwanda of Services: Indian Experience in Perspective.” Based on the TRADE-DSM Approach.” IGC Working Paper 12/102, National Institute of Policy Paper, International Growth Centre, Public Finance and Policy, New Delhi. Kigali. English, P. 2017. “Rwanda: Raising Exports and CCTTFA (Central Corridor Transit Transport Attracting Foreign Direct Investment: Some Facilitation Agency). 2013. Central Corridor Ideas for Action.” International Growth Transport Observatory Annual Report 2013. Centre, London. Kigali: CCTTFA and TradeMark East Africa. Farole, T., and D. Winkler, eds. 2014. Making http://observatory.centralcorridor-ttfa.org​ Foreign Direct Investment Work for Sub- /media/Annual_Reports/Central_Corridor​ S aharan A f r ic a : L oc al S pillove rs an d _­Transport_Observatory_Annual_Report_2014​ Competitiveness in Global Value Chains. .pdf. Washington, DC: World Bank. ———. 2014. Central Corridor Transport Frazer, G. 2012. “The EAC Common External Observatory Annual Report 2014. Kigali: Tariff (CET) and Rwanda.” IGC Policy Paper, CCTTFA and TradeMark East Africa. International Growth Centre, London. ———. 2015. Central Corridor Transport ———. 2016a. “The Boring Infrastructure That Observatory Annual Report 2015. Kigali: Rwanda Needs.” International Growth Centre CCTTFA and TradeMark East Africa. blog, August 6, ———. 2016. Central Corridor Transport ———. 2016b. “Imports and Skill Utilization in a Observatory Annual Report 2016. Kigali: Low-Income Country.” University of Toronto. CCTTFA and TradeMark East Africa. ———. 2017. “Examining the Impact of the Dihel, N., A. M. Fernandes, and A. Mattoo. Common External Tariff of the East African 2012. “Developing Professional Services in Community in Rwanda.” IGC Policy Paper, Africa through Regional Integration.” In International Growth Centre, London. De-Fragmenting Africa: Deepening Regional Freund, C., and T. Moran. 2017. “Multinational Trade Integration in Goods and Services , I nve stor s as E x por t Sup er st a r s: How edited by P. Brenton and G. Isik. Washington, Emerging Market Governments Can Reshape DC: World Bank. Comparative Advantage.” Working Paper 17-1, Dihel, N., and A. G. Goswami. 2016. “The Peterson Institute for International Economics, Unexplored Potential of Trade in Services Washington, DC. in Africa: From Hair Stylists and Teachers Freund, C., and M. Pierola. 2012. “Export to Accountants and Doctors.” World Bank, Surges.” Journal of Development Economics Washington, DC. 97 (2): 387–95. 132 FUTURE DRIVERS OF GROWTH IN RWANDA Gelb, A., C. J. Meyer, V. Ramachandran, the World Development Report 2016: Digital and D. Wadhwa. 2017. “Can Africa Be a Dividends, World Bank, Washington, DC. Manufacturing Destination? Labor Costs in Kenya Association of Manufacturers. 2017. Comparative Perspective.” CGD Working “KAM Study on the Review of the East African Paper 466, Center for Global Development, Community Common External Rates.” Kenya Washington, DC. Association of Manufacturers, Nairobi. Ghani, E., and H. Kharas. 2010. “The Service Keyser, J. 2014. “Avoiding the ‘Harm’ in Revolution in South Asia: An Overview.” Harmonized Standards for Food Staples In The Service Revolution in South Asia, in Africa.” World Bank Trade Post blog, edited by E. Ghani, 1–32. New York: Oxford December 1. University Press. K han, M. 20 07. “G overnance, E conomic Hallward-Driemeier, M., and G. Nay yar. Growth, and Development since the 1960s.” 2017. Trouble in the Making: The Future DESA Working Paper 54, United Nations of M a n u fa c t u r i n g - L e d D e ve l o p m e n t . Department of Economic and Social Affairs, Washington, DC: World Bank. New York, August. ST/ESA/2007/DWP/54. Harding, T., and B. S. Javorcik. 2013. “Investment Lalui, A. 2016. “Analysis of Cross-Border Trade Promotion and FDI Inflows: Quality Matters.” in Services: Issues and Trends with a Focus on CESifo Economic Studies 59 (2): 337–59. Rwanda and DRC.” TradeMark East Africa, Herrendorf, B ., R. Rogerson, and Á . Kigali. Valentinyi. 2013. “Growth and Structural Loungani, P., and S. Mishra. 2014. “Not Your Transformation.” NBER Working Paper Father’s S er vice S ector.” I nternational 18996 , Nat iona l Bu reau of E conom ic Monetary Fund, Washington, DC. Research, Cambridge, MA. Maloney, W. F., and F. Valencia Caicedo. 2016. Hoekman, B., and B. Shepherd. 2015. “Services, “The Persistence of (Subnational) Fortune.” Firm Performance, and Exports: The Case of Economic Journal 126 (598): 2363–401. the East African Community.” IGC Paper, Mayer, T., and M. Thoenig. 2016. “Regional I nternational Grow th Centre, London, Trade Agreements and the Pacification of October. Eastern Africa.” International Growth Centre, Hook, A. 2013. “Mutual Recognition Agreements London. in Services Trade in Africa: Lessons from McMillan, M., D. Rodrik, and C. Sepulveda. the East African Community.” World Bank, 2017. “Structural Change, Fundamentals, and Washington, DC. Growth: A Framework and Country Studies.” ICPAR (Institute of Certified Public Accountants Policy Research Working Paper 8041, World of Rwanda). 2017. “ICPAR Five Year Strategy: Bank, Washington, DC. 2017–2021.” ICPAR, Kigali. Melitz, M. 2003. “The Impact of Trade on Intra- InterVistas Consulting. 2016. “Costs and Benefits Industry Reallocations and Aggregate Industry of ‘Open Skies’ in the East African Community Productivity.” Econometrica 71 (6):1695–725. (EAC): Key Findings.” InterVistas Consulting, M I N E AC OM (M i n ist r y of E ast A f r ic a n Washington, DC. Community). 2017. “Made in Rwanda Policy.” Jaud, M., and C. Freund. 2015. Champions Government of Rwanda, Kigali. Wanted: Promoting Exports in the Middle MINICOM (Ministry of Trade and Industry), E a st an d Nor th Af ric a . Di rec tions in Rwanda. 2015. “Revised National Export Development. Washington, DC: World Bank. Strategy.” MINICOM, Kigali. doi: 10.1596/978-1-4648-0460-1. ———. 2016. “Strategy to Transform the Textile, Javorcik, B. S. 2004. “In Search of Spillovers Apparel, and Leather Sectors.” MINICOM, through Backward Linkages.” American Kigali. Economic Review 94 (3): 605–27. ———. 2017. “Private Sector Development ———. 2014. “Does Foreign Direct Investment Strategy 2 (PSDS2) 2018–2024.” MINICOM, Increase the Productivity of Domestic Firms? Kigali. In Search of Spillovers through Backward Morjaria, A ., and V. Steenbergen. 2017. Linkages.” American Economic Review 94 “Understanding Constraints to Value Addition (3): 605–27. in Rwanda’s Coffee Sector.” IGC Policy Kelly, T., and C. Kemei. 2016. “One Network Brief, International Growth Centre, London, Area in East Africa.” Background paper for February. TR ANSFORMATION THROUGH TR ADE 133 Nalule, V. 2016. “Energy in the East African ht t p s: //op e n k nowle d g e .worldb a n k .or g​ Community: The Role of the Energy Charter /handle/10986/2467. Treaty.” International Energy Charter, East Shepherd, B., J. de Melo, and R. Sen. 2017. Asian Community. “Reform of the EAC Common External Tariff: Nayyar, G. 2017. “The Future of Manufacturing- Evidence from Trade Costs.” IGC Policy Paper, Led Development in Rwanda.” Background International Growth Centre, Kampala. note for the “Future Drivers of Growth” proj- Spray, J. 2017a. “Assessing the Influence of EAC ect, World Bank, Washington, DC. Trade Costs Reductions in Exports and Export NCTTCA (Northern Corridor Transit Transport Supply-Chains in Rwanda.” International C oord i n at ion Aut hor it y). 2014. “T he Growth Centre, Kigali. Northern Corridor Transport Observatory ———. 2017b. “Exports and Promoting Backward Report.” Issue 5, NCTTCA, Kigali. Linkages: Ideas and Lessons for the Made in ———. 2015. “The Northern Corridor Transport Rwanda Policy.” International Growth Centre, Observatory Report.” Issue 6, NCTTCA, Kigali. Kigali. Spray, J., and S. Wolf. 2016. “Industries with- ———. 2016a. “The Northern Corridor Transport out Smokestacks in Uganda and Rwanda.” Observatory Report.” Issue 7, NCTTCA, International Growth Centre, London. Kigali. Steenbergen, V., and B. Javorcik. 2017. “Analysing ———. 2016b. “The Northern Corridor Transport the Impact of the Kigali Special Economic Observatory Report.” Issue 8, NCTTCA, Zone on Firm Behaviour.” IGC Working Kigali. Paper, International Growth Centre, London, ———. 2017. “The Northern Corridor Transport August. Observatory Report.” Issue 9, NCTTCA, S t e e nb e r g e n , V. , w it h J . S ut ton . 2 017. Kigali. “Establishing a Local Content Unit for Newfarmer, R., J. Page, and F. Tarp, eds. Rwanda.” Policy Note, International Growth Forthcoming. Industries without Smokestacks: Centre, London. African Industrialization Revisited. Oxford, Sutton, J. 2014. “Gains from the Natural Gas: U.K.: Oxford University Press. Local Content and Tanzania’s Industrial Newfarmer, R., and M. Soderbom. 2012 . Development.” Seventh Gilman Ruthinda “Rwanda and the East Africa Monetary Memorial Lecture, Bank of Tanzania, June. Union.” IGC Policy Brief, International Szirmai, A. 2012. “Industrialisation as an Engine Growth Centre, London, March. of Growth in Developing Countries, 1950– Newman, C., J. Page, J. Rand, A. Shimeles, 2005.” Structural Change and Economic M. Soderbom, and F. Tarp. 2016. Made in Dynamics 23 (4): 406–20. Africa: Learning to Compete in Industry. TradeMark East Africa. 2015. “Setting Standards Washington, DC: Brookings Institution Press. in Rwanda’s Food Industry.” TradeMark OECD (Organisation for Economic Co-operation East Africa, Nairobi, August 30. https:// and Development) and WTO (World Trade w w w.t rad e m a rke a . c om / i mp ac t- s tor ie s​ Organization). 2017. Aid for Trade at a Glance /setting-standards-in-rwandas-food-industry/. 2017: Promoting Trade, Inclusiveness, and Tybout, J., J. de Melo, and V. Corbo. 2015. Connectivity for Sustainable Development. “The Effects of Trade Reforms on Scale Geneva: WTO; Paris: OECD Publishing. and Technical Efficiency: New Evidence Rastogi, C., and J.-F. Arvis. 2014. The Eurasian from Chile.” In Developing Countries in Connection: Supply-Chain Efficiency along the World Economy, edited by J. de Melo, the Modern Silk Route through Central Asia. 237–56. Hackensack, NJ: World Scientific Directions in Development Series: Trade. Publishing. Washington, DC: World Bank. United Nations. Various years. UN Commodity Rodrik, D. 2008. “The Real Exchange Rate and Trade St at ist ic s (C omt rade) dat abase. Economic Growth.” Brookings Papers on New York: United Nations. Economic Activity 39 (2): 365–439. UNIDO (United Nations Industrial Development Schlumberger, C. E . 2010. Open Skies for Organization). 2010. “Making Private Africa: Implementing the Yamoussoukro Standards Work for You: A Guide to Private Decision. Directions in Development Series: Standards in the Garments, Footwear, and Infrastructure. Washington, DC: World Bank. Furniture Sectors.” UNIDO, Vienna. 134 FUTURE DRIVERS OF GROWTH IN RWANDA Un iversit y of M i n nesot a. Va rious yea rs. World Bank. 1993. The East Asian Miracle: Integrated Public Use Microdata Series E conomic G row th an d Public Polic y. (IPUMS) International database. University of New York: Oxford University Press. Minnesota. ———. 2017. “Rwanda E conomic Update: Vanguard Economics. 2017. “Measuring the Sustaining Growth by Building on Emerging Economic and Social Impact of Northern Export Opportunities.” World Bank, Kigali. Corridor Integration Projects (NCIP) in ———. 2019. Doing Business 2019: Training for Rwanda.” Vanguard Economics, Kigali. Reform. Washington, DC: World Bank. Wagh, S., A. Lovegrove, and J. Kashangaki. 2012. ———. Various years. Doing Business database. “Scaling up Regional Financial Integration Washington, DC: World Bank. https://www​ in the EAC.” In De-Fragmenting Africa: .doingbusiness.org/en/custom-query. Deepening Regional Trade Integration in WTO (World Trade Organization). 2018. World Goods and Services, edited by P. Brenton and Integrated Trade Solution online database. G. Isik. Washington, DC: World Bank. Geneva: WTO. 3 Faster Urbanization, Greater Agglomeration “No country has grown to middle income in Africa. Rural workers have migrated to without industrializing and urbanizing. urban areas in droves, where their productiv- None has grown to high income without ity is double that in rural areas; urban areas vibrant cities. The rush to cities seems have accounted for 48 percent of national chaotic, but it is necessary” (World Bank ­ labor productivity growth over the past 2009). 15 years. Nevertheless, urban areas in Rwanda, Introduction and Kigali in particular, have not generated the kind of productivity gains that ­ rapidly Rapid and well-managed urbanization is growing cities in successful East Asian econ- essential for Rwanda to achieve its aspira- omies have achieved. This is largely because tions for growth. Urbanization generates the agglomeration economies from urban- enormous economic benefits by increasing ization have been limited. To rectify this, density, which facilitates the transmission of Rwanda should focus more on delivering knowledge and ideas, increases economies of the fundamental drivers that link urban- scale and opportunities for specialization by ization to robust economic growth and less reducing transport costs, and improves firms’ on setting targets for levels of urbanization access to critical services and a large pool of itself. In essence, Rwanda should focus on labor with varied skills. Workers who are reshaping public policy, prioritizing invest- closer to a wider range of goods and services ments, and strengthening institutions along benefit from such opportunities to deepen three dimensions— spatial , sectoral , and their human capital and strengthen their bar- institutional: gaining power. These agglomeration econo- mies mean that the more workers and firms 1. Spatial. A bullish view is needed to are added to an urban location, the higher unleash the potential of Kigali to con- their individual productivity becomes. nect Rwanda with the regional and global Rwanda’s urban population has more economy, enabling the city to be built than doubled since 2002, and from 2002–12 and grow in response to international Kigali was one of the fastest-growing cities demand. 135 136 FUTURE DRIVERS OF GROWTH IN RWANDA 2. Sectoral. To enhance efficiency of capi- Investments tal investment in infrastructure, hous- Another priority is to provide efficient, ing, and commercial structures, there is affordable, and integrated public transport urgent need to strengthen land markets. and to boost investment in roads. Public 3. Institutional. Stronger intergovernmen- transit routes to Kigali’s center are con- tal coordination of economic and spatial gested and likely to become more so. To planning processes is needed to over- limit sprawl, excellent central public trans- come coordination challenges in urban port should be combined with quality pub- development. lic transport to a small number of strategic peripheral locations. Grids should be pro- Regulatory Policies vided (in Kigali and in secondary cities) to For the rapid economic growth targeted until ensure that rapid peripheral development 2050, Rwanda needs an economic engine to takes place in a planned way. Greater invest- deliver dynamic agglomeration economies of ment in roads also is necessary to improve the kind seen only in large cities. Only Kigali’s the integration of Rwanda’s cities with the economy, which is already seven times the size countryside and with neighboring countries of the next largest urban area, can perform and to strengthen the connections between this role. However, Kigali’s urban fabric is urban centers. fragmented by small-scale, patchy land devel- Kigali’s growth needs to be accompanied opment, which disconnects firms, labor, and by a more calibrated approach toward devel- input and product markets, undermining the oping a complementary set of secondary agglomeration economies and raising the costs cities and small towns. International expe- of private inputs and public infrastructure. rience shows that such investments should The main priority is to strengthen land mar- center on increasing external connectivity kets so that private developers face the appro- and laying urban grids to plan for future priate incentives to increase density. Rwanda’s expansion. In addition, investments in urban planning has been inappropriate within health and education would raise workers’ cities, leading to fragmented development, as productivity without tying them to a par- peripheries are less strictly regulated than the ticular place, so that markets can pick the center. This creates perverse incentives to build location and form of private sector develop- in peripheries, in exactly the scattered and ment more efficiently. Kigali needs to oper- inefficient way that planners wish to avoid. ate more like an effective, integrated labor Bedrock reforms that strengthen land markets and product market, which would increase and urban planning should be in place first its potential for the realizing agglomeration to enhance the efficiency of investment and economies and reduce the cost of private generate high economic returns. In dynamic, inputs and public infrastructure. fast-changing cities, the economic vocations of The government has pursued a policy of places are transformed regularly as the econ- grouped villages (imidugudu), which, in some omy grows in sophistication and land markets cases, has implied large sunk-cost invest- facilitate the transformation of structures to ments that, with effective urbanization, often meet the economic needs and opportunities of become redundant. The program facilitates the time. rural-urban links as well as indirectly prepar- Rwanda should move from highly pre- ing future urban dwellers for urban life while scriptive zoning requirements for structures they are in rural areas. Improving planning to a system that allows structures to evolve to and establishing settlements, while minimiz- serve the changing economy. However, regu- ing huge infrastructure costs and potential lations on structures should be used to miti- future financial losses, are recommended to gate negative externalities like environmental ensure that they serve the purpose. Policy also degradation. needs to prioritize rural-urban connectivity F aster U r b ani z ation , G reater A gg l o m eration 137 and anticipate migration to larger urban well-managed urbanization. Urbanization is areas. This approach can be complemented the spatial corollary of the process of eco- with “off-grid” rural infrastructure, which nomic structural transformation that involves delivers key services while placing smaller the movement of workers from less to more “bets” in the long-run vocation of more productive sectors—typically from farming remote locations. to off-farm sectors in the early stages of devel- opment. Although high birth rates make the Institutions natural increase of the population an impor- tant source of urban growth, rural-urban These changes will require stronger institu- migration is an even more significant force tions. Urban planning capabilities at the city driving the urbanization process. Migration and district levels need strengthening. Plans depends on forces that “pull” migrants should respond to markets and community to their destinations as well as forces that needs, while safeguarding against capture “push” them to leave their origins. Better eco- by special interests through strong public nomic opportunities in urban areas, partly oversight and engagement. due to agglomeration economies, are a major Urban planning also needs to be rational- pull factor, often providing the main motiva- ized. “Economic” planning is often discon- tion for rural-urban migration (Brueckner nected from “spatial” planning. As Rwanda and Lall 2014). In China’s urbanization works to strengthen the economic dividend process, for example, employment in manu- of urbanization, economic planners should facturing and services grew from less than start to think spatially, and spatial planners 30 percent of the labor force 30 years ago to should become more aligned with economic more than 60 percent now. The reallocation planning and goals. of labor to manufacturing and services—­ Further, economically connected dis- sectors in urban areas that rapidly increased tricts must be encouraged and required to their productivity—accounted for almost coordinate land use plans, transport and one-fifth of China’s gross domestic product service provision, major infrastructure like (GDP) growth during that period (World special economic zones, tourism infrastruc- ture, and so on, to reduce waste and exploit Bank and Development Research Center of complementarities. the State Council, China 2014). Finally, stronger institutions are needed to Agglomeration economies strengthen value land and to assign and protect property the links between urbanization and eco- rights. Rwanda has a strong basis for efficient nomic development. The enormous ben- and transparent valuation, with centralized efits of urbanization are seen most vividly and digitized landownership and transaction in East Asia, where growth in the last half records, a professional valuation body, and century is among the most spectacular epi- credible institutions for oversight. Credible sodes of development in recorded human land valuation will also enable urban areas to history. Figure 3.1 shows changes in urban- fiscalize public investments in land through ization and GDP per capita from 1985 to land value capture. 2010, when the processes of urbanization and economic development were intertwined. Countries such as China, the Republic of The Promise of Urbanization and Korea, Thailand, and Vietnam have sup- Agglomeration ported the development of vibrant cities that allowed frenetic, productive interactions Urbanization, Agglomeration, and among people and businesses, marked by Economic Development the transmission of knowledge and ideas. Few countries have attained upper-­ m iddle- Increasing returns to scale were driven by income status without substantial, rapid, and economic density, triggered primarily by 138 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE 3.1  Changes in urbanization and income high-income economies. Simply put, big cit- in East Asia and Pacific, 1985–2010 ies deliver extremely dense and varied labor and product markets, multiplying the forces Korea, Rep. of agglomeration, while enabling economies 80 of scale in the provision of infrastructure, Malaysia 70 Mongolia public spaces, and services. Seoul provides a powerful example of how a major city in a small country can put in place policies and % of population, urban 60 Fiji investments to deliver agglomeration econo- 50 Indonesia China mies (box 3.1). Thailand 40 Lao PDR Faster Urbanization, Greater 30 Agglomeration, and Greater Growth Vietnam Dividends for Rwanda 20 Rwanda has started its journey to urbaniza- 10 tion, and the government has picked urban- 200 2,000 20,000 ization as one of the key areas for economic GDP per capita (constant 2005 international $) transformation and job creation. Rwanda’s population has become considerably more Source: Based on World Development Indicators data (World Bank, urban since 2002. The urban population various years); Economist 2012. Note: Bottom of arrow = 1985; top of arrow = 2010. almost doubled during this time (box 3.2). Kigali is the largest urban agglomera- tion (1,133,000 people in 2012) and home lower transport costs for goods and peo- to a major share of Rwanda’s urban popula- ple. Well-functioning cities also can create tion. With population growth of 6.4 percent markets for specialized services, including a year from 2002 to 2012, Kigali is one of logistics, advertising, legal support, and the fastest-growing cities in Africa and the management consulting. These services are most favored destination for rural migrants. critical for entrepreneurs, who can focus on In 2014, 57 percent of Kigali’s population their “big idea” or core competencies without had migrated to Kigali at some point in their having to worry about developing in-house lives, and 14 percent had migrated from rural supporting functions. Cities are also instru- areas in the past three years. mental in matching skilled people with jobs Sectors bordering Kigali have also seen that value those skills. International evidence higher than average population growth.1 suggests that a 1 percentage point increase Half of the urban population outside Kigali in the urban population is associated with a is found along two urban corridors: the 3–8 percent increase in a country’s per capita Musanze–Rubavu corridor (one-third)2 and income (Rosenthal and Strange 2003). Each the Muhanga–Huye corridor (18 percent). 3 doubling of its size raises a city’s productivity Much of Rwanda’s remaining urban popula- by 5 percent on average. tion is spread between the roads connecting Successful urban-led development has Kigali to Bugesera, Kayonza, the Burundian often involved intense spatial concentra- border, and the more isolated settlements tion of economic activity, typically in loca- of Nyagatare and Rusizi. Almost a quar- tions that are well connected to domestic ter of the urban population resides in small and external markets. A common feature settlements of fewer than approximately of rapidly growing Asian countries has been 15,000 people (Diao, Randriamamonjy, and the heavy concentration of economic activity Thurlow 2017). in and around coastal metropolitan regions Urbanization in Rwanda has contrib- (map 3.1). This pattern is no less striking in uted significantly to economic growth and F aster U r b ani z ation , G reater A gg l o m eration 139 MAP 3.1  Visualizing economic concentration across five countries a. Republic of Korea b. Japan c. China Seoul Sapporo JAPAN Ansan Tokyo Chongju Taejon Daegu Ulsan Nagoya Osaka Chonju Pusan Kwangju Hiroshima Changwon Kitakyushu Fukuoka Mokpo d. Belgium e. United States Portland Seattle Chicago New Brussels Washington, DC York BELGIUM Antwerp FLANDERS Denver Boston Gent WALLONIA Phoenix Dallas- Leuven San Ft. Worth Brugge Francisco Miami Los Houston Angeles Sources: For panels a and d, WDR 2009 team and World Bank Development Research Group, based on subnational GDP estimates for 2005. See also Nordhaus 2006. For panels b and e, WDR 2009 team. For panel c, World Bank 2015. BOX 3.1  Seoul becomes a global city Home to more than 10 million people, produc- to constrain Seoul’s growth and promote 14 cities ing one-quarter of national GDP on 1 percent across the country. of its land area, Seoul is the Republic of Korea’s These controls were not successful. Investors and i nterlocutor with the global economy. The broader ­ citizens led a rapid succession of new projects regard- S eou l Met ropol it a n A rea (SM A) compr ise s less of policies, including residential complexes, 29 towns (including Incheon and Seoul) on 12 per- metropolitan highways, new towns, and a new cent of the country’s land; this area contains half international hub airport—bypassing the objective of the national population and generates about of limiting growth. However, the controls did hurt half of national GDP. efficiency: more than 200,000 factories in the SMA Over the past 70 years, tension has arisen over are unregistered, contributing to unmanaged urban how best to manage the SMA. Some policy mak- development. And urban growth is fragmented, exac- ers, concerned that rapid SMA growth would lead erbating congestion and environmental degradation. to regional imbalances, argued that such growth As Korea sought to position Seoul as a 21st-­ should be constrained with strict regulations. century world city, the government relaxed quantity Regulations restricted development in certain limits that restricted the location of new colleges, areas, imposed fines on buildings exceeding given firms, industrial estates, and housing sites. It adopted parameters, limited allocations of land to industry, price instruments to regulate urban construction, limited the industrial output allowed in the SMA, levying a development charge on new commercial prohibited various classes of activities, and required buildings. By recalibrating city management through national government approval for land develop- deregulation and market instruments, Korea is mak- ment projects exceeding 1 million square meters. ing Seoul’s quest to become a global city more likely In the 1970s a greenbelt strategy was employed to succeed. 140 FUTURE DRIVERS OF GROWTH IN RWANDA BOX 3.2  Measuring urbanization in Rwanda A change in the definition of urban areas has made population, density, administrative, size, and zon- it difficult to compare urbanization levels in 2002 ing aspects. It defines an urban area as a built-up and 2012 censuses. agglomeration if it exceeds 20 square kilometers The 2002 and 2012 population censuses show an and has a population of more than 10,000 inhabit- almost stagnant level of urbanization in Rwanda— ants, resulting in a population density of more than with the urbanization rate falling slightly from 16.9 500 persons per square kilometer. to 16.5 percent. This decline is at odds with the Law no. 10/2012 of 02/05/2012, “Governing urbanization that is observed on the ground. Urban Planning and Building in Rwanda,” defines This counterintuitive result is partly explained three categories of urban settlements according by a change in the definition of urban areas between to population size. It categorizes urban areas into the two censuses. For the 2002 census, 15 cities (1) a city with at least 200,000 inhabitants, (2) a were delineated and all settlements within these municipality with at least 30,000 but less than catchment areas were considered urban, even those 200,000 inhabitants, and (3) an agglomeration with that previously would have been classified as rural. at least 10,000 inhabitants but less than 30,000 The 2012 census defines urban-rural status based inhabitants. on the smallest administrative entity: the village. To An alternative approach to measuring and qualify as urban, a village has to possess an impor- comparing past and recent urbanization levels in tant built-up area and important infrastructure Rwanda. It is possible to study urbanization over (education facilities, electricity and water, markets, time in Rwanda by applying a simple approach to and banks and other financial institutions). As a classifying settlements on the basis of population result of the change in definition, the two census fig- size and density. Settlements are commonly classi- ures cannot be compared directly to evaluate urban fied as urban if they contain a minimum of 5,000 population trends. persons living at a minimum density of 1,000 per- Classifying settlements on the basis of the pres- sons per square kilometer. ence or absence of infrastructure also risks over- Applying this definition to an analysis of geo- looking true urbanization patterns, such as the coded population data indicates that Rwanda’s emergence of large, densely populated areas on the urban population increased from 15.8 percent in edges of formally planned areas that are lacking 2002 to 26.5 percent in 2015. This reflects growth infrastructure but are nonetheless functionally part of the urban population of just under 2 million of the broader urban landscape. people, from 1.49 million to 3.46 million between Different definitions of “urban” used in Rwanda. 2002 and 2015, at an average annual rate of Beyond the change of definition between the two 6.7 percent. These estimates are based on an analy- censuses, different institutions and policies have sis of population data from the WorldPop project defined “urban” in Rwanda differently. and an analysis that classifies settlements by break- The National Land Use Master Plan defines ing the national territory into grid cells of 1 square “urban” and “built-up” areas from a land use kilometer, as opposed to villages or sectors, which planning perspective, which considers spatial, have variable areas. structural change through the reallocation of A comparison of wages for similarly educated labor off farm. The contribution of urbaniza- workers provides a broad estimate of the tion to growth depends on the productivity urban-rural productivity gap; urban work- gap between similar sectors in rural and urban ers in Rwanda earn, on average, twice the areas, which occurs because (1) urban work- rural wage for similar work in similar sec- ers are generally better educated and higher tors of employment. Using these estimates, skilled and (2) capital-labor ratios tend to be analysis done for this report shows that higher in urban sectors where capital invest- urbanization has accounted for 37 percent ments are more likely to be concentrated. of national structural change (GDP growth F aster U r b ani z ation , G reater A gg l o m eration 141 through labor reallocation across sectors), and FIGURE 3.2  Contribution of urbanization to urban areas have accounted for 48 percent of structural transformation and labor productivity in Rwanda, 2002–12 national labor productivity growth over the past 15 years (Diao, Randriamamonjy, and Thurlow 2017; figure 3.2). 60 55.3 Rwanda’s economy could be 20 percent 51.9 47.7 larger by 2050 with faster urbanization. 50 42.4 A spatial computable general equilibrium 47.2 (CGE) model developed for this report exam- 40 34.5 42.5 Percent ines how alternative urbanization scenarios 37.0 30 would affect key economic and welfare out- 30.1 comes in Rwanda until 2050 (see annex 3A; 20 Diao, Randriamamonjy, and Thurlow 2017 21.1 for details). On the basis of past trends in 10 migration and population growth, the urban population’s share of total population (using 0 current official definitions) is expected to 1.0 1.5 2.0 2.5 3.0 grow to 41.2 percent in 2050. Simulations of Ratio of urban to rural average labor productivity faster urbanization are developed by increas- Urban economy’s contribution to total ing rural-urban migration above historical labor productivity growth rates. In 2050, the urban share of the popu- Urbanization’s contribution to between-sector growth component lation is increased to 54.6 percent, up from 41.2 percent in the baseline.4 This scenario Source: Diao, Randriamamonjy, and Thurlow 2017. is consistent with the targeted pace of urban- Note: The x axis shows the assumed ratio of urban to rural labor ization outlined in Vision 2020. The model productivity in the same sector. For example, how much more productive is an urban versus a rural construction worker? These shows that, compared with on-trend urban- assumptions are combined with data on the productivity of ization, faster urbanization raises the total different sectors and changes in the sectors in which people are employed in rural and urban areas. GDP growth rate by 0.5 percentage point per year over the 2015–50 period. Cumulatively, FIGURE 3.3  Simulated changes in growth outcomes with faster this means that by 2050 the economy is urbanization in Rwanda almost a fifth larger than it would be with slower urbanization (figure 3.3). Almost all of this additional GDP growth is generated National 17.8 in Kigali, where value added increases by 63 percent compared with the baseline. This Kigali 63.4 positive growth effect is due entirely to more people leaving rural areas for urban centers—­ Peri-urban 2.5 GDP per capita that is, the total population and workforce –17.9 remain unchanged from the baseline. Remote rural To achieve the rapid economic growth targeted until 2050, Rwanda will need to Agriculture 7.5 harness agglomeration economies through well-managed cities. Kigali is undoubt- Industry 23.7 edly the prime economic hub of Rwanda. Its output exceeds that of the six second- Services 18.4 ary cities plus Bugesera, Gicumbi, and Rwamagana-Kayonza combined. Firms –20 0 20 40 60 80 registered in Kigali accounted for 92 per- % deviation from baseline, 2050 cent of all turnover declared to the Rwanda Revenue Authority in 2015, and more than Source: Diao 2017. 142 FUTURE DRIVERS OF GROWTH IN RWANDA half of all formal firms and half of formal scale. They arise because of fixed costs of employment are located in Kigali. Per capita production (internal to a firm) and proxim- consumption in Kigali is five times that in ity to workers, customers, and people with rural areas, and, although regional GDP new ideas (external to a firm and even an is not calculated, nightlights-based esti- industry). However, places do not have to be mations suggest that Kigali accounted for huge to generate some of these gains. The 40 percent of GDP in 2012. Kigali is also size of settlements matters less than their the most urbanized economy, with 80 per- function: with reasonable transport costs, cent of nonagricultural employment even towns can be large enough to facilitate inter- in 2002, compared with just 30 percent in nal scale economies. Medium-size cities are other urban areas. A dynamic Kigali is cen- often large enough for “localization” econo- tral to harnessing agglomeration economies mies that come from thick input markets. in Rwanda. If managed well, Kigali has the Nevertheless, “urbanization” economies— potential to become a much larger economy especially those involving knowledge spill- and to provide most of the gains from faster overs—are generated mainly by large cities. urbanization; it will be home to a much The implication is that policy makers should larger population, who will need well- focus on the functions of different types of planned neighborhoods, affordable hous- cities and support agglomeration though ing, services, physical mobility, and digital economic density. connectivity. The nature of agglomeration economies differs by the function of different settle- ments. Rwanda’s small towns, secondary A Vision for Rwanda’s Urbanization cities, and Kigali should be performing Faster urbanization and greater agglom- complementary functions, each differenti- eration hold the potential to accelerate eco- ated by the type of scale and agglomeration nomic growth and structural transformation economy they can deliver. It would be coun- in Rwanda. The movement of workers off terproductive to slow the growth of Kigali farm into urban areas and the spatial con- and disperse people and businesses to sec- centration of economic activity are central to ondary cities. Businesses locate in Kigali Rwanda’s long-term growth. because of the range of diverse financial The World Development Report 2009: and related services not available elsewhere Reshaping Economic Geography provides a in the country (urbanization economies). framework for realizing the potential gains Indeed, several firms are located in Kigali from urbanization (World Bank 2009). to enjoy improved trade facilitation and Building on research in industrial orga- business services and to benefit from nization, economic growth, international strong transport links from Kigali to all of trade, and economic geography, the frame- Rwanda’s borders. Substitutes for Kigali are work explicitly incorporates the effects of not going to be found in Rwanda’s second- scale economies in production. The main ary cities or smaller towns. Many countries insight is that rising economic density—­ have tried to slow the growth of their large concentration of people, investment, and cities to balance development outcomes economic activity—is critical for faster eco- spatially, but they have had limited success. nomic development. Economic density is For example, policy makers in the Arab supported by shortening economic distance Republic of Egypt have encouraged the through lower transport and trade costs and development of secondary cities in order to increasing incentives to enhance scale and reduce population pressure in Cairo, which specialization. accounts for 50 percent of Egypt’s GDP but Economic density is a major driver of pro- just 0.5 percent of its land. These towns, ductivity. Large plants that operate in dense meant for 5 million people, have drawn environments generate increasing returns to barely 800,000. F aster U r b ani z ation , G reater A gg l o m eration 143 From a long-term perspective, policy 2000, attracted graduates at twice the rate makers in Rwanda need to take a bullish of the other 75. The reason is that skilled view toward unleashing the potential of workers gain from proximity to others. The Kigali to nurture agglomeration economies implication is that policies should not fight through scale and specialization, inter- the ­m arket forces that pull skilled people twining Rwanda’s growth engine with the together; migration is the most effective way regional and global economy. A differen- to reduce distance and thus to promote eco- tiated approach is required for each type nomic density. of settlement, informed by its function, Investments in mass education across opportunities, and challenges. Market Rwanda can encourage faster rural-to-urban forces—reflected by the location decisions migration and support opportunities for of people and businesses—have chosen more advanced specialization and knowl- Kigali. Durable investments in infrastruc- edge spillovers. Education encourages rural- ture, housing, and other amenities will be urban migration: only 1.5 percent of rural needed to ensure that the city is productive residents who did not complete primary and livable. The prospects for growth of school migrated to urban areas, compared different locations are difficult to predict, with 2.8 percent who completed primary, because they depend on the future of agri- 7.6 percent who completed secondary, and culture, peace, and conflict in neighbor- 16.2 percent who had some higher educa- ing countries, the development of mineral tion (World Bank 2017a). Poor learning out- deposits, major infrastructure investments, comes in Rwanda today have implications and many other factors. Thus, potential for the sequencing of urban-led growth; investments in other settlements should be urban areas may gain the most dynamism divided into those that can be made at low after only 10–15 years, once higher invest- risk before market demand has emerged ments in education today filter through to versus those that are better suited after it the labor market. Returns to education are has emerged. Here, choices will need to high in urban areas in Rwanda. Incomes be made between investments in place- are five times higher in Kigali than in rural specific durable assets (such as infrastruc- areas, and 57 percent of this difference is ture, industrial complexes, and housing) explained by differences in the returns to and people-focused portable assets (such education (33 percent is due to higher levels as health, education, and water and sanita- of education). tion). Portable assets are important both for Finally, overcoming frictions from high improving the welfare of people in situ and transport costs can help to amplify scale for facilitating long-term mobility. Further, economies in Kigali, as well as to provide broad institutional and governance reforms the potential for other cities to develop such as clarification of land and property their vocations. Falling transport costs tend rights set the foundations for urbanization to increase trade more with neighboring, and urban development. not distant, regions or countries; trade has Reducing the distance between people and become more localized than globalized. economic density can amplify agglomeration Places trade more with regions or countries economies. While conventional economic that are similar, because falling transport analysis implies that people should move to costs increase specialization. Thus, the basis where their skills are scarce, in reality human of trade increasingly becomes the exploita- capital tends to move to where it is abun- tion of economies of scale rather than the dant: educated migrants seek places where differences in natural endowments. The many others have similar skills. Among implication is that falling transport costs the 100 largest metropolitan areas in the change the composition of trade and make it United States, the 25 cities with the highest even more sensitive to such costs. Policies to share of college graduates in 1990 had, by reduce trade and transport costs should be 144 FUTURE DRIVERS OF GROWTH IN RWANDA a big part of growth strategies for low- and FIGURE 3.4  Puga Index of connectedness in middle-income economies. Kigali and select comparators Greater agglomeration supported by faster migration and specialization are the Kigali (Rwanda) fundamental drivers linking urbanization Mombasa (Kenya) with faster economic development and struc- tural change. Policies and investment should San Salvador (El Salvador) help firms and workers to benefit from Monrovia (Liberia) these transformative forces. Institutions and regulations should strengthen land mar- Managua (Nicaragua) kets and business development, support the spatial mobility of workers and capital Durg-Bhilainagar (India) (factor markets), and integrate product mar- Tegucigalpa (Honduras) kets. Infrastructure—for example, trans- port and information and communication Niamey (Niger) technology—is critical to connect regional Islamabad (Pakistan) product markets and urban labor markets. Finally, spatially targeted interventions may Kathmandu (Nepal) be needed to support the development of untapped resources. But these interventions 0 5 10 15 20 25 need to be implemented with great care, Puga Index because they often do not solve underlying (for distance of 10 kilometers) market or coordination failures and some- Source: Lall, Henderson, and Venables 2017. times are implemented at the expense of Note: The Puga Index considers the share of city residents accessible national economic growth. within 5 or 10 kilometers of each location in the city. Source locations are weighted according to their share of the urban The next section assesses key constraints, population. followed by a section outlining priorities for policy and investment. experienced large population increases from 2002 to 2012 (for example, net migration over this period was equal to 21 percent Constraints to Successful of the 2012 population in Gasabo), while Urbanization in Rwanda some more central areas retained similar, or even smaller, populations (net migration Low Density of Urban Settlement was only 3 percent of the 2012 population Development in Rwandan cities is spatially in Nyarugenge). From 2000 to 2012, more fragmented. Neighborhoods within Kigali than 40 percent of Kigali’s land development are poorly connected with one another: was in small patches disconnected from the development is common in many, sepa- preexisting urban fabric. Moreover, Kigali’s rate parts of the city rather than concen- built area expanded greatly from 1990 to trated in the center and along transport 2010, consistent with increasing fragmen- corridors. This translates into scattered tation (map 3.3). Rapid physical expansion peak densities beyond the central business also has been observed around second- district instead of clustered densities that ary cities, particularly Musanze, where the could enhance scale economies (map 3.2). urban footprint increased by 6 percent a Kigali scores very low on the Puga Index, year from 1999 to 2015, although less so in a measure of the degree of connectedness Huye (map 3.4). of people in a city (Lall, Henderson, and Spatial fragmentation reduces the ben- Venables 2017) compared with cities of efits of urbanization for productivity. The similar size (­ figure 3.4). Fewer central areas large distances between firms and workers F aster U r b ani z ation , G reater A gg l o m eration 145 reduce workers’ access to jobs by raising MAP 3.2  Fragmentation of densities beyond the central business commuting costs, thus limiting firm size; district of Kigali prevent job market pooling and match- Kigali ing; and ­ h inder the transfer of skills and Population per sq km2 0–265 knowledge. A 1 percentage point increase 266–605 606–1056 in connectedness as measured by the Puga 1057–1698 1699–2529 Index is associated with urban costs that 2530–4047 are 12 percentage points lower, controlling 4048–6456 6457–9662 for income levels and city population size. 9663–15490 15491–27094 To a considerable extent, high urban costs are driven by poor land management prac- tices, as efficient land use requires efficient city and neighborhood grids and appropri- ate zoning regulations that encourage den- sity, affordability, and formality (Wainer, Ndengeyingoma, and Murray 2016). Such high costs play an important role in reduc- ing the international competitiveness of labor in African cities5 by putting pressure on firms to pay higher wages; although Source: Henderson and Nigmatulina 2016. Note: The figure represents (through tower heights) the population density across Kigali wages, in absolute terms, are typically Province. High densities are found in the central business district, but also are scattered lower in Sub-Saharan Africa than in com- across the city. peting destinations, they are higher than expected considering income and skill lev- With the rising complexity of the urban els (for example, Gelb et al. 2017). environment, institutional adaptation is Fragmented urban land development now required to build flexibility into plan- may encourage informality, because the ning. In dynamic, fast-changing cities, the most accessible urban jobs are those pur- economic vocations of places are trans- sued at and around the home. For example, formed regularly as the economy grows in in Kigali, the home is the usual headquar- sophistication, and land markets facilitate ters for the 72 percent of firms that are the transformation of structures on land to run by a single individual (table 3.1). To be meet the economic needs and opportuni- sure, formal jobs are concentrated in cen- ties of the time. The use and form of struc- tral Kigali, just as they are in most cities; tures undergo successive cycles of change other examples include Kampala (Uganda) and renewal, as structures depreciate and and Lusaka (Zambia) (map 3.5). But 97 productive opportunities evolve. In other percent of all firms in Kigali are informal. countries, master plans communicate the Of those informal sector firms, 90 percent development vision for locations, coordi- engage in nontradable activities domi- nate investments, and raise confidence in nated by retail, personal services (such the future environment, but they also are as domestic work) and local construction “living” documents that are updated regu- labor. Such nontradable activities are not larly in response to market signals and citi- concentrated in the center; rather, they zen feedback and typically are not legally are scattered throughout Kigali’s urban binding. expanse. Regulatory inconsistencies across neigh- Fragmented development is driven partly boring jurisdictions and within cities also by building requirements tied to the use of have contributed to fragmented develop- master plans. This is due to inadequate local ment. Within the boundaries of master urban planning capacity. plans, zoning regulations are often far 146 FUTURE DRIVERS OF GROWTH IN RWANDA MAP 3.3  Built land in Kigali, 1990, 2000, and 2010 less demanding, inspections are less thor- ough, and master plan implementation is scheduled for later stages the closer to the periphery. Combined with “futuristic” cen- tral regulations, this discrepancy creates perverse incentives to build in the peripher- ies, in exactly the scattered and inefficient way that planners wish to avoid. New con- struction in Kigali from 2009 to 2015 was strongly concentrated in peripheral areas; even building improvements were more common in the suburbs than in the cen- ter (map 3.6). Likewise, areas just outside Kigali’s border have less stringent require- ments and have seen considerable develop- ment. These areas include Kamonyi District, parts of which are just 7.5 kilometers from Kigali’s central business district and whose population increased 28 percent annually from 2002 to 2012. Poor transport infrastructure exacer- bates the impact of spatial fragmentation. In Kigali, the connectedness of the local Source: Jones, Murray, and Ferreira 2016. Note: Data generalized from Landsat images showing built cover in 1990 and newly population by public transport is already built-up land in subsequent decades. strained: public transit routes to the center MAP 3.4  Building footprint in central areas of three secondary cities in Rwanda a. Central Musanze, 1999 and 2015 b. Central Muhanga, 2001 and 2015 c. Central Huye, 2002 and 2015 Huye bu er - 2 km Urban mask 1999 Urban mask 2001 Urban mask 2001 Urban mask 2015 Urban mask 2015 Urban mask 2015 0 0.25 0.5 1 1.5 0 0.25 0.5 1 1.5 0 0.25 0.5 1 1.5 Kilometers Kilometers Kilometers Total expansion 1999–2015: 120%; Total expansion 2001–15: 81%; Total expansion 2002–16: 34%; Average yearly expansion: 6% Average yearly expansion: 4.4% Average yearly expansion: 2.2% Source: World Bank 2017e. Note: The maps show the footprint of buildings for three secondary cities in 2001 or starting year (red) and 2015 or ending year (yellow). F aster U r b ani z ation , G reater A gg l o m eration 147 TABLE 3.1  Share of firms in Kampala, Uganda; Lusaka, Zambia; and Kigali, Rwanda, by size, 2011 % of firms Size of firm Kigali Lusaka Kampala Self-employed 72.23 50.65 55.5 Micro (2–4) 22.49 28.36 36.03 Small (5–49) 5.05 18.74 6.25 Medium (50–99) 0.14 1.15 0.13 Large (100+) 0.08 1.09 0.09 Total number of firms 123,364 17,117 184,335 Sources: Jones, Bird, et al. 2016; Jones, del Frari, et al. 2016; Jones, Murray, and Ferreira 2016. Data from Rwanda, Uganda, and Zambia economic censuses (2011). MAP 3.5  Location of formal jobs in Kampala, Uganda; Lusaka, Zambia; and Kigali, Rwanda Employment density No rm Low High Sources: Jones, Bird, et al. 2016; Jones, del Frari, et al. 2016; Jones, Murray, and Ferreira 2016. MAP 3.6  Changes in building stock in Kigali, by cell, 2009–15 a. Greenfield construction in peripheral northeast b. Rebuilt or upgraded buildings in suburbs Study area Study area Number of newly Number of build buildings buildings 0 upgraded to a 1–10 higher building 11–30 class 31–80 0–9 81–125 10–31 126–200 32–60 201–400 61–101 401–500 102–164 501–1000 165–306 1001–5112 N N Kilometers Kilometers 0 1 2 3 4 0 1 2 3 4 Source: Bachofer and Murray 2018. Analysis conducted through remote sensing of aerial and satellite images, plus manual correction. Note: These maps show the frequency of key changes in the building stock (new greenfield buildings, panel a, and building upgrades, panel b) by cell in Kigali. In a period of rapid population growth, there has been very little transformation in central Kigali. Construction on previously unbuilt land was strongly concentrated in peripheral areas (cells shaded dark brown, in panel a, saw between 1,000 and 5,000 new buildings on greenfield sites from 2009 to 2015). Even building improvements were more common in suburbs (panel b). 148 FUTURE DRIVERS OF GROWTH IN RWANDA are congested, and significant central areas Slow Pace of Rural-Urban Migration are unplanned or underdeveloped settle- Limited rural-to-urban migration constrains ments with few vehicular access routes. urbanization and the resulting benefits Further, models show that transport will from agglomeration. Rural-urban migra- become extremely restricted with rising tion has been a key component of urbaniza- congestion, deeply fragmenting urban labor tion in many countries, especially in East markets. As in many African cities, Kigali Asia. For example, rural-urban migration allocates little land to roads. In a sample of accounted for 41 percent of urban residents 30 cities around the world, 8 African cities in China during the 1990s (Cai, Park, and rank in the bottom 12 spots for road density, Zhao 2008). By contrast, only 14 percent including Kigali (ranked 19), Addis Ababa of Rwanda’s urban population were recent (24), and Nairobi (27). African cities devote migrants from rural areas in 2014.6 Of all less than 16 percent of their land to roads, internal migration between 2011 and 2014, compared with more than 20 percent in most only 20 percent went from rural to urban cities in high-income countries. The lack of areas (figure 3.5); rural-rural migration was roads, coupled with low affordability of the dominant form of internal population motorized public transport, means that loca- movement.7 Population movements included tions across the city are difficult to reach. both moves toward density, with Kigali Low affordability of transport increases the attracting 29 percent of recent migrants number of people who must walk, while the (including migrants from other urban areas), lack of roads increases congestion and pol- and parallel moves away from density, with lution. Quiros, Murray, and Bajpai (2014) Eastern Province (the least densely populated found that, when bus speeds drop 50 per- province in Rwanda) attracting 33 percent of cent because of congestion (for example, at recent migrants. Investing more in education peak hours), residents in Kigali’s periphery in rural areas could play an important role in (shaded red in map 3.7) lose access to 50–66 increasing rural-urban migration. percent of jobs in the city (where access is Even if growth is spatially unbalanced, defined as a job that can be reached within growth in Kigali can be expected to improve an hour by bus and walking). household incomes throughout the country, through migration, remittances, and demand MAP 3.7  Simulated change in access to employment in Kigali with for rural products, not to mention tax revenue a 50 percent drop in bus journey speed for broad-based public service provision. Data from the Integrated Living Conditions Survey 2013–14 (EICV4) show that Kigali’s house- holds send more than twice the value of remit- tances to rural areas as all other Rwandan towns combined (NISR 2015) (­ figure 3.6).8 Rwanda has pursued a policy of grouped villages (imidugudu) (map 3.8), with the goal being to improve access to health and edu- cation services as well as broader infrastruc- Legend ture and to preserve agricultural land. The Public transport Changes in employment policy also lays a fundamental foundation accessibility –10 to 0 that reduces land fragmentation, improves –20 to –10 –30 to –20 land use efficiency, and encourages creation –40 to –30 of agglomeration economies that stimulate –50 to –40 –66 to –50 rural productivity and consequently rural OSM base layer transformation. The investments made have Source: Quiros, Murray, and Bajpai 2014. had a positive impact on land consolidation F aster U r b ani z ation , G reater A gg l o m eration 149 FIGURE 3.5  Share and scale of rural-urban migration in Rwanda, 2014 9% of people moved between Among them, 20% moved from 2011 and 2014 in Rwanda rural to urban areas These rural-to-urban migrants account for 20% 14% of the urban 9% residents as of 2014 80% 14% 86% 91% Rwanda urban population 2014 Rwanda population 2014 Source: Calculation based on NISR 2015. FIGURE 3.6  Remittances from Kigali (and other locations) to rural areas of Rwanda, by province 45 Share of remittances to each location (%) 40 35 30 25 20 15 10 5 0 Rural eastern Rural northern Rural southern Rural western Recipient location Kigali Other countryside Other town Same village or town Source: Integrated Living Conditions Survey (EICV4) data (NISR 2015). and productivity. They have also had an many highly scattered rural settlements. The immediate impact on quality of life (decent government needs to ensure that this invest- housing and access to infrastructure and ser- ment does not become ­ r edundant in the vice, among others). future under a higher-growth strategy char- However, there is a need for a clear rural acterized by rapid growth in urban areas and settlement planning and development strategy driven by investments in higher-demand loca- linked to the national development agenda. tions, improvements in rural-urban connec- This requires large public investments in tivity, and more rapid rural-urban migration. 150 FUTURE DRIVERS OF GROWTH IN RWANDA MAP 3.8  Location of planned imidugudu (grouped villages) in 12 districts of Rwanda, 2016–17 400000 450000 500000 550000 600000 4880000 4880000 MININFRA Nyagatare 4840000 4840000 Burera Musanze Gicumbi Rubavu Nyabihu Gatsibo Gakenke Rulindo 4800000 4800000 Kayonza Rutsiro Ngororero Gasabo Nyarug Muhanga Kamonyi Rwamagana Kicukiro enge 4760000 4760000 Karongi Ruhango Ngoma Bugesera Kirehe Nyanza Nyamasheke Nyamagabe Legend Districts 4720000 4720000 Huye Lakes Rusizi National Parks Gisagara Human Settlements Projects Nyaruguru IDP Model Villages Elaborated by: Rural settlements site zoning plans 1. Eugene M, GIS Specialist/MININFRA Houses for relocated households from HRZs 2. Theophile M, GIS Specialist/RHA Existing roads 3. Basile I, R&D Specialist/RHA RHA access roads 4680000 4680000 0 15 30 60 90 Km Affordable land acquired 4. Innocent M, Logistics Officer/RHA 400000 450000 500000 550000 600000 Source: MININFRA 2017. Note: The map, made available by the Ministry of Infrastructure, shows in red the location of planned imidugudu in the 12 districts planned so far. Developing so many imidugudu (one is required by law for every 5 kilometers of the countryside) may be extremely costly and consume considerable agricultural land (because of large plot size and high servicing requirements). Imidugudu are not located on economically strategic locations, for example, along roads or around secondary cities. HRZ = high-risk zones; IDP = Integrated Development Programme; RHA = Rwanda Housing Authority. Planning for imidugudu should be strate- countries. Half of the rural population gic and reliant on well-studied spatial aspects lacks access to a road network in good con- that will trigger effective rural productivity, dition within a walking distance of 2 kilo- while minimizing huge infrastructure invest- meters (World Bank 2017d). The impact ment costs and potential financial losses. As of poor-quality rural roads is evident in such, settlements could be located in places the wide variation in food prices within with higher connectivity to urban areas, and across districts, implying poorly inte- which would improve rural residents’ access grated national food markets (for example, to services, product markets, and employ- map 3.9). The lack of rural-urban links ment in small towns, secondary cities, and limits the opportunities for rural workers Kigali and, in turn, raise demand for clus- (including ­off-farm rural workers) to raise tered settlements. their efficiency, specialization, and ulti- mately productivity, which can slow rural poverty r­ eduction and constrain the sup- High Transport Costs and Limited ply of inputs to urban areas. Poor quality Market Access is due to very low spending on infrastruc- The poor quality of rural roads impairs ture repair and maintenance by districts the integration of Rwanda’s cities with containing secondary cities. The level of the countryside and with neighboring spending is an order of magnitude lower F aster U r b ani z ation , G reater A gg l o m eration 151 MAP 3.9  Variation in tomato prices in western raise the cities’ impact on the surrounding Rwanda environment. Connectivity between urban areas is Legend also important for integrating the national Tomato prices above mean economy. Korea used incremental invest- Tomato prices below mean ments to improve connectivity between Markets unavailable cities, increasing specialization, reaping Feeder roads economies of scale, and helping major cit- District_boundaries ies to lift the entire population (map 3.10). In Rwanda, a clear trend of urbaniza- tion along major transport routes (urban corridors) is also emerging (World Bank 2017c). By strengthening such urban-urban links (for example, through improved conditions for night transport, more effi- cient routes, and the effective location of industrial parks), Rwanda can leverage its compact geography to deliver an inte- grated portfolio of urban places. Such connectivity encourages the efficiency- driving specialization of productive activi- ties between locations. However, it does not guarantee the greater development of secondary cities and other towns versus Kigali. Connectivity and specialization could also entail greater concentration of activities in Kigali. Priorities for Faster Urbanization, Greater Agglomeration Getting on the path to faster urbanization and greater agglomeration will involve determining where to prioritize urban development, what policies and invest- Source: World Bank 2017d. ments would yield high dividends, and who needs to be in charge of implementing solutions. Although the optimal choices than it should be, based on international on the where, what, and who of urban comparisons relevant to Rwanda (World development differ across countries, inter- Bank 2015). national experience during episodes of A particular missed opportunity lies high growth and assessment of Rwanda’s in the fact that a third of Rwanda’s poor u rba n i zation h ig h lig ht t he follow i ng live within 20 kilometers of a second- priorities: ary city, yet the productivity and welfare advantages of these cities are undetectable • Where. A bullish view is needed to unleash more than 5 kilometers from their centers the potential of Kigali, which market (World Bank 2017b). Strengthening rural- forces—reflected in the choices made by urban links around secondary cities could people and businesses—have chosen as the both improve their access to inputs and main center of production and commerce. 152 FUTURE DRIVERS OF GROWTH IN RWANDA MAP 3.10  Expressways in the Republic of Korea, 1970–2010 Source: Wang 2011. Note: The colors represent locations within 10, 20, and 30 minutes of an expressway station for the given year. For rapid national growth, policies and approach toward a complementary (and investment need to follow market signals, well-networked) set of secondary cities accelerate Kigali’s development, and inter- and small towns. twine its economy with that of Africa and • What. Investment in infrastructure, hous- the world. Kigali is Rwanda’s main inter- ing, and commercial structures is needed face with regional and global economies to manage urbanization; but bedrock and needs to be allowed to respond to reforms that strengthen land markets and market demand and fulfill its significant urban planning should be in place first to economic potential. These policies would enhance the efficiency of investment and be accompanied by a more calibrated generate high economic returns. F aster U r b ani z ation , G reater A gg l o m eration 153 • Who. Stronger intergovernmental coordi- and districts are therefore allowed the highest nation of economic and spatial planning densities. Outlying areas adjacent to transit processes is needed to overcome coordina- stations or where highways intersect also have tion challenges in urban development. higher density allowances to increase uptake of the transit system and limit the use of cars. Such regulations are intended to ensure that Supporting Agglomeration Economies the urban form does not become costly and in Kigali and Other Cities inefficient to service, rather than to force den- Master Planning for Density sity in locations where market demand does Efforts to increase density require a detailed, not yet exist. flexible approach. There is no “optimal” Such density regulations are developed density, but defragmenting the urban fabric iteratively, to work with rather than against allows for density to change in response to market demand, nudging rather than over- the needs of businesses and people. Cities hauling investors’ plans. Regulations that need Goldilocks density—not too high or fight against market demand, by contrast, low, but just right (Guardian 2014). And inevitably result in informality and sprawl in what is “just right” changes over time and less regulated areas and underinvestment in varies by neighborhood. high-potential locations. Likewise, an exces- Density should not overwhelm infrastruc- sively detailed and rigid quantitative regula- ture; nor should regulators keep density below tion of land use (such as zoning for detailed what the infrastructure can support. Planners building characteristics and functions) in New York, Seoul, and Singapore have makes inevitable the misallocation of capi- taken the “just right” approach to heart by tal, reduces the adaptive capacity of the city, developing a fine-grain pattern of small den- and considerably dampens returns to private sity zones (map 3.11), guided by infrastruc- investment. tural (particularly transit) capacity. Areas of Private investments in urban structures in these cities that are economically dynamic, Rwanda’s urban areas in 2017 need not, and such as central business districts, can be should not, be regulated to deliver the spatial served efficiently by transit systems to accom- form for the 2040 economy. Moving closer modate large daytime populations; these cities to the allocation of structures according to MAP 3.11  Goldilocks density: Getting density “just right” through fine-grain differentiation and adaptation over time a. Singapore: FAR values in downtown area b. Seoul: Variations in FSI are linked to the location of metro stations and to the network of main streets Metro stations Land outside downtown boundary New_seoul_zoning.shp Water Community facilities and utilities 0.5 Conservation areas 0.5–1 Streets 1–1.2 Park and open space 1.2–1.5 FAR 1.5–2 1.6–3 2–2.5 3–3.8 2.5–4 3.8–5 4–6 Metro stations 5–7.6 6–8 7.6–11 8–10 11.2–25 Vacant reserve sites 500 0 500 1000 1500 2000 2500 Meters 1 0 1 2 3 4 5 6 Kilometers Source: Vishwanath et al. 2013. Note: FAR = floor area ratio; FSI = floor space index. 154 FUTURE DRIVERS OF GROWTH IN RWANDA market demand (prices) is recommended reducing minimum plot sizes and raising rather than relying on centrally planned maximum ground cover ratios in strate- zoning (quantity). Land prices reflect the gic locations. To achieve high densities in demand for and scarcity of locations; they lower-cost structures—such as terraced can efficiently allocate private investments housing with back yards as opposed to high across space and allow structures to serve the rises—small plots with high ground cover evolving economy, delivering the “Goldilocks ­ are necessary. Yet several master plan zones density” for each neighborhood. in Kigali promote large minimum plot sizes To achieve this vision, investors should be or a low maximum ground cover, even in guided to invest in structures with no s ­ pecific central areas.9 Revising the minimum plot building heights, floor area ratios, and so on sizes downward and revising the maximum that match neither the existing nor the poten- ground cover requirements upward, com- tial market demand. Regulations should bined with more regular adjustments of focus on mitigating negative externalities requirements in response to market demand, like environmental degradation, ensuring would go a long way to encouraging densifi- structural safety, and reserving space for the cation, affordability, and formality. existing or planned infrastructure grid. With Deregulation would not promote infor- a price-based allocation of structures, scat- mality and poor standards of housing and tered and sprawling land development can structures, as some fear; international evi- result if no area offers particularly higher dence shows that, with strong property rights amenities or more accessibility. Such develop- and infrastructure provision, deregulation ment is avoided by privileging certain areas tends to promote more formalization (see, for for higher density through the strong public example, box 3.3). Stringent regulations raise provision of transport corridors and public costs above affordability, leading people to spaces. live in informal dwellings. Legalizing struc- Key institutional requirements for the tures brings them into the formal ­ m arket price-based allocation of land include the where investments are secure, incentivizing strong protection of property rights, trans- owners to upgrade dwellings over time, while parent valuations processes, and oversight. also, typically, mandating government service Within the existing master plan frame- provision. Together, these are powerful mech- work, planning standards could be amended anisms to raise living ­conditions, encourage to encourage density and affordability by formality, and mobilize investment. BOX 3.3  Avoiding slums through property rights and infrastructure Hanoi, Vietnam. Despite rapid and low-income with infrastructure development. The Programa development, Hanoi has, uniquely, largely avoided de Mejoramiento Integral de Barrios targeted 26 the development of large slums. Faced with rapid of the poorest city areas, called the unidades. The urban expansion and mass illegal building, its city unidades included 1,440 informal settlements, authorities legalized all past unpermitted construc- 300,000 untitled plots, and about 500,000 struc- tion that met building standards. This effort incen- turally substandard dwellings. The program legal- tivized owners of previously illegal houses to invest ized homes and neighborhoods and expanded in their buildings and, in turn, committed the gov- infrastructure (roads, rainwater traps, and sani- ernment to providing roads, sanitation, and basic tary and aqueduct networks) and urban facilities services, ensuring an adequate standard of housing. (stairs, parks, and community rooms). Living con- Bogotá, Colombia. City leaders in Bogotá simi- ditions improved for about 650,000 people (World larly succeeded by coordinating regulatory reform Bank 2012). F aster U r b ani z ation , G reater A gg l o m eration 155 Servicing Land for Density but also for the wide urban expansion The market-based allocation of investments areas around secondary and emerging cit- across space will not provide the public space ies, which, as noted, have expanded rapidly and infrastructure essential to raise the pro- and with fragmented and largely unplanned ductivity of land. Investments like roads, land use. They may require some readjust- sanitation, and laying out urban grids require ment of plots, in conjunction with improved huge coordination and deliver returns over infrastructure services, in areas where devel- long horizons, so they must be provided by opment has been haphazard. It is critical for the state. Two priorities are (1) providing Kigali to “get ahead” of its rapid expansion dense and good-quality central infrastructure with a more intensive program to protect the to crowd in capital and (2) reserving space for grid and regularize plots in fast-developing roads in a wide peripheral urban expansion neighborhoods. A useful example of this area through grid planning. process is New York City, where the 1811 Kigali could increase the density of cen- grid plan envisioned an infrastructure net- tral building and investment by providing work for a city 10 times its size at the time. quality infrastructure services for central Infrastructure was not laid directly, but the areas. Demand for a location is typically city worked with communities to readjust related to its level of amenities, accessibility, plots to fit the new plan around dirt roads and security of investing there (particularly and stone way-markers. Expansion of the low risk of future expropriation). Broad-use city’s buildings responded to market forces infrastructure—particularly, roads and road and was orderly, with minimal expropriation grids— serve these three functions. costs and high formality (box 3.4). Similar While developing central servicing, grids practices have been tried in Rwanda, with should be provided for the wider urban promising results, but at a fairly low scale expansion area to ensure that rapid periph- (box 3.5). eral development takes place in a planned way, reducing future expropriation and Enhancing Urban Mobility adjustment costs and raising the security of Investment in public transport would help new investments. Such investments in the to reduce spatial fragmentation in Kigali. grid would be beneficial not just for Kigali Weak public transport around the central BOX 3.4  The New York grid: Transformation of structures and the economy around a planned grid The 1811 Plan of Manhattan mapped roads for an high demand from landowners for their areas to be area more than 10 times the area of the city at the readjusted. Property taxes created strong incentives time (map B3.4.1). This plan became the skeleton for implementation on the side of the city govern- around which the city expanded, supporting con- ment: as the roads raised land values, property tax nectivity, formality, and density and reducing future revenue in the city rose fivefold from 1830 to 1837. expropriation costs. A large share of road openings was funded through The grid was not imposed on a blank slate: in addi- direct benefit charges to adjacent landowners. tion to challenging topography, many l ­andowners The economic vocation of each neighborhood lost large parts of their plots or had them totally has evolved dramatically since 1811, transforming reshaped to fit the new road grid. However, land val- its structures and their uses. Wall Street has pro- ues rose dramatically after road openings, creating gressed from mostly residential in the mid-1800s, (Box continues next page) 156 FUTURE DRIVERS OF GROWTH IN RWANDA BOX 3.4  (continued) MAP B3.4.1  Plan of Manhattan, New York City, 1811 to industrial (residents gradually moved uptown as the noise and congestion of industry rose), to financial by 1920. The Meatpacking District mixed residential and a range of heavy industries in the mid-1800s, coming to focus specifically on meat fac- tories by 1900. The industry declined with changes in meat cooling and transport technologies, and the district evolved into at first a low-end and then a highly fashionable and gentrified entertainment dis- trict. Now it is also home to many of New York’s tech companies. Even the 1811 grid plan was adjusted over time, however. The now iconic Broadway, cutting diago- nally across the city, was not in the original plan; neither were Central Park or the superblocks created for larger buildings like Grand Central Terminal. From 1916 to 1961, the city added zoning regula- tions, limiting the mass of building allowed on each lot. Realizing that these restrictions were holding back investment, productivity, and tax revenues, the revised 1961 Zoning Resolution switched to reward- ing property owners with extra height allowances in return for reserving public land on the floor of their plot (such as plazas or mid-block passageways); this resolution allowed density and investment to climb again, while improving streetscapes and pedestrian Source: New York City Planning Department. connectivity. BOX 3.5  Land readjustment for urban renewal Land readjustment refers to a process whereby agency returns to each owner a site of at least equal neighboring landowners collectively adjust their value in the vicinity of the original parcel. The size parcels to support servicing and regularization of parcels is typically reduced, but their value is typ- (figure B3.5.1). It has been used to manage urban ically higher due to formality and servicing, leaving expansion into rural peripheries, upgrade underser- landowners better off (photo B3.5.1). viced land in inner cities, or facilitate reconstruction Land readjustment has been piloted in several after disasters and conflicts. sites in Rwanda, including for central redevelopment Residents typically transfer their land to a gov- (agatare) and managed expansion on the periph- ernment agency or other body (a cooperative or ery (gahanga) (photo B3.5.2). In gahanga, district corporation), which invests in the proposed upgrad- authorities estimated that most plots were worth ing (road widening, public facilities, parks, or regu- FRW 500,000 before readjustment and afterward sold larization) and may reserve some land for sale to for FRW 3 million–FRW 6 million, with 20 percent of recover costs. On completion of the project, the project costs financed by landowners themselves. (Box continues next page) F aster U r b ani z ation , G reater A gg l o m eration 157 BOX 3.5  (continued) FIGURE B3.5.1  Land readjustment Land Share re-plotting development profit UR Block Land contribution Land value increase (create public land Park Before After land re-plotting and reserve land) Sales to UR participants recover construction Road budget Reserve floor Reserve land Floor for Before After original land owners Development UR building by Source: Matsui 2017. land right conversion Note: UR = urban renewal. PHOTO B3.5.1  Way-markers map plot boundaries Credit: Sally Murray. PHOTO B3.5.2  Land readjustment sites in Gahanga, Kigali Credit: Sally Murray. 158 FUTURE DRIVERS OF GROWTH IN RWANDA city reduces the returns to residing or locat- provide ways to manage the overall transit ing firms in and visiting (for example, for network more efficiently in real time and job search) central areas. As with road infra- provide on-demand services for mobility structure, a priority is to bring the quality of needs. Some novel mass transit concepts have transport service up to a high level in central emerged, such as on-demand buses developed areas and in areas already densely populated by the University of Tokyo, which replace to encourage and deliver the benefits of den- fixed-route bus lines by dynamically routing sity. An extensive public transport network pickups and drop-offs based on user demand. also encourages a degree of sprawl. Successful Key urban mobility trends are summarized in urbanization has often combined excellent box 3.6. central public transport with selected quality public transport to a small number of stra- Mobilizing Scale and Specialization in the tegic peripheral locations such as satellite Informal Sector towns, as the city has grown. In every country in the world, as urban for- Providing enough space for roads through mal sector jobs expanded, informality also the grid (discussed above) is important for rose. The expectation of securing formal sec- avoiding congestion and upholding fluidity, tor jobs pulls far more rural migrants into but, if it is done alone, households quickly cities than the formal sector can employ, respond by buying more cars, further con- and the surplus ends up in the informal sec- gesting roads. In Rwanda, rising incomes will tor. The informal sector is also a common increase the demand for cars. A priority is to destination for workers suffering temporary reduce people’s tendency to drive by privileg- unemployment, which may enable them to ing public transport. hold out longer to find a more suitable job, The right public transport solutions for improving the matching of firms with work- Kigali will evolve with urban density and ers. Thus, more broadly, the informal sector institutional capacity. For example, at pres- provides an important safety net that facili- ent densities, technologies like dedicated bus tates the growth of urban populations and lanes and safe walkways and cycleways are increases the fluidity of their labor markets. appropriate. Improving institutional capac- The informal sector also offers important ity will allow the introduction of more flex- benefits for urban productivity. Low cost ible, demand-responsive route and schedule and flexible, small firms support the com- planning. Bus rapid transit requires very petitiveness of more productive sectors and high-density catchment areas around stops, workers by keeping urban input costs (food, and light rail even more so; hence, these housing, services, and so on) low. For some approaches become appropriate at much later firms, remaining small and informal can stages. With strong institutional capacity, be a rational strategy to uphold efficiency more effective mass public transport manage- and flexibility; remaining small can occur ment and charges to deter the use of cars can in response to regulations, but it can also also be considered.10 occur in response to rising urban density and Options for mobility will evolve dramati- changes in technology, which make operat- cally by 2050. Rwanda needs to invest in ing at smaller scale less inefficient than was mobility that connects today’s residents with previously the case. Experiences from other opportunities, while taking advantage of new countries show that, if properly harnessed, technologies in the medium term. Important the informal sector can also breed power- here is improved management of traffic and ful human capital development through travel demand. Ubiquitous data flows will “learning by doing” in the off-farm sector. facilitate more effective use of existing transit Scale and specialization, which are facili- capacity, through better matching of supply tated by dense urban environments, matter of and demand for mobility across time and just as much in the nontradable, informal space for all types of trips. New technologies sector as in the formal economy. Enabling F aster U r b ani z ation , G reater A gg l o m eration 159 BOX 3.6  How will urban mobility change by 2050? Several emerging trends are expected to help cities to energy efficiency. Autonomous vehicles will to reduce pollution and congestion and to increase also be able to follow one another closely, improv- urban mobility over the next 35 years: ing synchronization with traffic signals and reduc- ing stop and start delays. Some experts claim that •• Last-mile solutions. Electric-powered last-mile driverless technology may reduce traffic flow by solutions will transport people from their home as much as 40–50 percent, greatly reducing emis- to mass transit stations and from stations to their sions. Driverless cars will also reduce the need for final destination. They may involve self-­stabilizing parking space in cities. two-wheel scooters and transporters and public •• Cleaner technologies. Cleaner technologies will electric bike systems, which will be shared and increase fuel economy and greatly reduce emis- available on demand. Mass use of these solutions sions. Electric vehicles and new engine technolo- will require barrier-free infrastructure and protec- gies will improve both driving and environmental tion from vehicular traffic. In combination with performance. The cost of electric vehicle batteries mass transit, they will offer a green urban mobil- is expected to decline by about 8 percent a year, ity option for most city dwellers. Cities will also making electric vehicles competitive with conven- encourage walking. Singapore is already build- tional cars within the next decade. ing covered walkways from homes and offices to •• Vehicle sharing and vehicle pooling. The combi- mass transit stations to encourage people to walk nation of better data and autonomous vehicles to stations. will increase vehicle sharing and pooling. Vehicle •• Increased use of low-energy and active transpor- sharing allows users to rent vehicles by the hour. tation options. Cities are likely to reclaim part It provides users with access to mobility whenever of the streetscape from cars, reallocating it to they need it without the burden and expense of walking and biking. Bike sharing will continue car ownership. Car sharing has obvious benefits to expand, improving both human and environ- for cities. Zipcar estimates that every shared vehi- mental health. Bike networks will likely be shared cle replaces up to 20 private automobiles, reduc- with electric bikes, with some separation to ensure ing both total vehicle miles and the land allocated safe travel. to parking. Carmakers, including BMW, Daimler, •• Autonomous vehicles. Autonomous vehicles are and Ford, are developing their own programs. likely to be widespread by 2050. Following lob- Data sharing will enable groups of people to pool bying by Google, which produces a driverless car, autonomous vehicles, supporting mobility where the U.S. states of California, Florida, and Nevada density of population may not warrant mass tran- allow such vehicles. The development of autono- sit systems. The line between the public and pri- mous vehicles with car-to-car communication is vate provision of mobility services will become expected to reduce traffic accidents and eliminate increasingly blurred as capacity in private vehicles the need to build heavy cars, which will contribute is used more effectively. Source: Lall 2017. small businesses to develop clusters of eco- productivity in this sector should be a high nomic activity creates the density of demand priority. Cities that successfully harnessed necessary to facilitate productivity gains the power of the informal sector to mobi- through specialization in nontraded goods lize savings, provide jobs, and raise human and services, while scale helps to keep costs capital addressed factors that kept firms and down. By specializing and keeping costs low, households in informality unnecessarily; they informal and small-scale businesses can have did so through deregulation, technical sup- better options to connect with citywide and port, and provision of public goods (such as broader value chains. infrastructure). The government can learn Given that most businesses in Kigali are in from cities that developed successful relation- the informal and nontraded sector, enhancing ships with small firms to enhance their role 160 FUTURE DRIVERS OF GROWTH IN RWANDA in absorbing rural-urban migrants, raising reduce the fragmentation that forces people human capital, and keeping inputs into the to cities. formal sector competitive (box 3.7). Anticipating Demand: People-Focused, Supporting Inclusive Urbanization Portable Assets In a fast-changing landscape like Rwanda’s, Investments in human capital can have a making the most of rural-urban links particularly large role in facilitating migra- requires anticipating migration to cities. tion and raising earnings in large cities like Chengdu in China has pursued coordination Kigali. Several studies show that workers in of its rural and urban areas (box 3.8), empha- more educated cities typically earn more than sizing support for early rural-urban migrants workers in less educated cities, even holding to succeed in the city through skills training their own years of schooling constant. This and micro loans and raising agricultural pro- connection between a city’s education level ductivity through land exchange platforms to and earnings appears to be even stronger in BOX 3.7  Examples of successful engagement by large cities with the informal sector Medellín, Colombia. In Medellín, the local gov- infrastructure for traders; provided business sup- ernment’s policy had been to seek out and punish port, capacity building, and market assistance; and (usually through fines) businesses that operated enhanced communication through an Informal illegally. Thus, the small business community saw Economy Forum. the state as an added nuisance, rather than as a Singapore. Singapore’s Skills Development Fund provider of useful services, and low-income entre- (SDF) offers lessons for supporting human capital preneurs struggled in the city. Federico Restrepo development in informal and small firms. Singapore Posada, the newly appointed secretary of planning established the SDF in 1979 to support e ­ mployers in Medellín, chose a strategy to change the govern- in developing a more skilled, adaptable workforce. ment’s relationship with the informal sector, raising In international experience, small employers typi- its productivity, tax contribution, and goodwill. He cally do not benefit from such schemes, because made the city budgeting process more transparent releasing employees for training harms production, and began a publicity campaign that explained to the small scale of training raises unit costs, and citizens where their tax money was being spent. smaller firms are less able to assess training needs The municipal government also changed the law to select appropriate programs. Recognizing this, so that informal microbusinesses were no longer the SDF incorporated special incentives to encour- illegal; made visible and effective urban invest- age participation by small firms. The SDF offered ment in the poorest neighborhoods, demonstrating vouchers to companies with fewer than 50 workers, that tax money supports the community; and cre- which supported 30–50 percent of training costs. ated ­business centers using expertise and staffing This helped small enterprises to ease cash flow from local nongovernmental organizations and the problems and reduced administrative procedures. ­ Chamber of Commerce, to provide basic skills in With many smaller firms employing workers with- business registration, accounting, marketing, and out education, the SDF also supported programs management. providing the equivalent of a sixth-grade (P6) level Durban, South Africa. Durban has taken one of functional literacy and numeracy and increasing of the oldest and most innovative approaches access to English and mathematics at the second- to the informal sector in Africa. The munici- ary education level. The voucher helped the SDF to pality adopted a policy on informal trading to reach 65 percent of enterprises with 10–49 workers clarify the rules of the game; established a depart- and 14 percent of those with fewer than 10 workers ment of street vendor management and provided (Arvil 2009; Yaushi 1997). F aster U r b ani z ation , G reater A gg l o m eration 161 BOX 3.8  Mobilizing rural-urban links: Chengdu’s coordinated development model Following rapid urban growth, Chengdu, China, has of land use and improved land use efficiency. Key pursued a model of urban-rural coordination since mechanisms include farmers voluntarily giving up 2003. The model aims to replace previous impedi- land for urban settlement packages, rural land- ments to rural migrants’ moving to the city, shifting for-land exchanges to reduce fragmentation, and the emphasis to a combination of support for early land-use rights for shareholding, whereby farmers rural-urban migrants to succeed in cities and “reten- exchange their land for shares in a specialized agri- tion effects” that raise the welfare of rural residents cultural enterprise (which reorganizes the land to who remain or move to rural towns. encourage large-scale production). The coordinated development model consists of 3. Agglomerating firms in industrial areas to encour- three main policies: age the creation of nonfarm jobs. Chengdu’s 21+10 Industrial Restructuring Plan reorganized 1. Helping rural residents to settle in urban areas. 116 industrial development areas into 21 areas To help rural migrants to find employment in and designated 10 major periphery towns with the urban sector, government offered subsidies greater resource endowments for new industrial to urban firms to train rural residents, micro projects. loans to a small set of migrants to encourage self- employment in the city, preferential employment Over the course of these policies, income has conditions for migrants who gave up their land grown in Greater Chengdu, and the urban-rural to become urban residents, and subsidies to firms disparity has narrowed at a faster rate than in the that provided the same social insurance to rural rest of China. The model has been successful in migrants as to urban workers. encouraging rural migrants to move to periphery 2. Desegmenting land to realize scale economies. townships and in creating off-farm jobs in these Improved land exchange has reduced f ­ ragmentation urban areas. Sources: Chen and Gao 2011; Yumin and LeGates 2013. China and India than in the United States, They prioritized basic education at an early suggesting that big gains in developing and stage, as the necessary foundation for effec- emerging markets may be available. Senator tive urban labor markets, and later invested Moynihan once claimed, “If you want to in higher-level skill development. Perhaps the build a world-class city build a great univer- most important urban dividend came from sity and wait 200 years.” Although cognitive speeding up the flow of ideas; in the 1980s skills—like the ability to draft a complicated the new growth theory, led by scholars such legal contract—have long been recognized in as Paul Romer and Robert Lucas, focused on the theoretical and empirical literature as an knowledge spillovers as the source of increas- important determinant of productivity and ing productivity that enabled sustained eco- wages, new research also reveals that large cit- nomic growth. The focus on the flow of ies pay a wage premium to workers with bet- knowledge pointed both toward education ter social skills, which is not surprising: urban and toward cities. dividends reflect the benefits of ­ interactions Human capital investments are a priority (Fujita and Thisse 2002). In France, human for rural areas and secondary cities. A general capital has been shown to affect earnings principle is to deliver a basic standard of ser- via greater communication at the workplace, vices to all people, while letting markets pick an effect larger in bigger and more educated the pace and form of private sector develop- cities (Duranton and Charlot 2006). ­ ment by location. This principle is exem- Successful urbanizers invested heavily, plified by Korea, which allowed the dense and from early stages, to raise human capital. concentration of economic activity in Seoul, 162 FUTURE DRIVERS OF GROWTH IN RWANDA while delivering basic services like education, as effective change-managers in an environ- health care, and clean water evenly across the ment in considerable flux. population (map 3.12). Investments in human capital, in particular, are “safe” bets, regard- Fiscalizing Public Investments in Cities less of location. This is because human capi- Successful, rapid-growth cities made huge tal is a portable asset that can have powerful investments in public infrastructure. These effects on people’s welfare and productivity investments, in turn, dramatically raised in situ as well as on their ability to succeed in the productivity of urban land, crowding higher-productivity areas. in large private investment and raising land Step-changes in Kigali’s productivity will values many fold. be possible from 2025 to 2030, as invest- Given its high cost, such investment is only ments made today to raise the quality of mass attainable when public investments in land education across Rwanda begin to filter into are fiscalized through land value capture. Rwanda’s labor market, delivering opportu- A range of tools exist to capture the land nities for more advanced specialization and value increments accompanying investment— knowledge spillovers. such as land value and property taxes, capital gains taxes, betterment fees for landowners or developers, special assessment districts Institutional Requirements for (with temporary additional taxes applied High-Growth Urbanization in return for infrastructure improvements), Emerging and growing cities will benefit land purchases by infrastructure agencies from strong and adaptive investments in (for example, box 3.9), and the direct com- human resources and institutions. These missioning and financing of infrastructure by investments should deepen the decentraliza- property developers. tion agenda, enabling local leaders to mobi- Effective land value capture should ensure lize local knowledge and resources and to act that tax rates track values in a way that is MAP 3.12  Concentrated urban development and universal basic services in the Republic of Korea a. Urban land (km2) in 2009 b. Access to piped water (% of population) in 2005 Access to piped water Urban land (% of (km2) population) ~10 ~10 ~20 ~20 ~30 ~30 ~40 ~40 ~50 ~50 ~60 ~60 ~80 ~70 ~100 ~80 ~200 ~90 Over 200 ~100 Source: Lall et al. 2013. F aster U r b ani z ation , G reater A gg l o m eration 163 BOX 3.9  Land value capture: The rail + property program in Hong Kong SAR, China Hong Kong, SAR, China, has financed a large por- •• MTRC then tenders out its development rights tion of its transit system using land value capture. Its to private developers (dividing them into smaller, metro system—Mass Transit Railway Corporation more manageable lots). (MTRC)—operates without government subsidy •• Private developers bear the construction and com- and is highly profitable, thanks to profits generated mercialization costs of residential and commercial by its real estate business. MTRC invests in and properties. shares profits with private developers in real estate •• Profit-sharing mechanisms are included in the sales and property management. agreements with private developers. For exam- MTRC’s property development program uses the ple, for residential units, the MTRC receives an following process: agreed portion of profit generated by the sales if the private partners sell all units before a con- •• MTRC, with the city government, assesses the tractual deadline. Otherwise, MTRC obtains the cost of construction and prepares a master plan to unsold units either to sell or to lease in the open identify property development along railway sites. market. •• MTRC purchases the right to develop property above underground stations and land adjacent to the The transport company is not in charge of con- railway (development rights). This land premium is struction; instead, it supervises the work and carries paid to the government but does not include the out civil works required for integrating the railway increased value resulting from the transport project. and property development. Sources: Cervero and Murakami 2009; Salon 2014. both acceptable to taxpayers and able to revenue mobilization is low—estimated at deliver the revenue required by districts. 10 percent of district budgets in 2013–14.11 Governments can soften contestation by Districts therefore rely on central revenues communicating the approximate nature of and do not capture many of the gains of valuations: in certain U.S. states, the valu- their investments, which flow largely to ation for property tax purposes is by law private landlords (many of whom acquire 70 percent of the real assessed property value land speculatively to reap the gains of pub- to soften contestation, whereas in the United lic investment) and the national government Kingdom, properties are assigned to broad (in the form of centralized corporate taxes). value bands to avoid making precise claims This reliance both reduces district resources liable to contestation. Rwanda has struck a for investment and reduces their incentives sensible balance with its new ­ decentralized to identify and prioritize investments where taxes law, avoiding communicating explicit they are most likely to unlock growth. plot valuations that may be contested, Rwanda has a strong basis for efficient but preparing ground for area-based fees and transparent valuation, with centralized that track market values. Some countries and digitized landownership and transac- in East Asia recognized the importance of tion records, a professional valuation body, appropriate land taxes to generate the rev- and credible institutions for oversight like the enues required for adequate infrastructure Ombudsman’s Office. However, significant investment (box 3.10). additional investment is needed in human Land valuation is critical for effective land and institutional resources for valuation. value capture. District land value capture has Optimal systems combine the analysis of the capacity to strengthen decentralization large-scale transaction price (and geo) data in the future. Outside of Kigali, own-source with on-the-ground audits and assessments 164 FUTURE DRIVERS OF GROWTH IN RWANDA BOX 3.10  Land value capture in East Asia Property—especially land—taxes were central to the that captures those elevated land values for reinvest- urban development strategies of East Asian develop- ment in broad-use infrastructure unlocks a virtuous mental states. In Taiwan, China, by 1975, property- cycle of self-financing urbanization. related taxes were 70 percent of local government Taxes on land were also used to align investors’ collections. By 2000 property taxes constituted a incentives with the manufacturing sector, in the higher proportion of tax revenue in the Republic of context of high land speculation. Whereas property Korea than in any other high-income country. In and land taxes were high, taxes on productive sec- Singapore, property taxes constitute 8 percent of tors were typically lower than those seen today in national tax revenue. Rwanda and other African countries; ratios of final These states saw that investments in urban infra- tax to GDP were similar to those in contemporary structure are expensive, but crowd in private invest- Sub-Saharan Africa. ment, translating directly into higher land and The fact that revenues flowed back to the cities property values. It is not uncommon, internationally, meant that local governments were not just financed for the combined effects of public infrastructure invest- to make high and appropriate investments, but also ment and in-migration from rural areas to appreciate highly motivated to do so. Incentives to select the urban land values 10- to 20-fold in less than a decade. right investments were created by the promise that For example, in Seoul, Korea, land prices rose 33-fold more effective investments would raise land and between 1974 and 1996 (Goodfellow 2017). A tax property values, and hence city budgets, more. to derive and publish transparent valuation combining the high skill and credibility of formulas. Lessons from Bogotá, Colombia planning teams with institutions that forced may be useful for Rwanda (box 3.11). transparency about planning revisions both within government and with the public. Strengthen Urban Planning Making the process more transparent and A limited local planning capacity explains participatory can ensure that plans respond the problems affecting the use of mas- to markets and community visions, while ter plans in Rwandan cities. In all cities, safeguarding against special interests through skilled urban planning teams are required strong public oversight and engagement. to ensure that managed urbanization sup- Rwanda should begin to experiment with ports growth, investment, and community developing institutions that require and aspirations. As urban complexity rises, a facilitate regular, responsive, and transpar- particularly skilled, credible, and empow- ent planning adjustments. These adjustments ered team is required to carry out adaptive, should include agreed and transparent proce- market-responsive planning. The ongoing dures for investors and citizens to apply for restructuring of local government bod- changes in regulations and for review of these ies should ensure adequate urban planning applications (Huang 2016). Examples from capacities in all locations. To this effect, India, Japan, and a sample of global cities urban planners will be integrated in the dis- are provided in boxes 3.12, 3.13, and 3.14, trict structures. respectively. There is a fear that more flexible and market-responsive land use regulation plan- Cross-District Coordination ning would open the floodgates for special M any u rban areas are developing as interests, on the one hand, and accusations cross-district “corridors” (such as the of complicity by planners, on the other. Musa n ze – Nyabi hu – Rubav u corri- Other cities have combatted these risks by dor, K ig a l i – Mu h a n g a – Huye – Ru si z i , F aster U r b ani z ation , G reater A gg l o m eration 165 BOX 3.11  Updating land values in Bogotá, Colombia A pioneer in land valuation, Bogotá updated its forward, construction permits were to be used to cadastral database over 2008–10 by revaluing its update physical data automatically; another option 2.1 million urban properties. The update yielded is to use satellite data with manual identification a 30 percent (US$171 million) increase in property or remote sensing of buildings. Given a shortage tax revenues, bringing these to 40 percent of own of market information in Bogotá, initial price esti- revenues by 2010; the cost of the improvements was mates were collected from expert appraisers who US$15 million. The update was achieved by improv- combined sales price, construction cost, and income ing human resource management, introducing (rental) valuation approaches. Today valuation data information technology, enacting policies to soften are shared throughout the government to improve property tax jumps (to manage political implica- planning across agencies and sectors. tions), engaging stakeholders and career civil ser- Because the cadastral values had long lagged vants, reviewing the project’s results in a public and behind actual market values, the property tax transparent manner, and adopting technical innova- base was expected to increase sharply after the tions to improve assessments. update. To prevent excessive resistance by property Hundreds of temporary workers surveyed plots ­ owners—while nevertheless retaining the assump- and buildings, but technological innovations helped. tion that property values are strongly correlated Spatial information from Geographic Information with ability to pay—the city limited property tax Systems, such as a property’s distance from key sites increases ­ proportionally to the logarithm of its and amenities, was incorporated into the statistical value (with differentiated ceilings for residential and techniques for estimating property values. Going ­nonresidential uses). Sources: Ruiz and Vallejo 2010; Uribe 2010. BOX 3.12  Institutions for market-responsive urban planning: Yokohama, Japan A visionary mayor, Ichio Asukata, and his appointed meetings to resolve blockages and discuss and urban planner, Akira Tamura, helped to transform approve adjustments to plans. The planning bureau Yokohama, Japan from a declining, dirty port town also held regular public meetings with residents in the 1960s to a thriving, attractive, modern city. and businesses to receive feedback. A city official This transformation was achieved through the from the time was asked what share of the city’s following: plans were revised; he responded that 80 percent were revised, but that the share should have been •• Leadership of an elevated and skilled planning even higher, because planners can never get every- bureau. A new Bureau of Planning and Coordi- thing right in advance. nation was given a “slightly higher” status than •• Urban planning as “community-building.” A seri- other institutions and empowered to adjust plans ous Citizen Councils initiative was launched, led and coordinate institutions. It was staffed with by a seasoned civil society campaigner and staffed 15 rising young officials from across institutions, with former student activists, that engaged resi- hand-picked by Tamura. Tamura himself had three dents and businesses to contribute to and update degrees (law, politics, and architecture) and a strong district plans. Plans were then presented to citizens reputation from experience across government. through a robust campaign. The citizen ownership •• Transparent implementation and adjustment of this engendered meant that plans were pursued plans. Tamura chaired monthly cross-institutional long after Mayor Asukata left office. Source: Tokyo Development Learning Center 2017. 166 FUTURE DRIVERS OF GROWTH IN RWANDA BOX 3.13  Participatory planning and budgeting: Kerala, India In 1996, India’s Kerala State launched a remark- council, which consolidated them to produce an able experiment in participatory planning. The annual plan and budget. This plan was presented People’s Plan Ca mpaig n for the Ninth Plan back to citizens and then sent to district plan- devolved 35 percent of the state budget to the ning committees, which scrutinized it to iron out lowest tier of local government structures and inconsistencies and fill in gaps. developed a series of local-level planning includ- The plan radically improved the delivery of pub- ing seminars, task forces, local council meetings, lic infrastructure and services, raised caste and eth- implementation and monitoring committees, and nic equality, facilitated the participation of women so on. Local meetings, facilitated by a popular in public life, and improved state-people relation- nongovernmental organization, discussed and ships. The scheme was so popular and comprehen- prioritized the development needs of the commu- sive that even a change of government in 2001 could nity and presented them to the local government not overturn the model. Source: Franke 2007. BOX 3.14  Involving stakeholders in city development: Lessons on participatory planning from global cities More and more cities have shifted their plan- these parameters guided the individual parcels ning practice toward creating a sense of place and and mixed-use plans. The procedure entailed four engaging communities in the planning process. rounds of public consultation with some 30,000 Recognizing that certain places hold special mean- people that allowed a balancing of local benefits ing for particular groups, planners engage local against profitability for the developer. people in transformation. New York City, United States. In Manhattan, Hong Kong SAR, China. In Hong Kong SAR, a nongovernmental organization called Friends of China, for example, people-oriented planning fol- the High Line used participatory planning meth- lows a process of “Communicate—Understand— ods, such as focus group interviews and commu- Interpret—Plan,” using citizens’ views to inform the nity campaigns, to boost the involvement of local design. The planning process also pays attention to communities. Their efforts eventually fostered the the scale of the plan, with a sense of place generally transformation of an abandoned elevated urban rail identified for small neighborhoods. into a vibrant public space. This linear park project London, United Kingdom. In London, the devel- upgraded the physical and social landscape of the opment of King’s Cross entailed major engagement neighborhood along its route. based on the publication of proposals and discus- Seoul, Republic of Korea. Seoul’s 2030 plan was sions with residents. This process led to 10 design the first to build on wide engagement of local residents. principles for a human city as well as parameters for Through onsite discussion and online platforms, regeneration based on lessons learned from other citizens helped the planning authorities to shape the ­ projects in London. Design frameworks embedding vision and identify 5 core issues and 17 specific goals. Kigali– Rwamagana– Kayonza, and the major infrastructure like special economic newer Kivu Belt). Kigali’s bordering dis- zones, tourism infrastructure, and so tricts are also increasingly integrated with on, to reduce waste and exploit comple- Kigali’s economy. Economically connected mentarities. Box 3.15 shows how Korea’s districts should be encouraged and required central government supported cross-juris- to coordinate land use plans (to avoid diction collaboration through “happy liv- sprawl), transport and service provision, ing zones.” F aster U r b ani z ation , G reater A gg l o m eration 167 BOX 3.15  Cross-district governance: The Republic of Korea’s happy living zones Korea experienced rapid urbanization following features (such as commuting routes, industries, the Korean War, with a high number of displaced geography, and culture) and to develop policies rural peasants leaving villages for Seoul in search tailored to broader “places” rather than to specific of more economic opportunities. The growth of jurisdictions. Since its inception, 63 zones have been Seoul and its surrounding capital region has driven established. Each zone has established an asso- the Korean economy, but regional disparities have ciation of local governments, which prepares and persisted as a primary concern for Korean policy implements development projects tailored for the since the 1980s. zone, including economic revitalization, education, Since 2013, the central government has pursued culture, welfare, and basic infrastructure. Central a new spatial concept, happy living zones, as a basis governments have supported the preparation and for regional development policy. These zones are implementation of happy living zones by coordinat- intended to enhance happiness and quality of life ing local governments, funding initiatives through outside the capital city region to retain and attract the special account for regional development, con- people. The primary goal is to encourage coopera- sulting and offering technical assistance, and dereg- tion among local governments that share common ulating to support initiatives. Source: Roberts 2014. Unite Spatial and Economic Planning As Rwanda works to strengthen the eco- Economic planning, as carried out by nomic dividend of urbanization, economic the Ministry of Finance and Economic planners should start to think spatially, and Planning, Ministry of Agriculture and spatial planners should become more aligned Animal Resources, Ministry of Trade and with economic planning and goals. In Korea, Industry, Rwanda Development Board, for example, a dedicated spatial develop- and so on, is sometimes not aligned with ment research entity supports higher-level “spatial” planning, as conducted by the structures to marry spatial and economic M in ist r y of I n frast r uc t u re, M in ist r y planning. of Local Government, and Ministry of Environment. Annex 3A  The Rwanda Spatial CGE Model and Scenarios This study commissioned a CGE model of financed by urban taxation, meaning that Rwanda to examine how alternative urban- rural investment stays the same ization policy scenarios would affect key eco- • Scenario 5. Scenario 4, with public urban nomic and welfare outcomes until 2050. Five investment that is effective in attracting urbanization scenarios were considered: private urban investment and raises pro- ductivity in urban areas • Scenario 1. Business as usual • Scenario 2. Faster rural-urban migration The researchers drew on a spatially dis- • Scenario 3. Faster migration, with higher aggregated social accounting matrix (SAM) public investment in urban areas and constructed in 2015. A SAM is a consistent lower public investment in rural areas (due accounting framework that captures all to a fixed government budget) income and expenditure flows in an econ- • Scenario 4. Faster migration, with higher omy during a given year; it includes all sec- public investment in urban areas that is tors, households, and the government, as 168 FUTURE DRIVERS OF GROWTH IN RWANDA well as the economy’s interactions with the public infrastructure and services. The latter rest of the world. The SAM for Rwanda was must be offset by public urban investment. built using national accounts, censuses, and Modeling such outcomes for the five agricultural and economic surveys from the urbanization scenarios considered, the study National Institute of Statistics of Rwanda; shed light on how fundamental policy deci- government revenue and expenditure data sions regarding urban migration, public from the finance ministry; and balance investment, and taxation may affect growth, of payments data from the central bank welfare, and equity across Rwanda over the (see Diao, Randriamamonjy, and Thurlow course of Vision 2050. 2017). Inputs and results were disaggregated for Kigali, other urban and peri-urban areas, and Notes remote rural areas; and interactions between 1. A 40 percent increase in Kigali’s population all three locations also were considered. from 2002 to 2012, compared with a 30 per- Trends were also distinguished according to cent increase in the national population. household poverty and education levels. 2. Musanze, Nyabihu, and Rubavu districts. The model assumes that labor and their 3. Huye, Muhanga, Nyanza, and Ruhango dependents move across Rwanda on the districts. basis of wage differentials. It assumes that 4. The annual flow of migrants leaving rural the migration effect of wage differentials areas rises from 75,700 in the baseline to observed in 2015 is equal to that observed 107,900 in the faster urbanization scenario, with 22,300 arriving in (peri-)urban areas in the 2012 population census data. This and 85,600 arriving in Kigali City. results in an average net annual outflow 5. Food, housing, and transport prices are higher from rural areas of 75,700 migrants, with in African cities than in competing low- and 50,200 moving to Kigali City and 25,500 middle-income countries elsewhere (by moving to other urban areas. around 35, 55, and 42 percent, respectively). Labor productivity and GDP are key vari- “Overall, urban households pay 20 to 31 per- ables of interest. Disaggregating these for urban cent more for goods and services in African and rural locations is challenging, because gov- countries than in other developing countries” ernment GDP statistics are only published at (Lall, Henderson, and Venables 2017). the national level. To overcome this challenge, 6. Integrated Household Living Conditions the model combines data on national sec- Survey 2013–14 (EICV4) district-level analy- sis (NISR 2015). toral labor productivity and regional sectoral 7. Internal migration is defined as moving from employment trends with assumptions about one district to another, so changes of residence how labor productivity in each sector differs within districts are not captured as internal in rural versus urban areas. Overall, labor migration. That is, neither rural-to-urban productivity gains are constituted by within- migration within districts nor urban-to-rural sector productivity gains (the productivity in migration within districts is counted as part of each economic sector, weighted by initial labor the population movement. For details on meth- share in the national economy) and structural odology and results, see World Bank (2017a). change (gains caused by workers moving from  8. Analysis of EICV4 data, exempting high less to more productive sectors). outliers. The comparative impacts on urban welfare   9. Kigali’s R1 residential zone, for example, cov- ers many central areas, enforcing a minimum of rural and urban public investments were plot size of 600 square meters and requiring modeled. This modeling examined how rural that 60 percent of the plot be left unbuilt. outflows of food and migrants both help and The R2 zone (also common in central areas) hinder urban areas. As households migrate to demands that 40 percent of each plot be left urban areas, productivity gains from agglom- undeveloped. It allows plots of 90 square eration economies are assumed, as well as meters—better, but still high as a minimum; “congestion effects” through pressures on consider that Philadelphia allowed 30 square F aster U r b ani z ation , G reater A gg l o m eration 169 meters as it was settled, whereas Singapore’s edited by C. Spannos, 130–35. Oakland, CA: minimum plot size today is 90 square meters. A. K. Press. 10. The key requirements for effective congestion Fujita, M., and J.-F. Thisse. 2002. Economics of charging are not technological, but the author- Agglomeration: Cities, Industrial Location, ity and capacity of the implementing institu- and Regional Growth. Cambridge, U.K.: tion, for which Rwanda has some advantages. Cambridge University Press. 11. For Kigali, own-source revenues are closer to Gelb, A., J. Meyer, V. Ramachandran, and 50 percent. D. Wad hwa. 2017. “C a n A f r ic a B e a Manufacturing Destination? Labor Costs in Comparative Perspective.” CGD Working References Paper 466, Center for Global Development, Arvil, A. 2009. Skills Development in the Washington, DC. Informal Sector of Sub- Saharan Africa. Goodfellow, T. 2017. “Taxing Property in a Neo- Washington, DC: World Bank. Developmental State: The Politics of Urban Bachofer, F., and S. Murray. 2018. “Remote Land Value Capture in Rwanda and Ethiopia.” Sensing for Measuring Housing Supply in African Affairs 116 (465): 549–72. Kigali.” International Growth Centre, Kigali. Guardian. 2014. “Cities Need Goldilocks Brueckner, J., and S. V. Lall. 2014. “Cities in Housing Density—Not Too High or Low, but Developing Countries: Fueled by Rural-Urban Just Right.” Guardian, April 16. Migration, Lacking in Tenure Security, and Henderson, Vernon, and Dzhamilya Nigmatulina. Short of Affordable Housing.” In Handbook 2016. “The Fabric of African Cities: How to of Regional and Urban Economics. Vol. 5, Think about Density and Land Use.” Draft, ch. 21. Amsterdam: Elsevier. London School of Economics. Cai, F., A. Park, and Y. Zhao. 2008. “The Huang, C.-Y. 2016. “Enabling Structure: Role of Chinese Labor Market in the Reform Era in Public Sector and Institutions.” Background China’s Great Economic Transformation.” In paper for Africa’s Cities: Opening Doors to The Chinese Labor Market in the Reform Era, the World, World Bank, Washington, DC. edited by L. Brandt and T. Rawski, 167–214. Jones, P., J. Bird, C. Beck, and A. Haas. 2016. Cambridge, U.K.: Cambridge University Press. Kampala: A Policy Narrative. Washington, Cervero, R., and J. Murakami. 2009. “Rail DC: World Bank. and Property Development in Hong Kong: Jones , P., K . del Fra ri, L . Ferrei ra, and Experiences nd Extensions.” Urban Studies M. Sladoje. 2016. Lusaka: A Policy Narrative. 46 (10): 2019–43. Washington, DC: World Bank. Chen, A., and J. Gao. 2011. “Urbanization in Jones, P., S. Murray, and L. Ferreira. 2016. Kigali: A China and the Coordinated Development Policy Narrative. Washington, DC: World Bank. Model—The Case of Chengdu.” Social Science Lall, S. 2017. “How Will Urban Mobility Change Journal 48 (3): 500–13. by 2050?” Unpublished paper, World Bank, Diao, X. 2017. A Growth Analysis. Washington, Washington, DC. DC: International Food Policy Research Institute. Lall, S., B. Dasgupta, N. Lozano, and H. Wang. Diao, X., J. Randriamamonjy, and J. Thurlow. 2013. “Cities: A New Habitat for Tigers.” 2017. “Urbanization and the Future of Unpublished presentation, World Bank, E c onom ic D e velopm e nt i n Rwa nd a .” Washington, DC. International Food Policy Research Institute, Lall, S., V. Henderson, and A. Venables. 2017. Washington, DC. Background paper for Future Africa’s Cities: Opening Doors to the World. Drivers of Growth in Rwanda. Washington, DC: World Bank. Duranton, G., and S. Charlot. 2006. “Cities Matsui, Minoru. 2017. “Case Study: Land and Workplace Com mu nication: Some Readjustment in Japan.” Knowledge Product, Quantitative French Evidence.” Urban Studies Tokyo Development Learning Center, Tokyo, 43 (8): 1365–94. and World Bank Group, Washington, DC. Economist. 2012. “Daily Chart: The Urbanization MININFRA (Ministry of Infrastructure and Trap.” Economist (online), October 2. Water and Sanitation Cooperation), Rwanda. Franke, R. W. 2007. “Local Planning: The 2017. Site Distribution of Rural Settlement Kera la E x p er i ment .” I n R e a l Ut o pi a : Layout Plans since 2017 to May 2017. Kigali: Participatory Society for the 21st Century, MININFRA. 170 FUTURE DRIVERS OF GROWTH IN RWANDA NISR (National Institute of Statistics of Rwanda). Principles for Low- Cost Housing. Kigli: 2015. Rwanda Poverty Profile Report 2013– International Growth Centre. 2014: Results of Integrated Living Conditions Wa ng , H . 2011. “L e s son s f rom Kore a’s Sur vey ( E IC V ). Kigali: Government of Urbanization Policies to Facilitate Rapid Urban Rwanda, September. Transition.” Unpublished presentation, World Nordhaus, William D. 2006. “Geography Bank, Washington, DC. and Macroeconomics: New Data and New World Bank. 2009. World Development Report Findings.” Proceedings of the National 20 09: Reshaping Economic Geography. Academy of Sciences 103 (10): 3510–7. Washington, DC: World Bank. Quiros, T., S. Murray, and J. Bajpai. 2014. Managing ———. 2012 . Pl anning , C onnecting , and Accessibility for Sustainable Development of Financing Cities—Now: Priorities for City Kigali. Washington, DC: World Bank; London: Leaders. Washington, DC: World Bank. International Growth Centre. ———. 2015. “Background Study for the Rwanda Roberts, B. 2014. Managing Systems of Secondary Urban Development Project (RUDP).” World Cities: Policy Responses in International Bank, Kigali. Development. Brussels: Cities Alliance. ———. 2017a. “Note 2: Internal Migration in Rosenthal S., and W. Strange. 2003. “Geography, Rwanda.” Reshaping Urbanization in Rwanda, Industrial Organization, and Agglomeration.” World Bank, Washington, DC. Review of Economics and Statistics 85 (2): ———. 2017b. “Note 3: Urbanization, Job 377–93. Creation, and Poverty Reduction in Rwanda.” Ruiz, F., and G. Vallejo. 2010. Using Land Reshaping Urbanization in Rwanda, World Registration as a Tool to Generate Municipal Bank, Washington, DC. Revenue: Lessons from Bogotá. Washington, ———. 2017c. Reshaping Urbanization in DC: World Bank. Rwanda: Economic and Spatial Trends and Salon, D. 2014. “Location Value Capture Proposals; Synthesis Report. Washington, DC: Opportunities for Urban Public Transport World Bank. Finance.” White Paper prepared for the 2014 ———. 2 0 17d . “ Rw a n d a F e e d e r R o a d Transit Leadership Summit, Regional Plan Development Program: Impact Evaluation.” Association, London, May. Unpublished paper, Development Impact Tokyo Development Learning Center. 2017. Evaluation, World Bank, Washington DC. Yokohama: Reinventing the Future of a City. ———. 2017e. “Urbanisation and the Evolution Competitive Cities for Jobs and Growth. of Rwanda’s Urban Landscape.” Reshaping Washington, DC: World Bank. Urbanization in Rwanda, World Bank, Uribe, M. C. 2010. “Land Information Updating, Washington, DC. a De Facto Tax Reform Bringing up to Date the World Bank and Development Research Center Cadastral Database of Bogotá.” In Innovations of the State Council, China. 2014. Urban in Land Rights Recognition, Administration, China: Toward Efficient, Inclusive, and and Governance, edited by K. Deininger, Sustainable Urbanization. Washington, DC: C. Augustinas, S. Enemark, and P. Munro- World Bank. Faure, 180–207. Washington, DC: World Bank. Yaushi, H. 1997. Skills Development Fund: Vishwanath, T., S. Lall, D. Dowall, N. Lozano- A Preliminary Assessment of a Financing G r a c i a , S . S h a r m a , a n d H . Wa n g a . Alternative for Enterprise-Based Training in 2013. Urbaniz ation be yon d M unicipal the Context of APEC. Nagoya, Japan: Graduate B o u n d a r i e s :  N u r t u r i n g M e t ro po l i t a n School of International Developoment, Nagoya Economies and Connecting Peri-Urban Areas University; Asia-Pacific Economic Cooperation. in India. Directions in Development: Countries Yumin, Y., and R. LeGates. 2013. Coordinating and Regions. Washington, DC: World Bank. Urban and Rural Development in China: Wainer, L. S., B. Ndengeyingoma, and S. Murray. Learning from Chengdu. London: Edward 2016. Incremental Housing and Other Design Elgar Publishing. 4 Competitiveness and Enterprise Development for Innovation-Led Growth Introduction sector entrepreneurs can provide the dyna- mism, innovation, and risk taking required for In the immediate aftermath of the geno- a modern, sophisticated, and rapidly growing cide against the Tutsi, the level of economic ­economy. activity was extremely l ­ow. Private enter- The government has built a strong track prises were almost non­ existent, and even the record of reforms to support the development government was severely constrained in the of viable ­enterprises. This is evidenced, for resources it could m­ obilize. Over the course example, in Rwanda’s rapidly improved rank- of the reconstruction years, the government ing in the Doing Business I ­ ndex. The work was able to channel capital into more pro- is by no means finished, ­ however. Many pri- ductive uses and to kick-start entire sectors vate sector firms are small and informal; they of the economy through the establishment of have limited scale economies critical for com- state-owned enterprises ­ (SOEs). The govern- petitiveness and have limited export presence ment also took steps to enable the develop- and capacity for i ­nnovation. The number of ment of a vibrant private sector, including by formal sector firms is expanding gradually, liberalizing some industries previously run and the pace of job creation in the sector has ­ onopolies. Rwanda’s as state or quasi-state m ­slowed. Vision 2020 states, “The emergence of a The limited private sector presence viable private sector that can take over as reflects two high-level challenges that need the principal growth engine of the economy to be at the center of policy responses: key.” is absolutely ­ As Rwanda’s economy develops and 1. High economywide costs faced by enter- matures, the need for the private sector to play prises in Rwanda relative to costs in other a leading role in achieving the rapid growth economies at similar stages of their devel- required to reach Rwanda’s ambitious income opment, most importantly in energy, aspirations only i­ncreases. The massive rise transport, and finance in investment required to achieve Rwanda’s 2. Productivity that has not grown at a rate growth aspirations has to come predominantly capable of sustaining rapid economic from the private sector, since public invest- growth because of a suboptimal allocation ment is constrained by rising debt and lim- of resources and insufficient technological its on external ­ assistance. Moreover, private ­innovation 171 172 FUTURE DRIVERS OF GROWTH IN RWANDA To address these constraints, the following FIGURE 4.1  Share of firms in Rwanda, by firm are key areas for the government’s continued size, 2011 and 2014 reform agenda: 80 • Reducing business costs by tackling cross- 70 cutting constraints 60 • Improving the effectiveness of the govern- Share of firms (%) 50 ment’s tax incentives and other industrial policy interventions by shifting from gen- 40 eral support to performance-based target- 30 ing of successful enterprises 20 • Defining the future role of SOEs and 10 further strengthening their corporate governance 0 1 2–4 5–9 10–19 20–99 100+ • Building an effective national innovation Firm size (employees) system ­(NIS). 2011 2014 These constraints and reform areas are Sources: Calculations based on 2011 and 2014 Rwanda Census of explored in further detail in the remainder of Business Establishments data (NISR 2011, ­ 2015). this chapter, after a general overview of the current state of Rwanda’s enterprise ­ sector. FIGURE 4.2  Share of employment in Rwanda, by Current State of Rwanda’s firm size, 2011 and 2014 Enterprise Sector Rwanda’s enterprises, together with its people 40 (chapter 1) and institutions (chapter 6), are 35 Share of employment (%) key to harnessing the significant opportunities 30 that arise from modern-day technologies, 25 global trade and investment patterns, and 20 urban ­agglomeration. The number of nonfarm enterprises, 15 although still low in absolute terms, has 10 grown rapidly, from a state of decimation 5 after the genocide against the Tutsi to about 0 150,000 firms in ­ 2014. In that year, these 1 2–4 5–9 10–19 20–99 100+ firms employed about 360,000 workers, or Firm size (employees) 6 percent of the overall ­workforce. A major- 2011 2014 ity of these firms are small, although the size distribution is beginning to shift up gradu- Source: Calculations based on 2011 and 2014 Rwanda Census of 2015). Business Establishments data (NISR 2011, ­ ally: 65 percent of the firms in 2014 were one-person firms, down from 73 percent in 2011 (figure ­ 4.1). The number of large firms (those employing 100 workers or more) more enterprise employment (4 percent) remains than doubled (to 216) from a small base i nsignificant. There is strong international ­ between 2011 and 2014, and their share of evidence that small firms are less productive total employment increased to 20 percent than large firms (Bartelsman, Haltiwanger, (figure ­4.2). The number of medium firms and Scarpetta 2009), so the lack of scale and (51–99 employees) also increased sharply size is an important impediment to growth (by around 60 percent), but their share of that will need to be ­ addressed. C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 173 FIGURE 4.3  Share of firms and employment in Rwandan firms, for the most part, are Rwanda, by year of establishment young and concentrated in the nontradable ­sector. About 90 percent of the enterprises 100 that were operational in 2014, accounting 90 for 80 percent of employment in the enter- 80 prise sector, were established after 2006 70 (­ ­ figure 4.3). The two main sectors—­ wholesale Share (%) 60 and retail trade and accommodation and 50 food services—accounted for 75 percent of 40 all firms and 45 percent of total employment 30 in 2014 (figure ­ 4.4). Although the number of 20 manufacturing firms increased between 2011 10 0 and 2014, the sector’s share of employment Firms Employment ­declined. Pre-genocide (before 1995) As in other low-income countries, many Recovery (1995–2005) firms are informal (figure ­ 4.5). The gov- Postrecovery (2006–14) ernment has taken a policy stand to pro- mote formalization of ­ enterprises. Informal Source: Calculations based on 2014 Rwanda Census of Business Establishment data (NISR ­ 2015). establishments added less than 24,000 jobs FIGURE 4.4  Share of firms and employment in Rwanda, by sector, 2011 and 2014 a. Share of firms b. Share of employment Real estate Real estate Electricity Water supply Water supply Arts, entertainment Arts, entertainment Information Construction Electricity Information Professional services Transportation Transportation Mining Construction Administration Administration Professional services Finance and insurance Agriculture Mining Finance and insurance Agriculture Manufacturing Other services Other services Manufacturing Accommodations Accommodations Wholesale and retail Wholesale and retail 0 10 20 30 40 50 0 10 20 30 Share of firms (%) Share of employment (%) 2011 2014 2015). Source: Calculations based on 2011 and 2014 Rwanda Census of Business Establishments data (NISR 2011, ­ firms. Note: Excludes health, education, public administration, and defense sectors and one-person formal and informal ­ 174 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE 4.5  Informality as a share of total firms between 2011 and 2 ­ 014. Yet job creation in and employment in Rwanda, 2011 and 2014 the informal sector can be a major contribu- tor to the country’s structural transformation 100 and thus productivity, providing oppor- 90 tunities to close to 70 percent of the work- 80 force still in low-productivity subsistence agriculture. Box ­ ­ 4.1 takes a fresh look at the 70 potential of informality on the basis of inter- 60 national ­experience. Share (%) 50 Between 2013 and 2016, the pace of 40 new firms’ entry and job creation in the 30 sector slowed (figure 4 ­ ). Job cre- ­ .6, panel b 20 ation has been driven mainly by the expan- sion of incumbent formal firms, not by the 10 entry of new ­ fi rms. Over the 2011 to 2016 0 period, only 3 percent of formal firms expe- Informal Informal firms employment rienced high employment growth (more than 2011 2014 20 percent a year on average) for three years in a row, but the number of these “high- Source: Calculations based on 2011 and 2014 Rwanda Census of growth firms” ­ declined. Business Establishments data (NISR 2011, ­ 2015). Rwanda. Note: NISR = National Institute of Statistics of ­ BOX 4.1  Is informal normal? In other countries, the informal nonfarm s ­ ector While informal is normal across the low- and has of ten been a remarkably persistent and middle-income world, government responses have important driver of early-stage development, even varied ­considerably. Some countries (Bangladesh, during periods of high GDP ­ g rowth. A major Cambodia, and Ghana) have been more forthcom- contributor to the structural transformation ing in embracing it, while others (including several process, the informal nonfarm sector provides countries in Sub-Saharan Africa) have been more a valuable pathway for surplus labor to move ­ cautious. Resistance to informality, however force- out of agriculture, with opportunities that are fully applied, has not succeeded in stemming its ris- often more remunerative and productive than in ing tide, including in Sub-Saharan ­ A frica. What it a griculture. Reviewing the experience of certain ­ has done is to forgo some of the sector’s significant Latin American and Southeast Asian economies economic and job-creating potential in the early (­ B 4.1.1), a 2009 Organisation for Economic fi gure ­ stages of ­development. This loss particularly affects Co-operation and Development study concludes, mobility and opportunities for young people who “Informality is increasingly becoming normal, lack the education and skills required for a wage not least in middle- and even high-income coun- ­job. tries.… The development in selected countries Policy makers need to recognize the informal in Southeast Asia and Latin America is telling sector’s significant potential and then follow up in this respect: over the last 30 years, growth in with supportive ­ policies. To identify the appropri- these countries was accompanied by increasing, ate policies, it is important for the sector to be rep- not falling, informal employment” (Jütting and resented in consultations between the government de Laiglesia ­ 2 009). and the private sector (as in Ghana, for ­ example). (Box continues next page) C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 175 BOX 4.1  (continued) It is also important for policy makers not to view have proved willing to pay for them through fees informal sector enterprises as small and medium and ­taxes. enterprises in the ­ m aking. Both have a role to The second priority is to provide access to play in early-stage economic expansion and job finance, a key requirement for business, whether ­creation. formal or i­ nformal. Finance is needed for household Whatever their specific form, policies targeted at enterprises that otherwise suffer from lack of finan- enhancing the productive potential of informal sec- cial inclusion (Filmer and Fox ­ 2014). tor firms need to address three key ­ areas. The third priority is to provide skills and train- T he f i rst priorit y is for gover n ments to ing ­opportunities. Such opportunities can go a long in­corporate the growth of informal sector firms in way in supporting youth employment in house- their planning processes and then provide them with hold ­ enterprises. Trainings can consist of private adequate space in key areas of foot traffic in the c ­ ity. courses and a ­pprenticeships. Governments should This can avoid overcrowding and congestion, while not attempt to deliver training directly; rather, they ensuring good market access for the ­ fi rms. Informal should focus on market-enhancing programs that sector firms (just like those in the formal sector) also disseminate information about training opportuni- need key services (such as security, sanitation, elec- ties and enable disadvantaged youth to access train- tricity, transport, and water supply) to operate and available. ing that is already ­ FIGURE B4.1.1  Informal employment as a share of GDP per capita in Latin America and Southeast Asia, various years a. Latin America b. Southeast Asia 58 6.2 75 3.9 57 6.0 3.7 Informal employment (% of total Informal employment (% of total nonagricultural employment) nonagricultural employment) 70 56 (US$, thousands, PPP) (US$, thousands, PPP) 5.8 3.5 Real GDP per capita Real GDP per capita 55 65 5.6 3.3 54 5.4 60 3.1 53 5.2 2.9 52 55 51 5.0 2.7 50 4.8 50 2.5 1990–94 1995–99 2000–07 1985–89 1990–94 1995–99 Share of informal employment in total nonagricultural employment Real GDP per capita (right scale) 2009. GDP data from Penn World Tables (Feenstra, Inklaar, and Timmer ­ Source: Jütting and de Laiglesia ­ 2015). Note: PPP = purchasing power ­ parity. Rwanda’s currently small domestic mar- exporters (mostly involved in agropro- ket makes it necessary for exports to be a cessing) in 2015 (­ 4.7). From a pol- fi gure ­ g rowth. There is still quite some driver of ­ icy perspective, increasing integration in room for improvement on this, because regional and global markets is an impor- the country had only 69 manufacturing tant objective for achieving high growth, 176 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE 4.6  Number of formal firms and total employment in Rwanda, by type of firm, 2011–16 a. Number of firms b. Total employment 2011 2011 2012 2012 2013 2013 2014 2014 2015 2015 2016 2016 0 2,000 4,000 6,000 8,000 10,000 12,000 0 50,000 100,000 150,000 Private State-owned enterprise Joint venture data. Source: Calculations based on Rwanda Revenue Authority ­ FIGURE 4.7  Number and share of exporters in Rwanda, by sector, 2015 a. Number of exporters b. Share of exporters Manufacturing (meat) Furniture and other Bakery products Nonmetal minerals Furniture and other Grain milled Nonmetal minerals Wood and paper Printing Printing Wood and paper Bakery products Grain milled Textiles Textiles Beverages and tobacco Beverages and tobacco Metals Manufacturing Manufacturing (meat) (vegetables, animals) Metals Chemicals Chemicals Other food Other food Manufacturing (vegetables, animals) 0 5 10 15 20 0 20 40 60 80 100 Number of exporters Share of exporters (%) 2017. Source: Calculations based on NISR ­ a discussion taken up in chapter 2 of this present in many sectors, including in air ­report. transport, financial services, agriculture, SOEs, either fully or nearly fully owned utilities, real estate, cement, services, met- by the government or held by public-private als, dairy, and tea h­ oldings. In the absence investment groups in which government has of a strong private sector, SOEs have played invested along with private investors, are a significant role in the enterprise ­ s ector. C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 177 They have been used to de-risk strategic price trajectories of the Asian high-growth sectors, paving the way for private sector economies whose long-term performance entry. Under well-defined conditions, their ­ Rwanda aspires to surpass (figure 4 ­ .8). It presence can remain beneficial for future was only when these economies reached growth, if the level of competition and com- income per capita of more than US$5,000 petitiveness of SOEs versus private firms (2010 ­U.S. dollars purchasing power par- is strategically managed. In sectors where ity) that the ratio of their price levels the role of SOEs is no longer or will soon to those of the United States exceeded cease to be necessary, clear paths need to Rwanda’s current ­ ratio. be drawn for the government’s withdrawal, W hat d rives Rwanda’s price levels alongside efforts to build up private sector higher? In the past, high price levels partly ­capacity. reflected overvaluation of the Rwanda f ranc. That overvaluation has been cor- ­ rected through depreciation of the franc Constraints Faced by Rwanda’s over 2015–17 (IMF ­ 2017). The high costs Enterprise Sector now stem from limited access to credit, high cost of energy generation, remaining Rwanda’s private sector must play a lead- impediments to competition, and distance ing role in the country’s drive for high to markets (by virtue of being landlocked) g rowth. But the sector is generally small- ­ exacerbated by weak transport links and scale, informal, and has a small foothold in costly logistics (see chapter ­2). There are external markets, despite the impressive set also still some areas of improvement for of business-friendly reforms that Rwanda Rwanda’s generally well-regarded busi- implemented. What, then, are the bind- has ­ ness regulations, which can lower business ing constraints faced by the private sector? transaction costs ­indirectly. Several factors, principally high costs, low economic returns, suboptimal allocation of resources, and a low capacity for innovation, FIGURE 4.8  Price levels and GDP per capita in Rwanda and comparator countries and economies, 2011 impede the competitiveness of Rwanda’s pri- vate ­enterprises. 0.9 0.8 High Costs That Undermine 0.7 Index (level of U.S. GDP in 2011 = 1) Competitiveness of Rwandan Enterprises 0.6 Price levels and thus input costs are high in 0.5 Rwanda, which constrains the competitive- 0.4 ness of its e­ nterprises. Low- and middle- income countries tend to have lower price 0.3 levels than high-income economies because 0.2 of the lower costs of nontradable goods and services (the prices of tradable goods 0.1 tend to be similar across all ­ c ountries). 0 This divergence in prices tends to close 500 5,000 50,000 as countries’ income levels ­ r ise. Rwanda’s GDP per capita (2011 prices, PPP) relative prices have also followed such pat- Vietnam Rwanda China terns, but at more elevated levels compared India Korea, Rep. Taiwan, China with other economies at similar stages of 9.0 data (Feenstra, Inklaar, and Source: Calculations based on Penn World Tables ­ ­ d evelopment. The contrast is especially Timmer ­2015). stark when compared with the relative Note: PPP = purchasing power ­ parity. 178 FUTURE DRIVERS OF GROWTH IN RWANDA High Cost of Finance banks in Rwanda are compared with those As in many other low- and middle-income in neighboring ­ countries. Because of a larger economies, the cost of finance is high in domestic market, Kenyan banks, for exam- Rwanda. With an average nominal lend- ­ ple, are able to reach greater economies of ing rate of about 17 percent (figure ­ 4.9), or (table 4.1). scale, reflected in their lower costs ­ 12 percent in real terms, borrowing costs Similarly, insurance penetration (ratio of are a constraint for many private enterprises, gross written premium to GDP), also ham- especially smaller ­ ones. This reflects low pered by high costs and low scale, remains domestic savings, the small size of most firms below 2 percent compared with ­ 3.6 percent (which makes it harder for banks to recoup for the overall African insurance ­ market. the fixed costs of lending), as well as higher Capital markets, which can be a vehicle operational costs (due to small market size for raising medium- and long-term financ- and a comparatively large number of com- ing, continue to grow rapidly, but they banks). This latter challenge becomes peting ­ reflect similar scale, cost, and low-base evident when the profitability indicators of issues. Stock market capitalization was esti- ­ mated at about 26 percent of GDP as of end- FIGURE 4.9  Average lending and deposit rates in Rwanda, 2016, the second lowest in the East African 2000–17 Community ­ ( EAC). The stock market is also thinly traded, with just seven listed 20 companies, including four local companies 18 and three ­ cross-listings. The bond market, Average nominal interest rate (%) 16 introduced in 2008, is ­ shallow. The market 14 also lacks intermediaries (primary dealers) 12 and is illiquid, because most investors hold 10 securities to maturity, so there is limited 8 presence of a secondary ­ m arket. Although the size of the government primary market 6 has increased, corporate sector participa- 4 tion remains limited, with only two issu- 2 ances in ­2016. 0 2000 2002 2004 2006 2008 2010 2012 2014 2016 High Cost of Infrastructure Lending rate Deposit rate Linear (lending rate) Linear (deposit rate) The cost of electricity in Rwanda remains high because of high unit costs for generation Source: National Bank of Rwanda d ­ ata. US$0.27 per kilowatt-hour in fiscal (about ­ TABLE 4.1  Efficiency and profitability of banks in Rwanda and other East African Community countries, 2013 percentages Ratio of overhead to Country Return on assets Return on equity total assets Cost-to-income ratio Net interest margin Rwanda 1.8 10.5 8.3 67.7 10.3 Burundi 1.3 9.3 6.9 70.2 7.3 Kenya 3.4 21.2 5.2 50.3 8.3 Tanzania 1.6 14.0 5.8 64.1 7.5 Uganda 2.7 16.6 6.8 55.6 10.2 Expected median 1.8 17.5 4.5 58.1 6.0 Low income 1.8 16.7 5.3 62.4 6.3 2015. Source: FinStats and Bankscope data for ­ comparisons. Expected median is the predicted level Note: To ensure comparability, Bankscope figures are used for all countries in the peer ­ FinStats. based on a set of country characteristics, reported by ­ C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 179 2016–17) (figure ­ 4.10). The government has FIGURE 4.10  Cost of electricity in Rwanda and other Sub-Saharan put in place measures to reduce the impact African countries, 2016 of high energy costs on enterprises through a subsidized rate for industrial users ­ (US$0.11 0.30 per kilowatt-hour on average), which has fis- 0.25 Cost (US$ per megawatt-hour) cal implications (Kojima and Trimble 2 ­ 016). There are also concerns about the access to 0.20 and reliability of electricity supply. Almost (31.5 ­ a third ­ percent) of firms p ­ articipating 0.15 in the Integrated Business Enterprise Survey (IBES) reported access to reliable electricity 0.10 as a major challenge (NISR 2 ­ 017). These con- cerns are also reflected in Rwanda’s ranking 0.05 (68) on the “getting electricity” component 0 of Doing Business indicators (­ figure 4.11). Rwanda Kenya Burundi Ethiopia Tanzania Malawi Uganda The government has embarked on an impressive rollout of telecommunications 2016. Source: Calculations based on World Bank ­ infrastructure, which now reaches more than 90 percent of the ­ p opulation. It will FIGURE 4.11  Rwanda’s global rankings on Ease of Doing Business now be important to improve affordability to indicators ensure widespread use of the ­ infrastructure. Looking at cost relative to gross national income per capita, mobile cellular prices Registering property 2 amount to 10 percent, whereas fixed broadband prices amount to 166 ­ p ercent Getting credit 3 (ITU ­2016). Protecting minority investors 14 High Cost of Transport Overall 29 Rwanda’s landlocked geography continues Paying taxes 35 to result in high transport ­ c osts. Despite sustained efforts to reduce transport costs Starting a business 51 through improved logistics and reduced time duration of transport (with a fall of Resolving insolvency 58 72 percent since 2013 in the time it takes Getting electricity 68 for a container to reach Kigali from Dar es Salaam, for example), the average cost of Enforcing contracts 78 moving a container from Mombasa to Kigali is ­US$3,633.1 Trading across borders 88 Trade continues to rely wholly on road Dealing with 106 and air transport (discussed in more detail in construction permits report). To address this, two chapter 2 of this ­ 0 20 40 60 80 100 120 regional railway projects are in the ­ pipeline. Global ranking Kenya–Uganda–Rwanda (Northern Corridor) Source: World Bank 2019. and Tanzania–Rwanda (Central Corridor) will connect Rwanda to East Africa’s major ­ports. These projects are very complex Business Regulations and will require consensus on their feasibil- Despite a high aggregate score on the Doing ity and financing framework; in this regard, Business Indicators, Rwanda still has room recent developments regarding the Central to decrease the regulatory costs associated Corridor have been ­ encouraging. with starting a business, resolving insolvency, 180 FUTURE DRIVERS OF GROWTH IN RWANDA enforcing contracts, trading across borders, its income (figure 4 ­ .14), which undercuts its and obtaining construction permits and elec- labor ­ productivity. TFP growth has fallen tricity ­connections. since 2008, widening the gap with other ­countries. Low Productivity Growth An additional concern for Rwanda is the perception of low returns (especially among Low returns (real and perceived) contribute foreign investors) because of its small mar- to the slow uptake in private sector ­ activity. ket s­ ize. More than half of the companies Trends in productivity are an important indi- responding to a recent Investor Perception cation of economywide ­ returns. Rwanda’s Survey cited access to national markets as a labor productivity (output per worker) is still key reason to invest in A ­ frica. 3 In a survey 4.12), explained by neg- relatively low (figure ­ of potential investors in Rwanda, 40 percent ative or weak growth in within-sector labor cited the small market as a weakness of productivity across most sectors ­ figure 4.13). (­ the country as a foreign direct investment Three factors can explain this outcome: 2 (FDI) location—more than double any other slow accumulation of physical capital, slow response. Rwanda has the potential to over- ­ accumulation of human capital, and low lev- come this barrier by targeting larger markets (TFP). On the els of total factor productivity ­ and supply chains outside its borders, an first, Rwanda’s investment rate, currently at agenda discussed in chapter ­ 2. 26 percent, has been relatively healthy for What explains Rwanda’s weak TFP per- a while (although investment will need to formance? A country’s TFP has two key increase sharply to meet Rwanda’s aspira- determinants. The first is the allocative effi- ­ tions for growth, as discussed in the overview ciency of its resources (land, labor, ­ capital)— of this r­ eport). Therefore, physical capital that is, the extent to which these public and accumulation does not explain Rwanda’s private resources get channeled to their most low labor ­ p roductivity. Challenges with productive ­ u se. The second is the pace of human capital accumulation are discussed in technological innovation in the economy— chapter ­ ­ 1. Rwanda’s TFP levels are low for that is, the pace at which the frontiers of technology and good business practices FIGURE 4.12  Labor productivity and GDP per capita in Rwanda expand. The significance of technological ­ and other countries innovation will gather pace as the Rwandan economy moves toward middle-income status 100 and beyond, becoming an increasingly press- 90 ing subject of reform for future ­ growth. Labor productivity (2010 US$, PPP) 80 70 Resource Misallocation 60 Resource misallocation refers to inefficient enterprises commanding more resources 50 (land, labor, and capital) than warranted 40 by their productivity l ­evels. Analysis of 30 Rwanda’s manufacturing sector suggests 20 that TFP in Rwanda’s manufacturing sector 10 could be doubled by reallocating resources to 0 4.15).4 The scope more efficient firms (figure ­ 400 4,000 40,000 for productivity improvements through more GDP per capita (2010 US$, PPP) efficient resource allocation is likely to be All countries Rwanda even larger in other sectors: this is because it is typical to witness less dispersion between 9.0 data (Feenstra, Inklaar, and Source: Calculations based on Penn World Tables ­ Timmer ­2015). countries in TFP in the manufacturing sector Note: PPP = purchasing power ­ parity. than in other sectors of the economy, largely C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 181 because of the sector’s tradability (see annex FIGURE 4.13  Average growth of labor productivity in Rwanda, 4A for a technical ­ explanation). Aggregate by sector, 2001–16 productivity could be improved further through more efficient allocation of resources Agriculture across ­sectors. Transportation services D e cision s on publ ic i nve st ment —­ accounting for more than half of total invest- Health and social services ment in the economy—need to be guided Hotels and restaurants by analysis of economic r ­ eturns. Rwanda Wholesale and retail trade has made good progress in this regard, for Mining and quarrying example in developing a robust public invest- Culture and other services ment management (PIM) system, which can be applied more rigorously across priority Financial services and real estate projects. Moreover, the government’s focus in ­ Administrative and support services its industrial policy needs to shift toward dif- Utilities ferentiating firms and sectors by performance Manufacturing (past or ­potential). Finally, remaining gaps in Construction market institutions (as discussed in chapter 6) likely contribute to the inefficient allocation –3 –2 –1 0 1 2 3 4 5 Average growth rate in of ­resources. output per worker (%) Role of the State in Resource Allocation. 2017. Source: Calculations based on Diao, Randriamamonjy, and Thurlow ­ Rwanda has developed a strong PIM sys- tem overall, but gaps remain, especially with respect to project selection and ex post evaluation. Moreover, the rigors of PIM are ­ not always applied to investments under- FIGURE 4.14  Total factor productivity and GDP per capita in taken by nonbudgetary public agencies (such Rwanda and other countries as special investment vehicles for large infra- structure ­projects). Often, decisions through 1.5 off-budget processes target long-term stra- tegic priorities where expected benefits 1.3 take time to materialize and can therefore be ­u ncertain. At the same time, less than 1.1 TFP (United States = 1) 10 ­percent of total investments are targeted 0.9 at agriculture, where the scarcity of infra- structure is more clearly ­ visible. 0.7 Rwanda’s industrial policy involves active government interventions in many a ­ reas. 0.5 The government offers firms a wide range 0.3 of industrial policy incentives to develop the enterprise ­sector. The largest recurrent 0.1 item consists of tax incentives introduced by the 2015 Investment Code, estimated –0.1 400 4,000 40,000 at 1 percent of GDP (Bode, Lohmann, and GDP per capita (2011 US$, PPP) Steenbergen ­2017). Large resources have also gone into meetings, incentives, conferences, All countries Rwanda Log of all countries and exhibitions and the “Made in Rwanda” Source: Calculations based on Penn World Tables ­ 9.0 data (Feenstra, Inklaar, and plan, which is in the process of scaling ­ up. Timmer ­2015). Rwanda should monitor and assess the costs Note: PPP = purchasing power ­ parity; TFP = total factor productivity. 182 FUTURE DRIVERS OF GROWTH IN RWANDA of such incentives and ensure that that these These incentives can be made more and other incentive regimes are not overlap- ef fec tive i n several ­ w ays. Fi rst , t hey ping and are being directed toward the most appear to prioritize export promotion efficient firms likely to increase their returns and import substitution activities ­ equally. incentives. as a result of the ­ Past attempts at import substitution, for the most part, have led countries down FIGURE 4.15  Resource misallocation in formal the path of inefficiency and low growth, manufacturing in Rwanda and select countries with the lost decade of the 1980s for Latin America as a good e ­ xample. A stronger focus on firm productivity, scale, and Kenya export promotion in areas aligned with Rwanda the country’s comparative advantages would be more rewarding for Rwanda, Ghana just as it was for the high-growth East Ethiopia Asian ­e conomies. Second, as part of industrial policy, fis- India cal incentives are granted to a range of sec- Côte d’Ivoire tors rather than built around performance China criteria targeting successful ­ f irms. For example, all manufacturing firms are uni- 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 formly given value added tax exemptions Measure of misallocation regardless of ­ p erformance. Accelerated depreciation is similarly available for any Source: Calculations based on NISR ­2017. Note: Resource misallocation is measured by the dispersion of substantial capital investment, while elec- marginal products of inputs across ­firms. Large dispersions suggest tricity subsidies are available to any indus- that frictions in input and output markets prevent the movement of productive resources across firms and that inefficient firms trial ­u ser. command more resources than warranted by their ­ productivity. Ta x i n c e n t i v e s a r e e x p a n s i v e . Figu re 4.16 shows the probabilit y of receiving tax incentives, broken down by FIGURE 4.16  Probability of receiving tax incentives in Rwanda, by firm size. The larger the firm, the higher size of firm revenue, 2013–16 the likelihood of receiving tax incen- tives; firms in the bottom 50 percent of 0.15 the revenue distribution have a negative likelihood of receiving tax ­ i ncentives. Probability of receiving tax incentives F igure 4.17 compares the sectoral com- ­ 0.10 posite annual growth rate (in revenue) with the amount of tax incentives as a share of revenue between 2010 and 2 ­ 016. 0.05 Sectors like information and communica- tion technology (ICT), financial services, construction, and manufacturing receive 0 a disproportionate share of tax incentives; but these sectors have also grown less –0.05 quickly in terms of overall ­ revenue. Further, current incentives are not nec- 10 0 0 0 0 0 0 0 0 00 essarily targeted optimally across firms: –2 –3 –4 –5 –6 –7 –8 –9 0– –1 10 20 40 50 60 70 80 30 90 Firm deciles according to revenue only about a third of expenditures go to firms whose investment decisions are likely Source: Bode, Lohmann, and Steenbergen ­ 2017. influenced by the incentives (for current C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 183 purposes, these are firms whose rate of to continue strengthening firm capabili- return on investment without the incentives ties for productivity-led growth to meet was between 10 and 20 percent) (table ­ 4.2). its long-term income ­ a mbitions. Although At the same time, 28 percent of all incen- structural transformation and improved tives are given to firms that would have allocation of resources can stimulate labor achieved high rates of return on invest- productivity growth for another decade or ment (in excess of 20 percent) even with- so, sustaining it over a longer period will out the ­i ncentives. Similarly, 38 percent of only be possible through innovation and the incentives go to firms that achieved low technology d ­ iffusion. The strong positive returns (below 10 percent) even after ben- correlation between a country’s innovation efitting from the i ­ncentives. These failing capacity and its TFP illustrates this point firms (roughly half of which had negative 4.18). Global technological advance- (figure ­ returns) are unlikely to survive under nor- ments have raised the stakes even more mal market conditions, with or without the tax ­incentives. FIGURE 4.17  Sectoral revenue growth and tax incentives in Resource Misallocation Caused by Rwanda, 2010–16 Shortcomings in Market Institutions.  Resource allocation through market signals is efficient Agriculture only to the extent that the signals are generated 200 in competitive and well-regulated m ­ arkets. Other services Despite having made major gains in the past 150 Revenue CAGR (%) Wholesale and retail trade 20 years, Rwanda still faces several challenges Mining in this ­ regard. Its factor markets and core mar- Financial services ket institutions for competition and contract 100 Manufacturing and property rights enforcement are still in an Construction ICT early stage of development and face a signifi- 50 cant agenda for further reform (discussed in Transportation and storage chapter ­6). Utilities 0 0 5 10 15 20 Technological Innovation Tax incentives/revenue (%) Rwanda is in the early stages of building Industrial sectors Weighted regression its innovation capacity, which is the abil- 2017. Source: Bode, Lohmann, and Steenbergen ­ ity to introduce new products, processes, Note: CAGR = compound annual growth rate; ICT = information and communication solutions. It needs ideas, technologies, and ­ ­technology. TABLE 4.2  Share of total tax revenue for firms, by return on investment with and without incentives % of total tax revenue ROI without incentives (%) All firms FDI firms Domestic firms ROI with incentives (%) <0 0–10 10–20 > 20 <0 0–10 10–20 > 20 <0 0–10 10–20 > 20 <0 21.0 14.8 22.5 0–10 13.8 3.2 0.0 2.8 17.7 2.6 10–20 0.4 6.8 16.9 0.0 2.7 0.1 8.2 0.2 20.5 > 20 5.2 0.0 4.8 27.8 0.0 0.0 4.3 75.2 6.4 0.8 4.7 16.2 Source: Steenbergen and von Uexkull ­ 2018. investment. Yellow: unviable incentives (firm is still not profitable with tax ­ Note: ROI = return on investment; FDI = foreign direct ­ incentives). incentives). Orange: redundant incentives (firm is profitable even without tax ­ Green: marginal firms (effective tax ­ incentives). 184 FUTURE DRIVERS OF GROWTH IN RWANDA for Rwanda (and other low- and middle-­ Developing the capacity to innovate is income countries), by revolutionizing the not an overnight endeavor; it requires an world economy and presenting major ecosystem that takes decades to ­ develop. It opportunities and challenges that need to involves sustained investments to cultivate 4.2). be managed proactively (box ­ and upgrade a dynamic private sector by building higher-order human capital (chap- FIGURE 4.18  Correlation between innovation capacity and total ter 1) and a general societal and institu- factor productivity tional openness to new ideas, learning, and risk taking (chapter 6­ ). An effective NIS— 1.4 an ecosystem for innovation—is needed to ensure strong uptake of learning, innova- 1.2 United States tion, and technology diffusion from both 1.0 the demand and the supply ­ sides. Countries have found the NIS framework to be a use- Relative TFP 0.8 ful conceptual anchor for developing and 0.6 implementing innovation policies and for building the necessary infrastructure and 0.4 institutions (box ­ 4.3). Rwanda 0.2 The good news is that the innovation capacity of Rwandan firms has been on 0 10 20 30 40 50 60 70 an upward trend, as reflected in Rwanda’s Global Innovation Index rankings. For example, on the Global global ­ Innovation Index, Rwanda moved from Sources: Cornell University, INSEAD, and WIPO 2017; Jones ­ 2016. Total factor productivity (TFP) 112 in 2013 to 99 (out of 126) in 2017 is computed for 2010 using Penn World Tables ­ 2015). 8.0 data (Feenstra, Inklaar, and Timmer ­ Note: TFP is calculated as ­ labor-augmenting. TFP = 1 for the United ­ States. (­ ­f igure 4.19). That improvement made BOX 4.2  The global context and technological transformation The world is undergoing a significant technologi- These technologies also create new complementar- cal ­ t ransformation. Some are calling it the fourth ities across sectors (for example, manufacturing, industrial revolution, or Industry 4 ­ .0, while oth- agriculture, and services) that open up new areas ers see it as a continuation of the information and of comparative advantage for both high-income communication technology ­ r evolution. This new and low- and middle-income ­ e conomies. These industrial revolution is characterized by the adop- complementarities create both opportunities and tion of cyber-physical systems such as robotics and important challenges for Rwanda, as well as other drones, three-dimensional printing, artificial intel- low- and middle-income economies, as traditional ligence, and machine learning across all sectors of labor cost advantages become less relevant with the economy, reshaping both the way in which and advanced automation and manufacturing (Cirera, where production is ­ done. Frias, and Hill ­2017). Industry ­ 4.0 has the potential to make a large In the low- and middle-income world, benefits impact on employment and income distribution thus far have been most visible in East Asia, where (Autor and Dorn 2013; Frey and Osborne ­ 2 016). countries have accelerated economic growth by This impact is already becoming apparent, with harnessing technological advancements to evolving robotics and artificial intelligence enabling the opportunities in global ­ t rade. Other parts of the reshoring of advanced manufacturing (as well as world—Sub-Saharan Africa in particular—would services) to mature economies, although typically do well to learn from the Asian responses, includ- with fewer jobs that require more ­sophisticated skill ing by upgrading their technological and innova- sets (Acemoglu and Restrepo 2017; Ford ­ 2 015). tion ­c apacity. C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 185 BOX 4.3  The national innovation system In its most basic sense, the national innovation On the demand side of NIS are firms that apply system (NIS) encompasses an ecosystem of orga- new knowledge to business and production activi- nizations and policies that allow workers and ties in order to generate economic and commercial firms to produce in new and better ways, con- returns. High returns are necessary incentives for ­ tinuously adding knowledge to products (figure firms to continue their quest for the best available ­B 4.3.1). knowledge and technology, including by investing On the supply side of NIS are institutions that in the necessary managerial and human capital generate knowledge to feed the innovation ­ process. ­capabilities. These institutions include universities, research But to engage in innovation activity, firms institutes, technology centers, business develop- need an enabling environment that includes effec- ment providers, skills providers, small and medium tive institutions and infrastructure to support the enterprise support services, and domestic and for- NIS and sound innovation ­ p olicies. Other neces- eign firms that generate k­ nowledge. For low- and sary complementary factors include availability middle-income countries, innovation for the most of skills, access to technology and finance, com- part consists of adopting (localizing) and applying petition, economies of scale, favorable conditions existing knowledge and technology, largely gener- for business, and good managerial and organiza- ated in higher-income ­ countries. tional ­practices. FIGURE B4.3.1  The national innovation system as an ecosystem for economywide innovation Government oversight, resolution of market and systemic failures, coordination SUPPLY ACCUMULATION/ALLOCATION DEMAND Universities / think tanks / The rm K Physical capital technology extension centers H Human capital Human capital Incentives to accumulate Support to rm capability A Knowledge – Macro context NIS INSTITUTIONS upgrading – Competitive structure – Productivity/quality extension Barriers to all accumulation – Trade regime and international services Credit networks – Process/best practice Entry/exit barriers Business/regulatory climate Firm capabilities dissemination – Advanced consulting services Rule of law – Core competencies (management) – Production systems Domestic science and technology Barriers to knowledge accumulation – Technological absorption and system Rigidities (labor, etc.) production International NIS Seed/venture capital Innovation externalities Source: Maloney ­2017. Note: NIS = national innovation system. Rwanda the best-performing low-income Rwanda 44 (out of 137) on a measure of country, with some of the subindicators innovation, ahead of many Asian and Latin not far from the top performer in Africa American ­ e conomies. Yet in some areas (South Africa), though still some distance Rwanda appears to be doing less well, espe- from Switzerland (the top performer in the cially in areas related to knowledge and ­ .20). The 2017–18 Global world) (figure 4 4.20). This technological readiness (figure ­ Competitiveness Index similarly ranked ranking can be traced to gaps in human 186 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE 4.19  Ranking of Rwanda and other East African ICT infrastructure, and creative outputs Community countries on the Global Innovation Index, (for example, industrial designs and online 2013 and 2017 ­creativity). 140 Demand for Innovation 120 The demand side of Rwanda’s NIS is com- 112 posed of a small universe of firms that 99 100 are mostly informal, small-scale enter- prises, as highlighted ­ e arlier. As a result, Global ranking 80 Rwandan firms, for the most part, are 60 not yet strong demanders of i ­nnovation. Demand for innovation is hampered by 40 low capabilities for innovation among firms (even larger ones), which generally 20 do not utilize the quality of research being done in the country ­­ (figure 4.21, panel ­ a). 0 The capabilities gap relative to other coun- a a e da da a an ny bi qu an an tries also is reflected in the low number m Ke tsw bi Rw Ug Za am Bo of patent applications per capita (figure oz M 2013 2017 4 .21, panel ­ ­ b). However, Rwanda has Sources: Cornell University, INSEAD, and WIPO 2013, ­ 2017. made significant progress in increasing the Note: Ranking is out of 140 ­countries. number of trademarks that originate from the country (figure ­ 4 .22). 5 This increase was mostly on account of nonresidents, FIGURE 4.20  Frontier of innovation capacity in Rwanda, South Africa, and Switzerland underscoring the importance of import- ing external ­t alent. Low firm capabilities Institutions stem from skills constraints and short- 120 falls in management practices (PSF ­ 2013). Almost 20 percent of firms report that 80 Human capital access to skills is a constraint to their busi- Creative outputs and research ness (NISR ­ 2017). Large businesses report 40 more constraints regarding access to skills than do small ­ businesses. 0 Innovation is also hampered by the Knowledge and technology outputs Infrastructure scarcity of links with foreign firms and export markets (chapter 2). FDI is another important external source of technological ­ catch-up. FDI inflows to Rwanda have risen Business sophistication Market sophistication in recent years, but they are concentrated mostly in the nontradable sectors, where Rwanda South Africa Switzerland the scope for technology transfers can be ­limited. 2017. Sources: Cornell University, INSEAD, and WIPO 2013, ­ Note: Main components of the Global Innovation ­ Index. Supply of Innovation The supply side of Rwanda’s innovation sys- capital and research (chapter 1), as well as tem is only beginning to emerge and will to low levels of knowledge creation and have to address two key challenges: (1) a technology outputs (for example, number skilled workforce and (2) the quality and of patents and number of scientific and relevance of research and advanced training technical publications per unit of GDP), chapter 1 for ­ (see ­ details). C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 187 FIGURE 4.21  Quality of research institutions and number of patent applications in Rwanda and other countries a. Quality of research institutions, collaborations, b. Patent applications per capita and investments in R&D Israel India United States Finland China Australia Singapore Chile Japan Korea, Rep. Mexico India China Colombia Chile Mexico Rwanda Colombia Rwanda Sub-Saharan Africa Sub-Saharan Africa 2 3 4 5 6 7 0 5 10 15 20 Index PCT patent applications Quality of scientific research institutions per 1 million people University-industry collaboration in R&D Company spending on R&D Source: World Economic Forum ­2017. Note: The quality of research institutions assesses the prevalence and standing of private and public research institutions, calculated as the sum of the inverse ranks of all research institutions of a country included in the SCImago Institutions ­Rankings. Collaboration in R&D is the average score on the World Economic Forum’s Executive Opinion Survey (1–7) on the question: In your country, to what extent do business and universities collaborate on R&D (1 = not at all; 7 = to a great extent)? Investments in R&D are the expenditure on research and development as a percentage of ­ GDP. Patent applications are the total number of patent families filed in at least two of the five major offices in the world, from the PATSTAT database (EPO, various ­ years). EPO = European Patent Office; PCT = Patent Cooperation Treaty; R&D = research and ­development. The Enabling Environment for Innovation strong ­ i mprovements. On intellectual A n ef fe c t ive i n novat ion system a lso property rights, the 2009 law is in line involves an enabling environment that with good practices established by the allows the demand and supply sides to World Intellectual Property ­ Organization. efficiently. The factors that shape interact ­ Tighter implementation is now the ­ k ey. the enabling environment include pro- The Rwanda Standards Board was estab- tection of intellectual property rights, lished in 2013 to provide standardization quality IC T and other infrastructure, ­ s ervices. It is a quality institution that is availability of start-up finance, and the well regarded by investors and consumers overall business c ­ limate. Rwanda’s per- at home and abroad, providing a compara- formance on these measures has shown tive advantage to ­ Rwanda. 188 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE 4.22  Number of trademark applications and registrations in Rwanda, 2001–15 a. Trademark applications b. Trademark registrations 1,200 1,200 1,000 1,000 800 800 Number Number 600 600 400 400 200 200 0 0 20 1 02 20 3 20 4 20 5 20 6 20 7 20 8 20 9 20 0 20 1 12 20 3 14 15 20 1 02 20 3 20 4 20 5 20 6 20 7 20 8 20 9 20 0 20 1 20 2 20 3 20 4 15 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 20 20 20 20 20 20 Resident Nonresident Source: WIPO ­2017. 2011. Note: No information is available on patent applications between 2008 and ­ G over n ment has made subst a nt ia l Reform Agenda for Reaching investments in ICT infrastructure, but con- Upper-Middle-Income Levels tinued efforts are needed to upgrade the quality (and uptake) of ICT ­ infrastructure. Emergence of a vibrant, competitive, and ICT infrastructure needs to be high-speed, innovation- d riven enterprise ­ s ector — reliable, available, and accessible, and con- bot h pr ivate a nd publ ic f i r m s — t h at tinued investments are required to improve responds with agility to emerging domes- bandwidth and infrastructure r ­ eliability. tic, regional, and global opportunities The Networked Readiness Index is a com- is essential for future ­ g rowth. Private prehensive composite index that assesses enter pr ise s cre ate jobs a nd generate a country’s “preparedness to reap the incomes, drive economic transformation, benefits of emerging technologies and to compete in global export markets, and, capitalize on the opportunities presented ultimately, drive aggregate productivity by the digital revolution and beyond” growth and ­ i nnovation. Moreover, the (World Economic Forum, INSEAD, and private sector will have to provide much Cornell University ­ 2 016). Rwanda per- of the huge investment needed to support forms least well in readiness (115) and is future growth ambitions, because public ranked 106 in ­ i nfrastructure. In terms of investment is already at the limits set by digital infrastructure, Rwanda is lagging financing ­ o ptions. Efforts to boost pri- because of the lack of investment and inad- vate investment need to be accompanied equate metropolitan and last-mile access by measures to scale up domestic ­ s avings. ­ n etworks. The high cost of broadband Ongoing efforts to boost long-term sav- lines, combined with low computer owner- (box 4.4). ings are crucial in this regard ­ ship and an erratic supply of electricity, put The state will have an equally impor- the service beyond the reach of most pri- tant, but complementary, role to play vate ­users. in correcting various (clearly identified) C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 189 BOX 4.4  Rwanda’s Long-Term Saving Scheme: Opportunities and challenges The Long-Term Saving Scheme went into effect in rates per income group per ­ m onth. In monthly 2 018. What its impact on household savings will ­ ­ muganda meetings, people could contribute the u be is still ­ u nclear. Because of the large proportion money they committed to saving into a savings of Rwandan workers whose income varies signifi- scheme, which could be attached to the Long- cantly throughout the year or who work in the Term Saving ­ S cheme. This type of policy allows informal sector, participation is not ­ m andatory. people to make an (informed) choice about their Rwandans will need to be sensitized on the benefits actions and nudges them in a positive direction of this type of scheme; otherwise savings could be with the understanding that they know their own ­suboptimal. b est. Thus, the costs of failing to meet the choices ­ Using public com mitment as an incentive savings target should simply be the knowledge of can be ­ u seful. When people make a public com- s hort. A heavier sanction—for example, falling ­ mitment to how much they will save, they are some degree of social ostracism—­ would under- more likely to save (Gugerty 2005); self-help mine the ­ p olicy. peer groups increase the number of deposits in Policy makers also may wish to consider a formal savings account ­ 3 .5-fold and almost other complementary measures to boost ­ s avings. double the average savings balance (Kast, Meier, One approach would be to adjust macropru- and Pomeranz ­ 2 012). This approach may work dential regulations for the banking sector to for Rwanda because of widespread respect for discourage households from overborrowing for authority, the importance of reputation, and the ­ c onsumption. In 2014, more than 70 percent of informal nature of most people’s w ­ ork. Using the population had an outstanding loan or had umuganda — the last Saturday of each month repaid a loan within the past 12 ­ months. Only when Rwa nd a n s b e t we en 18 a nd 65 come one-third of households with loans reported together to do a variety of public works such as borrowing for investment or education-related infrastructure development and environmental expenses, including buying agricultural equip- protection—could provide behavioral economic ment or inputs or expanding a business, while insights to increase the savings r ­ ate. The gov- two-thirds of households reported borrowing for ernment could recommend a range of savings consumption (NISR ­ 2 016). Source: ­MINECOFIN. market failures and defining its role in pri- ­ ncreases. SOEs will also remain crucial for i vate investment—by implementing well- ­ rivate sector several years to come, as the p targeted public investments, strengthening needs time to ­build. bureaucratic capacity, withdrawing gradu- The policy response will have four main ally from productive activity while enhanc- elements: ing its regulatory and facilitating functions, and developing capable and accountable 1. Strengthening the overall business envi- market ­ institutions. Public investments will ronment by continuing to tackle key cross- remain important for several years to come, cutting constraints (such as finance and given the vast infrastructural and social other backbone services, infrastructure, needs of the country; that said, with grow- and market institutions) ing financing needs under the First National 2. Improving the effectiveness of the gov- Strategy of Transformation and tightening ernment’s tax incentives and other indus- limits on public borrowing, the relative trial policy interventions by shifting from share of public investment will decline over ­ general support to performance-based tar- time as the share of private investment geting of successful enterprises 190 FUTURE DRIVERS OF GROWTH IN RWANDA 3. Defining the future role of SOEs, clarify- services content—­ services supply the major- ing their objectives, and improving their ity of added value within value chains— corporate governance which calls for policies that emphasize 4. Building an NIS to support effective developing a strong financial sector (dis- uptake of innovation from both the cussed below), developing other professional demand and supply ­ sides services (chapter 2), and building the skills base (chapter ­ 1). Rwanda’s financial sector has taken Reduce Business Costs by Tackling important strides toward better serving the Cross-Cutting Constraints economy’s growing complexity and ­ needs. Rwanda has made substantial progress Apart from banks and other deposit-taking toward reforms that improve the business institutions, insurance companies, pension environment, and the government has put in funds, and other nonbank financial insti- place a variety of programs to provide sup- tutions have established a presence and are port for ­fi rms. Nevertheless, more effort is providing an expanding range of products needed in three key areas: (1) developing the and services, including for payments, sav- financial sector and other backbone services, ings, credit, insurance, and retirement prod- (2) building accessible and affordable infra- ucts, with different levels of sophistication structure, and (3) strengthening market insti- and ­i nnovation. 6 The positive trends are tutions and fiscal ­ prudence. indicators. visible in several financial sector ­ For example, in 2016, 89 percent of adults Develop the Financial Sector and Other had an account at a financial institution, Backbone Services compared with 72 percent in 2012 and Development of modern services (includ- far higher than comparator countries in ing finance) is crucial for competitiveness the r­ egion. Private sector credit equaled and long-term g ­ rowth. A country’s ser- 1 5.7 percent of GDP as of end-2016, a ­ vice ­sector typically grows in size (relative four-point increase over 2009, but the ratio to GDP) and complexity as its income level remains on the low s ­ ide. Strong progress r ises. But income is not the only factor at ­ has been made in the use of electronic pay- play. Structural characteristics such as the ­ ments and mobile ­­ banking. As of end-2016, quality of institutions and the urbanization ­ more than ­ 9.7 million users had subscribed ­ atter. For example, rate of the country also m to mobile payment systems, nearly 1 ­ million in India services that were the most underde- users had subscribed to mobile banking veloped were also the most heavily regulated services, and the value of total e-payments (Eichengreen and Gupta ­ 2010). Countries had risen to more than 30 percent of GDP with better institutions have larger and more within five ­years. dynamic service sectors (Amin and Mattoo The banking sector’s assets have almost tri- 2008). Relatively slow urbanization has been ­ pled in nominal terms since 2008, as existing identified as a reason why the service sector banks have grown larger and several new ones has developed more slowly in China than in have ­entered. The rising presence of foreign- India (Wu ­ 2007). owned commercial banks (which accounted Rwanda has been actively develop- for 46 percent of banking sector assets as ing its service sector, an emphasis that of end-2016) reflects the increasing attrac- will remain important for the foreseeable tiveness of the Rwandan market to foreign f uture. An efficient service sector not only ­ investors, particularly from within the ­EAC. makes direct contributions to a country’s In an especially welcome move, some for- GDP and export capabilities but also serves eign banks have started to cover underserved as an important input to manufacturing market segments (low-income households and ­ a griculture. High-value-added seg- and small and medium enterprises [SMEs]) ments of value chains are especially rich in using innovative lending techniques (such as C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 191 cash-flow-based lending and agent ­ banking).7 have stagnated or declined over the last two The National Bank of Rwanda (BNR) and the years, which may indicate the limitations government have played an important role in of efficiency gains in the context of a small this regard by providing an accommodating economy, increased competition from new legal and regulatory framework and business market entrants, or a more challenging eco- ­environment. nomic ­environment. Despite these positive developments, sev- Sustained reforms in the financial sector eral challenges remain, reflecting the low will remain important for Rwanda’s high- base and small size of financial institutions growth a ­ mbitions. The backdrop of shrink- and the high costs of ­ f inance. Rwandan ing external assistance raises the stakes even enterprises often cite limited access to and more, because it requires Rwanda to mobilize high cost of long-term finance among their more domestic resources and use them more key ­challenges. The lack of adequate long- ­ efficiently. Moreover, only long-term finance term financing is a key concern, which may (defined as maturities of at least five years have been exacerbated in recent years by the or stable financing sources with no specific crowding-out impact of the rise in lending to maturity) can effectively substitute for devel- large-scale infrastructure or hospitality proj- opment ­assistance. ects (often backed by sovereign g ­ uarantees). Four key areas are on the reform agenda The rise in large exposures in the banking constraints. to alleviate long-term financing ­ sector may be an indirect indicator of this phenomenon (table ­ 4.3). Preserve and Strengthen the Stability and Although Rwanda’s banking sector is Performance of the Financial Sector. The sound on a systemwide basis, on the basis success of the financial sector as a catalyst of typical stability indicators, the sector’s for Rwanda’s ambitious growth aspirations liquidity, ­efficiency, and profitability indi- is contingent on its ability to sustain and cators should be monitored ­ carefully. As of consolidate its role as an efficient and reli- March 2017, capital in the sector was well able provider of financial ­ services. The rapid above regulatory requirements, and balance growth of the sector has realized demonstra- sheet vulnerabilities from liabilities denomi- ble results, but also revealed vulnerabilities nated in foreign currency were low (BNR that need to be monitored and ­ mitigated. The 2016). But the ratio of nonperforming loans ­ need to address these vulnerabilities holds for to gross loans has been on the rise, increas- all parts of the financial sector, from banking ing from ­ 6.2 percent in December 2015 to to pensions, insurance, and capital m ­ arket. 8.1 percent in March 2017, albeit still lower ­ Investors are attracted to markets that are than that of regional ­ p eers. 8 Indicators of sound, stable, and efficient, including a stable banking sector efficiency (cost-to-income macroeconomic ­framework. ratio, overhead-to-income ratio) and profit- The government, through the National ability (return on assets, return on equity) Bank of Rwanda, has undertaken major TABLE 4.3  Large exposures in the banking sector in Rwanda, 2015–17 2015 2016 2017 Indicator March June September December March June September December March Insider loans to 2.1 2.8 3.0 4.0 3.1 3.5 3.6 2.3 4.4 core capital Large exposures 86.0 107.2 98.6 112.6 99.2 122.7 141.0 154.2 166.8 to core capital Large exposures 20.1 22.5 20.9 23.4 22.6 26.5 29.9 31.7 33.1 to gross loans ­ ata. Source: National Bank of Rwanda d 192 FUTURE DRIVERS OF GROWTH IN RWANDA reforms of the legal and regulatory frame- funds, it has to ensure that the funds are work for the financial sector, including kept safe and, at the same time, earn a suf- strengthening the system to deal with a finan- ficient ­return. occur. These reforms cial crisis, if one were to ­ A modern insurance sector will be able will enable Rwanda to become a more attrac- to play its part in strengthening the investor tive destination for local, regional, and inter- ­ base. The split-up of the composite insur- national ­ i nvestments. Underpinning these ers9 is increasing the demand for longer-term developments is the need for financial edu- investments, a trend that will only grow over cation and awareness to build a culture sup- ­ time. It is also expected that longer-term sav- portive of long-term savings and ­ investment. ings products will become more widely avail- A sound debt management policy, already in able as new market entrants offer different place in Rwanda, is similarly important to product ­ menus. The to-be-introduced risk- proactively identify and mitigate debt r ­ isks. based approach to regulation and supervision Fiscal and industrial policies also have a role; also will encourage insurers to diversify their tax breaks and other policy incentives for investments and invest in assets with better investments in strategic sectors would need credit ratings and l ­iquidity. This approach to be balanced better against the objectives could increase the demand for government of financial stability and fiscal ­sustainability. and corporate bonds as well as encourage banks to seek credit ratings, which would Broaden the Base of Institutional help to develop the market ­further. Investors.  Institutional investors are an impor- tant source of long-term financing, yet the base Deepen Regional Market Integration.  Tapping of institutional investors is quite narrow in into regional sources of funding will be a nec- Rwanda (as in the rest of the ­ EAC). In 2014, essary complement to domestic resources, institutional investors had about ­ US$1.05 given the limited size of the domestic financial billion in assets, which was less than the net market. Global experience shows that smaller ­ overseas development assistance received by countries tend to benefit more from regional the country that ­ year. Although some of these integration by leveraging the position of the funds were invested in long-term projects, largest market in the region to create advan- a more efficient allocation of the remaining tages of greater scale, efficiency, and visibility funds could contribute to the investment needs with institutional ­ investors. country. of the ­ Reducing barriers and costs to attract a Supporting the Rwanda Social Security wider range of investors will require action Board (RSSB), for example, in strength- at the regional (EAC) ­ level. The 2014 EAC ening its investment policy and fund Common Market Scorecard calls for a sig- management is a necessary precursor to nificant rollback of laws, regulations, and establishing a larger base of institutional investment codes that impede and limit i nvestors. Rwanda is still at an early stage ­ the benefits from regional ­ i nvestment. of reforming its predominantly public pen- There is a need to address several gaps in sion sector with the establishment of a new the legislation and regulations of partner regulatory regime and other measures aimed states regarding trading in securities and at strengthening the management and gov- derivatives and to develop regional mar- ernance of the ­ institution. The RSSB’s fund ket ­i ntermediaries. Wagh, Lovegrove, and management capacity also requires further Kashangaki (2012) mention five other key strengthening. Investments in equity have ­ priorities for regional integration of finan- grown over time, but they need to be more cial markets: d iversified. At the same time, the invest- ­ ments need to be commensurate with the 1. H a r m o n i z i n g t h e l e g a l a n d RSSB. Because the RSSB is risk profile of the ­ ­r eg u l a t o r y ­f ra m e wo rk s . T he E AC responsible for managing all public pension still faces ­ d ifferential tax regimes and C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 193 regulatory ­ d iscrepancies. Further align- demonstrated that the Rwandan capital mar- ing supervisory frameworks and report- ket can facilitate long-term, local-currency ing requirements is also ­ i mportant. financing from beyond the ­ region. The bond 2. Adopting a single-licensing regime and helped to attract attention to the Rwandan mutual recognition among r ­egulators. market and may entice other potential issu- Single licensing can reduce cross-border suit. The IFC issue considerably ers to follow ­ transaction costs and barriers to entry increased the value of nongovernment bonds and should be made available for banks, outstanding and contributed to building the brokers, and insurance ­ companies. Single capacity of regulators, market intermediar- licensing should be accompanied by Rwanda. The regulatory ies, and investors in ­ mutual recognition among regulators, framework for nongovernment bond issuance converging around international principles is considered supportive, but more work is (for example, Basel ­agreements). needed to develop a pipeline of credible issu- 3. Building up regionally compatible finan- ers (including subnational issuers) and build a cial ­i nfrastructure. Critical parts of credit culture and capacity among ­ investors. national financial infrastructure need to be compatible at the regional level, including Build Accessible and Affordable central securities depositories and trading Infrastructure platforms for national ­exchanges. Expanding and improving infrastructure 4. Strengthening cross-border supervisory ­ wanda. Public invest- are key priorities for R ­p ractices. Consolidated supervision is ments have greatly improved Rwanda’s infra- important to ensure that weaknesses in structure and service ­ delivery. To ensure the one financial institution do not put the successful upgrading of the country’s infra- regional financial system at ­ risk. structure for meeting the demands of the next 5. Strengthening data ­ gathering. Information phase of sustained growth, effective planning on current volumes of cross-border trade and prioritization of infrastructure and effi- in financial products is often incomplete or cient allocation of infrastructure financing inaccurate. Data gaps limit the possibilities ­ resources (public, donor, and private) will be for supervision, while also obscuring the ­ required. Four reform priorities are impor- extent of cross-border links (thus hiding regard. tant in this ­ the potential opportunities from regional integration). These issues are compounded ­ Strengthen the Public Investment Framework by high volumes of unrecorded informal for Infrastructure.  Rwanda’s national PIM trade and the widespread use of physical system establishes the processes and bench- cash. Stronger capacity for data collection ­ marks to assess trade-offs and prioritizes and analysis would be beneficial both at the among public investment o ­pportunities. national level and for the EAC ­ Secretariat. To improve it further, the first imperative is to continue strengthening the public invest- Although the EAC Council of Ministers ment planning process to prioritize public has set out several key directives that would infrastructure spending, select appropriate increase the regional scope of financial laws structures and financing sources, and man- and regulations, little progress has been made age multiyear fiscal commitments and finan- in the last few years (EABC 2 ­ 016). Using a cial implications that can materialize in the presidential forum to fast-track these initia- medium to long ­ term. This would help to tives could help to transform this sector, achieve a better balance between long-term which is critical for Rwanda’s ­development. strategic priorities and areas where returns are more clearly i dentified. The second ­ Attract Finance beyond the Region. The imperative is to broaden the coverage of the success of the International Finance PIM framework to include public nonbudget- Corporation’s (IFC’s) umuganda bond10 ary i­nvestments. Rwanda’s strengths in PIM 194 FUTURE DRIVERS OF GROWTH IN RWANDA need to be extended to budgetary agencies, to the long-term sustainability and manage- misallocation. eliminate any potential risks of ­ ment of the government’s financial com- The third imperative is to prioritize ex post mitments to PPPs are integrated into the evaluation and external audit, while fur- current medium-term debt management ther strengthening the cost management strategy. framework. These issues are discussed in ­ • Clarify accounting and budgetary treat- 6. more detail in chapter ­ ment of PPPs to enable the government to deliver long-term financial commitments Leverage Private Finance in Support of to PPPs (which is not possible under the Infrastructure Investment.  Rwanda has had current framework). mixed experience with implementing public- • Prepare municipal PPP guidelines to facili- private partnership ­ projects.11 On the one tate the development of municipal-level hand, 39 projects that can be broadly defined PPP projects. as public-private partnerships (PPPs) have • Design and deliver a communications already been implemented: 29 in the energy strategy to demonstrate government com- sector, 3 in ICT projects, 1 in manufacturing, mitment to and ownership of the PPP and 6 in ­ mining. On the other hand, these program. transactions were processed on a project- • Develop a framework to deal with unsolic- by-project basis, outside of a dedicated PPP ited proposals to ensure value for money, law and were procured under sector-specific for instance, by allowing the evaluation of or project-specific laws that are designed for alternative ­proposals. traditional public procurement processes and lack the detailed multistage, PPP-specific pro- High-level government support for PPPs cesses that can ensure transparency and com- is shown by enactment of the PPP Law in petitiveness (box ­ 4.5). 2016. Attention must now focus on building ­ The following recommendations are over time a robust PPP pipeline, in parallel to intended to strengthen the PPP framework: building the required institutional processes and capacity (especially among implementing • Finalize and approve the PPP guidelines agencies). Project identification, preparation, ­ to create the enabling framework for procurement, implementation, and overall PPPs. contract and fiscal oversight methodologies • Develop a PPP pipeline via a clearly need more w ­ ork. In Rwanda, the central PPP defined project-screening process. entity is placed in the Rwanda Development • Develop a fiscal commitments and contin- Board, which serves as the chief n ­ egotiator gent liabilities framework and ensure that for government on strategic investment d ­ eals. BOX 4.5  Institutional framework for PPPs: An international perspective The institutional framework for public-private part- •• Ministry of Finance: responsible for assess- nerships (PPPs) varies across countries that have ing and managing the fiscal implications of enacted a PPP law, dependent on each country’s specific PPPs political and bureaucratic ­ context. However, the most •• Centralized PPP entity: responsible for pro- common institutional framework lays out clear roles moting and coordinating PPP activities and for for three entities, in particular: supporting line ministries or contracting agen- cies in the development of ­PPPs. This entity is •• Line ministries or contracting agencies: respon- most typically located within the Ministry of sible for identifying, preparing, procuring, and ­Finance. implementing PPPs C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 195 This means that RDB and the Ministry of is fully implemented as planned, the gover- Finance and Economic Planning need to nance framework in the power sector could closely coordinate on the assessment and become a model for sector governance in management of fiscal commitments. Rwanda more ­ broadly. Specific actions are needed to ensure a reli- Tackle the High Cost of Electricity.  Rwanda able and affordable supply of energy: has implemented a suite of restructuring mea- sures to improve governance of the electric- • Institutionalize least-cost sector planning utility. The private sector has become a ity ­ through adoption and regular updates of strategic partner for Rwanda’s power sector the power sector master plan, based on through its investments in power generation reliable updated demand forecasts; com- and off-grid ­ access. Blackouts have become petitive implementation of the resulting frequent. Nevertheless, the high cost of less ­ least-cost projects is essential. electricity poses a constraint on Rwanda’s • Institute a process to assess, measure, economic and industrial development, as dis- monitor, and manage contingent liabilities cussed ­earlier. from the power sector. Electricity supply is expensive because • Conduct a comprehensive assessment Rwanda has limited low- cost energ y of the potential of indigenous energy resources, and it does not have a large enough resources available for electricity genera- market to benefit from scale e ­ conomies. tion and options for their optimal exploi- As firms are caught between the high cost of tation, with a focus on dissolved methane electricity and limited affordability, the gov- in Lake Kivu. ernment has stepped in with fiscal support, • Promote regional electricity trade to take at some risk to the b ­ udget. Going forward, advantage of the regional energy potential the negative impact of high production costs of lower-cost sources of supply through could be exacerbated if generation capacity bilateral contracts (consistent with keeping far exceeds expected ­ demand. If the power adequate levels and secure supply), closely plants in the pipeline are completed accord- coordinate with neighboring countries ing to the government’s schedule, the supply to ensure that the planned cross-­ border of electricity will far exceed the expected t ransmission infrastructure projects ­ peak demand by ­ 2025. are not delayed further, and participate Recognizing these challenges, the gov- actively in developing the East Africa ernment is preparing a framework to insti- Power Pool platform to balance regional tutionalize least-cost sector planning and short-term mismatches between demand competitive procurement, together with and supply. reliable demand ­ forecasts. Finalizing and • Take full advantage of off-grid private operationalizing the Least-Cost Power Sector sector investment to boost rural electrifi- Development Plan will be an important step cation in areas where extending the grid in this ­direction. As part of least-cost plan- is not financially viable, while closely ning, the government also should consider monitoring the private sector’s adher- accelerating its preparations for coordinat- ence to the International Electrotechnical ing energy imports from and exports to Commission standards for off-grid solar neighboring countries, building capacity to technology. procure and implement power-purchasing • Promote energy efficiency in consump- agreements, strengthening demand fore- tion to take advantage of available energy-­ casting, and better integrating system plan- efficient technologies (energy-efficient ning and system operations functions in the street lighting, energy-efficient appliances) utility. If the plans to make the utility holding ­ to mitigate the effect of the country’s company (Rwanda Energy Group) financially l imited energy resources and resulting ­ independent play out and the new PPP law high costs of service delivery. 196 FUTURE DRIVERS OF GROWTH IN RWANDA • Define and put in place a trajectory resources necessary to set the regulatory toward the permanent application of tar- framework and conditions for wholesale iffs that would allow for the recovery of data services regarding nondiscrimina- efficient operating costs from all electricity tory access at regulated prices that mimic consumers who can afford them, accom- the results of a competitive market and to panied by a social safety net to protect grant spectrum rights in an efficient ­ way. low-income users, and adopt arrange- With respect to mobile telephony and short ments to cover the funding gap between message services, revised mobile termina- sector costs and tariff revenues while the tion fees have encouraged price competi- completed. trajectory is ­ tion and created a more level playing field among ­ o perators. The full implementa- Boost Access to and Reduce the Cost of tion of lower termination rates by January Broadband Connectivity. Rwanda’s tele- 2019 also should improve affordability communications market is one of the fastest- and ­ a ccess. Monitoring implementation growing wireless markets in Southern and and its impact will showcase the power Eastern ­ A frica. Mobile cellular telephone of pro-competition rules to achieve sec- subscriptions are the rule in Rwanda, with goals. The European Union’s state aid tor ­ hardly any landline ­ subscriptions. In 2016, control framework for designing govern- close to 70 percent of inhabitants had a ment support for broadband deployment mobile cellular subscription; the mobile cel- provides a useful framework for effective lular penetration rate has increased rapidly PPPs in telecommunications that would in Rwanda in recent ­ years.12 allow the government to refocus its direct The 4G LTE network was launched in participation in the sector toward setting 2014 with the target of covering 95 per- an enabling framework for full private cent of the population by ­ 2 018. Rwanda solutions (European Commission 2 ­ 003). has adopted a unique wholesale model RURA and the Bank of Rwanda should set for the nationwide deployment of 4G LTE rules for mobile financial services, particu- s ervices. In 2013, the government entered ­ larly regarding essential services by mobile a 25-year joint venture arrangement with providers, is key to supporting the develop- Korea Telecom to design and deploy a 4G ment of these ­ services. LTE network using the fiber backbone that government had ­ built. Yet most of the coun- Strengthen Market Institutions try’s connectivity is still through 2G and Accelerated private sector development 3G ­ connectivity. The household penetration will require robust market ­ i nstitutions. rate of fixed high-speed Internet subscrib- The evidence is compelling that function- ers is less than 1 percent, well below the ing markets require well-defined rules regional average of 6 ­ percent. The high cost of the game, enforced transparently and of a broadband connection, combined with ­ predictably. Rwanda has already made sig- low computer ownership and an erratic elec- nificant advances in building the basic struc- tricity supply, puts the service beyond the tures of its market economy, as reflected reach of most private ­ users. in its strong standing on the 2019 Doing The development of the telecommunica- Business Indicators (in which Rwanda tions sector calls for effective competition ranked 29th in the world overall, above i n key tele com mu n ic at ions m ­ a rket s. all other low-income countries and second Publication of a detailed regulatory frame- only to Mauritius on the ­ continent). It has work for managing the single wholesaler made good progress in strengthening its of high-speed wireless broadband and pro-competition regulations, but consid- its rights on radio frequencies is a prior- erable scope remains for further improve- a rea. The Rwanda Utilities Regulatory ity ­ ment, as noted earlier and discussed in more Authority (RURA) should have all of the ­ . Significant progress has detail in chapter 6 C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 197 also been made in improving the regula- government support for specific industries or tory regime for property ­rights. The proper effective. sectors is ­ enforcement of expropriation procedures, Incentives need to be linked more clearly further improvements in the efficiency of to the productivity and export performance the land management system, and provi- of beneficiary fi­ rms. Direct support should sion of more resources to implement the be focused on export promotion rather 2008 law on intellectual property would than on import substitution, given the abso- further strengthen property r ­ ights. There lute necessity of expanding nontraditional is also scope to further improve efficiency exports and the checkered record of import of the commercial dispute resolution pro- substitution in other low- and middle- cess, including contract enforcement—an income ­ countries. The government should important aspect of a functioning market encourage those industries in which Rwanda economy—on which Rwanda ranked 78th has a clear comparative a ­ dvantage. It should globally (among 190 economies) on the set clear policy objectives and performance Doing Business Indicators (World Bank targets for beneficiary firms from the begin- 2019). ning, ensure that policies are closely coor- dinated across government entities, and include a rigorous system to monitor prog- Improve Effectiveness of Tax and ress and enforce sanctions and incentives so Industrial Policy Incentives through as to reward success and punish failure—a Better Targeting model that the Republic of Korea followed Rwanda has already established some of the closely in its early years of development key preconditions for an effective industrial (box ­4.6). policy, including strong macroeconomic What is lacking are tax expenditures management, institutional coherence, rea- targeted at “efficiency-seeking” inves- sonable implementation capacity, and good tors looking for low-cost destinations sector. In addition dialogue with the private ­ for labor-intensive export p ­ roduction. to these, closer monitoring of incentives for Although efficiency-seeking FDI firms are results is needed to ensure that any direct highly demanding on a country’s business BOX 4.6  Targeted interventions in the Republic of Korea Korea, a notable example of a successful develop- enter new business ­ l ines. Various kinds of prefer- mental state, emphasized targeted i ­nterventions. ences were linked to export performance; firms In 1960, Korea was a poor country with GDP per that did not perform were encouraged to exit the capita of just US$160 (US$1,103 in 2015 dollars) ­market. and a large but noncompetitive private s ­ ector. After Korea also pursued financial policies that were some initial missteps, Chunghee Park’s military highly dedicated to exports and allocated credit in government of 1961–79 systematically provided a ruthless fashion to achieve export ­ results. Export government privileges to private firms that dem- promotion incentives tended to be general for success- onstrated success, first in import substitution and ful exporters, while incentives for targeted industries later in export promotion or targeted industries were sector specific and sometimes product or proj- (for example, heavy or chemical industries of mili- ect specific (Kim and Vogel 2011, especially ­ ch. 2–4, tary r­ elevance). These privileges included directed 7–9; Lim ­ 2016). In Japan as well as other countries, credits from state-owned banks, government guar- the effects of interventions on exports where larger antees, tax incentives, customs exemptions, pref- when domestic sectors where more competitive (less erential access to foreign exchange, and licenses to concentrated) (World Bank ­ 2015). 198 FUTURE DRIVERS OF GROWTH IN RWANDA environment and incentives alone are rarely export guarantee ­ facilities. One example of enough to “seal the deal,” global evidence a performance-based approach stipulated in suggests that these investors are highly the export strategy is that the government responsive to special incentives (Andersen, supports firms with the potential to repeat Kett, and von Uexkull 2017; IMF, UN, and past ­s uccesses. In this context, the focus World Bank 2015; James ­ 2009). on the so-called “64 million dollar export- Reforms that enhance the predictability ers” (64 MDEs) is w ­ elcome.13 The “Made of incentives and reduce up-front costs to in Rwanda” policy identifies many priority obtain information and gain approval for sectors (box ­ 4.7). Mining is an example of incentives could help to reduce the market- a sector that would benefit from a focused distorting effects. Ideally, these reforms strategy for its development (box 4 ­ .8). would go hand in hand with (1) implementa- Such incentives should be time bound tion of a framework to evaluate the effects and revocable to encourage sustainable on markets and competition of investment gains and avoid ­ d ependency. Ten-year incentives and any other government sup- incentives could, for example, be phased port for specific firms and (2) effective out during the last five ­ years. For example, implementation of competition rules to deter an enterprise that engages in irresponsible anticompetitive ­practices. investment or borrowing or in mismanage- Incentives should also be linked with ment should have its privileges ­ revoked. In export p­ erformance. Already, the Export Brazil, for example, if a company is found Growth Facility provides matching grants to have infringed competition law (partic- for new market entry; a subsidy to nar- ipated in a cartel or abused its dominant row the gap between the interest rate on position), it can be withdrawn from the list export financing and prime access rates, of companies eligible for government subsi- through an investment catalyst fund; and dies or tax ­exemptions. BOX 4.7  “Made in Rwanda” policy The “Made in Rwanda” policy, launched in 2017, •• Give local preference in public procurement. aims to improve competitiveness, enhance demand •• Upgrade the quality of infrastructure. for Rwandan value-added products, narrow the •• Boost training, certification, and inspection of trade balance, and generate ­ jobs. It brings together standards. existing government interventions under a clear pol- •• Facilitate small and medium enterprises with high icy framework and addresses supply-side bottlenecks growth potential to access land in special eco- via targeted interventions aimed at improving qual- nomic zones. ity, boosting cost competitiveness, deepening domes- •• Establish a local content unit in the Rwanda tic supply chains, and developing action plans for Development Board. specific high-potential value ­ chains. •• Support industrial research and development. The policy has five main pillars: (1) changing •• Boost access to finance through partnership with mind-sets, (2) improving quality, (3) reducing the commercial banks and through creation of the cost of production, (4) promoting backward links, Industrial Development ­ Corporation. and (5) undertaking sector-specific s ­ trategies. •• Improve access to industrial inputs through sup- The following interventions are key: port for regional value chains, simplified pro- cedures for value added tax, and import duty •• Conduct a national communication campaign exemption for and monitoring of the supply of about the policy. materials. critical raw ­ C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 199 BOX 4.8  Developing Rwanda’s mining sector Minerals, represented mostly by tin, tantalum, incentives to conduct and share prospecting sur- and tungsten, traditionally have been an impor- veys through competitive auctioning of potential tant export commodity for Rwanda and, in recent resource “blocks,” perform a detailed value chain years, have accounted for 15–20 percent of total analysis of Rwanda’s minerals to identify the par- exports of goods and s ­ ervices. Despite considerable ticular stage of value addition that is economically potential, the volume of traditional minerals has viable for Rwanda, further adapt the legal and increased little recently, with the recent resurgence regulatory framework to the geology and mineral in mining revenues in the fourth quarter of 2016 potential of Rwanda, and strengthen the capacity and first quarter of 2017 driven mainly by increased of the Rwanda Mining Board to promote, sup- exports of other high-value m ­ inerals. The general port, and regulate the mining sector ­ properly. upswing in mineral prices has raised expectations of •• Fiscal regime and management of mineral ­ wealth. the sector’s prospects (New Times ­ 2017). Developing an optimal fiscal regime so that the Attracting investments in the sector and enabling government can capture an efficient and fair share the government to collect a fair share of the wealth of mineral wealth is a policy p ­ riority. Tax admin- from mineral exports to reinvest in human capital, istration should be the focus of reforms, espe- resilience, and sustainability are important policy cially in the context of greater formalization in objectives for R­ wanda. Given the current structure the ­sector. So far, profit tax performance has been of the sector in Rwanda, these efforts must take into weak, in part due to transfer pricing and defi- account artisanal miners, larger-scale operators, cient export pricing that often does not reflect the and the national interest: value of other minerals accompanying the main mineral. To address this, the government needs to ­ •• Artisanal mining ­ (ASM). Artisanal mining develop the capacity of Rwanda Revenue Author- accounts for 80 percent of total minerals produc- ity to administer taxation in the ­ sector. Rwanda. As much as 30 percent additional tion in ­ value can be obtained from Rwanda’s mining sites The government is also considering playing if mechanized processing is implemented instead a direct role in the mineral sector by having an of manual methods (Heizman and Libetrau ­ 2017). equity participation in production and processing Production at some sites is reaching limiting depths companies. This is a valid policy option and would ­ of “easy to mine” resources using artisanal meth- strengthen government’s capacity to collect taxes ods, which highlights the importance of having on the ­s ector. Minority equity participation can better tools and methods to maintain and increase be used to fund geological ­ studies. Larger partici- production. In addition, miners are not always ­ pation to gain a greater share of mineral revenues able to capture a fair share of export receipts should be exercised with caution to avoid under- because of their weak bargaining power relative mining private investments and knowledge ­ transfer. to licensed traders, which discourages ASM oper- Stabilization is another important function that ators from investing in new tools and ­ learning. is needed to achieve improved fiscal management of Creating incentives that promote investments in mineral w ­ ealth. As mining exports grow and the new tools and skills, organizing fragmented opera- government captures its fair share, mechanisms that tions into cooperative operations, creating strong can insulate the budget from price volatility, such as links between ASMs and processors, and provid- introducing countercyclical principles in the use of ing basic skills training and extension services to mineral revenues, will be ­important. ASMs are all important ways the government can Depending on the scale of new discoveries, support the artisanal mining ­ sector. Rwanda should make intertemporal choices with •• Larger-scale private investments in production, respect to the consumption of current and future processing, and value ­ addition. Developing an generations. The government has established a sov- ­ industrial mining sector will require foreign capital ereign wealth fund, which can serve the purpose of and knowledge of mineral processing with accu- smoothing intergenerational consumption of min- rately targeted i­ncentives. To attract investments eral resources and stabilizing countercyclical spend- into Rwanda, the government should provide ing of mineral ­revenues. 200 FUTURE DRIVERS OF GROWTH IN RWANDA Define the Role of SOEs in the Economy sector. Enterprise development the private ­ and Strengthen Corporate Governance in Rwanda should continue to contribute to alleviating market failures and encouraging SOEs played an important role in the early private sector development, especially given stages of Rwanda’s economic ­ development. that investment capacity among Rwanda’s Investments in them were a response to the private sector remains ­ low. SOEs are also massive investments needed in the after- seen as playing a useful role in maintaining math of the genocide against the Tutsi, social and political stability. when a viable domestic private sector was absent. Observers have debated the rela- ­ tive merits and development attributes of Strengthen Corporate Governance of SOEs Rwanda’s ­ S OEs. Some have asserted that Effective corporate governance achieves a they contribute to economic development necessary balance between autonomy for and social stability (Behuria 2012; Booth enterprise “agents” and oversight by enter- and Golooba-Mutebi 2012), while not- prise owners ­ (principals). Enterprise agents ing that, as in Indonesia and Malaysia, normally include enterprise management and rents have been “deployed in ways that a board of directors, the latter overseeing the assisted national development” (Booth and ­ former. It is essential to determine the appro- Golooba-Mutebi 2012; see also Behuria and priate authority at each level—management, Goodfellow ­ 2016). Others have concluded directors, and ­ owners. that SOEs cause a fiscal drain and use public The guidelines of the Organisation for capital inefficiently, while constraining pri- Economic Co-operation and Development vate sector development (Gokgur ­ 2012). (OECD) for the corporate governance of There was strong justification for state SOEs14 could usefully be adapted—with involvement in the economy, given the large any necessary modifications—for Rwanda’s reconstruction needs in the aftermath of the SOEs, which are already following some of genocide against the Tutsi, an incapacitated guidelines. Following international good the ­ private sector, and ready availability of for- practices, the government has already imple- eign ­assistance. This justification holds in mented most of the corporate governance the medium term, and SOEs have a crucial principles detailed ­below. First, government role to play in Rwanda’s future d­ evelopment. should consistently act as an informed ­ owner. The practical question is how they can do Proper ownership functions include so in the most effective way, so that public capital is preserved and put to good use while • Being represented at shareholder meetings the development of the private sector is not and voting owned shares; ­undermined. • Establishing merit-based and transpar- Ultimately, Rwanda’s First National ent processes for nominating the board of Strategy of Transformation and Vision directors in full- or majority-owned SOEs, 2050 establish a pathway for the private actively participating in the nomination of sector, over time, to play a leading role in all SOE boards, and contributing to board Rwanda’s growth; to achieve this goal, diversity; resources have to be allocated through • Setting and monitoring the implementa- efficient ­m arket ­ m echanisms. As Rwanda tion of broad SOE objectives and man- approaches middle-income levels, the com- dates (for example, financial targets, plexities of its market economy will grow capital structure, risk tolerance); almost ­ exponentially. Strong government • Setting up reporting systems that allow support will still be needed, but its role will the owner to monitor, audit, and assess increasingly be to facilitate private invest- SOE performance regularly and to oversee ment and build strong market institutions and monitor SOE compliance with corpo- that ensure predictable rules of the game for rate governance standards; C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 201 • Developing a disclosure policy for SOEs full International Financial Reporting that identifies what information should Standards (IFRS) for all domestic com- be publicly disclosed, appropriate chan- panies whose securities trade in a public nels for disclosure, and measures to ensure ­ market. In addition, “companies with pub- information quality; lic accountability (banks or other financial • Maintaining appropriate dialogue with I FRS.”15 institutions) must use full ­ external auditors and government control In addition to financial reporting, SOEs organs; and should publicly disclose other important • Establishing a clear remuneration policy ­ information. This information includes the for SOE boards that reflects the long-term following: interests of the SOE and can attract and motivate qualified professionals to serve • Governance, ownership, and voting struc- on SOE boards of d ­ irectors. tures of the SOE, including the content and implementation of any corporate gov- If a SOE is listed on the Rwanda Stock ernance code Exchange or otherwise includes nongov- • Remuneration of board members and key ernment investors among its owners, the executives relevant laws and regulations should be • Board member qualifications, selection pro- f ollowed. The participation of minority ­ cess, roles of other company boards, and shareholders in shareholder meetings should whether these other boards are considered be facilitated so that they can take part in as independent by the SOE board key decisions, such as elections to the board • Any material foreseeable risk factors and of ­directors. Transactions between the gov- risk mitigation measures ernment and SOEs or between SOEs should • Any financial assistance, including guar- take place on market ­ t erms. If an SOE is antees, received from the government and required to pursue a public policy objective, SOE commitments arising from public- information on this requirement should be private partnerships available to all shareholders (OECD 2015, • Any material transactions with the gov- ch. I ­ ­ V). The government’s ownership rights ernment or government-related entities and oversight should be exercised only peri- • Any material issues relating to SOE odically (for example, at shareholder meet- employees or other stakeholders (OECD ings) and through normal mechanisms (for 2015, ­ch. ­V I) example, voting of s ­ hares). Autonomy for SOEs needs to be maintained while ensuring Rwandan SOEs already follow many of that SOEs have the necessary competencies these corporate governance ­ practices. For and ­accountability. example, All listed SOEs should regularly issue audited financial statements to the ­ public. • SOEs typically have a seven-person board First, the government needs reliable finan- of directors; cial information to guide economic ­ policy, • Separate individuals are appointed to serve regulation, and investment ­ d ecisions. as the SOE’s chief executive officer and as S econd, leading international private the board chair; i nvestors who might provide ­ ­ a dditional • The board of directors often maintains equity or debt financing require such specialized board committees, such i nformation. Third, Rwandan citizens ­ as for audit and human resources or are all SOE stakeholders, and regular and remuneration; reliable financial reporting would serve • The board of directors is generally obliged to maintain the public’s t ­rust. Per cur- to meet at least four times a year and may rent practice, the government requires have a predetermined meeting calendar; 202 FUTURE DRIVERS OF GROWTH IN RWANDA • The chief executive officer and board chair private ­i nvestors. This policy should define, often sign performance contracts; among others, the rationale for ownership, • Many SOEs have an internal audit depart- the government’s role in the corporate gov- ment that reports to the board or its audit ernance of SOEs, how the government will committee; implement ownership, and the respective • An independent external auditor audits roles and responsibilities of any ministries SOE annual financial statements; or agencies involved in ­ i mplementation. • Some SOEs require periodic rotation of For SOEs with noncommercial objectives, external auditors; the economic or social rationale for each • Each SOE is expected to have a multiyear “public service mandate” should be stated strategic plan and a risk management pol- ­ c learly. In general, SOEs should be reim- icy (MINECOFIN 2017); and bursed at least for the costs of any public • At least some SOEs have budget for cor- service ­ m andate. This ownership policy porate governance training for new should be subject to appropriate procedures ­directors.16 of political accountability, disclosed to the public, and reviewed at regular intervals Clarify the Government’s Ownership Policy (OECD 2015, ­ ch. ­I ). Several OECD juris- The Ministry of Finance and Economic dictions have recently undertaken state P l a n n i n g ’s G o v e r n m e n t P o r t f o l i o ownership policy initiatives along these Management Unit is tasked with ensuring lines (table ­4.4). that the government acts as an informed An accompanying strategy for pub- ­ owner. In its annual Fiscal Risk Review of lic share offerings in successful SOEs Government Investment, the Government would offer multiple b ­ enefits. Even just a Portfolio Management Unit highlights cor- minority public share offering facilitates porate governance issues for the board’s SOE access to f ­ollow-on capital market attention, tracks the disposition of issues ­ f inancing. Oversight by public shareown- previously identified, analyzes financial per- ers would strengthen the corporate gover- formance and position for each state-owned nance of listed S ­ OEs. A regular program of enterprise, and summarizes key financial SOE public share sales and listings would risks for the government (MINECOFIN energize the Rwanda Stock Exchange and ­ 2017). Overall, the Financial Risk Review raise Rwanda’s profile among international represents an excellent foundation for fur- ­ i nvestors. Other economies have success- ther development of an SOE ownership pol- fully followed similar approaches with their icy, establishing greater clarity for potential state-owned ­enterprises.17 TABLE 4.4  State-owned enterprise policy initiatives in the OECD since 2006 Country Policy initiative Finland Government Resolution on State Ownership Policy (May 2007), establishing key principles and operating policies for the state’s ownership function Norway Government Ownership Policy for SOEs, published since 2007 Poland Draft legislation identifying state companies of “key importance” to the Treasury Portugal Council of Ministers 2007 best practices for public companies to increase transparency and improve corporate governance Spain 2009 Royal Decree on General Rules on the Assets of the Central Government, stating an ownership policy for SOEs, establishing guidelines for commercial and noncommercial objectives, outlining the role of shareholder meetings, and establishing transparency and other good practices for SOEs Switzerland Federal Council Corporate Governance Reports since 2006, establishing criteria for outsourcing tasks to SOEs and SOE governance guidelines Source: OECD ­2011. enterprise. Note: OECD = Organisation for Economic Co-operation and Development; SOE = state-owned ­ C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 203 On the basis of these analyses, the govern- demand a competitive societal climate that ment might usefully categorize Rwanda’s real values openness to new ideas and stimu- economy by four sector ­ types: lates risk taking and technology ­ upgrading. All three were crucial to the rapid growth of 1. Sectors in which SOEs will retain a Japan, Korea, and Singapore—examples that ­monopoly Rwanda holds up for its own growth and 2. Sectors in which SOEs will remain pres- innovation ­ambitions. ent and compete with private ­ firms Developing a strong NIS, therefore, is 3. Sectors from which SOEs will withdraw the fourth reform priority for Rwanda in when efforts to build up private sec- its efforts to build a competitive enterprise tor capacity prove successful (discussed sector. NIS development involves develop- ­ below) ­ ing an ecosystem that ensures strong uptake 4. Sectors from which SOEs will withdraw of learning and innovation from both the immediately (for example, through a demand and the supply ­ sides. well-planned and well-supported priva- The first requirement of a functioning tion program) because the private sector NIS is a critical mass of dynamic and com- is already sufficiently capable and there is petitive firms that provide the demand side no compelling social rationale for S­ OEs for ­innovation. These enterprises also must have the absorptive capacity to adopt and For type 2–4 sectors in which private firms carry out new research and hold longer- are competitive with SOEs, state support for term strategic views on firm and product enterprises could be based on common sec- upgrading and human resources necessary tor goals, such as demonstrated market-based innovation. The lack of such abilities in for ­ achievements in export promotion; any such many low- and middle-income economies privileges should be time bound and r ­ evocable. reflects limited competition in product mar- For type 2–4 sectors in which private firms kets and the low capabilities of firm manag- cannot generally compete with SOEs, the gov- ers (Bloom and Van Reenen 2007; Maloney ernment could give private firms some oppor- and Sarrias ­ 2014). Addressing management tunities to catch ­ up. This could involve, for capability gaps was a crucial element of the example, time-bound and revocable privileges, “East Asian Miracle” (Cirera and Maloney for example, lower performance thresholds for 2017). As discussed earlier, firm capabilities ­ certain incentives, some private sector set-asides in Rwanda require significant ­ improvement. for government procurement contracts, or par- Management practices also tend to be bet- SMEs.18 tial credit guarantees for private ­ ter in older firms (those that have been in Public-private boundaries can shift as the business for more than 20 years; Cirera and private sector gains s­ trength. The priority now Maloney 2017), which is a challenge for is to lay strong economic foundations and to Rwanda given that 90 percent of its firms design, implement, and enforce a sector-­ based are less than a decade ­ old. Firms that par- enterprise strategy that considers short-term ticipate in foreign trade or are part of global competitiveness, long-term economic develop- value chains also tend to have better man- ment goals, and social stability ­ needs. agement practices (Bloom and Van Reenen 2010; Maloney and Sarrias 2017), earn higher returns on innovation, and face lower Develop an Effective NIS research and development (R&D) costs An innovation-driven economy will demand (Guadalupe, Kuzmina, and Thomas 2 ­ 012). an unambiguous commitment to cultivat- Rwandan enterprises are nascent in this ing and upgrading a dynamic private sector area, with about 8 percent of formal manu- and far-reaching reforms in the education facturing firms exporting their products and and training systems for generating higher- very few embedded in global value chain order human ­capital. More generally, it will production networks (chapter ­ 2).19 204 FUTURE DRIVERS OF GROWTH IN RWANDA The supply side of the NIS (discussed in The first imperative is to develop a strong, chapter 1) also has strong requirements, espe- competitive, and innovation-driven enter- cially with regard to a skilled workforce and the prise ­s ector. Otherwise, all efforts on the quality and relevance of research and advanced innovation side will be supply-push and, in institutes. training in universities and research ­ the worst case, costly and valueless high-tech The capabilities escalator provides a useful ­ ventures. But the presence of such firms is not framework for conceptualizing and building ­ enough. To show greater urgency in press- an effective NIS (figure ­ 4.23). A good start- ing for new technologies, firms must also ing point for economies with incipient inno- see a strong business case for innovation and vation systems (stage 1, which is Rwanda’s believe that they can manage any potential current situation) are policies and invest- challenges and ­ risks. This can be encouraged ments to strengthen the capabilities of local by providing (1) technological extension ser- firms to adopt new technologies, improve vices to build the managerial and organiza- their access to technologies, develop human tional capabilities of firms and (2) incentives capital (especially in science, technology, for technology absorption and ­ diffusion. engineering, and mathematics), and put in place quality ­ i nfrastructure. The emphasis Provide Technological Extension Services to on building firms’ technological capabili- Build Managerial and Organizational ties, improving the quality and relevance of Capabilities research in the country, and developing effec- Innovation cannot develop without adequate tive research collaborations between industry sophistication among firms and farmers, and academia comes next (stage 2), followed beginning with the ­ basics. Throughout the by development of cutting-edge R&D and high-income world and in dynamic low- and technological programs when the country middle-income economies, governments have reaches a mature state of innovation (stage 3 ­ ). employed extension services to facilitate tech- nological transfers to enterprises (box ­ 4.9). FIGURE 4.23  The capabilities escalator: A framework to support Integrated support programs—providing innovation and entrepreneurial activity support for companies to improve their man- agement and other internal capabilities—are especially well suited in this ­ regard. The Rwandan government already has • Long-term R&D and STAGE 3 technological programs taken important steps to raise the innovation Mature NIS of firms and the bureaucracy—­ significant • Minimize innovation gap leaders and laggards among them has been the establishment • Collaborative innovation projects of the National Industrial Research and Development Agency (NIRDA), the National • Building technological capabilities Commission of Science and Technology Level of development STAGE 2 • Incentivize R&D projects Maturing NIS (NCST), and a few other o ­ rganizations.20 The • Link industry and academia • Improving quality of research, task at hand now is to consolidate the activities innovation, and export infrastructure of these institutions to ensure effective coordi- nation of the ­ N IS. This effort is important to • Building managerial and organizational identify where to invest scarce resources and STAGE 1 capabilities Incipient NIS • Start collaborative projects who the key stakeholders are and to define • Need to develop STEM skills and engineering clear roles for how these institutions could col- • Need for basic infrastructure—NQI and incubation laborate to strengthen the s ­ ystem. NIRDA can • Elimination of barriers to physical, human, and knowledge capital play a key role in facilitating technology adop- tion by firms, improving their managerial and organizational capabilities, facilitating their Source: Adapted from Cirera and Maloney 2017. Note: NIS = National Innovation System; NQI = national quality infrastructure; R&D = research access to innovation and technology adop- and development; STEM = science, technology, engineering, and mathematics. tion funds, and, in close coordination with the C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 205 BOX 4.9  The importance of firm capabilities for productivity: Examples from Brazil, India, and Singapore Brazil. PEIEX (Industrial Extension Project for correlated with export performance (Cruz, Bussolo, Exporting) is a program from the Brazilian Export and Iacovone ­2018). Promotion Agency (Apex-Brasil) that focuses on India. An experiment in India involved evaluat- enhancing the competitiveness of Brazilian small and ing the impact of free consulting on the manage- medium enterprises by providing coaching and con- ment practices of large Indian textile ­ fi rms. Firms sulting on management and production best ­ practices. that adopted the recommended management prac- The initiative was set up as supplementary assistance tices raised productivity 17 percent in the first year for firms interested in taking part in export promotion by improving quality and efficiency, reduced inven- services already provided by the agency (for example, tory, and opened new production plants within participation in trade fairs and business rounds or three years (Bloom et ­ 2 013). About 10 years a l. ­ meetings with foreign b ­ uyers). Once these firms con- later, the gap in practices between those firms that firm their interest in participating in the program, they received support and the others was still large and receive a standard competitive strength assessment al. ­ significant (Bloom et ­ 2018). covering different areas of the enterprise (for example, Singapore. The Singapore Standards, Productiv- strategic organization, human capital, finance and ity, and Innovation Board was established in the costs, sales and marketing, international trade, prod- 1980s with support from the Japanese Productivity uct design, production, and innovation), with a final Center. It plays a key role in the extensive system ­ report identifying their strengths and w ­ eaknesses. of small and medium enterprise support ­ services. This assessment is followed by a plan with suggestions Similarly, the Economic Development Board of focusing on improving firms’ ­ competitiveness. The Singapore introduced the Local Industry Upgrad- program offers consulting services in partnership with ing Program to improve the exchange of knowledge universities and institutes of technology in fields such between local firms and foreign ­ multinationals. In as marketing, human resources management, finance, addition to providing extended loans to buy machin- product design, and ­ trade. An evaluation found that ery and equipment, the Local Enterprise Technical the PEIEX has been particularly effective in promot- Assistance Scheme run by the Economic Develop- ing the reorganization of small and medium firms ment Board provided grants to hire external experts and that changes in firms’ organization are positively to upgrade operations and ­ management. NCST, facilitating the interaction between the management capabilities have proved effec- demand for and supply of ­ knowledge.21 tive and s ­ ustained. In Togo, personal initia- A good starting point for shaping spe- tive training programs to develop behavior cific policy guidance would be to con- associated with a pro­ active, entrepreneurial duct a survey to measure the management mind-set proved to be effective in boosting capabilities and practices of Rwandan women-owned small businesses (Campos et ­ f irms. Several countries have implemented ­al. ­2018). the World Management Survey, includ- On the basis of this evidence, the follow- ing a few countries in Sub-Saharan Africa ing interventions could be adopted in Rwanda: (Ethiopia, Ghana, Kenya, Mozambique, (1) provide training and coaching services Nigeria, Tanzania, and Z ­ ambia).22 Findings to diffuse good business managerial practices; from management surveys can shape the (2) complement business support programs programs supporting businesses by incor- with personal initiative training; (3) expand porating components to provide manage- and improve the quality of graduate-level rial coaching and assistance, such as in the management courses; (4) provide managerial example of PEIEX in Brazil (Cruz, Bussolo, training to cooperatives that support farm- and Iacovone ­ 2018). Interventions to build ers; and (5) increase interaction between 206 FUTURE DRIVERS OF GROWTH IN RWANDA multinational firms and local ­ s uppliers. • They build capabilities that help firms to Whereas the first two examples find strong compete in those markets and meet that support in previous experiments, such as in ­demand. India and Togo, the other suggestions adapt • They usually build firm management these interventions to the constraints and capabilities and look to engage in process opportunities faced by ­ Rwanda. It would be innovation and new product ­ development. useful to test these interventions in Rwanda • They try to expose firms to markets with small-scale experiments before invest- through targeted efforts to connect with ing large amounts of r ­ esources. Such experi- potential buyers (for example, trade ments would enable policy makers to learn missions and fairs or “meet the buyer” more about how to adapt them to Rwanda. ­events). Similar experiments are in place in other • They encourage cooperation and group countries, including Argentina, Colombia, learning between firms and provide a and ­Mozambique. structure to reach out and work with Strong market demand for their products innovation ­providers. can be an important motivator for firms • They use a variety of innovation policy to try and ­ upgrade. Interventions that seek tools, including advisory services, vouch- to build both firms’ capabilities and their ers, matching grants, and services like access to markets—as opposed to address- national quality infrastructure and intel- ing these needs separately—tend to be espe- lectual property in an integrated ­ fashion. cially ­effective. There are various models for Rwanda to consider, ranging from supplier These types of programs can become good development programs to export competi- tools for embedding capability ­ upgrading. By tiveness programs, depending on the specific combining support for SMEs to upgrade with upgrading needs of the firms and charac- a tangible market opportunity, they can be teristics of their main buyers (Chavhan, much more effective than supply- or demand- Mahajan, and Sarang 2 ­ 012). The PEIEX only initiatives, although they are more com- is an export competitiveness program pro- plex to ­deliver. viding support for building managerial Rwanda already has many initiatives to capabilities. Supplier development programs ­ support incubator or accelerator programs, 23 are usually provided by firms (buyers) with and further scaling up may be better left to government ­ support. Such programs work the private sector in many c ­ ases. Rather than by providing funding or getting involved building new facilities, the government can in SME ­ upgrading. Examples of programs coordinate these actions and, perhaps more include building local value chains, partic- important, support the managerial capabili- ularly in agribusiness (such as in Haiti), or ties of these p­ rograms. 24 It also can seek to building international value chains, in some strengthen the organizational capabilities of cases tied to investment of multinational the public enterprises i­nvolved. For example, corporations, to strengthen the local sup- Rwanda has recently updated the laws on plier base (such as electronics in Vietnam) intellectual property rights and established a or to strengthen exports (such as special standards ­board. economic zones in ­ M exico). In this case, It will be critical for Rwanda to advance the government could facilitate the match- on the coordination, monitoring, and evalu- ing between local providers and exporting ation across policy instruments that sup- ­fi rms. port ­i nnovation. Any new initiative should Successful programs of this type have sev- be started at a relatively small scale and be eral aspects in common: evaluated before being scaled ­ up. A good example is in the field of social policy, where • They are always focused on understanding Mexico’s successful conditional cash trans- and serving a known market ­demand. fer programs were designed on the basis of C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 207 rigorous randomized controlled trials before that are new to the country) need to over- up. being scaled ­ come the wariness of traditional financiers and potential ­ buyers. Pre-incubator, incuba- Incentivize Technology Adoption and tor, and accelerator programs can be impor- Diffusion tant resources for such s ­ tart-ups. Many of Firms’ capabilities alone are not ­ e nough. their interventions are focused on innovative Low innovation can also reflect lack of incen- companies or sectors and, in some cases, are tives for firms and other barriers to technol- linked to public sector research organizations ogy adoption, such as insufficient early-stage that support the commercialization of knowl- financing, weak intellectual property rights edge, as in Colombia (box 4 ­ .10). Incubators protection, and difficult risk ­ management. can combine mentoring, business plan For example, firms may worry that competi- preparation, and links to potential finance tors will copy their innovations, eliminat- in an integrated fashion, which encourages ing much of the potential profit of investing networking and mutual ­ learning. Shorter- ­ nowledge. Financial incentives from the in k term accelerator programs may help in the government, in the form of direct subsidies, adolescence phase where technical issues tax write-offs, matching grants, or patents, are consolidated, but strategic guidance is can raise the private return to equal the social demand. Ongoing initiatives in this area in ­ return, particularly on the grounds of mar- in Rwanda include the Incubator Center of ket failure arising from the appropriability the Integrated Polytechnic Regional College ­externality. and Milicom’s “Think” initiative as well as Business incubator and accelerator pro- the recently approved Rwanda Innovation grams for the provision of infrastructure and Fund, but government also has a role to play, advisory services are increasingly popular for example, in mapping these initiatives and instruments for innovation capacity and have coordinating among ­ them. already been initiated in ­ Rwanda. Many start- R w a n d a ’s a m b i t i o u s e f f o r t s i n ups have potentially marketable ideas, but developing the Kigali Innovation City are ­ they lack the experience to start a company c rucial in this regard (box 4 ­ ­ .11). It now up. In addition, new businesses or to scale it ­ needs to strengthen the entrepreneurship (particularly those that are built around ideas ecosystem around the innovation city to BOX 4.10  The experience of Apps.co Apps.co is a program for start-ups managed by In the incubator phase, ­ Apps.co provides guid- the Ministry of Information Technologies and ance from mentors and consultants for teams of Communications of ­ Colombia. It comprises activi- entrepreneurs on developing sustainable business ties focusing on a range of start-up development models with solutions that segment needs in these phases (pre-incubator, incubator, and accelerator), ­industries. including boot camps to increase awareness of Apps.co provides coach- In the accelerator phase, ­ business opportunities and activities to accelerate ing for selected companies in the information and ­start-ups. communication technology industry (digital content In the pre-incubator phase, A ­ pps.co fosters the and web and mobile applications) that require advice creation of businesses in digital content, more spe- regarding how to expand and consolidate their busi- cifically creative digital industries, by providing ness to reach break-even or investment s ­tatus. At services and methodological training for teams of this stage, companies must have a digital product, entrepreneurs to develop a validation process for a a validated business model, and demonstrable trac- product and a sustainable business ­model. tion in a defined segment of real ­ customers. 208 FUTURE DRIVERS OF GROWTH IN RWANDA BOX 4.11  Kigali Innovation City The Kigali Innovation City (KIC) is Rwanda’s flagship Rwanda’s Centre of Excellence for Biomedical initiative for innovation-led economic ­ transformation. Engineering and e-Health), which are operating in This initiative aims to bring together a highly con- Kigali, will commence construction of their KIC nected ecosystem of growth-stage local and interna- campuses in early 2 ­ 018. The International Centre tional technology companies, “special innovation for Theoretical Physics will continue to reside out- centers” nurturing a network of early-stage compa- side the KIC site, but it will remain a key member nies, a community of world-class learning institutions community. of the ­ providing an annual supply of high-aptitude pan-­ A key objective of the KIC initiative is to accel- African talent, an industry skills academy compris- erate the growth of human ­ c apital. The higher ing the global technology academies, a professional learning institutions will provide an annual supply services cluster, and the Rwandan Innovation Fund of “pre-intern” high-aptitude pan-African talent (target fund size of US$250 million) for growth of comprising 150 engineers, 150 mathematicians, companies within Rwanda’s technology ­ sector. 150 scientists, and 150 business ­ entrepreneurs. To at tract foreign technolog y companies, Employers will take advantage of the co-domiciled Rwanda offers a package of investment ­ i ncentives. industry skills academy to turn their talented These incentives include preferential corporate employees into world-class professionals able to income tax rates, income tax holidays, exemptions innovate and produce highly competitive products from specific customs taxes, accelerated depre- and to provide the professional services required to ciation, and a value added tax ­ refund. The main deploy the products in many economies of ­ A frica. attraction is the pan-African talent pool accessible This talent pool will contribute significantly to the to technology companies for “continuous intern- creation and commercialization of ­ start-ups. Other ship” until they eventually join as ­ e mployees. projected outcomes of the KIC initiative include Carnegie Mellon Universit y-A frica has been the creation of 250 new technology companies by operating in Rwanda for the last six years and in 2022, graduation from the Rwanda Innovation early 2018 will occupy its newly finished campus Fund portfolio of at least 10 corporations with a within the KIC s ­ ite. Other higher learning institu- combined market valuation of more than US$500 tions (African Institute for Mathematical Sciences, million, and an increase in export earnings of at African Leadership University, and University of least US$180 million by ­ 2025. ensure its commercial success; 25 to do Access to finance plays a key role in tech- so, it is important to attract talented and nology adoption and i ­nnovation. Rwanda experienced entrepreneurs from across the has several initiatives to provide science, country and abroad, in addition to the stu- technology, and innovation funds (table ­ 4.5), dents who come out of the various educa- including NIRDA, RDB, and the Business tional entities within the innovation ­ c ity. Development ­ F und. Finance for innova- The government should bring in a range tion should extend beyond just start-up of domestic and international players to finance; innovation finance also should be run start-up, incubator, and accelera- made available for established companies tor events; investment readiness services; that may need to be induced to undertake equity investment ­ m odels; and local and risky innovation projects for which there is international support and advisory net- support. It is impor- no traditional financial ­ works from the best ­ e cosystems. It should tant to improve coordination, monitoring, also regularly map the ecosystem, identify and evaluation across these institutions and areas of weakness, and coordinate actions ­ instruments. A key starting point would be for addressing these issues though policy to ensure the functionality of existing sup- ­i nterventions. port programs to determine whether their C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 209 TABLE 4.5  Use of science, technology, and innovation funding initiatives in Rwanda Fund Implementing agency Purpose Green Fund (National Climate and Environment Environment and Climate Change Fund and Fund environment and climate change Fund) Ministry of Natural Resources of Rwanda investment Rwanda Innovation Endowment Fund Ministry of Education Fund R&D for innovation and for priority economic and social areas Skills Development Facility Workforce Development Authority Expand the number of individuals with the relevant skills in critical sectors Rwanda Innovation Fund Rwanda Development Board Fund KIC-related science, technology, and innovation programs National Research and Innovation Fund National Commission of Science and Align research with national priorities Technology Business Development Fund Business Development Fund Provide SME support credit guarantees, credit lines, matching grants, quasi-equity, and advisory services Fund for African Private Sector Assistance African Development Bank Provide untied grants for technical assistance and capacity building ICT-SME Fund Ministry of Youth and ICT and Rwanda Provide public-private funding for job creation Development Board National Industrial Research and National Industrial Research and Fund R&D for industry needs Development Agency Development Agency Export Growth Facility Rwanda Development Bank Catalyze investment through matching grants for market entry costs and export guarantees Not applicable Rwanda Development Bank Provide student loans for graduate programs abroad Skills, Employability, and Entrepreneurship Workforce Development Authority and Fund skills, entrepreneurship development, Program (SEEP III) Fund African Development Bank and job creation program Student loans for graduate programs Rwanda Development Bank Support skills development Source: UNCTAD ­2017. technology. Note: R&D = research and development; KIC = Kigali Innovation City; SME = small and medium enterprise; ICT = information and communication ­ objectives are clear, whether they have a to make up the national quality infrastruc- sound business model, whether they are ture, which extends firms’ market access efficiently run, and whether they reach the beyond the national borders to include the market. Moreover, it would be impor- target ­ availability of and access to decent private tant to have a unified database to identify accreditors. Meeting global labs, testers, and ­ the main beneficiaries, the efficiency, and the standards is also important for export firms, effectiveness of these programs, in order to because many governments as well as con- impact. track their ­ sumers want products to meet certain qual- International standards can play an ity and safety ­ requirements. With all this important role in technology a ­ doption. New in mind, Rwanda established a standards technologies tend to follow global stan- board. The availability and affordability of dards, so firms may find it more challeng- key testing, assessment, and accreditation ing to adopt new technologies if they do not services are key, as is increasing awareness already meet global ­ s tandards. Examples about ­standards. of global standards include internationally Encouraging collaboration among firms recognized safety standards for household or between firms and knowledge providers is products, food testing and labeling, and innovation. another effective policy to foster ­ construction materials, among ­ others. These Grants, matching grants, and vouchers standards are important for Rwanda (given are common types of direct support used the role of agricultural ­exports). They help to increase business innovation a ­ ctivities. 210 FUTURE DRIVERS OF GROWTH IN RWANDA As part of this effort, Rwanda could con- knowledge organizations and providers and sider developing voucher ­ schemes. Applicants to develop ­ innovations. Most of the evidence typically receive vouchers when they meet regarding the impact of these instruments is preset eligibility requirements, which can be from OECD countries, including Austria, used for firm-specific innovation needs, and Ireland, Scotland, and Switzerland, among can take the vouchers to innovation provid- others, as highlighted by Becker (2015), ers on a preapproved ­ list. Voucher schemes García-Quevedo (2004), and Zúñiga-Vicente have also been used to stimulate innovation al. ­ et ­ (2014). Vouchers are often valuable in in service sectors, where formal R&D may inducing firms to innovate and so have been be less ­ common. Global evidence on voucher used to target sectors (for example, in ser- instruments suggests that the most promi- vices) that have not traditionally been seen nent impacts are follow-up projects for some as highly innovative (or at least are not R&D SMEs and a change in attitude toward collab- intensive). However, to be effective, a good ­ oration on innovation p ­ rojects. Collaborative supply of useful innovation expertise is matching grants are another instrument to needed for SMEs to tap into; if this exper- consider, where the voucher is conditional on tise is not available, then the innovation eco- collaboration (for example, with a research system should seek to raise capacity on the ­institute). supply ­side. Grants are a direct allocation of funding Finally, ensuring intellectual prop - from public agencies to firms to finance all or erty rights (including their protection and part of an innovation project or innovation-­ enforcement) will become more impor- related firm ­ upgrading. In matching grants, tant in the medium t ­erm. The recent law public agencies match a percentage contri- on intellectual property rights enacted in bution of the project made by the applicant 2009 is in line with good practices estab- to ensure that the applicant is committed to lished by the World Intellectual Property the a­ ctivity. The main objective is usually ­Organization. to entice firms to start collaborating with Annex 4A  Measuring Resource Misallocation Misallocation of resources is measured Yi = Ai Kiα Li1−α (4A.1) as the dispersion of the marginal revenue product of factors of production across where Y is value added of firm i, and K and firms within an industry, following the L are capital inputs, r­ espectively. A i is the framework proposed by Hsieh and Klenow producer’s productivity level ­ ( TFP). The ­ (2009). The key assumption underlying economy is endowed with labor and capital this approach is that, in the absence of stocks, indexed L = Si Li and K = Si K ­ i. any distortion, the marginal revenue prod- A firm maximizes its ­ profit: uct of factors should be equalized across firms within an ­ i ndustry. The counterfac-   pi = (1 + tyi)PiYi − (1 + tki)RKi − w (4A.2) tual productivity gains of eliminating the misallocation are calculated by equalizing function subject to its demand ­ total factor productivity of revenue (TFPR), σ which is a summary measure of distortions, P  Yi = Y  , (4A.3) across the existing set of producers in an  Pi   ­industry. More formally, each firm produces a dif- where R and w are the common rental cost ferentiated product according to the produc- of capital and the wage rate, r ­ espectively. tion ­function t ki denotes the firm-specific capital-labor C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 211 distortion, which increases the cost of capi- Notes labor. The capital-labor distor- tal relative to ­ 1. Rwanda’s ranking on the Logistic Performance tions may reflect government policy such as Index has improved remarkedly, moving from credit market imperfections or labor regula- 148 in 2007 to 62 in 2016, driven almost tions that favor particular ­fi rms. t yi denotes exclusively by public ­ investments. the output distortion, which may stem from 2. Labor Productivity = TFP * Human Capital government restrictions on size, output taxes, * [a / (1 − a)] * Capital-Labor ratio, where or high transport ­ costs. a = share of capital in ­ production. Profit maximization yields the firm’s out- 3. Rwanda Investor Perceptions Survey was put ­price: undertaken by Adam Smith International in late 2017 in conjunction with the World Pi ∝ (1 + τ ki )α , (4A.4) Bank Group (Adam Smith International ( Ai 1 − τ yi ) ­ 2017). The survey identified the perceptions of investors and the constraints of export- where the elasticity of price with respect to ing firms in Rwanda’s priority economic productivity is ­ −1. The distinction between ­sectors. quantity-based productivity Ai(TFPQi) and 4. Misallocation of resources is measured as the revenue-based productivity TFPR i is criti- dispersion of the marginal revenue product of cal for measuring misallocation, especially in factors of production across firms within each differentiated-product ­ industries. To illustrate, sector, following the framework proposed by TFPRi =Pi × Ai, where Pi is the price that a Hsieh and Klenow ­ (2009). The key assump- firm c­ harges. Firms with higher productivity tion in this framework is that, in the absence charge lower prices, resulting in ­ constant of any distortion, the marginal revenue prod- uct of factors should be equalized across ­ TFPRi = Pi Ai ∝ (1 + τ ki )α . (4A.5) firms within ­ i ndustries. The counterfactual productivity gain is calculated by equalizing (1 − τ ) yi TFPR across the existing set of firms in each ­industry. In a frictionless environment, where all 5. Trademarks may be viewed as a proxy for firms face the same level of distortion, that innovation in marketing and in services, is, ­τ ki = τ k and ­ τ yi = τ y TFPR should be which is usually not covered by other tra- equalized across fi ­ rms. Therefore, TFPR dis- ditional indicators (Lemarchand and Tash persion across firms can be interpreted as a 2015; Millot ­ 2009). 6. Rwanda’s financial sector is composed of 504 misallocation of ­ resources. supervised financial institutions, of which 16 The implicit distortions facing each firm are banks (11 commercial banks, 3 micro- can then be inferred from the firm’s profit finance banks, 1 development bank, and maximization conditions, including factor 1 cooperative bank); 472 are microfinance (labor and capital) distortions and output institutions (17 limited liability companies and d istortions. A firm faces a higher capital- ­ 455 cooperatives, of which 416 are umurenge labor distortion if the ratio of labor com- savings and credit cooperatives and 39 are pensation to the stock of capital is larger non- umurenge savings and credit coopera- than what is warranted by the industry tives); 15 are insurers (13 private and 2 public); elasticity of output with respect to factors and 1 is a public pension ­ scheme. of ­production. A firm faces a higher output 7. Agency banking in that context refers to a bank contracting a retail outlet such as a distortion if the labor compensation is low shop or post office to facilitate client trans- compared with what is implied by industry actions, including payments, deposits, and elasticity of output with respect to ­ l abor. w ithdrawals. These agents are not full- ­ The output distortion does not distort fledged banks and therefore offer a smaller the optimal combination of factor inputs, range of services, but they are less costly to because it affects the marginal products of establish, helping the bank to reach more both inputs ­ proportionally. ­ remote ­ a reas. They are common in many 212 FUTURE DRIVERS OF GROWTH IN RWANDA countries across the world, including Brazil, received from authorities in the following India, and ­ Kenya. countries: Cabo Verde, Ecuador, the Arab   8. Loan loss provisions relative to nonperform- Republic of Egypt, India, Indonesia, Iraq, the ing loans stood at ­ 4 4.5 percent; ­80.2 percent Lao People’s Democratic Republic, Malaysia, of total assets were classified as earning assets Mauritania, Morocco, Myanmar, Paraguay, in 2017, reflecting strong engagement in lend- Suriname, Thailand, Uruguay, and ­ Vietnam. ing ­activities. 15. See ­h ttp://www.ifrs.org /-/media /feature​   9. Composite insurers are companies offering a /around-the-world /jurisdiction-profiles​ full range of insurance services, such as life, /­rwanda-ifrs-profile.pdf. property, and motor vehicle ­ insurance. 16. Based on ­ interviews. 10. The umuganda bond is a five-year IFC bond 17. For example, Singapore’s Temasek state denominated in Rwandan francs, first issued shareholding fund progressively reduced its in May ­ 2014. It marked the first placement holdings in government-linked corporations by a nonresident issuer in Rwanda’s domes- through the sale of shares on the Singapore tic capital markets, with the objective of Stock Exchange and reinvested its cash in expanding the availability of long-term local- new investments (see ­ https://en.wikipedia.org​ currency finance for local businesses, while /wiki/Temasek_Holdings). Austria’s OIAG strengthening the country’s domestic capital state shareholding fund progressively reduced ­markets. See ­https://ifcext.ifc.org/ifcext/press- its holdings in strategic state companies to a room/IFCPressRoom.nsf/0/9548E38CE0309 “blocking minority,” in order to maximize 58F85257CDA00424FC1. access to capital market financing and cor- 11. PPPs are defined as long-term contracts porate governance discipline (see h ­ ttp://www​ between a public party and a private party for .oebib.gv.at/en/mandate/history/). In China, the development or management of a public the government undertook about 1,100 pub- asset or service, in which the private party lic sales of minority shareholdings in state- bears significant risk and management respon- owned enterprises, which eventually improved sibility through the life of the agreement and their corporate governance and raised China’s remuneration is linked to performance—­ international investment profile (Mako and the demand for or use of the public asset or Zhang ­2008). ­ ­service. 18. According to the World Bank, Doing Business 12. I T U Wo r l d Te l e c o m m u n i c a t i o n / I C T 2018, Rwanda scores reasonably well with Indicators database (ITU, various ­ years). regard to creditors’ legal rights and credit 13. The “64 MDEs” are Rwandan firms that information as well as speed of contract export more than US$64 million per year, enforcement (World Bank ­ 2 018). But the many of which grew from being small export- credit bureau covers only 20 percent of adults, ers to being “64 MDEs” within a five-year contract administration costs, on average, period. The government of Rwanda looks to ­ 83 percent of claim value, and the creditor can support firms with the potential to replicate expect to spend ­ 2 .5 years on an insolvency this s ­ uccess. See MINICOM ­ (2015). case and recover just 19 percent of its ­ loan. 14. OECD first published its OECD Guidelines Fixing these issues could facilitate SME access on Corporate Governance of State-Owned to ­credit. Thus, the government may wish to Enterprises in ­ 2005. These were revised in provide more credit information before estab- 2015 by OECD countries in cooperation lishing a partial credit guarantee facility for with a large and diverse group of partners ­SMEs. and stakeholders (OECD ­ 2015). Colombia, 19. Calculated using the number of manufactur- Latvia, and the Russian Federation have par- ing firms reporting that they export in the ticipated in this review on an equal footing IBES and the total number of formal manu- with OECD countries and associate them- facturing firms in the Census of E ­ nterprises. selves with its ­ outcome. The World Bank Census of Enterprises data are used because participated as an ­observer. Argentina, Brazil, small firms are underrepresented in the IBES China, Costa Rica, Kazakhstan, Lithuania, data because of the sampling ­ strategy. Hence, Peru, the Philippines, South Africa, and the percentage of export manufacturing firms Ukraine participated directly in discussions based solely on the IBES data may lead to an of the ­revision. In addition, comments were upward ­bias. C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 213 20. T hese other organizations include the entrepreneurship,’ number of serial entre- Rwanda Development Board, the National preneurs, degree of sell-out mentality within Agricultural Export Development Board, the firms and levels of entrepreneurial ambition) Rwanda Agriculture Board, the Development which formally and informally coalesce to Bank of Rwanda, the Business Development connect, mediate, and govern the performance Fund, the Rwanda Information Society within the local entrepreneurial environment” Authority, the Rwanda Utilities Regulatory (Mason and Brown ­ 2014). Authority, the Capacity Development and Employment Services Board, and the Higher Education ­Council. References 21. The NCST has four key mandates: (1) provide Acemoglu, ­ D., and P Restrepo. ­ ­. ­ 2017. “Robots the government with strategic advice on sci- and Jobs: Evidence from US Labor ­ Markets.” ence, technology, innovation, and research Working Paper 23285, National Bureau of policy matters; (2) identify and promote new Economic Analysis, Cambridge, M ­ A. high-impact areas of technology innovation; Adam Smith ­ I nternational. ­ 2 017. “Investors’ (3) support the use of science and technol- Perceptions and Const raints Faced by ogy in sustainable development of national E xpor ting Fi rms in Rwanda’s Priorit y resources and infrastructure; and (4) mobilize Economic S ­ ectors.” Final Report, Adam Smith funds and manage the National Research and International, ­London. Innovation ­Fund. Amin, ­ M ., and A ­. ­ M attoo. ­2 008. “Human 22. See World Management Survey ­ (http://world​ Capital and the Changing Structure of the managementsurvey.org/academic-research​ Indian ­ E conomy.” Policy Research Working /­field-experiments/). Paper 4576, World Bank, Washington, ­ DC. 23. There are several examples of this kind of Andersen, ­ M ., ­ B . Kett, and ­ E . von ­ U exkull. initiative. kLab is a start-up incubation cen- ­ ­ 2 017. “Corporate Tax Incentives and FDI ter launched and managed in close coordina- in Developing ­ C ountries.” In 2017–2018 tion with the Private Sector Federation’s ICT Global Investment Competitiveness Chamber. The African Entrepreneur Collective ­ R e por t: Foreig n Investor Pe rs pec tives (Germany) is a pan-African nonprofit based in and Policy Implic ations, ­ c h. 3, 73 –99. Kigali that provides training and resources to Washington, DC: World B ­ ank. ­ h ttps://doi​ local ­ entrepreneurs. 42Kura (Israel) is a com- .org/10.1596/978-1-4648-1175-3_ch3. mercial development that houses co-working Autor, ­ D., and ­ D. ­ Dorn. ­2013. “The Growth of space for ICT start-ups and provides incuba- Low-Skill Service Jobs and the Polarization of tion s ­ ervices. FabLab Rwanda (Massachusetts the US Labor ­ M arket.” American Economic Institute of Technology) boasts of being Africa’s Review 103 (5): ­ 1553–97. first fabrication lab, providing innovators and Bartelsman, ­ J. Haltiwanger, and ­ E ., ­ S. ­Scarpetta. designers with access to three-­ d imensional 2 0 09. “C ros s - C ou nt r y Di f ferenc e s i n ­ printers and ­ more. The list of programs also Productivity: The Role of Allocation and includes Impact Hub Kigali, Inkomoko, Spring S election.” NBER Working Paper 15490, ­ Rwanda, and Think Rwanda (Kapil and Aridi National Bureau of Economic Research, ­2017). Cambridge, ­M A. 24. Evidence from an evaluation of incubators Becker, ­ B. ­2 015. “Public R&D Policies and and accelerators in Kenya shows that lack of Private R&D Investment: A Survey of the managerial capabilities adversely affected the Empirical ­Evidence.” Journal of Economic effectiveness of these ­ programs. Surveys 29 (5): ­ 917–42. 25. The entrepreneurship ecosystem is defined as B e hu r i a , ­P. ­ 2 01 2 . “C e nt r a l i s i n g R e nt s “a set of interconnected entrepreneurial actors and Dispersing Power W hile P u rsuing (both potential and existing), entrepreneur- Development? Exploring the Strategic Uses of ial organizations (for example, firms, venture Military Firms in R ­ wanda.” Review of African capitalists, business angels, banks), institutions Political Economy 43 (150): ­ 630–47. (universities, public sector agencies, finan- Behuria, ­ P., and T ­. ­G oodfellow. ­ 2 016. “The cial bodies), and entrepreneurial processes Political Settlement and ‘Deals Environment’ (for example, business birth rate, numbers in Rwanda: Unpacking Two Decades of of high growth firms, levels of ‘blockbuster Growth.” ESID Working Paper 57, Effective ­ 214 FUTURE DRIVERS OF GROWTH IN RWANDA States and Inclusive Development Research Fontainebleau, and Geneva: Cornell University, Centre, University of Manchester, Manchester, INSEAD, and ­ W IPO. ­U.K., ­April. ———. ­2 017. Global Innovation Index 2017: Bloom, ­ N., ­ B . Eifert, ­ A. Mahajan, D ­ . McKenzie, Innovation Feeding the ­ W orld . Ithaca, and ­ J. ­ Roberts. ­ 2 013. “Does Management Fontainebleau, and Geneva: Cornell University, Matter? Evidence from I ­ndia.” Quarterly INSEAD, and ­ W IPO. Journal of Economics 128 (1): ­ 1–51. Cruz, ­ M ., ­M . Bussolo, and ­ I acovone. ­ L. ­ 2018. Bloom, ­ N ., ­ A . Mahajan, D ­ . McKenzie, and “Organizing Knowledge to Compete: Impacts ­ J . ­ R ob e r t s . ­ 2 018 . “ D o M a n a g e m e nt of Capacity Building Programs on Firm Interventions Last? Evidence from I ­ndia.” ­O rganization.” Journal of International NBER Working Paper 24249, National Bureau Economics 111 (March): ­ 1–20. of Economic Research, Cambridge, ­ M A. Diao, X., J. Randriamamonjy, and J. Thurlow. Bloom, ­ N., and ­ J. Van ­ Reenen. ­ 2007. “Measuring 2017. “Urbanization and the Future of and Explaining Management Practices across E c onom i c D e velopm e nt i n Rwa nd a .” Firms and ­ C ountries.” Quarterly Journal of International Food Policy Research Institute, Economics 122 (4): ­ 1351–408. Washington, DC. Background paper for ———. ­ 2010. “Why Do Management Practices Future Drivers of Growth in Rwanda. Differ across Firms and Countries?” Journal of EABC (East African Business ­ Community). ­ 2016. Economic Perspectives 24 (1): ­ 203–24. “East African Common Market ­ S corecard.” BNR (National Bank of ­ Rwanda). ­ 2016. Annual EABC, ­Nairobi. Financial Stability Report: June 2015–June Eichengreen, ­ B., and ­ Gupta. ­ P. ­ 2010. “The Service ­2016. Kigali: ­ BNR. Sector as India’s Road to Economic Growth?” Bode, ­ M., ­ J. Lohmann, and ­ V. ­ Steenbergen. ­ 2017. Working Paper 249, I ndia Council for “Attracting Investments While Raising Tax Research on International Economic Relations Revenue: Recommendations for Reviewing (ICRIER), New ­ Delhi. Rwanda’s Tax I ­ ncentives.” IGC Policy Note, EPO (European Patent O ­ ffice). Various y­ ears. International Growth Centre, ­ Kigali. PATSTAT ­ database. Munich: ­ EPO. Booth, ­ D ., and ­ F. ­G olooba-Mutebi. ­ 2 012. European ­C ommission. ­2003. The Broadband “Developmental Patrimonialism? The Case of State Aid Rules Explained: An eGuide ­Rwanda.” African Affairs 111 (444): ­ 379–403. for Decision M ­ akers. Brussels: European Campos, ­ F., ­M. Frese, M ­ . Goldstein, ­ L . Iacovone, ­Commission. H. Johnson, ­ ­ D. McKenzie, and M ­ .­ Mensmann. Feenstra, R ­ .­ ­ . Inklaar, and M C ., R ­ .­ Timmer. P. ­ ­ 2018. “Personal Initiative Training Leads to ­ 2 015. “The Next Generation of the Penn Remarkable Growth of Women-Owned Small World ­Table.” American Economic Review Businesses in ­ Togo.” Gender Innovation Lab 105 (10): ­ 3150–82. ­ w ww.ggdc.net/pwt. Policy Brief 22, World Bank, Washington, D ­ C. Filmer, ­ D., and ­ Fox. ­ L. ­ 2014. Youth Employment Chavhan, ­ R ., ­S. ­ K . Mahajan, and ­ J. ­S arang. in Sub-Saharan ­ A frica. Africa Development ­ 2012. “Supplier Development: Theories and S eries. Washing ton, DC: World ­ ­ B an k. ­P ractices.” IOSR Journal of Mechanical ­doi:10.1596/978-1-4648-0107-5. and Civil Engineering (IOSR-JMCE) 3 (3): Ford, ­M . ­2015. Rise of the Robots: Technology ­37–51. and the Threat of a Jobless ­ Future. New York: Cirera, ­ X., ­J. Frias, and ­ J. ­ Hill. 2017. “Innovation Basic ­Books. Policy Guide and Technology Adoption Frey, ­ B ., and ­ C. ­ M. ­ Osborne. ­ 2016. “The Future in Developing C ­ ountries.” Unpublished of Employment: How Susceptible Are Jobs to ­ manuscript, World Bank, Washington, ­ DC. Computerization?” Technological Forecasting Cirera, ­ X ., and ­ W. ­ F. ­ M aloney. ­ 2 017. The and Social Change 114 (1): ­ 254–80. Innovation Paradox: Developing-Country García-Quevedo, ­ 2004. “Do Public Subsidies J. ­ Capabilities and the Unrealized Promise of Complement Business R&D? A Meta-Analysis Technological ­C atch-Up. Washington, DC: of the Econometric ­ Evidence.” Kyklos 57 (1): World ­Bank. ­87–102. Cor nell Un iversit y, I NSE A D, a nd W I PO Gokgur, N ­ .2­ 012. “Rwanda’s Ruling Party-Owned (World Intellectual Property O ­ rganization). Enterprises: Do They Enhance or Impede ­2 013. Global Innovation Index 2013: The Development?” Institute of Development Policy Local Dynamics of I ­nnovation. Ithaca, and Management, ­ Antwerp. C o m p etitiveness and E nter p rise D eve l o p m ent for I nnovation - Led G rowth 215 Guadalupe, ­ M ., ­ O. Kuzmina, and ­ C. ­T homas. Paper 18417, National Bureau of Economic 2012. “Innovation and Foreign ­ ­ O wnership.” Research, Cambridge, ­ M A. A m e r ic a n E c o n o mic R e v i e w 102 (7 ): Kim, ­ B .-K., and ­ E . Vogel, ­eds. ­2011. The Park ­3594–627. Chung Hee Era: The Transformation of South Gugerty, ­ M. ­ 2 005. “You Can’t Save Alone: ­Korea. Cambridge, MA: Harvard University Commitment in Rotating Savings and Credit ­Press. Associations in K ­ enya.” Daniel J ­ . Evans School Kojima, ­ M . and C Trimble. ­ ­. ­ 2 016. Making of Public Affairs, University of Washington, Power Affordable for Africa and Viable ­Seattle. for Its ­ U tilities. Washington, DC: World Heizman, ­ J., and ­ M. ­Libetrau. ­ 2017. “Efficiency ­B ank. ­https://openknowledge.worldbank.org​ of Mineral Processing in Rwanda’s Artisanal /­handle/10986/2509. and Small-Scale Mining S ­ ector.” Federal Lemarchand, ­ G . A., and A. Tash, ­ 2 015. e ds. ­ I n st it ute for G eo s cienc e a nd Nat u ra l Mapping Research and Innovation in the Resources, ­B erlin. Republic of ­ R wanda. GO-SPIN Country Hsieh, ­ C ., and ­ P. ­ J. ­ Klenow. ­ 2009. “Misallocation P rof i le s i n S c ie nc e , Te ch nolo g y, a nd and Manufacturing TFP in China and I ­ ndia.” Innovation Policy, ­ 4. Paris: United Nations vol. ­ Quarterly Journal of Economics 124 (4): E duc at ion a l , S c ient i f ic , a nd Cu lt u ra l ­1403–48. Organization ­(UNESCO). IMF (International Monetary ­ F und). ­ 2 017. Lim, ­ W. ­ 2 016. “The Development of Korea’s “Article IV for ­ Rwanda.” IMF, Washington, Electronic Industry during Its Formative Years, ­DC. 1966–1979.” Korea Development Institute, ­ IMF (International Monetary Fund), UN (United School of Public Policy and Management, ­ Seoul. Nations), and World ­ Bank. ­ 2015. “Operations Mako, ­ W., and ­ C. ­Zhang. ­2008. “Why Is China So for Low-Income Countries’ Effective and Different from Other Transition Economies?” Efficient Use of Tax Incentives for ­ Investment.” In Privatization in Transition Economies: In International Monetary Fund Report to The Ongoing ­ Story. Amsterdam: ­ Elsevier. the G-20 Development Working ­ G roup. Maloney, ­W. ­F. ­2 017. Revising the National Washington, DC: ­ I MF. Innovation System in Developing ­ C ountries. ITU (International Telecommunications ­ Union). Washington, DC: World ­ Bank. ­2 016. Measuring the Information S ­ ociety. M a l o n e y, ­ W. , a n d M ­ . ­ Sarrias. ­ 2 0 14 . Geneva: ­I TU. ­ “Convergence to the Managerial F ­ rontier.” ——— . V a r i o u s ­ ye a r s . I T U Wo rld Policy Research Working Paper 6822, World Telecommunication/ICT Indicators d ­ atabase. Bank, Washington, ­ DC. Geneva: ­I TU. ———. ­ 2 017. “Convergence to the Managerial James, ­ S. ­ 2 009. “Tax and Non-Tax Incentives ­Frontier.” Journal of Economic Behavior and a nd I nvest ments: Evidence a nd Polic y Organization 134 (C): ­ 284–306. I mplications.” Foreign Investment Advisory ­ Mason, ­ C., and ­R. ­Brown. ­ 2014. “Entrepreneurial Service, World Bank, Washington, ­ DC. E cosystems and Grow th- Oriented Jones, ­ C. I ­. 2­ 016. “The Facts of Economic E ntrepreneurship.” Final ­ ­ version of a back- ­Growth.” In Handbook of Macroeconomics, ground paper prepared for “Entrepreneurial vol. 2A, edited by ­ ­ J. Taylor and ­ H. Uhlig, ­ ch. ­ 1. E cosystems and Grow th Oriented Amsterdam: Elsevier ­ North-Holland. Entrepreneurship” workshop, OECD LEED Jütting, ­ J., and ­ J. de ­ Laiglesia. ­2009. “Is Informal Programme and the Dutch Ministry of Normal? Towards More and Better Jobs in Economic Affairs, The Hague, Netherlands, Developing ­ Countries.” Development Centre November 7, ­ 2013. Studies, OECD, ­ Paris. Millot, V ­.2 ­ 009. “Trademarks as an Indicator of Kapil, ­ N ., and ­ A. ­ A ridi. ­2 017. “Innovation Product and Marketing I ­ nnovations.” OECD A g e n c i e s: Ta ke away s f rom E m e r g i n g Science, Technology, and Industry Working Innovation ­ Systems.” Unpublished paper, Paper 2009/06, OECD, ­ Paris. World Bank, Washington, ­ DC. M I N EC OF I N (M i n ist r y of Fi na nc e a nd Kast, F ­ . Meier, and D ­ ., S ­. P ­ omeranz. 2 ­ 012. Economic Planning), ­ Rwanda. ­ 2 017. Fiscal “Under-Savers Anonymous: Evidence on Self- Risk Review of Government I ­ nvestment. Help Groups and Peer Pressure as a Savings Kigali: MINECOFIN, Government Portfolio Com mitment ­ D evice.” N BE R Working Management Unit, D ­ ecember. 216 FUTURE DRIVERS OF GROWTH IN RWANDA MINICOM (Ministry of Trade and Industry), in the E ­ AC.” In De-Fragmenting Africa: ­ Rwanda. ­ 2015. “National Export Strategy ­ I I.” Deepening Regional Trade Integration in MINICOM, ­Kigali. Goods and Services, edited by P ­ . Brenton and New ­ T imes. ­ 2 017. “How Smelting Firm Lost ­ G. ­ Isik. Washington, DC: World ­ Bank. Licence to Refine Tin ­ O re.” New Times , WIPO (World Intellectual Property O ­ rganization). June ­20. ­http://www.newtimes.co.rw/section​ ­2 017. World Intellectual Property Report /­read/214579/. 2017: Intangible Capital in Global Value NISR (National Institute of Statistics of ­ Rwanda). ­Chains. Geneva: ­ W IPO. ­2 011. Establishment Census ­ 2 011. Kigali: World ­ 2 015. “Export Competitiveness: B ank. ­ NISR, ­September. Why Domestic Market Competition ­ Matters.” ———. ­ 2015. Establishment Census ­ 2014. Kigali: Viewpoint, ­June. ­https://openknowledge.world​ NISR, ­June. bank.org /bitstream / handle/10986/23658​ ——— . ­2 016. Int eg rat ed Hou sing Liv ing / ­E xport0competi00Competition0Matters​ Conditions Survey 2013/ 2014 ( ­EICV4). .pdf?sequence=1&isAllowed=y. Kigali: ­N ISR. ­ ——— . 2 016 . M ak i ng Po we r A f fo rd a bl e ———. ­2 017. Integrated Business Enterprise for Afric a and Viable for Its Utilities. Survey (IBES) ­ 2015. Kigali: NISR, ­ July. Washington, DC: World Bank. OECD (Organisation for Economic Co-operation —— —. 2018. Doing Business 2018: Reforming to and D ­ evelopment). ­ 2 011. “State- Owned Create ­Jobs. Washington, DC: World ­ Bank. Enterprise Governance Reform: An Inventory ———. 2019. Doing Business 2019: Training for of Recent ­ Change.” Paris: ­OECD. ­ Reform. Washington, DC: World Bank. — — 2015. Guidelines on Corporate Governance —. ­ World Economic ­ F orum. ­ 2 017. The Global of State-Owned ­ Enterprises. Paris: ­OECD. Competitiveness Report ­ 2017–2018. Cologny, PSF (Private Sector ­ Federation). ­2013. “Business Switzerland: World Economic F ­ orum. Investment Climate ­ Survey.” Kigali, ­ Rwanda. World Economic Forum, INSEAD, and Cornell Steenbergen, ­ V., and ­E . von ­ U exkull. ­ 2 018. ­University. ­2016. Global Technology Report “Raising the Cost-Effectiveness of Rwanda’s ­2016. Geneva: World Economic ­ Forum. Tax I ­ ncentives.” Unpublished paper, World Wu, ­ 2007. “Service Sector Growth in China Y. ­ Bank, Washington, ­ DC. and India: A ­ C omparison.” C hin a: An UNCTAD (United Nations Conference on Trade International Journal 5 (1): ­ 137–54. and ­Development). ­2017. Science, Technology, Zúñiga-Vicente, ­ A ., ­ J. ­ C . Alonso-Borrego, F J. ­. ­ and Innovation Policy ­ R eview. Geneva: Forcadell, and ­ I. ­ J. ­ G alán. ­2014. “Assessing ­U NCTAD. the Effect of Public Subsidies on Firm R&D Wagh, ­ S., ­A. Lovegrove, and J­. ­ Kashangaki. ­ 2012. Investment: A ­ Survey.” Journal of Economic “Scaling up Regional Financial Integration Surveys 28 (1): ­6 –67. 5 Transitioning Agriculture and Food as an Engine of Growth Introduction with regional and global ­ m arkets. Its con- tribution will need to shift from primarily Agriculture has been a major source of supplying commodities for domestic use to national income and growth for ­ Rwanda. It producing higher-value-added goods as an accounts for close to 70 percent of employ- integral part of food supply chains linked ment, more than 30 percent of gross domestic to regional and international ­ markets. This product (GDP), and more than 50 percent of will require a much larger role for the pri- exports of ­goods. Considerable government vate sector in determining the allocation of effort has been devoted to developing the agricultural ­ resources. It also will involve a sector, especially after the food price shocks continuing large, but evolving, role for the 2007–08. Agricultural value added has of ­ public ­s ector. The chapter makes the case risen more than 5 percent a year over the past that Rwanda is well placed to make this 15 years, and productivity in agriculture has ­happen. increased ­strongly. The main question for this chapter is how long and to what extent High growth in the future will come will agriculture remain an engine of growth from productivity-increasing innovations in the Rwandan economy? Does Rwanda’s and technical and allocative efficiency in highly ambitious economic future continue resource ­use. Only a few initiatives should to have agriculture at its core, or will the sec- receive public financial support, accom- tor launch high long-term growth and then panied by a rigorous evaluation to deter- gradually take a back seat? mine whether adequate returns are being The agriculture sector will remain a achieved. Initiatives with persistently low ­ major source of comparative advantage for returns should be reallocated to higher- Rwanda, even as it climbs the income lad- return ­ a ctivities. Public sector activities der over the coming ­ decades. This suggests in agriculture will nevertheless need to that the sector should remain a major part increase, but be targeted more effectively to of any high-growth ­ strategy. To accomplish providing key public ­ goods. this, agriculture will need to continue to First, Rwanda’s effective research and modernize, become more responsive to mar- regulatory institutions should continu- ket signals, and integrate more effectively ally be adapted to evolving opportunities 217 218 FUTURE DRIVERS OF GROWTH IN RWANDA and ­ t hreats. The high technical expertise Fourth, Rwanda needs to profit from of the agency in charge of agriculture will the big-data revolution in innovation, mak- be increasingly important as improving ing the benefits of big data accessible at standards and Rwanda’s efforts to increase reasonable cost to ­ smallholders. Big data exports increase the frequency and impor- approaches are critical to both increasing tance of sanitary and phytosanitary ­ issues. innovation and private sector investment and Rwanda’s reputation for probity and account- dealing with the smallest farms and most ability is one of its most precious assets and densely settled rural areas in A ­ frica. “Smart is especially important in agriculture, where farming,” where just the right amounts of quality and food safety issues are dominant the right inputs are used for each parcel on concerns of ­ consumers. the basis of information gathered by hand- Second, a more consistent and selective held devices with the right sensors, offers one policy needs to be formulated regarding such ­ opportunity. Big data approaches such vertical coordination through value chains as blockchain technology similarly offer the and horizontal coordination of production potential to lower the costs of small financial across f ­arms. More effective vertical coordi- transactions that require secure record keep- nation in agriculture would be where private ing and decentralized input, such as land reg- sector partners provide skills, capital, and istration and ­ mortgages. entrée into international markets and farmers Fifth, public investments in ­ infrastructure— acquire higher incomes and new ­ resources. especially rural roads, electricity, and Rwanda benefits from an active cooperative water—should be linked more effectively system, but outcomes from contractual rela- to agricultural development and higher- tionships between farmers and private sector value-added ­ products. Irrigation is clearly aggregators vary w ­ idely. The success of such ac­ onstraint. Notably, current yields of rice relationships from the point of view of both are insufficient to cover the cost of irrigation farmers and aggregators is essential to small- investment. Therefore, while emphasizing an ­ holder inclusion in high-value-growth busi- increase in rice productivity, other high-value ness models (Delgado ­ 1999). crops like horticulture could also be consid- Third, Rwanda should engage more effec- ered in irrigated ­ areas. tively and opportunistically with neighbor- Sixth, to facilitate the transmission of ing countries on a regional division of labor, technical knowledge and skills to smallhold- both in production based on comparative ers, there is a need for more proactive and advantage and in knowledge generation consistent training of managers and ­ farmers. and ­diffusion. Rwanda has a major stake in In particular, a more consistent approach increasing the importation of raw materials to training managers and technical experts such as cereals for its food-processing and working for aggregators is needed for verti- animal protein industries, while increasing cal i­ ntegration. Companies might be involved its exports of processed foods and high- in this process through training programs to value calories—for example, from milk and build up nationally needed skill sets in agri- meat ­ products. Ways should be explored business, high-value supply chains, and agri- to exploit the huge market potential of cultural ­ technology. The best multinational neighboring countries by increasing food companies have a track record of making production. Another initiative might include ­ these investments and provide examples of a more common approach to certification how such investments can be done to mutual of food safety, sanitary, and phytosanitary ­advantage. conditions. Regional integration also should ­ Seventh, land degradation and global cli- be pursued through continuing efforts to mate change, which threaten the future of adopt transparent science-based standards, Rwandan agriculture, need to be addressed product registrations, and certification of more ­ comprehensively. Although consider- agricultural ­inputs. able progress has been made in constructing T ransitioning A gricu l ture and F ood as an E ngine of G rowth 219 robust “bench” (wide) terraces, much more new demand for local consumption items like needs to be done to secure the land a ­ sset. processed fresh food and drinks and local Fortunately, the soil and water management services; the size of the “growth multiplier” interventions central to halting land degrada- from this effect depends on the extent to tion are also the ones most useful to increase which the new income coming into remote the ability of farmers to adapt to climate areas is spent on locally produced items that change through better water retention and have no outlet outside the local area and quality. Landscape restoration increased soil ­ that can be produced by formerly underused and conservation are community-level activi- resources (Delgado et ­ al. 1998; World Bank ones. Local government, ties, not individual ­ 2007b). The Integrated Household Living ­ producer groups, and national technical Conditions Survey of 2013/14 (EICV4) shows expertise and funding all have a role to play that farmers work, on average, only 20 hours in ­implementation. a week in Rwandan agriculture, although this likely reflects periods of both labor slack and labor scarcity (MINAGRI 2017c; NISR Policy Support for Strong Growth 2014). In any event, Diao et ­ ­ al. (2017a) esti- mate that such multipliers are still evident in in Agriculture more isolated parts of R ­ wanda. However, The agriculture sector plays an important this effect, based on demand constraints in role in the Rwandan ­ economy. Agricultural rural areas, cannot be expected to last in the GDP grew 5 percent a year from 1999 to longer term as the economy becomes more 2016, which is a high rate of growth by global fully ­connected.3 standards. Despite slow growth in sectoral ­ Supportive policies have helped to boost employment over this period ­ (1.1 percent a agricultural p ­ roduction. The National year), agriculture still accounted for close Agricultural Policy of 2004 and its four- to to 70 percent of total national employment five-year implementation plans (Programs for in ­2016.1 Employment in other sectors rose Support to Agriculture, PSTA1 to PSTA3) ­ 9.8 percent a year from 1999 to 2 ­ 016. The emphasized growth in staple food outputs rapid movement of workers out of agriculture for the domestic market and widespread and into more productive ­ sectors accounted improvement in the income of s ­ mallholders. for ­3.2 percentage points of the country’s The emphasis on self-sufficiency in staple 4.4 ­ ­ percent annual rise in labor productivity food crops was reinforced when the doubling from 1999 to ­ 2016. Further, growth in agri- of international food prices in 2008 was culture has likely stimulated growth in other accompanied by the shutting off of cereal sectors through backward links (for example, exports from Tanzania and U ­ ganda. The by encouraging input industries such as fer- food price crisis drove a severe compression tilizers) or forward links (for example, by of welfare (on average, at that time, Rwandan increasing food processing), although these households spent 60 percent of their income links tend to be smaller in agriculture than in on food), especially for the ­ poor.4 other sectors (Hirschman ­ 1958). Government efforts to develop new irri- Widespread growth in the purchasing gated land and improve existing land through power of farm households in remote areas improved terraces contributed to agricultural can increase demand for local goods and ­ growth. From 2008 to 2012, irrigation was services, which can mobilize unemployed introduced in 23,000 hectares of margin- local ­resources. Consumption growth links ally used marshlands and 2,500 hectares arise from bringing outside spending power of hillsides, and more than 46,000 hect- into remote areas—whether from the sale ares of “bench” terraces (wider and more of export crops or from unilateral transfers level terraces typically requiring substan- such as ­remittances.2 The key is that under- tial labor or other power to construct) were employed resources are mobilized to meet added (MINAGRI 2012; World Bank ­ 2014). 220 FUTURE DRIVERS OF GROWTH IN RWANDA However, the cost of irrigation investments is out by individual landowners, albeit on a high, about ­ 7.5 to 10 times the annual earn- communal schedule (Pritchard ­ 2013). Land ings (including subsidies and after costs) of consolidation has also facilitated the devel- rice farmers benefitting from irrigation in opment of new irrigated agricultural lands the ­ m arshlands. Moreover, potato farm- in former marshlands and joint soil and ers with nonirrigated land in the northern or water management in other sites to reduce southern provinces earn some four times, and erosion. Thus, consolidation has been a key ­ coffee growers earn more than three times, part of the strategy to increase productivity the amount earned by rice farmers in the and expand the area under food production irrigated ­ marshlands. Increasing scarcity of through state leadership (Kathiresan 2012; land and water suggests that such growth in World Bank ­ 2014). area cultivated will be difficult to sustain in The government made a major effort the longer run and will need to be in higher- to register agricultural land to individual value items, especially at the current cost owners from 2009 to ­ 2013. This program, levels for irrigation in Rwanda (US$6,000 more extensive than similar efforts in most to US$8,000 per hectare for irrigation in the other African countries, has achieved the marshlands and three to four times that on demarcation, adjudication, and digitiza- ­hillsides). tion of ­ 10.7 million parcels of agricultural The government has made significant land, resulting in coverage of more than progress on land ­ consolidation. The aver- 90 ­ p ercent of agricultural parcels in the age farm size is only ­ 0.6 hectare, with per- ­c ountry. 8 Providing secure title has argu- haps 0 ­ .12 hectare per ­worker. 5 Production ably made the job of consolidating produc- on numerous small plots of land can tion decisions more palatable to farmers impose technical inefficiencies related to (Pritchard ­ 2 013). Titling improved land water control and distribution, crop choice security, especially for women, and led to and interactions, and planting materi- considerable inceases in farmers’ investment als used, while limiting the potential for in their own ­ soils. However, it also has been mechanized ­ plowing. Land consolidation associated with a decrease in land sales, was launched with the organic land law in suggesting that titling has not promoted 2005, 6 institutionalized through the Land land market activity or commercial con- Use Consolidation Program (LUCP) in solidation, at least not in the short run (Ali, 2007, and implemented through the Crop Deininger, and Goldstein ­ 2014). Intensification Program (CIP) starting in Cooperative development has been a key 2 008. Land under consolidated use rose ­ component of state-led collective ­ a ction. from 28,788 hectares in 2007 to 502,916 Cooperatives are organized along commod- hectares in 2012, primarily through the ity lines corresponding to value chains—for consolidation of production rather than of example, rice, maize, potatoes, beans, cof- ownership (World Bank ­ 2014). At this rate fee, tea, horticulture, and ­ dairy. Crop and of increase, some 39 percent of farm house- livestock cooperatives account for roughly holds would have participated by ­ now.7 half of Rwanda’s 8,000 primary coopera- The land consolidation program has been tives at the present time (MINICOM ­ 2018). state led, with some room for individual Most of the cooperatives are aggregated into ­ i nitiative. The government has determined federations along commodity lines, with rice which (food staple) crop varieties and seeds and tea being the most significant in terms of must be planted on adjoining plots, when members and value (World Bank ­ 2016). The they are to be planted, and, to a lesser degree, more organized federations, such as those what other inputs will be u ­ sed. However, for rice and tea, offer support services such consolidation has not involved mandatory as finance, aggregation, and ­ marketing. Yet consolidation of ownership, and planting, less than 20 percent of farmers belonged to cultivation, and harvesting are still carried cooperatives in 2016 (NISR ­ 2016b). In part T ransitioning A gricu l ture and F ood as an E ngine of G rowth 221 this may be due to the fact that some com- Government efforts also contributed to modities, such as rice and coffee, more growth in the livestock s ­ ector. Livestock naturally lend themselves to cooperative value added rose by 5 ­ .2 percent per year development in dealing with ­ aggregators. In from 2000 to ­2016. Efforts will continue part it is likely due to resource constraints being deployed to increase animal resources in government programs such as subsidized products, notably milk, meat, eggs, honey, seed and fertilizer inputs for maize or wheat and ­fish. cultivation that are distributed through ­cooperatives. More than 70 percent of farmers with more Emerging Policy Challenges than 10 hectares (large farms by Rwandan Achieving the government’s ambitious standards) were members of cooperatives, growth objectives will require a shift in the suggesting that membership was important sources of growth in ­ agriculture. Periods of for obtaining access to government services rapid growth in agricultural production in and other support (World Bank ­ 2016). At Rwanda have largely reflected either expan- the same time, a recent survey conducted sion of the land devoted to agriculture (1960s for the Ministry of Trade and Industry also and 1970s) or increased use of inputs (2000s) indicated dissatisfaction among coopera- 5.1). However, the scope for further (figure ­ tive members with regard to accountability expansion of agricultural land is limited, as is and transparency in the system (MINICOM the potential for achieving worthwhile gains ­2018).9 A copious literature on good prac- from applying ever larger amounts of labor tice in development of agricultural coopera- hectare. There is significant potential for per ­ tives addresses typical sticking points such continuing to improve farmer skills, extend as ­ t ransparency, clarity of ownership, and accountability (Zeuli, Cropp, and Schaars ­ 2004). In Rwanda, interviews suggest that FIGURE 5.1  Decomposition of sources of agricultural growth in linking becoming a member of a specific Rwanda, 1961–2014 cooperative, receiving benefits through it, and actually residing in the location concerned 6 5.55 4.01 5.15 will be a particular issue in future discussions 5.04 5 regarding cooperative ­ governance. Average annual net agricultural Government promotion of the use of fer- 4 output growth (%) tilizer and improved seeds contributed to 3 1.84 growth in a ­ griculture. The share of farms 2 applying inorganic mineral fertilizer, asso- ciated with substantial improvements in 1 0.05 yields, increased from 12 percent in 2008 to 0 30 percent in 2012 (MINAGRI 2014; World Bank 2014), although use rates have fallen –1 since then, as discussed b ­ elow. Promotion –2 of fertilizer use has been an important part 1961–70 1971–80 1981–90 1991–2000 2001–10 2005–14 of the CIP launched in 2007 (MINAGRI TFP Area expansion Inputs intensification 2 015b). The government procured fertil- ­ izer abroad in bulk and in 2010 created Source: Compiled from U.S. Department of Agriculture 2017 data. Note: The three sources of growth listed sum to output growth in the period in question. a network of small-scale private dealers Output growth may be different than column height because of negative contributions to for distribution, located throughout the growth in the period in question. Growth not explained by area expansion or increased use per hectare of inputs (including labor) is attributed to total factor productivity (TFP). TFP is ­ country. Privatization of fertilizer imports a combination of increased allocative and technical efficiency and technological change. was implemented in 2013, rising to eight Because the numerator of TFP (output growth) is highly affected by weather outcomes in any given year, it makes sense to consider TFP outcomes over at least a 10-year period active importing companies in 2015–16 that includes both good and less good weather. This explains the strategy for handling (MINAGRI ­2015b). differences between the early and later parts of the period 2001–14. 222 FUTURE DRIVERS OF GROWTH IN RWANDA irrigated area, stabilize and expand terraces, outside the country, actively pursuing scale boost the use of more adapted fertilizers, economies to increase the competitiveness of and increase farmers’ use of better s ­ eeds. exports, and benefitting from the experience Extending these programs to more farms and networks of successful firms elsewhere would likely continue to promote a rate of in the ­world. This approach implies creating growth of 5 percent and improve the welfare an institutional, infrastructural, and policy of large numbers of smallholders for perhaps environment led by market signals about a few years, as farms yet to be reached adopt where, when, and how agriculture produces more modern ­ practices. However, the impact and trades and shifting the public role gradu- of such improvements will decline over ally to facilitating the fair implementation of time as the “low-hanging fruit” are picked those private sector ­ decisions. Public sector and more difficult cases are taken ­ on. Most efforts to improve the technical efficiency of important, this approach would not generate smallholders would ­ continue. the agricultural growth necessary to achieve The shift in some responsibilities from the rapid increases in output envisioned the state to the private sector will build on under the government’s ­ strategy. recent policy changes in this d ­ irection. The More rapid growth will require a faster first implementation plan for the National increase in total factor productivity ­ (TFP). Agriculture Policy envisions further boost- Total factor productivity growth reflects ing the productivity of staple foods while improvements in technical efficiency, made attracting more private sector investment possibile by the redeployment of inputs, into agriculture (MINAGRI ­ 2017b). At the land, and labor to increase physical output; same time, the National Agricultural Export allocative efficiency, which considers the Development Board has presided over priva- costs of using different inputs and ­factors in tization of the tea sector, begun privatizing addition to technical issues, in order to maxi- parts of the coffee sector, and promoted mize profitability; and technological change, exports by private entities of fruits and veg- which embodies scientific and technical etables (NAEB ­ 2017). innovation to get more from less (annex ­ 5A). Rapid increases in productivity will TFP growth in Rwandan agriculture has require more mechanization (MINAGRI exceeded 2 percent a year since 2005 (on par 2017a). Despite substantial progress in land ­ with countries in Asia and Latin America), consolidation, most of Rwandan agriculture reflecting continuing payoffs from public is carried out under conditions of ­ abundant— research and extension to support staple sometimes surplus—labor on small plots, food ­production.10 Increases in TFP have mostly on hillsides (MINAGRI ­ 2017c). As been based largely on improved technical a result, the use of labor-saving machinery, ­ efficiency. For Rwanda to become an upper- such as tractors and combine harversters, middle-income country by 2035, further is confined to a few small n ­ iches. However, increases in TFP in agriculture will have to mechanical solutions are needed to expand be based more on improvements in allocative bench terraces and extend irrigation at lower efficiency and technological ­ change. costs per unit, in both cases involving whole More rapid growth in allocative effi- ­ c ommunities. Postharvest mechanization ciency and technological change, in turn, will needs—including for smallholdings as much require greater reliance on private ­initiative. as for large farms—are for transport, pro- Investments will have to be based more on cessing, and packaging of agricultural com- expected economic returns, cutting losses on modities along the value ­ chain. In some cases, losing ventures, adding value through pro- equipment will need to be owned ­ collectively. cessing where it is profitable, specializing in It will be critical to match the form of mecha- the most profitable opportunities, importing nization to actual needs in the system and to those products that are cheaper in neighbor- ensure that the machines used are serviceable ing countries, tapping into larger markets in Rwanda at reasonable ­ cost. T ransitioning A gricu l ture and F ood as an E ngine of G rowth 223 Further efforts are necessary to promote lack access to the forage or pasture required the use of inorganic ­ fertilizer. The share of to ensure that the cow is well nourished farms using inorganic fertilizer fell from (World Bank 2 ­ 014). Moreover, payments 30 percent in 2012 to below 20 ­ p ercent received by farmers for raw milk appear in ­2 017.11 This decline may reflect the low, about half to one-third of the level in fall in the level of subsidy from 100 percent neighboring countries, even though the in 2007 to between 16 and 50 percent in price of ultra-high-temperature processed 2016, depending on the specific formula- milk in Kigali shops is comparable with tion (AGRA 2 ­ 016).12 Moreover, limited that in neighboring ­ c ountries.13 Farmers access to financing and extension services are paid prices jointly set by the Ministry may be constraining the growth of fertilizer of Agriculture, Ministry of Industry, farmer use (IFDC 2 ­ 014). Continuing gains could cooperatives, milk processors, and milk col- be promoted by expanding the coverage of lection centers (MCCs). Interviews suggest locations, which there clearly is still scope that following their own assessments of to do, and improving fertilizer formulations business prospects, formal milk processors (IFDC ­ 2 014). The Rwanda Agriculture have been reluctant to pay higher ­ prices. The Board, for example, is experimenting with number of MCCs is high, following rapid fertilizer formulations for Irish potatoes in growth in their numbers driven by p ­ olicy. volcanic soils that are more suitable than Given the difficulty they face in getting the standard 17N-17P-17K fertilizer that higher prices from processors and farmers’ Rwanda now imports for use everywhere reluctance to supply to them, many MCCs (MINAGRI ­2017c). are operating at low capacity, in the vicinity The government needs to continue pro- percent on ­ of 25 ­ average. Low turnover leads moting the use of improved seeds and to high unit costs, further compounding the modernizing its seed r ­ egulations. The CIP problem of low producer prices (Abdulsamad typically purchases improved seeds for cere- and Gereffi 2 ­ 016). Because the roots of the als in bulk in other countries and distrib- problem seem to be rapid growth in the sup- utes them to ­ farmers. Only maize, soybean, ply of raw material combined with a limited and wheat seeds were subsidized at the farm domestic market, especially at the high end, level in 2016; subsidies equaled 75–85 per- Rwanda’s sales of milk will need to expand cent of the cost of imported seeds and about into neighboring countries if the sector is to 50 percent of the cost of seeds produced thrive (Abdulsamad and Gereffi ­ 2016). locally (AGRA 2 ­ 016). As part of a move Lack of agricultural finance remains a toward a market-oriented seed system, in major constraint on transformation of the 2015 Rwanda launched an effort to harmo- s ector. Transforming agriculture requires ­ nize Rwanda’s seed trade regulations with i nvestment. Rwandan farmers’ “financial ­ norms of the Common Market for Southern inclusion” in the formal financial sector and Eastern Africa; harmonization was com- is comparable with that of the rest of the pleted in 2017 (COMESA 2017; MINAGRI population and high compared with farm- 2015a). Rwanda is also taking the first steps ­ ers’ access in the rest of Africa (Access to to legalize agricultural genetically modi- Finance Rwanda–Finscope ­ 2016). Financial fied ­organisms. The Rwanda Environment inclusion here is determined by interviews Management Authority drafted a law that with farmers and other agricultural actors will govern genetically modified organisms; such as small and medium enterprises the law would implement the Cartagena (SMEs), as well as individuals and SMEs Protocol, to which Rwanda is a signatory in other s ­ ectors. The survey asks ques- (Ntirenganya ­2018). tions about ease of physical access, ease of Important constraints to development opening an account, convenience, services of the milk sector have emerged over ­ time. on offer, access to loans and interest rates Many beneficiaries of the Girinka program charged, and other issues; and a score is 224 FUTURE DRIVERS OF GROWTH IN RWANDA computed from the ­ a nswers. Agricultural widely considered as a way to manage finan- entities, including individual farmers, have cial transactions, asset transfers, and regu- mean scores similar to participants from latory o­ bligations. Land registration, where other ­ s ectors. However, the actual use of Rwanda has made huge recent progress in formal financial services in the agriculture clarifying individual data points, is a natural sector in Rwanda is quite low and occurs venue for reducing the costs of registration mainly through savings and credit coop- and record keeping through decentralization eratives (SACCOs) and mobile money pro- of updates and safekeeping and transpar- viders (World Bank ­ 2 016). Although the ency of r ­ ecords. Such approaches are now overall ratio of credit to domestic GDP being tried or considered in Georgia, Ghana, for Rwanda was 20 percent in 2016, the the Russian Federation, Ukraine, and the comparable ratio for agricultural credit United Kingdom, in addition to some states in relation to agricultural GDP was only of ­India.16 4.6 percent, up from 3 ­ ­ .6 percent in 2012 The technology offers promise for reduc- (World Bank ­ 2016). ing the costs of registering small mort- Agricultural SMEs—critical for agricul- gages and other ­ l oans. In this regard, tural transformation—in particular make the Rwandan Blockchain Project was less use of formal financial services through established in 2017 in partnership with banks (as opposed to other loan-making t he W I S eKey’s Blockchai n C ent re of financial institutions such as SACCOs) than Excellence, a major Swiss information other SMEs (World Bank ­ 2 016). Reasons technology ­ c ompany. It aims to position identified include low availability and qual- the country as a key player in digital trans- ity of data to decrease lenders’ risks; limited formation by providing citizens and busi- financial experience of farmer and producer nesses with access to policy, technical, and organizations; limited operational capac- business ­ e xpertise. In October 2017, the ity among formal sector banks to serve the government of Rwanda, W ISeKey, and agriculture sector; limited availability of Microsoft announced a partnership to medium- to long-term liquidity; and fiscal adapt Rwanda’s land registry specifically disincentives to ­ borrowers. With regard to to ­blockchain.17 fiscal disincentives, improvements in the Both the state-owned Development Bank land tenure system and collateral registry of Rwanda (BRD) and private banks are have resulted in increased use of land as increasingly investing in agribusiness as collateral. However, the high costs of reg- ­ opposed to agricultural ­ production. BRD istering mortgages for relatively small loans is the largest formal provider of credit to act as a disincentive to financial i ­ nstitutions. agriculture, primarily for tea production These extra costs are typically added to the and ­ agroprocessing. Overall, nonperform- costs facing b ­ orrowers. Furthermore, less ing loans equaled 18 percent of lending for than ­ 0 .5 percent of farmers surveyed in agricultural production in 2016, compared Rwanda report having access to agricul- with 8 percent of total agribusiness l ­ending. tural insurance (World Bank ­ 2016).14 Further, 57 percent of the formal banking Blockchain technology provides high sector’s agricultural loan portfolio in 2016 potential for reducing the cost of r ­ egistering (FRW 70 billion or US$86 million) sup- f i n a nc i a l ly sig n i f ic a nt ­t ra n s ac t ion s . ported agroprocessing and agrotrading as Blockchain is a digital database (or ledger) opposed to production (World Bank ­ 2016). distributed across a network of ­ computers. The implication is that any policies designed The records are protected by cryptography to increase formal sector bank finance for and thus largely protected from human error, agricultural production will need to consider editing, and ­ removal.15 The technology can how to reduce the risk of default, but the enable secure transactions, digital authenti- default risk may be less of a concern for lend- cation, and legally binding s ­ ignatures. It is ing to agroprocessing and ­ trading. T ransitioning A gricu l ture and F ood as an E ngine of G rowth 225 Targeting Larger Markets p ­ racticed. Revenues from tea have risen in for Growth recent years, a large share of Rwandan tea traded at the Mombasa tea auctions captures Achieving and sustaining higher agricultural premium prices, and Rwandan tea has won growth will require identifying crops, live- competitions for quality there (NAEB ­ 2017). stock, and food products where Rwanda has Tea is a plantation crop, with production a comparative ­ advantage. Farmers should managed and bulked by cooperatives that not direct their production just to satisfying factories. The tea industry sell to private tea ­ rising demand from domestic consumers as currently comprises 14 operational factories, urbanization and incomes r ­ ise. High growth all of which were privatized between 2004 comes from scale economies and specializa- and ­2012. tion, which can only be achieved through Success in boosting agricultural exports regional and international trade in a small through multiple layers of value chains economy such as ­ Rwanda. As incomes rise, (for example, cooperatives, federations, fac- consumers will demand greater diversity in tories, and exporters) also requires ensur- product lines; for example, the average tra- ing that farmers get a decent share of the ditional food-oriented U ­ .S. supermarket final wholesale export p ­ rice. In the case of carries roughly 40,000 products originating tea, concern emerged in 2011 that farmers from around the world, and large megastores were not getting a high enough share of the that sell food carry far more p ­ roducts.18 “made-tea” price earned by the factory; the Eventually, diversity in product lines at the system in place was a form of “cost-plus” consumer level and specialization at the pro- pricing by factories in which producer prices ducer level will be the norm ­ everywhere. It were delinked from world ­ markets. In 2012, will be important for Rwanda to enter the Rwanda’s Cabinet mandated a producer price game as early as possible to capture desir- share of 30 percent of the “made-tea” auction able market shares for its specialty p ­ roducts. price received by the ­factories. This share was Producers will have to evaluate shifting raised to 40 percent in early 2016, despite patterns of demand and the country’s core opposition from the ­ factories. Evaluations advantages carefully and then assess how to by the World Bank found that, after the deploy them to maximize income ­ gains. last price increase, farmer (and not just Competing in global food markets ­ cooperative) revenues per kilogram increased requires achieving high levels of food quality 40 percent relative to the pre-2011 ­ level. The and safety and receiving credit for product evaluation did not find any impact on qual- quality and safety through credible certi- ity, but increased farm investment did lead to fication, traceability, and harmonization higher ­quantity. Finally, the evaluation found of ­standards. Certification is important to that the 40 percent producer price share had reducing transaction costs in marketing and no significant impact on tea factory bought- reaping market recognition for quality (FAO leaf costs as a share of total c­ osts.19 Box 5­ .1, 2010; Jaffee, Henson, and Diaz Rios ­ 2011). relating to the upgrading of both quality and The Rwanda Standards Board’s global rec- brand in a Chinese case, illustrates the high ognition of integrity and competence is a potential for adding value to ­ tea. key part of Rwanda’s comparative advantage Unlike tea, coffee is truly a smallholder going ­forward. For example, five private com- crop grown on 400,000 diversified individ- panies presently sell value-added Rwandan ual farms throughout the country and sold tea, benefitting from high quality (single through cooperatives (NAEB ­ 2017). Since estate brands), Fair Trade certification, and 2002, Rwanda policy has been to improve certification as pesticide free by the Rwanda the quality as well as the quantity of cof- Standards B ­ oard. Food safety certification fee for ­export. International coffee prices (such as the International Organization for have fluctuated by a factor of 7 in nominal Standardization’s ISO 22000:2005) is also terms in the last two ­ decades. 20 The decline 226 FUTURE DRIVERS OF GROWTH IN RWANDA BOX 5.1  High-value tea development in Yunnan Province, China Yunnan Province in China is a mountainous, fertile small farm communities to shift a larger share of region that falls within the South China–Vietnam their bulk tea production to higher-value commer- subtropical evergreen forest ­belt. It is one of China’s cial ­outlets. Reforestation of the most degraded largest tea-producing regions, with 70 percent of land is actively promoted through cash ­ i ncentives. production sold outside the ­ province. Significant Testing and certification of tea for food safety has environmental challenges have come from erosion been pursued vigorously, especially after Yunnan on hillsides and overuse of pesticides, leading to tea exports were rejected in Europe and Japan in soil and water ­ pollution. A large share of produc- 2 011. Labeling programs for specific geographic ­ tion comes from small farms averaging ­ 0.3 ­hectare origin, dating, and “greenness” are also being e ach. The very best tea from small tea gardens ­ implemented to add ­ value. typically is marketed directly to wealthy urban At the extreme, raw tea from Pu’er (one of ­ c onsumers via specialized m ­ iddlemen. However, Yunnan’s subregions) commanded up to US$20,000 commercially branded teas form the large majority per pound in 2013, although this price could not be of high-end commercial supplies to cities and virtu- sustained. More usually, certified and branded qual- ­ ally all exports; quality and food safety are checked ity teas from Yunnan achieve a 300 to 400 p ­ ercent r igorously. Low-end tea is sold in bulk for domes- ­ markup over bulk tea from the same s ­ubregions tic use and encounters consumer concerns for food of ­Yunnan. Although tea from bulk operations safety as well as ­ quality. is grown at higher density per hectare, value per Government has been trying to reduce the envi- hectare is still higher for the branded and certified ronmental impacts of tea production and helping operations in the same ­ ­ locations. literature. Source: Havemann 2015, which includes a comprehensive survey of the ­ in revenues from coffee, shown in table ­ 5.1, The path that vertical coordination con- in large part reflects steep declines in global tinues to take in coffee is likely key to its coffee prices since the 2011 p ­ eak. As a f uture. Much has been achieved through ­ relatively small producer in East African focusing on enforcing zoning to improve terms, Rwanda sells mainly to traders who traceability of coffee, an element of brand- bulk up coffees from different sources and ing embraced by the private sector and criti- are poorly placed to pass on quality premi- cal for firms to benefit from their investments ums to only one s ­ upplier. Instead, Rwanda in farmers (FAO, MAFAP 2017; NAEB needs to increase the share sold directly to 2017). Further quality improvements will ­ roasters, particularly those that sell to spe- result from the transfer of more specialized cialty markets willing to pay a premium knowledge and inputs from export custom- for certain ­characteristics. 21 Rwanda pro- ers (such as international roasters) to coop- duces some of the best Arabica coffees in eratives, farmers, and washing s ­ tations. This the world, major players such as Starbucks transfer of knowledge and inputs also should are involved, and Fair Trade certification encourage companies to pay higher price pre- has been ­ a chieved. Under active promo- miums to cooperatives and encourage farm- tion from the National Agricultural Export ers to comply with quality s ­ tandards. Fair Development Board, the number of coffee- Trade certification opens the door to higher washing stations—full washing is key to retail prices in final markets, but certification improving the quality of coffee—has grown requires a minimum level of production to significantly, reaching 300 and adding a overcome the costs of certification and moni- final export price premium of U ­ S$0.45 per toring at the producer ­ level.23 These costs are pound (NAEB ­ 2017). 22 likely too high for most washing stations in T ransitioning A gricu l ture and F ood as an E ngine of G rowth 227 TABLE 5.1  Rwanda’s main agricultural exports, 2012/13 to 2015/16 current US$ realized, millions, free on board Product 2012/13 2015/16 Three-year change (%) Tea 63.9 73.0 +14 Coffee 69.4 52.3 −25 Livestock and products 54.7 73.9 +35 Fruits and vegetables 5.9 5.7 −3 Cassava and products 1.1 8.2 +720 (http://www.naeb.gov.rw). Sources: NISR 2016a; NAEB website ­ Note: Coffee and tea volumes from the National Institute of Statistics of Rwanda are converted at net average prices reported by the National Agricultural Export Development ­Board. “Livestock and products” includes meat, milk, hides, and ­skins. “Cassava and products” includes ­flour. isolation; it is difficult to foresee certification • H ig h- end f loricult u re destined for being extended without vertical coordination Europe is necessarily very capital and with multinational ­ firms. knowledge ­ intensive. It is highly competi- Trends in global demand for higher-value, tive, because it is largely a saturated mar- certifiably organic, green, and high-­ quality ­ oint. It is quite dependent on ket at this p branded agricultural products, such as being able to keep air transport costs as essential oils, herbs, spices, medicinals, and low as possible, when countries such as natural insecticides, also offer ­ considerable Kenya have already achieved major econ- potential for value ­ addition. In smallholder omies of scale in this r ­ egard. Vegetable systems in Rwanda that face long, wide, and enterprises can target less-demanding anonymous value chains, vertical integra- and less-contested growing markets tion with value chain actors and collective in low- and middle-income countries, action will be necessary to gain the inter- including in the ­ region. However, food national market recognition necessary for safety certification and its credibility ­success. become issues if higher-value markets Horticulture and floriculture products for prepared vegetables, such as prepared offer significant opportunities for value foods for hotels and supermarket supply addition through diversification of e ­ xports. chains, are ­targeted. Rwanda has already made substantial pub- • The government should retain its credible lic investments in horticulture and flori- and supportive orientation toward pri- culture incubators, but exports remain vate sector investment to reduce the risks modest—while growing rapidly year-on-year involved in new i ­nvestments. Focusing on (2015–17). African countries’ experiences ­ securing additional investment in existing with horticulture and floriculture exports to brownfield developments is more likely European markets highlight several elements to succeed than undertaking new green- critical for success (box 5­ .2). Their experience field ­ investments. Yet supportive govern- offers some lessons for Rwanda’s potential to ment interventions need to be balanced develop horticulture and floriculture exports: with allowing an entrepreneurial private sector to determine its own investment • Rwanda is under pressure to establish a ­ opportunities. This includes facilitating globally competitive industry ­immediately. the import of skilled labor and key ­ inputs. Alternative markets that are growing Given the limited support services present rapidly—Asia and the Far East—may pro- ­ in Rwanda, operators will initially depend vide some opportunities to gain experience on expertise and inputs from elsewhere before competing directly with Ethiopia in the ­region. Rwanda will need to facili- and Kenya in highly competitive European tate this technology transfer, especially markets. But fundamentally Rwanda will ­ the smooth and efficient importation of need to board a moving ­train. planting material and the application of 228 FUTURE DRIVERS OF GROWTH IN RWANDA BOX 5.2  Horticulture and floriculture exports from East Africa The experience of horticulture and floriculture to secure economies of scale and to gain access exports from East Africa provides useful lessons to well-established and efficient destination sup- concerning the prerequisites for the success of this ply chains (marketing costs within the destina- industry in Rwanda (table ­ B 5.2.1 provides sum- tion country are typically more than half of retail mary data on these ­ examples). Microclimates need ­prices). to permit year-round production in order to distrib- Reasonably priced airfreight services (account- ute fixed costs across a longer production p ­ eriod.a ing for about 20 percent of the ­ U.K. retail price) are This augurs well for ­ Rwanda. essential and can be offered through scheduled pas- Large-scale production is essential to achieve senger flights for limited volumes or through cargo competitive production costs,b gain access to the nec- volumes.d flights for larger ­ essary expertise,c and negotiate prices for support- Required off-farm infrastructure includes qual- ing services (for example, associations may be able ity roads, reliable power (essential for factories to negotiate discounts for bulk purchases, especially and refrigerated storage facilities), water distribu- airfreight ­ costs). Securing adequate land for large- tion, and port and airport f ­ acilities.e Absent this, scale production can be a challenge; even where the costs of operator provisioning can raise overall governments have sought to facilitate “land banks,” investment costs substantially and strain cash flow, investors often have sought to secure their own ­land. because these are essential up-front ­ services. Most successful horticultural-floriculture indus- Standards and traceability are essential to suc- tries developed with significant foreign investment cess, particularly in more remunerative markets, for TABLE B5.2.1  Exports to Europe of flowers, vegetables, and fruits from select African countries, 2017 Indicator Kenya Ethiopia Tanzania Uganda Zambia Ghana Senegal Total value of exports 1,050 250 63 98 19 180 130 (US$, millions) Floriculture 750 248 26 57 17 — — Vegetables 150 2 12 25 2 40 95 Fruit 50 — 7 16 — 140 25 Total tonnage By air 248,000 90,000 3,200 17,000 8,000 5,000 — By sea 52,000 — — — — 140,000 125,000 Main products Cut roses, Cut roses Cut roses Cut roses Cut roses Bananas, Cherry green beans pineapples tomatoes, bobby beans Unit value (US$ per kilogram) Floriculture 3.95 2.76 8.13 3.80 2.36 — — Vegetables 2.59 — 0.80 1.25 2.50 1.14 0.92 Fruit 0.96 — 0.78 2.00 — 1.27 0.92 4,000 (roses); Land area (hectares) 1,600 (roses), — 75 (cut roses) — — — 12,000 250 (summer (vegetables) flowers) On-farm employment 150,200 55,000 2,750 10,450 2,250 17,200 24,400 2017. Source: Accord Associates ­ Note: — = not ­ available. (Box continues next page) T ransitioning A gricu l ture and F ood as an E ngine of G rowth 229 BOX 5.2  (continued) example, supermarkets in high-income c ­ ountries. conditions by one or two exporters (as recently hap- The many competing standards regimes, includ- pened in G ­ hana). ing industry standards such as GlobalGAP, ethical Government has typically played a supporting standards appealing directly to consumer prefer- role by providing political stability and an efficient ences (for example, Fair Trade), and the standards regulatory ­e nvironment. f Investors are typically of individual retailers (for example, Sainsbury and wary of greenfield schemes, so government’s ability Waitrose in the United Kingdom), can present a to offer turnaround opportunities can be ­ important. barrier for new e ­ ntrants. Not surprising, a large Given the long gestation for investments, tax holi- share of investment in African horticulture and days rarely matter as an investment ­ incentive. floriculture is backward integration of European Finally, because initial, high-value horticulture and Union growers, importers, and distributors who are floriculture investments often fail—a trend consistent fully aware of these standards and how to achieve with the outcomes of commercial agribusiness invest- t hem. Traceability is necessary to avoid penalizing ­ ments more generally—the ability to recycle sunk entire industries for the violation of phytosanitary investments rapidly to new investments is ­ critical. a. Ethiopia, Ghana, and Kenya benefit from stable weather conditions, whereas Zambia and Zimbabwe face greater seasonal variability and have trouble reducing unit ­costs. b. A competitive rose production unit requires 40–50 hectares initially, rising to 100 h­ ectares. An open-field flower or vegetable farm requires 500–1,000 ­ hectares. c. Larger companies can afford separate specialist positions in enterprise and agronomic management, but smaller operations need to combine the two with some expertise. Established industries like Kenya have a strong pool of talent that can be hired by new entrants; Zambia established a research and training farm that loss of ­ provides a regular flow of well-trained supervisors and ­ operatives. d. Tanzania (especially from Kilimanjaro, the closest airport to the production areas), Uganda, and Zambia struggled because of a limited number of passenger ­ flights. Uganda reached critical volume for dedicated cargo flights once its fresh fish export industry ­ developed. Ethiopia directed its national airline to carry all cut flower exports to Europe, possibly at subsidized rates (with rates at about U­ S$1.80 per kilogram ­ currently). Opportunities for back-loading can help to keep costs ­ low. Kenya continues to cut costs and improve competitiveness as scale increases and unit air transport costs decline; Kenya now exports ­ 0.25 million tons of high-value perishable products annually by ­ air. e. On-farm infrastructure is best managed by individual ­ operators. Kenya has well-maintained infrastructure around the Lake Naivasha r ­ egion. Ethiopia invested heavily in four horticulture hubs, although private sector uptake has been skewed with slower uptake of the more heavily promoted Bahr Dar ­ zone. Successful government- sponsored cold storage facilities include airport storage in Lusaka and Nairobi and the multipurpose refrigerated cold storage facility in Tema port in ­ Ghana. f. For example, stable political institutions encouraged Kenyan horticulture and floriculture, whereas instability plagued Ethiopian and Ugandan ­ industries. Ethiopia has been an unusual case in state-led development of its floriculture sector, whereas other countries have had equally supportive but less interventionist public ­ policy. agrochemicals purchased elsewhere in factors such as regional integration agree- Rwanda. East Africa for use in ­ ments, the existence of regional transbound- • RwandAir’s direct flights from Kigali to ary business networks, and possible concerns London are helpful if freight rates remain about overdependence on a small number of comparable to those paid by Kenyan neighbors for strategic food supplies come exporters to Air Kenya—reported to be play. Furthermore, food safety, sanitary, into ­ U S$1.80–US$2.50 per kilogram, which ­ and phytosanitary issues that are so com- already imposes a cost disadvantage for plex in exporting to the Organisation for competitors who pay higher ­ rates. Costs Economic Co-operation and Development need to fall without undermining the countries are typically less evident in regional financial sustainability of the ­ airline. exports. However, this situation will change ­ over time, especially for products sold Issues involving regional market competi- through supermarkets to higher-income con- tiveness are, in some respects, different from sumers in regional ­ capitals. Rwanda already global issues, in that the consumers, products, has a regional comparative advantage given and markets in question are closer in nature the credibility of its certification procedures ­ wanda. In addition, institutional to those in R and ­ i nstitutions. It has every interest in 230 FUTURE DRIVERS OF GROWTH IN RWANDA maintaining this advantage and extending growing rapidly in Rwanda and is likely to it to increasingly complex and valuable food increase rapidly in neighboring countries if ­products. rates. Table ­ their incomes rise at similar ­ 5.3 Rwanda’s growth ambitions are con- shows that Rwandan exports of Irish pota- tingent on expanding its reach in trade, toes and green beans rose dramatically from as discussed in chapter 2, particularly in 2003 to 2 ­ 013. By contrast, Rwanda would ­ a groprocessing. Ultimately, the goal of do better importing lower-value cereal crops pursuing regional trade opportunities, and bulk starches (for example, cassava) like global ones, is to be able to specialize from neighbors and exporting higher-value and profit from economies of s ­ cale. Thus, return. items in ­ Rwanda would like to export high-value Livestock products face constraints in both processed items like certified-safe baby regional and international trade, linked to food, horticulture, and packaged perishable the cross-border transmission of animal dis- animal products using the credibility of its eases such as foot-and-mouth disease and to certifying procedures and to import lower- the transmission of potentially serious patho- value commodities, such as maize, that ben- gens with transmissibility to humans (zoono- efit from large consolidated land areas and ses), such as avian ­ influenza. Traceability is mechanized commercial ­ farming. key both to controlling these diseases and to Regional agricultural trade is likely to creating confidence in ­purchasers. As in the include trade in starchy food staples as well provision of extension services and banking, as higher-value items like processed foods, electronic solutions also have high potential horticulture, and animal p ­ roducts. Rwanda for improving monitoring and surveillance to is likely to continue to have a strong com- prevent animal disease and facilitate certifica- parative advantage (domestic resource cost status. Rwanda’s export tion of disease-free ­ ratios well below unity) in several products, revenues from meat reached US$23 million such as Irish potatoes, horticulture, tea, in the year ending June 2017 (NAEB 2 ­ 017). coffee, essential oils, dairy products, and Botswana provides an example of how radio meat (table ­ 5.2). For example, milk sold by monitoring technologies that are both fea- Rwandan farmers at FRW 100 per liter in sible for small-scale producers and cost-­ Rwanda is sold by Kenyan farmers in the effective overcame the natural constraints Nairobi milkshed at the equivalent of FRW keeping a major African beef exporter out of 350 per ­ liter. Demand for these products is the European market (Moreki et ­ 2012). al. ­ TABLE 5.2  Indicators of comparative advantage for select crops in Rwanda, 2016 Crop Best-performing provincea DRC ratio Rank among 8 selected crops Irish potatoes Northern 0.19 1 Coffee Western 0.30 2 Beans (runner) Western 0.31 3 Soybeans Eastern 0.40 4 Maize Southern 0.45 5 Cassava Eastern 0.50 6 Wheat Northern 0.85 7 Rice Eastern 1.00 8 Source: Cambridge Resources International ­2017. Note: DRC = domestic resource cost ­ ratio. DRC < 1 indicates likely comparative ­ advantage. DRCs here are the ratio of the shadow value of domestic resources used to produce one unit of the commodity divided by the shadow value of tradable outputs from one ­ unit. DRCs less than 1 are normally thought to indicate comparative ­ advantage. Values shown are for the best-performing ­ province. a. These are the best cases across provinces for each crop; a comparable value is ­ 1.24 for wheat in 1.23 for maize in Northern Province; ­ Southern Province; and ­ 1.13 for rice in Western ­Province. Given higher numerical findings for cereals, regional variations in results, and inevitable questions about the myriad assumptions used to produce these kinds of indicators, caution should be exercised in attributing a favorable comparative advantage for any DRC > ­ ranking. 0.75. The main value is in the ­ T ransitioning A gricu l ture and F ood as an E ngine of G rowth 231 TABLE 5.3  Exports of Irish potatoes and green beans by Rwanda and select African competitors, 2003 and 2013 tons Country Product 2003 2013 Change (times) Ethiopia Irish potatoes 5,539 118,019 21.3 Green beans 1,712 5,879 3.4 South Africa Irish potatoes 28,107 117,891 4.2 Green beans 216 5,474 25.3 Rwanda Irish potatoes 84 8,066 96 Green beans 0 11,715 n.a. Kenya Irish potatoes 490 2,981 6.1 Green beans 27,193 32,081 1.2 Tanzania Irish potatoes 2,306 827 0.36 Green beans 82 4,912 59.9 colleagues. Source: Data from FAOStat, courtesy of International Finance Corporation ­ applicable. Note: ­n.a. = not ­ Understanding major structural changes and urbanization are already having major in nearby trade partners provides some impacts on the quantity and quality of food insight into Rwanda’s prospects in regional items ­demanded. For example, the demand ­ markets. Farm sizes in countries in Eastern for high-value items like animal proteins, and Southern Africa outside of Rwanda rose fruits, vegetables, and dining out increases following the cereal price increase in ­ 2008. sharply as incomes rise, whereas in Kigali, For example, surveys show that the total land income elasticities for unprocessed starches area in Tanzania operated by farms of less other than Irish potatoes have already than 5 hectares fell by 6 ­ .1 percent between turned (faintly) negative (table ­ 5.4). Results 2008 and 2012, but land operated in farms are broadly as expected from work in other greater than 5 hectares rose by the same countries with income levels similar to ­amount.24 Even more striking, 43 percent of Rwanda’s: research using household data by all farmland in Tanzania was formally pur- Michigan State University in five countries chased or rented in 2014 (Tanzania National of East and Southern Africa estimates that Panel Survey 2014/15), 25 suggesting that demand for processed foods in urban areas markets were active in allocating agricul- will increase by a factor of 8 over the next tural ­land. Similarly, the percentage of culti- three ­decades.28 Thus, demand for Rwandan vated land in farms larger than 5 hectares in agriculture and food products has strong Zambia was 52 percent higher in 2012 than prospects in Rwanda and in the r ­ egion. in ­2001. 26 Early results in both cases sug- This augurs well for value addition strate- gest that this primarily reflected increased gies based both on diversifying production investment by urban elites in farming, which patterns into higher-value commodities such if sustained could lead to increased capital as animal and horticultural products and investment and faster technological c ­ hange.27 on processing cereals and other starches These results have direct implications for into products more convenient to use and of Rwanda’s ability to compete in the produc- more consistent ­ quality. tion of field crops like cereals, for which there Income growth and urbanization are also are economies of scale in ­ production. driving changes in the quality of products A n analysis of producers’ opportu- required and in wholesale and retail mar- nities in the domestic market can pro- structures. Kigali has already begun to ket ­ vide some insight into the prospects for experience the “Supermarket ­ Revolution.” Rwanda’s regional ­ exports. Rapid changes Events in neighboring countries such as in demand associated with income growth Kenya, Tanzania, and Uganda suggest that 232 FUTURE DRIVERS OF GROWTH IN RWANDA TABLE 5.4  Response of demand to a 1 percent increase in household income in Rwanda, by degree of urbanization demand response (%) Rural Urban Product Remote rural Peri-urban Small towns Kigali Unprocessed cereals 0.71 0.61 −0.09 −0.03 Processed cereals and products 1.18 1.16 0.88 0.77 Irish potatoes 0.71 0.36 0.50 0.39 Other roots and tubers and plantains 0.41 0.52 0.09 −0.04 Legumes 0.54 0.61 0.31 0.31 Vegetables 0.73 0.75 0.76 0.67 Fruits 1.23 1.24 0.84 0.73 Meat, poultry, and eggs 1.71 1.36 1.46 1.13 Fish 0.90 1.32 1.72 0.78 Dairy products 1.94 1.41 1.38 1.35 Vegetable oil 1.00 1.07 0.92 0.66 Sugar products 1.46 1.36 0.91 0.57 Dining out 1.16 1.15 0.89 1.00 Other manufactured foods 1.24 0.21 1.80 0.81 Nonfood 1.25 1.24 1.08 1.11 Sources: International Food Policy Research Institute estimates using Integrated Household Living Conditions Survey household-level data for 2013–14 and the approach developed by King and Byerlee ­ (1978). See Diao et ­ (2017b). al. ­ Note: Own-produced food is included in expenditures with purchased food, at market ­ prices. Plantains (cooking bananas) are included with roots and tubers, rather than with mangos, sweet bananas, and other ­ fruits. this revolution will take off in Rwanda in trade in higher-value, processed, and certi- the next decade and spread widely to sec- fied commodities in which it has a com- ondary and even tertiary towns, even as it parative ­advantage. ­ continues to spread throughout East and • Rwanda also has considerable compara- Southern A­ frica. Supermarket procurement tive advantage in a few higher-value systems involve the consolidation of pur- staples suitable for industrial processing, chases, a shift to specialized wholesalers, such as potatoes and ­ dairy. It should be and tough quality and safety ­ standards. To able to develop these products to capture meet these requirements, producers need to much larger and fast-growing regional invest and adopt new p­ ractices. This is hard- ­markets. ­ est for small producers, who risk exclusion • W ith the explosion in agricultural trade in from dynamic urban markets increasingly the last 15 years in the region, there are dominated by ­ supermarkets. Smallholders grounds for broader recognition of the will need to address these difficulties value of free markets and the need for reli- through collective action (Jaffee, Henson, able export markets for staple foods in the and Diaz Rios 2011; Weatherspoon and prices. ­ face of volatile international ­ Reardon ­ 2 003). The rise of supermarket • For those commodities and processed procurement in Rwanda is likely to spur products in which Rwanda has a compara- specialization and regional trade, as super- tive advantage, capturing regional markets market chains source products in least-cost will involve addressing logistics, quantity ­countries. and quality management, joint ventures to The following are the key takeaways smooth cross-­ border trade, and regional regarding regional competitive advantage in ­agreements. exports of food-related products for Rwanda:­ These issues are relevant to most sectors • Given a small domestic market, Rwanda in Rwanda and are covered extensively in has every interest in developing regional report. chapter 2 of this ­ T ransitioning A gricu l ture and F ood as an E ngine of G rowth 233 Securing the Natural Resource Rainfall patterns have become more Base for Innovation variable, with projections of a 20 percent increase in variability by the 2050s (REMA Agriculture is the most natural resource– 2011). The prevalence of small-scale, rain- ­ dependent of all economic ­ sectors. It is thus fed farming that relies on traditional tech- very susceptible to fluctuations in rainfall nologies and practices renders the sector and temperature and is dependent on the especially vulnerable to rainfall ­ variability. quality of soils (structure, mineral content, Heavier rains increase floods and land- and health) and on the management of water slides, resulting in crop losses, health risks, that reaches those ­ soils. Agricultural out- and damage to infrastructure, particularly comes are particularly vulnerable to climate in the north and west of the ­ country. Poor change and to soil degradation, both multi- smallholders in Rwanda typically have little year conditions that can limit production in access to the wealth, liquid assets, credit, the medium and long t ­ erm. Adapting to the knowledge support, and infrastructure that first and halting the second are critical long- wealthier farmers elsewhere rely on when term ­priorities. adapting to climate or other ­ shocks. In sea- At the best of times, Rwanda experi- sons A and B in 2015 in Rwanda, between ences high year-to-year differences in rain- 1.1 and ­ ­ 1.5 percent of all agriculture opera- fall and is affected by El Niño–Southern tors practiced irrigation, compared with Oscillation events (El Niño and La N ­ iña). between 24 and 28 percent of large-scale Consequently, the country experiences peri- ­operators. 29 Moreover, the combination of odic floods and ­ d roughts. Rwanda has two rain-fed, small-scale agriculture, high rainfall annual rainy seasons (seasons A and B for levels, and steep hillsides also leads to very reporting purposes, with a drier season high rates of soil ­erosion. A higher variability between), and average annual rainfall is of rainfall implies more frequent droughts, 1,250 millimeters, a relatively high ­ amount. which limits the availability of water and However, there are large differences across feed for livestock, particularly in the east and the ­ c ountry. Average temperatures differ parts of the south, and increases vulnerabil- significantly between the cooler, mountain- diseases. Production losses to the dairy ity to ­ ous northern regions and the warmer, low- value chain are most significant in major lying southwestern valleys and drier eastern drought ­ years. Although subsistence farmers ­ fl atlands. Average rainfall declines signifi- are most affected, climate variability affects cantly as one moves from the western to the all agriculture sectors and lowers annual pro- eastern part of the ­ country. duction and e ­ xports. Increasing the resilience Since 1970, Rwanda has experienced a of Rwanda’s productive system, including to mean temperature rise of 1 ­ .4°C, higher than climatic risks, is essential for sustainable pro- the global average, with projections of a fur- duction and higher productivity as well as for ther 1°C to ­ 2 .5°C increase by 2050 (Agri- greater food and nutrition ­ security. TAF 2016; Future Climate for Africa 2014; The development and use of agricul- MINIRENA ­ 2 012). Higher temperatures tural technologies adapted to abiotic and have led to the spread of pests and diseases, biotic stresses would help to cope with these impairing the health of livestock and humans, ­difficulties. 30 Such technologies include the lowering crop yields, harming food security, development of early-yielding crop variet- and decreasing export e ­ arnings. It is likely ies and of crops that are more tolerant to that gradual shifts in agroclimatic zones will extremes of heat and water availability; the affect some crops as well as husbandry prac- promotion of higher nutritive quality fodder; tices for livestock, cumulating to significant the use of expanded rotation and multicrop- changes by 2050 in addition to any other ping for more sustainable and integrated pest influences that may be felt (Agri-TAF ­ 2016). and disease management and sustainable 234 FUTURE DRIVERS OF GROWTH IN RWANDA yield; the use of early warning systems for and peri-urban areas (REMA ­ 2 015). pest and disease management and the devel- Considerable progress will need to occur opment of seasonal schedules for sowing in global carbon offset markets and certifi- based on weather forecasts (for example, pro- cation procedures before they can become motion of earlier sowing for specific crops); a viable source of future finance for miti- and the promotion of climate-smart agri- gating agricultural GHGs in R ­ wanda. cultural technologies to increase crop and Fortunately, innovation is starting to pro- livestock productivity while maintaining the vide ways to improve productivity while natural resource ­base. also reducing GHG emissions; two exam- Agriculture accounted for an estimated ples ­follow. 46 percent of Rwanda’s total annual green- house gas (GHG) emissions in 2010 (CIAT Big data technologies are quickly coming 2015). That same year, roughly 84 percent ­ on stream to link large numbers of individ- of the agriculture sector’s GHG emissions ual sample points providing a wide variety were from livestock, split evenly between of data with huge databases and analytical enteric fermentation in ruminants such as capacity to provide data-based recommen- cattle and manure management or manure dations on remedial a­ ction. In climate-smart left on pastures (CIAT ­ 2015). Not surpris- agriculture, the precise analysis of needed ing, agriculture featured prominently in soil nutrients and the ability to measure Rwanda’s “national determined contri- sequestration of carbon in soil down to bution” to global GHG mitigation at the 30 or 40 centimeters provide a promis- December 2015 Conference of Parties in ing ­e xample. A cost-effective technology Paris (REMA ­ 2015). would save on inputs for climate smartness, Technical approaches would mitigate increase productivity and profit, and create agricultural GHG emissions in Rwanda, but the ability to monitor soil carbon sequestra- it is hard to see how they would be imple- tion on smallholders’ ­ fields. A new genera- mented when farmers have no way of cap- tion of handheld soil scanners provides an turing a return for the extra e ­ ffort. These easy-to-use way to diagnose soil fertility in techniques include the root zone and needed additives, based on cellular transmission of scanned data to • Increased efficiency of manure manage- a large central database and a 10-minute ment through the promotion of collective automated report ­ back. 31 In the smallholder cattle keeping; context, ownership of the equipment and • Reduction of the livestock population and training would likely best be done through progressive replacement of local breeds existing ­cooperatives. with more productive breeds to maintain Addressing land degradation, a major and increase productivity per unit of GHG environmental problem for agricultural pro- emission; ductivity in Rwanda, also offers considerable • Expansion of less-emission-intensive types hope for mitigation of GHG e ­ missions. 32 of animal-source foods (for example, fish- Land degradation affects all types of land eries and poultry); use systems in Rwanda and is an explosive • More judicious use of fertilizer based on and largely neglected problem for long-term crop nutrient requirements (such as micro- development in most of Sub-Saharan Africa, dosing and burying pellets); where more than 65 percent of arable land, • Large-scale tree planting for reforestation 30 percent of grazing land, and 20 percent to invest in a future carbon sink to cover of forests regionwide are degraded (FAO 30 percent of land with planted trees; and ­ 2 017). It is especially difficult to ignore • P romot ion of energ y for cook i ng this danger given Rwanda’s erosion-prone stoves other than firewood in urban steep slopes and strong seasonal r ­ ains. T ransitioning A gricu l ture and F ood as an E ngine of G rowth 235 Fortunately, the government of Rwanda has and the inclusion of output in higher-value made tremendous efforts to address these ­ arkets. Increased allocative efficiency m challenges ­comprehensively. During PSTA2 will be needed in the sense of targeting (2008–12), for example, the land area pro- resources to their highest-value r ­ eturns. tected by trenches and “progressive” ter- Irrigated bench terraces, in particular, will races (essentially narrow, traditional, sloped need to have much higher returns than terraces) increased by nearly 300,000 hect- cereals such as maize can ­provide. ares, a 60 percent increase (World Bank In the future, more sustainable farming 2 017). Other core programs implemented ­ systems in Rwanda will likely be charac- to control land degradation included control terized by expanded areas of wide, flat ter- of erosion through bench terraces, reforesta- races stabilized by “multipurpose” trees tion of the tops of slopes, irrigation, mecha- and grasses, nutrient replenishment based nization, and fertilizer ­ subsidies. on crop needs, reduction of nutrient leach- Although initial progress was made on ing off farms, and rainwater harvesting and the construction of bench terraces under expanded tree and forest ­planting. China is PSTA 2 and beyond, these terraces are a successful example of how to handle major costly to build and ­ u nderexploited. Using changes in degraded agricultural land on these terraces to maximum advantage densely populated steep slopes, as exempli- involves complexities of management for fied by the Loess Plateau projects profiled in which many farmers may not be ­ r eady. box ­5.3. Further, involvement of the private sector Key elements of interest for Rwanda may not be profitable under current man- include the halting of activities that led to dates to produce food s ­ taples. Ultimately, degradation; the construction of wide ter- the return on investment for these terraces, races, an expensive undertaking that required whether public or private, will depend on the planting of high-value crops to gener- the choice of crops or livestock products ate adequate returns; a shift from grazing BOX 5.3  Successful restoration of the productive landscape on China’s Loess Plateau Deforestation, degradation of grasslands, overgraz- slopes, cutting of trees, and free-range grazing of ing, and cultivation of marginal land led to huge goats—and to encourage biomass ­ r egeneration. soil erosion problems in China, reducing grain Land tenure responsibilities and benefits were production by an estimated ­ 5 .7 million metric clarified. Earth-moving equipment was brought in ­ tons a year in the late 1980s and increasing the to replace the farmers’ hand-dug terraces, which risks of floods and ­l andslides. One of the most crumbled each year, with more stable terraces three degraded areas was the Loess Plateau, a region of or four times as wide (6–12 ­ meters). Land that was about 640,000 square kilometers covering four unsuitable for grain production was planted with of China’s poorest interior provinces and parts of trees or shrubs instead or was allowed to grow wild Inner Mongolia (photo ­ B5.3.1). again, resulting in large-scale reforestation and The Chinese Ministry of Water Resources and grasslands ­regeneration. All livestock were confined the World Bank worked together to produce two to ­pens. watershed rehabilitation projects spanning 1994 to To ensure local buy-in and sustainability of 2005, mobilizing US$491 million between t ­ hem. the projects, farmer groups and county-level gov- The key goals were to halt the activities that led to ernment entities were fully engaged in decision land degradation—in particular, planting on steep ­ making and ­ i mplementation. At the request of (Box continues next page) 236 FUTURE DRIVERS OF GROWTH IN RWANDA BOX 5.3  (continued) farm groups, loans and advice were mobilized to losses were reduced by 60 million–100 million introduce smallholder dairy and Kashmiri sheep, tons per ­year. Soil carbon storage also increased, new to the ­region. The World Bank estimates that mostly because of the restoration of forests and the projects lifted more than ­ 2 .5 million people out ­ g rasslands. The overall economic rate of return of of poverty and boosted incomes from about US$70 the two projects was estimated at 20 percent, but to about US$200 per person per year through the livestock components had an economic rate of higher farm productivity and ­ d iversification. Per ­ easure. return of 27, which is very high for this m capita grain output rose from 365 to 591 kilo- Moreover, the approaches developed through the grams per year, and the employment rate increased project have been applied more broadly across the from 70 to 87 ­ percent. Water retention increased, Loess Plateau—where, as of 2008, more than half making farms more resilient to ­ ­ d rought. Soil ero- the degraded area had been restored—and in the sion was curbed on 920,000 hectares, and soil Yangtze and Pearl River ­ basins. PHOTO B5.3.1  China’s Loess Plateau, 1990 and 2012 Credits: Left: Till Niermann, GNU free documentation license v1.2 (1990); right: Erick Fernandes (2012). al. 2014; Global Commission on the Economy and Climate 2014; Huang, Gallichand, and Zhang 2003; Xie et a Sources: Dang et ­ 2007a. ­ l. 2010. See also World Bank ­ for ruminants to cut-and-carry and fodder Policies to Achieve Rapid production systems; and the importance of and Sustainable Growth in active involvement by farmer groups and local ­government.33 Agriculture Rwanda should consider improvements Over the next 20 years, Rwanda’s agricul- in the allocation of scarce water supplies ture sector will aim to increase productiv- to agriculture. User fees should not be off ity while adapting to the ­f uture. State-led the table, especially for larger ventures; it efforts to increase agricultural productiv- will become harder, but even more impor- ity and maximize the sector’s potential tant, to tackle this issue as time ­ p asses. in Rwanda will continue in the medium Israeli experience illustrates how the alloca- term, serving a complementary role to more tive efficiency of agricultural water under ­ market-oriented agriculture over ­time. State very difficult conditions can be addressed efforts involve extending the approach to sustainably through a combination of pub- consolidation of production decisions and lic leadership, transparency, technology, and staple food crop–oriented intensification market allocation procedures (box ­ 5.4). that has already reached at least 20 p­ ercent T ransitioning A gricu l ture and F ood as an E ngine of G rowth 237 BOX 5.4  Israeli success at being more efficient with agricultural water Israel is a diversified, small cou ntr y with a •• Effective enforcement of a data-based and trans- h ig h ly produc tive ex por t ag ricu lt u re sec tor parent system of water allocation immune from under irrigation, despite being one of the most political intervention is ­critical. water-stressed countries in the world and facing •• Investing massively in new water infrastructure steadily increasing demands for water withdraw- needs to be done in ways that promote fiscal neu- als outside a­ griculture. Although many of the trality in the water ­ sector. These ways include structural water issues faced by Israel are quite both public-private partnerships and a clear sepa- different than those faced by Rwanda, Israel’s ration of roles among policy setting, regulation, experience offers lessons of direct relevance for planning, and operation of ­ infrastructure. Rwanda: •• Wastewater recycling is very expensive and requires high ongoing subsidies for treatment and storage; it •• Water needs to be recognized and treated as the is likely only viable in very high-return ­cases. scarce resource it i ­s. In Israel, large public aware- •• Water use efficiency can be promoted most cost- ness campaigns were carried out to emphasize effectively in the case of high-value crops, such the value of water, accompanied by the pricing of as fruits and vegetables, including on slopes, water at its actual ­cost. through drip irrigation ­ technology. al. ­ Source: Marin et ­ 2017. (level of cooperative membership) and at only be achieved by reaping the benefits of most 40 percent (extent of contact with the scale economies and specialization through CIP and LUCP) of f ­arms. Public support exports and the production of higher-value- should continue to improve the efficiency added g­ oods. Such efforts will require a rapid of small farms that are still not using fer- response to market signals, ready access to tilizer in appreciable amounts, lack access investment resources, technical expertise, to adequate agricultural water, are situated and ability to organize production and pro- on unprotected hillside slopes, or use infe- vide appropriate incentives for workers, gen- rior seeds, so that such farms can achieve erated by the private s­ ector. The state needs to the technical efficiency of Rwanda’s better retain a lead role as umpire, generator of pub- ­farmers. lic knowledge, and provider of public goods To produce the transformative growth such as infrastructure and basic r ­ esearch. rates sought by Rwanda, both public and Many key innovation public goods for small- private sector expertise in agriculture must holders and for basic food staples will likely be ­ leveraged. It is estimated that continued need to be provided by public actors for the state-led efforts would extend the 5 percent foreseeable ­future. annual agricultural growth experienced in Achieving higher growth rates in the recent ­ years. Moreover, these efforts soon future will require responding to major will approach various limits, including the changes in demand (locally, regionally, and mounting fiscal burden, rising unit costs as globally); gains from specialization, econo- less accessible agricultural land is exploited, mies of scale, and a larger division of labor competition from neighboring countries through trade; changes in industrial organi- in crops promoted by the CIP, and shifting zation (how farmers are linked with markets internal and external demand for animal and with each other and who decides what) proteins, fruits, vegetables, and highly pro- in order to take advantage of opportunities; cessed ­ foods. Higher agricultural growth can and changes in the contact that farmers have 238 FUTURE DRIVERS OF GROWTH IN RWANDA with infrastructure, information, and tech- private sector ­ aggregators. Beyond the tea nology and the sophistication of those tech- and coffee cases discussed earlier, aggrega- nologies, leading to higher productivity and tors have shown interest in partnering with ­sustainability. producers and government in animal feed, The demands faced by Rwandan farm- avocados, bananas, beans, dairy, eggs, essen- ers are changing rapidly, as demand grows tial oils, fish, flowers, various forms of fruit for higher-value products, food safety, con- and vegetable processing, meat processing, venience, and qualities such as Fair Trade cereal milling, and s ­ ugar.34 These firms often and organic certification (in global ­ markets). bring managerial skills, capital, extension of Major shifts in demand are occurring in knowledge to farmers, and entree into com- domestic markets, but even more so in mercial networks outside ­ Rwanda. They regional markets, including all of Africa, as have the expertise to achieve global certifica- urbanizing and increasingly richer people tions in desirable ­ t raits. Government could are eating more animal products, more fruits play a role in ensuring the enforcement of and vegetables, and especially more highly agreements—especially regarding recouping processed food items that better fit urban input credits that companies give to farmers, lifestyles. The large cities of Africa are the ­ securing adequate supplies from zones where contested markets of rapid growth in demand extension investments have been made, and for high-value items for which agricultural ensuring the traceability of supplies of raw interests all over the world are ­competing. materials, allowing farmers to garner pre- Rwanda should ramp up its efforts to miums for q ­ uality. In many cases, multina- promote trust in regional trade in Rwandan tionals seek the active financial involvement food products, including through continued of the government, which can be highly ben- engagement in the East African Community eficial for the promotion of high-value spe- and Common Market for Eastern and cialized ­ crops. Government should clarify Southern Africa agricultural p ­ rocesses. goals and define satisfactory performance as Rwanda has a major stake in increasing its well as provide an institutional structure to imports of raw materials such as cereals follow up transparently at regular intervals for its food-processing and animal protein and redeploy public resources where per- industries, while increasing its exports of formance has fallen ­ short. In the Republic processed foods and high-value calories— of Korea, for example, this system is highly for example, from milk and meat ­ products. developed for public-private partnerships for Ways should be explored to exploit the huge ­infrastructure. 35 Government financial par- market potential of neighboring countries ticipation should likely be limited to a very by increasing food production, for example, small number of activities that have a com- through the provision of financial guarantees pelling need for public ­ participation. under a regional ­ process. Another initiative Support for aggregator activities should might include a more common approach to reflect different requirements for different certification of food safety, sanitary, and phy- ­commodities. 36 In milk, for example, food tosanitary ­ conditions. Regional integration safety is paramount, asymmetries of informa- also should be pursued through ­ continuing tion are large (especially when dilution with efforts to adopt transparent science-based unclean water is easy and many small batches standards, registration of products, and cer- are mixed together), and market power tification of agricultural ­ i nputs. Rwandan changes radically from producer to buyer manufacturing has interests here and should depending on the s ­ eason. For these reasons, be ­engaged. dairy cooperatives work well, despite being Government could support vertical coor- high ­cost. Coffee requires a different form of dination arrangements by expanding par- industrial organization to ensure quality and ticipation in cooperatives and encouraging reliability and to achieve a verifiable process T ransitioning A gricu l ture and F ood as an E ngine of G rowth 239 of production (such as organic or Fair T ­ rade). managers and technical experts to fill ­ gaps. Typically, best results are achieved when Government should consider developing roasters can assist small farms or coopera- programs to increase the prevalence of these tives directly with inputs and expertise under skills ­domestically. Companies might be contract farming agreements and can be sure involved through instituting training pro- of getting the resulting produce for their own grams to build up nationally needed skill sets operations, which may require state zoning in agribusiness, high-value supply chains, and of washing and m ­ arketing. In horticulture, agricultural ­ technology. Cargill and Nestlé prices achieved will depend not only on the have been pioneers in Africa and beyond, right sorting of produce for retail delivery providing technical training for small farm of a perishable at the right time, but also suppliers, business training for cooperative on the absence of residue on items coming leaders, and managerial and higher technical from many farms, and so f ­orth. Outgrower training for their own national staff (Cargill schemes—where a large farm or agribusiness 2014; Nestlé ­ 2016). Where government finan- that is producing a commodity according to cial participation is sought in a joint venture, international norms provides management, such training might be a reasonable point to expertise, inputs, and marketing services to ­consider. small farmers in the same locality to produce One aspect of promoting increased reli- the same commodity on their own land— ance on markets postharvest involves the tend to work well here, particularly to con- development of more efficient wholesale trol and store chemicals that are toxic or give m arkets. Wholesale markets are curiously ­ rise to ­ residue. scarce in Rwanda, perhaps because of the Government should seek to develop a small size of the urban ­ p opulation. The knowledge platform regarding which forms bulking, grading, and transport functions of industrial organization work best for of value chains appear to be carried out addressing specific industrial organization either at the source (a cooperative or large problems in Rwanda’s smallholder ­ context. enterprise) or by the retail firm (a supermar- The platform could provide reliable data on ket, dairy store, or e ­ xporter). As special- vertical coordination, finance, prices, costs, ization, scale, and trade grow, the need for and weather; this information will be criti- independent intermediaries will likely grow cal to reducing uncertainties and risks for well. How this function develops and is as ­ investors and should encourage ­ competition. supported by regulators will be important Documenting and publicizing the impact of success. A first step was the establishment to ­ different private, public, and public-private in 2016 of a new wholesale potato market experiences with vertical coordination can in Kigali, 37 and this trend is expected to help to share experiences across ­ commodities. accelerate. South African experience yields ­ Doing so also will facilitate a key policy role of two important insights: first, wholesale government: to promote a level playing field of markets tend to develop first for items like knowledge and regulation that encourages for- potatoes, onions, fruits, and vegetables, as mal sector private investment while advancing urban demand ­ spikes. They facilitate reli- the interests of the farming ­population. able supplies for urban retailers and financ- Government policy should encourage ing, quality control, and price predictability skills development by private firms to com- for smallholder ­ suppliers. Second, over time plement public sector extension ­ activities. with the growth of supermarket and chain Vertical integration with aggregators is a key store retail outlets and growth in the size private sector method of transmitting skills of farm suppliers, direct retail-farmer ver- to smallholders for specialty crops destined tical coordination becomes more common for export; these firms in Rwanda generally (Seidler 2001; Vermeulen, Kirsten, and have brought in regional and non-African Sartorius ­2008). 240 FUTURE DRIVERS OF GROWTH IN RWANDA Improving SMEs’ access to finance is registry represents a major advance for both important for the formal private sector to efficiency and good governance under small- play a bigger role in agricultural decision holder ­conditions. Land registration is just a ­ making. Four areas of action could serve to beginning, however, and similar partnerships foster this process, while strengthening the will be vital to increasing financial inclusion role of SMEs in agriculture:38 and investment into relatively small farms and cooperatives of small ­ farms. Possible • First, strengthen the enabling environment applications might involve cutting the costs of for agricultural finance by improving the registering small mortgages on land and lev- quality, granularity, and accessibility of smallholders. eling the playing field a bit for ­ relevant data, including data on firms, Public investment in infrastructure such agricultural finance, and w ­ eather. Also as roads, electricity, and water is criti- needed are efforts to reduce the high cost cal to encourage private sector entry into of registering mortgages on land for small agribusiness. Reliable electricity is key for ­ ­loans. cold storage and processing, which are vital • Second, increase the financial inclusion of for the safe handling of horticulture, flori- agricultural SMEs and commercial farm- culture, and animal ­ products. It is striking ers by increasing the business transparency that agroprocessing, primarily of horticul- of cooperatives and commodity federa- ture, accounted for 55 percent of all man- tions, by promoting electronic payments ufacturing value added in Tanzania in for settlements and government transfers, 2015—­ horticulture used raw material from and by creating a public (electronic) direc- 450,000 farms from all over the country and tory of ­ enterprises.39 sold products all over the country—yet the • Third, deepen the agricultural financ- 287 firms involved were almost all located ing market by strengthening the opera- in Dar-es-Salaam (Kumar and Agarwal tional capacity of banks, microfinance ­ 2016). As shown in other chapters of this institutions, and SACCOs in agricultural report, Rwanda has made tremendous prog- finance to provide financing packages that ress in extending the reach of the electrical meet the needs of small-scale and com- grid and providing piped water to second- mercial ­ farmers. In particular, the credits ary and tertiary ­ towns. However, unit costs and matching grants of the BRD play a in this regard will remain key parameters in big role in shaping the formal allocation achieving competitiveness relative to other sector. BRD’s of capital to the agriculture ­ countries in the region, and agricultural agricultural capacity can be strength- development based on high-value perishables, ened. BRD loans might also transit from as opposed to starchy staples, will require retail to ­wholesale—for example, funding infrastructure that permits rapid process- SACCOs that would then fund the retail ing and ­ refrigeration after ­harvest. It hardly ­end. seems plausible that the least-cost option for • Fourth, scale up formal agricultural fast processing will always be truck trans- insurance. While a public-private partner- ­ port to the capital ­ city. Yet decentralization ship model will likely be needed, a formal of infrastructure into agroparks that com- assessment of options is necessary to target bine reliable electricity, water, and various funding effectively for capacity ­building. agricultural service industries is not without ­ dangers. A key lesson from experience with Rwanda’s agrarian structure based on agroparks oriented to high-value agribusiness smallholders reinforces the critical impor- in Latin America is that the choice of loca- tance for agriculture of innovations in tions needs to be based solely on commercial finance such as blockchain t ­echnology. reasons (FAO ­ 2010). Rwanda’s public-private partnership in Irrigation is clearly a constraint, especially applying blockchain technology to its land in the drier eastern regions of the country T ransitioning A gricu l ture and F ood as an E ngine of G rowth 241 (Malesu et a 2010). Less than 5 percent of ­ l. ­ staple food crops under the ­ C IP. However, all land used for production and less than only ­ 0 .7 percent of public expenditures 3 percent of cropped land (that is, agricul- on agriculture were allocated to research tural land not including pasture or forest) is and innovation in fiscal 2014/15–2015/16, irrigated at the present time, estimated to be whereas 64 ­ percent were allocated to crop equivalent to ­ 7.5 percent of irrigable potential intensification, ­ i ncluding input subsidies (MINAGRI ­ 2017c). But more than half of the (Policy Associates ­ 2017). Strategic training of land deemed irrigable is on hillsides, where scientists and development and maintenance irrigation costs can run very high (World of research infrastructure to target high-­ Bank ­ 2014). A more useful estimate might be technology research are ­ needed. Given an that 16 percent of potentially irrigable marsh- increasing international shift to private (and lands have been developed for ­ i rrigation.40 proprietary) funding of agricultural research In any event, there is potential for much and likely continued budgetary pressures greater investment in the irrigation of marsh- in Rwanda, the share of privately funded lands at US$6,000–US$8,000 per hectare research will need to increase in order to have (World Bank ­ 2014). Current yields of rice are an overall increase in funding and to ensure not sufficient to cover the investment c ­ ost. that Rwanda remains engaged in the premier Therefore, while emphasizing the increase in international research c ­ onsortiums. Private rice productivity, other high-value crops like research will likely be focused on high-value horticulture also should be considered in irri- crops or other items that provide major private gated ­areas. Horticulture for export would sector opportunities for ­ profitability. Public likely be able to cover these cost levels more research should also become more demand easily, including exports to the rapidly grow- driven. This can be accomplished by estab- ­ ing market in the Democratic Republic of lishing innovation platforms with the private Congo and possibly farther west by ­ air.41 The sector as a major player and demander or take-away lesson is that linking irrigation sponsor and interacting with leading research expansion plans with market development centers to develop necessary ­ expertise. is necessary for higher-value products like The big-data revolution can also cost-­ horticulture and dairy, which, especially the effectively promote collective action with market for horticulture products, are most advice tailored to the individual conditions likely to be regional ­markets. of small operations at low unit ­ costs. The Rwanda’s strong public institutions and application of “climate-smart” agricultural high level of governance provide important practices to a large number of small farms advantages in supporting rapid growth in on steep hillsides will be greatly facilitated ­ a griculture. The ability to provide cred- by fast-moving developments that allow even ible product certification in the organic, Fair small farms to interact with major regional Trade, and ISO 22000 areas needs to be pro- and global databases in order to access knowl- tected, particularly as the demand for certi- problems. This is edge to solve their particular ­ fication ­escalates. The Rwanda Agricultural likely to be one of the fastest and most prom- Board and its staff should be encouraged to ising areas of innovation in the sector, deliv- reach out to public and private colleagues in ering very high-tech solutions to users with neighboring countries and elsewhere in the only modest information technology s ­ kills. world to build technical links for resolving It will require active participation and exper- common problems through research, analy- tise from researchers and extension agents sis, and ­extension. at public organizations such as the Rwanda Government support for agricultural Agriculture Board and can benefit from the research and extension is essential to boost existing cooperative ­ organizations. It also has productivity growth in the ­ s ector. The great potential to lower the costs of private Rwanda Agricultural Board already invests investors to work with large numbers of dis- heavily in providing extension services for persed small ­farms. 242 FUTURE DRIVERS OF GROWTH IN RWANDA Another promising reform related to big and rural ­ a reas. The Farmer Field School data is the introduction of electronic deliv- approach, which involves farmers meeting ery of extension services (e-extension) to in local field settings with a trained facilita- create an electronic and interactive bridge tor to discuss common problems, should be network where farmers and other stakehold- ­expanded. ers meet and ­ transact. The use of high-tech Adapting to climate change, stopping land solutions for delivering more efficient exten- degradation, and managing groundwater sion messages (such as the use of drones remain a public ­ priority. Most agriculture in in crop surveys for assessing crop growth, Rwanda is practiced on steep slopes prone pests, and disease management) needs to to erosion during the rainy ­ season. While be ­encouraged. In an upper-middle-income considerable progress has been made in con- Rwanda, the traditional extension and structing robust “bench” (wide) terraces, advisory services will need to focus on much more needs to be done to secure the land farmer-promoter knowledge ­ development. asset. In conjunction with the private sector, ­ A continuous training program and e-​ investments will be required in infrastruc- f eedback system will also be needed to ­ ture, other public services, and processing deliver extension messages on existing chal- to support the restoration and conservation lenges, seed availability and sowing dates, of productive ­ landscapes. Whole-landscape rain forecasts, fertilizer rates, and so ­ forth. approaches emphasizing soil, water, and A strong and well-organized public exten- biomass management are critical to achiev- sion system will be needed to contribute to ing the triple wins of productive growth in messages and adapt as needed, especially agriculture, increased adaptation to climate in food staples of lesser interest to the pri- shocks, and mitigation of GHG emissions vate ­sector. This system could be achieved through carbon ­ sequestration. How livestock through delivering farmer-promoter knowl- is handled in hillside development is key to edge (exemplary farmers elected by the com- success. Confinement of stock with cut-and- ­ munity to facilitate the delivery of extension carry fodder systems will likely be ­ needed. messages at the village level); increasing Implementation of landscape approaches access to training and advisory services deliv- requires associating high agricultural sci- ered by the private sector; and enabling access ence expertise with farmer consultation and to improved extension service delivery by follow-up ­ activities. Local government capac- expanding feeder roads, in both peri-urban ity is ­ key. Annex 5A  A Graphic Presentation of the TFP Story in Rwandan Agriculture Rwanda has been highly successful at ­ rea. This was also the time of the begin- a maintaining an impressive rate of agricul- nings of TFP g ­ rowth. Actions undertaken tural growth, on the order of 4–5 percent by the government under the CIP had or more per year since the early 1960s, major impacts on the growth of output in with the exception of the “lost decade” staples through improved seeds and fer- of the 1980s (in com mon with many s ubsidies. Land development and tilizer ­ other African countries) and the 1 ­ 990s. improvement under the LUCP implemented However, success in boosting agricul- through CIP also contributed, including tural growth in earlier years was attribut- bench terraces and development of new irri- able largely to land expansion, as seen in ­ arshlands. Heavy emphasis gated lands in m figure ­ ­ 5.1. Success in the 2000s was largely on extension throughout undoubtedly pro- due to increased use of inputs, especially moted increased technical efficiency and from 2007 onward, and some expansion of thus ­T FP. Measured TFP growth took off T ransitioning A gricu l ture and F ood as an E ngine of G rowth 243 in the second half of the 2000s and contin- FIGURE 5A.1  Heuristic diagram of the ued through at least ­ 2014. production possibility frontier TFP is the gain in output that results from some combination of at least one of three Nonstaples improvements. The first is improved techni- ­ L2 cal efficiency, which comes from redeploying existing inputs, land, and labor regardless PPF of prices in a way that leads to net physical d gains using existing ­ technologies. The second is allocative efficiency, which considers the costs of using different inputs and factors in addition to technical issues, in order to maxi- c mize ­ profitability. The third is technological a change, which embodies scientific and tech- b L1 nical innovation to get more from ­ less. In economic parlance, technical efficiency Food staples is moving production to the production pos- sibility frontier (PPF), allocative efficiency is Note: The production possibility frontier (PPF) represents the moving along the frontier to the most prof- maximum level of productivity for a given state of technology. a, b, c, and d show Rwanda’s different possible positions with “a” as the itable point given relative prices of differ- starting point. L1 and L2 show different scenarios of relative prices ent inputs and outputs, and ­ technological of staples versus nonstaples. change is pushing the frontier ­ o utward. In smallholder farming systems such as those shown as moving from point a to point b in Rwanda, boosting technical efficiency is closer to the frontier defined by good agri- largely a matter of extension services, typi- cultural ­practice. Here Rwanda is producing cally public; raising allocative effciency is both more staples and more nonstaples with often a combination of policy, private sector the same set of resources, largely because of involvement, and functioning markets; and improved extension and input supply s ­ ystems. technological change is largely the product Continuing in this vein leads to point c on the of successful agricultural research and inno- frontier, in effect catching up with the most vation systems, in partnership with other technically efficient neighbors and favoring research bodies inside and outside the coun- staple ­foods. try and sometimes with the private sector, in However, although point c is technically the case of high-value specialty ­ items. efficient, it is only economically efficient if This complex set of relationships in TFP the relative prices of staples and nonstaples relevant to Rwandan agriculture can be in Rwanda are portrayed adequately by the illustrated simply in the heuristic diagram in slope of line L­ 1. If instead nonstaples become figure ­ ­ 5A.1. Consider a starting point from a more valuable relative to staples than implied base of low agricultural productivity follow- by L1 (that is, you get more staples for a given ing the events of 1994 (point ­ a). Agricultural amount of nonstaples), the price ratio may productivity in the late 1990s was lower in now be portrayed by line L ­ 2. In that case, Rwanda than in regional comparators, whose policies that keep the price ratio artificially situation is summed up in a theoretical ­ PPF. at L1 will succeed in producing more staples PPF is the maximum level of productivity for than otherwise but will have a cost in terms a given state of technology; as drawn, it sym- of lost economic (or allocative) ­ efficiency. In bolizes the trade-off in resource use between theory, producing at point d in a market sys- producing staple food crops and n ­ onstaples. tem would allow a trading country to have Concerted efforts by the government of more of both staples and nonstaples under Rwanda—not least the CIP and the LUCP— price regime L2 than producing at point ­ c. successfully increased technical efficiency, Under a true free-market system, aggregate 244 FUTURE DRIVERS OF GROWTH IN RWANDA production would move from point c to point Notes d, which would maximize the value of agri- 1. The growth rates and subsectoral decom- cultural output at price ratio L ­ 2. Allocative position are from Diao et ­ a l. (2017a), using efficiency is largely the result of improved MINECOFIN ­data. policies and better-functioning ­ markets. 2. This effect can produce growth multipli- The best solution is for true technologi- ers in many parts of Africa on the order of cal change to ­ occur. This implies that farm- 2 or more (meaning that every unit of agri- ers literally produce more with less than cultural growth induces at least an extra they could on the old ­ PPF. Technological unit of nonagricultural growth) (Delgado change would be represented by a move- et ­a l. 1998; Diao et ­ a l. 2017a; World Bank ment of the PPF outward and away from 2007b). Such growth in farm incomes needs ­ the o­ rigin. Technological change is made to originate from outside the local area if it is to have a true effect inside the local area (and possible by successful agricultural research not just redistribute purely local income); an and innovation ­ systems. Unlike efficiency, example is income from export crops or even which is ­ limited in scope once basic changes from crops and livestock sold elsewhere in are made, technological change can power ­Rwanda. productivity growth i ­ndefinitely. The sum 3. As local economies become more integrated of growth in technical efficiency, allocative with larger economic spaces, any addi- efficiency, and technological change is total tional local consumption from new income factor ­productivity. The technical efficiency sources tends to be met either by increases in component is where Rwanda has become imports from outside the locality in question quite successful since 2005, but where the or by decreases in exports from the locality components of allocative efficiency and in question, leading to no net gain in local production. See Delgado et a ­ ­ l. (1998) for a technological change will increasingly need fuller ­description. to lead longer ­ term. 4. This point was made on several occasions by Most of these forms of agricultural pro- Rwandan government officials interviewed in ductivity growth since 1999, which stressed September and December 2017; it is consis- expanding cropped area, improving land, tent with experience in several other African using more purchased inputs, and more countries at the time and with the evolution of recently improving technical efficiency, have Rwandan policy interventions after 2 ­ 008. The worked well, but may have decreasing repli- figure of 60 percent is from EICV2, which cability in the future as more land is brought pertains to 2005/06 (NISR, various y ­ ears). under existing ­ programs. Longer-term strat- EICV4, which pertains to 2013/14, yields an average of 50 percent (Diao et ­ al. 2017b; egies for addressing continued agricultural NISR, various ­ years). productivity growth under Rwandan con- 5. Only one-quarter of Rwandan farms are ditions will require continual processes for larger than ­ 0 .7 hectare in total, and in boosting TFP in ­ agriculture. Past success the 10 most densely settled rural districts, in moving a large share (20–40 percent) of 40–50 percent of farms are ­ 0.2 hectare or less the smallholder population closer to the (MINAGRI ­2017a). PPF defined by existing technologies and 6. More formally the Organic Law ­ no. 08/2005 resources has done the country great good of 14/7/2005, “Determining the Use and and demonstrably boosted national growth Management of Land in ­ Rwanda.” See overall. However, expanding agricultural ­ Pritchard ­(2013). output at the rate required to achieve upper- 7. Just under 30 percent of farm households participated in land consolidation accord- middle-income status by 2035 will require ing to the most recent household survey the more rapid growth in TFP that can only (EICV 2013/14); if progress has continued be achieved by moving the PPF outward, at the same rate since then (7 percent a year which is where innovation systems, institu- since the 2010/11 survey), some 39 percent of tional development, infrastructure, and the farm households would have been affected by private sector will need to come i ­n. 2017/18 (NISR, various ­ years). T ransitioning A gricu l ture and F ood as an E ngine of G rowth 245 8. Through the highly successful Land Tenure on farmers and factories were completed in Regularization Program prepared starting in 2017 and are f ­orthcoming. Although the 2005 (DfID ­ 2015). farmers’ share of bought-leaf costs rose, 9. The new cooperative strategy of the govern- bought-leaf costs remained a constant share ment is built around addressing these and of overall ­ costs. ­ other issues that have arisen during the mete- 20. With a low of U ­ S$0.40 per pound in fall 2001, oric expansion of primary cooperatives in a high of ­ US$2.80 per pound in spring 2011, Rwanda, which have grown from less than and a price of U ­ S$1.25 per pound in January 1,000 in 2005 to more than 8,000 presently ­2018. See ­http://www.macrotrends.net/2535​ (MINICOM ­2018). /­coffee-prices-historical-chart-data. 10. It is likely that the rise in food staple prices 21. Professor Ameet Morjaria (Kellogg School in Africa in 2008, which endured at least of Management, Northwestern University), through 2012–13, helped to improve alloca- ­personal ­communication. tive efficiency in staples production and 22. Professor Ameet Morjaria (Kellogg School boosted the value of output growth inde- of Management, Northwestern University), pendent of increased use of inputs or land ­personal ­communication. ­expansion. 23. For details of many certification experi- 11. Nineteen percent in season A and 15 percent ences, see ­https://www.fairtradecertified.org​ in season B (NISR ­ 2017). /­why-fair-trade/our-impact. 12. In 2016, subsidized fertilizers were available 24. Professor Thomas Jayne (Michigan State to smallholder producers of maize, wheat, University), personal communication, using soybean, rice, beans, cassava, Irish potatoes, Living Standards Measurement Survey ­ data. ­ and ­vegetables. 25. See ­http://microdata.worldbank.org/index​ 13. The reasons for these low payments are not .php/catalog/2862. ­ clear. The prices reported here were observed 26. See ­http://microdata.worldbank.org/index​ informally from a small sample during a field .php/catalog /2862 , but using Zambian trip in September ­ 2017. More research in this Ministry of Agriculture and Livestock Crop area is ­warranted. ­ Forecast ­Surveys. 14. The fiscal treatment of insurance discourages 27. See ­http://microdata.worldbank.org/index​ financial institutions and borrowers from .php/catalog/2862. seeking to expand agricultural i ­nsurance. 28. Professor David Tschirley (Michigan State Although agricultural insurance premiums University), personal ­ communication. Also have been exempted specifically from the see Tschirley et ­ al. (2015) for ­ background. value added tax of 18 percent, a 15 percent 29. Season A starts in September and ends in reinsurance withholding tax ­ remains. This February of the following year, season B makes insurance expensive for farmers and starts in March and ends in June of the same limits the revenue generated by insurance year, and season C starts in July and ends in companies, without making a substantial September of the same ­ year. The source is the contribution to the government’s tax revenues 2015 Seasonal Agricultural Surveys (NISR (World Bank ­ 2016). ­2016b). 15. For example, see ­ https://www.lexology.com​ 30. Biotic stresses include disturbances from /­l ibrary/detail.aspx?g=a57dec69-283d-4bb0​ ­ living organisms such as pests, bacteria, and -a684-61a798fbd6c4. ­ f ungi. Abiotic stress arises from factors such 16. See ­http://www.osborneclarke.com/insights​ as heat, sunlight, and ­ wind. /blockchain-and-land-registries-records-of​ 31. For a current example in the field, see ­ http:// -the-future/. www.soilcares.com/en/products/scanner/. 17. S e e ­h t t p s : / / w w w.w i s e k e y. c o m / p r e s s​ 32. Addressing agricultural land (that is, soil) /­w isekey​-a nd-m icrosof t- collaborate -to​ and forest degradation (loss of biomass) -support​-­r wandan​- government-make-secure​ jointly through conservation and restora- -transactions-using-blockchain-technology/. tion of productive landscapes is now seen 18. See the ­ U.S. Food Marketing Institute, a trade as central to agricultural adaptation to and group, at ­ https://www.fmi.org/our-research​ mitigation of climate change in the tropics /­supermarket-facts. and indeed ­ g lobally. This argument is laid 19. World Bank (2015) is the baseline evaluation out more fully in the Global Commission on ­ report. Two subreports focusing, respectively, the Economy and Climate Change (2014), 246 FUTURE DRIVERS OF GROWTH IN RWANDA which went on to have a seminal impact more experience and time for Rwanda to on the December 2015 United Nations work its way into highly competitive inter- Committee of Parties Summit in ­ Paris. It can national markets for ­ horticulture. Including also be seen in the work on agriculture and subsidies and after costs are paid, rice farm- climate change of the Consultative Group on ers might earn US$800 per hectare in high- International Agricultural Research ( ­https:// cost irrigation projects in the marshlands and ccafs.cgiar.org). maize farmers might earn US$250 per hect- 33. The source and original documentation for are on hillsides, compared with US$3,000 to this paragraph are from Global Commission US$3,500 per hectare by farmers with potato on the Economy and Climate ­ (2014). lands in the northern or southern provinces 34. All proposals for government participa- and US$2,000 for coffee farmers (Cambridge tion in agribusiness ventures go through Resources International 2 ­ 017). Returns to the Investment Promotion Coordination smallholder horticulture farmers producing C o m m i t t e e c h a i r e d b y t h e Rw a n d a for the domestic market or the Democratic Development B ­ oard. The commodities listed Republic of Congo have not been estimated all have had projects submitted at one time or comprehensively, but they are thought to be another, and this is substantiated by the back- at least on a par with the returns to potato ground material, by commodity or commod- farmers or four times as high as the returns ity group, prepared for the ­ committee. See farmers. to rice ­ RDB ­(2017). 35. See, for example, h ­ ttp://blogs.worldbank.org​ /­ppps/how-national-ppp-units-can-influence​ -regional-performance-korea-s-experience. References 36. This argument is set out in detail in Delgado Abdu lsa m ad , ­ A ., a nd ­ G. ­ 2 016 . G eref f i. ­ ­(1999). E ast Afric a Dair y Value Chains: Fir m 37. S e e ­ht t p: // w w w.m i n ag r i.gov.r w/ i ndex​ Capabilities to Expand Regional ­ Trade . .php?id=469&tx_ttnews%5Btt_news%5D​ IG C Repor t ­ F -38202-RWA-1. L ondon: =1175&cHash=0caddb1dfa0a480df74016a3 International Growth Centre; Durham, NC: d6872b23. Center on Globalization, Governance, and 38. The source for the recommendations that Competitiveness, Duke University, O ­ ctober. follow in this area is the World Bank’s Access to Finance ­ Rwanda–Finscope. ­ 2 016. “Agriculture Finance Diagnostic for Rwanda,” Financial Inclusion in Rwanda ­ 2016. Kigali: which provides details (World Bank ­ 2016). Access to Finance ­ Rwanda. 39. A comprehensive electronic business direc- Accord ­ A ssociates. ­2 017. “Future Sources of tory is a simple tool that is easily u ­ pdated. Growth in Rwanda Export ­ H orticulture.” It helps to keep track of a rapidly changing Draft report prepared for the World Bank, business environment, lowers search costs Washington, DC, ­ November. for foreign firms seeking partners in Rwanda, AGRA (Alliance for a Green Revolution in and helps to encourage increased communica- ­ A frica). ­2016. “Towards a Third Generation tions within the national business ­ community. Subsidy Model, Rwanda Country ­ R eport.” Government participation would lower the Draft report, AGRA, ­ Nairobi. costs of keeping the directory current through Agri-TAF (Agriculture Technical Assistance current information on ­ registrations. Facility). ­ ­ 2 016. “Climate Risk Assessment 40. Dividing estimated irrigated marshland of Agriculture in R ­ wanda.” International of 46,000 hectares (reducing the figure of Fund for Agricultural Development, Rome, 48,511 hectares of total irrigation in 2016–17 November ­26. in MINAGRI [2017c] by the figure of 2,500 D. ­ Ali, ­ K. Deininger, and ­ A., ­ M. ­Goldstein. ­2014. hectares on hillsides from World Bank [2014] “Environmental and Gender Impact of Land and dividing by an estimate of 277,164 hect- Tenure Regularization in Africa: Pilot Evidence ares of potentially irrigable marshland in from ­Rwanda.” Journal of Development MINAGRI [2017c]). Economics 110 (September): ­ 262–75. 41. Producing for regional horticulture markets Cambridge Resources ­ 2017. “Study International. ­ like the Democratic Republic of Congo will on Comparative Economic Advantage of Crop have a revenue advantage over producing Production in ­ Rwanda.” Cambridge Resources staples for the domestic market and provide International, Cambridge, MA, March ­ 31. T ransitioning A gricu l ture and F ood as an E ngine of G rowth 247 Cargill. ­ 2 014. “Cargill and IFC Invest ­ $ 2.5 Rwanda, Monitoring Price Incentives for Million to Expand Cocoa Farmer ­ C oop.” ­ Coffee.” Draft, FAO, MAFAP, ­ Rome. SCOPEinsight, Côte d’Ivoire, O ­ ctober. ­ http:// Future Climate for ­ Africa. ­2014. Report ­Rwanda. www.scopeinsight.com/news/cargill-and​-­i fc​ Cape Town: Future Climate for ­ Africa. -invest-2-5-million-to-expand-cocoa-farmer​ Global Commission on the E conomy and -coop/. Climate. ­ ­ 2014. “Land ­ Use.” In Better Growth, C I AT (I nternational C enter for Tropical Better Climate: The New Climate Economy ­ 2 015. “C l i m a t e S m a r t A g r i c u l t u r e). ­ Report 2014, ­ 3. Washington, DC: Global ch. ­ Agriculture in ­ Rwanda.” CIAT, ­ Calí. Commission on the Economy and C ­ limate. COMESA (Common Market for Eastern and ­ht t p: //newcl i m ate e conomy.repor t / 2014​ Southern A ­ frica). ­ 2017. “Country Harmonized / ­land-use/. COM E SA Seed Regulations Have Been Havemann, ­ T. ­ 2 015. “Tea Landscapes in L aunched.” COMESA, June ­ ­ 13. ­ http://www​ Yunnan, ­ C hina.” In Steps toward Green: .comesa.int/country-harmonized-comesa-seed​ Greening Export Agriculture in East and -regulations-have-been-launched. Southeast Asia; Policy Responses to the Dang, ­ Y., ­ W. Ren, ­ B . Tao, ­ G . Chen, ­ C . Lu, Environmental Footprint of Commodity J. Yang, ­ ­ S. Pan, ­ G. Wang, ­ S. Li, and ­ H. ­ Tian. Agriculture in East and Southeast Asia, 2 014. “Climate and Land Use Controls on ­ edited by ­ S . Scherr, ­ K . Mankad, ­ S . Jaffee, Soil Organic Carbon in the Loess Plateau and C­ . Negra, ­ 143– 66. Washington, DC: Region of C ­ hina.” PLOS ONE 9 (5): e ­ 95548. EcoAgriculture ­Partners. ­doi:10.1371/journal.pone.0095548. Hirschman, ­A . ­1958. The Strategy of Economic Delgado, C ­. ­ 1 999. “Sources of Growth in ­D evelopment. New Haven: Yale University Smallholder Agriculture in Sub-Saharan ­Press. Africa: The Role of Vertical Integration with Huang, M ­ ., ­ J . Gallichand, and P ­. ­ Z hang. Processors of High Value ­ Items.” Agrekon 38 ­ 2 003. “Runoff and Sediment Responses (Special Issue, May): ­ 165–89. to Conservation Practices: Loess Plateau Delgado, ­ C ., ­J. Hopkins, ­ V. Kelly, ­ P. ­ B. ­R. Hazell, of ­C hina.” Journal of the American Water A. ­ ­ A . McKenna, P ­ . Gruhn, B ­ . Johhati, J ­ . Sil, Resources Association 39 (5): ­ 1197–207. and ­C . ­Courbois. ­1998. Agricultural Growth ­doi:10.1111/j.1752-1688.2003.tb03702.x. Linkages in Sub- Saharan A ­ frica . IFPR I IFDC (International Fertilizer Development Research Report ­ 107. Washington, DC: ­Center). ­2014. Rwanda Fertilizer A ­ ssessment. International Food Policy Research ­ Institute. Muscle Shoals, AL: IFDC, ­ August. Df I D (Department for I nternational Jaffee, ­ S . Henson, and L S ., ­ ­ . Diaz R ­ ios. ­2011. ­ D evelopment). 2 ­ 015. “Rwanda Inclusive Making the Grade: Smallholder Farmers, Growth ­ Diagnostic.” DfID, Kigali, June ­ 19. Emerging Standards, and Development Diao, ­ X ., ­ P. Fang, ­ J . Randriamamonjy, and Assistance Programs in Africa: A Research J. ­ ­ T hurlow. ­ 2 017a. “A Growth Analysis— Program ­S ynthesis . Report ­ 6 2324-AFR. Drivers of Growth in R ­ wanda.” International Washington, DC: World Bank, J ­ une. ­ https:// Food Policy Research Institute, Washington, hubs.worldbank.org/docs/ImageBank/Pages​ ­DC. November ­21. ­ /­DocProfile.aspx?nodeid=442392. ———. ­ 2017b. “Patterns of Food Consumption K at h i re s a n , ­A . ­2 012 . “ Fa r m L a nd U s e in Rwanda: An Analysis of EICV4 ­ D ata.” Consolidation in Rwanda: Assessment from International Food Policy Research Institute, the Perspectives of Agricultural ­ S ector.” Washington, DC, November ­ 13. MINAGRI, Kigali, June ­ 29. FAO (Food and Agriculture ­ Organization). ­ 2010. K ing, ­ R ., and ­ D. ­ B yerlee. ­ 1 978. “Factor Agro-Based Clusters in Developing Countries: I ntensities and Locational Linkages of Staying Competitive in a Global ­ E conomy. Ru ral Consu mption Pat terns in Sierra Agricultural Management, Marketing, and ­L eone.” American Journal of Agricultural Finance Occasional Paper ­ 25. Rome: ­ FAO. ­ Economics 60 (2): 1 ­ 97–206. ­ h ttps://doi​ ———. ­2 017. Agroforestr y for L andsc ape .org/10.2307/1240048. ­R estoration. Forest Landscape Restoration Kumar, ­ R ., and A ­ .­Agarwal. ­ 2016. “Promoting ­Mechanism. Rome: ­FAO. More, Better, and Inclusive Jobs: Using a ‘Jobs FAO, MAFAP (Monitoring and Analysing Food Lens’ for the Tanzanian Horticulture ­ Sector.” and Agricultural Policies P ­ rogramme). ­ 2017. Jobs Group, World Bank, Washington, DC, “MAFAP Commodity Dashboard Template: ­January. 248 FUTURE DRIVERS OF GROWTH IN RWANDA Malesu, ­ M. ­ M ., ­A. ­R . Oduor, ­ K . Chrogony, ­ D. NA E B (National Ag ricultu ral E xpor t Nyolei, ­ D. ­K. ­ K . Gachene, ­ E. ­K . Biamah, Development ­ Board). ­ 2017. ­ “Coffee.” NAEB, ­ M . O’Neil, ­ M . Lyama, and ­ J. ­M ogoi. ­K iga l i. ­h t t p: // w w w. n aeb.gov. r w/ i ndex​ ­2 010. Rwanda Irrigation M aster ­ P lan . .php?id=48. Kigali: Ministry of Agriculture and Animal Nestlé. ­2016. Nestlé in Society: Creating Shared Resources, Ebony Company, and World Value and Meeting Our Commitments ­ 2016. Agroforestry ­Centre. Lausanne: ­N estlé. ­https://www.nestle.com​ Marin, P ­ ., ­ S . Tal, J­ . Yeres, and K ­ .­ B. ­ R ingskog. /­a sset-library/documents/library/documents​ 2 017. “Water Management in Israel: Key ­ /­c orporate_social_responsibility/nestle-in​ I n novat ion s a nd L e s s on s L e a r ne d for -­society-summary-report-2016-en.pdf. Water-Scarce ­ C ountries.” Water Global NISR (National Institute of Statistics of ­ Rwanda). P rac t ic e Work i ng Pap er, World B a n k , ­2 014. Four th Population and Housing Washington, ­D C. Census 2012: Thematic Report; Population MINAGRI (Ministry of Agriculture and Animal ­Projections. Kigali: NISR, ­ January. ­ Resources), ­ Rwanda. ­ 2012. “Strategic Plan for ———. ­2 016a. Rwanda Statistical Yearbook the Transformation of Agriculture in Rwanda, ­2016. Kigali: ­ N ISR. Phase 3 ­ (‘PSTA3’).” MINAGRI, ­ Kigali. ———. 2 ­ 016b. “Seasonal Agricultural Surveys— ———. ­ 2 014. “National Fertilizer P ­ olicy.” Season A ­ (May).” NISR, ­ Kigali. MINAGRI, Kigali, ­ June. ­ ———. ­ 2017. “Seasonal Agricultural Surveys— ——— . ­ 2 0 1 5 a . “ C O M E S A S e e d Tr a d e Season A (May) and Season B ­ (September).” Harmonisation Regulations Launched in NISR, ­Kigali. Rwanda.” Government of Rwanda, Kigali, ­ ———. Various ­ years. Integrated Household Febr ua r y ­2 5. ­h t t p: // w w w.gov.r w/news​ Living Survey ­ N ISR. ( EICV). Kigali: ­ _­d etail/?tx_ttnews%5Btt_news%5D=1082​ Ntirenganya, E 2 018. “Rwanda Takes Early ­. ­ &cHash=46ceb55f78e. ­ Steps towards Legalising G ­ MOs.” New Times, ———. ­ 2015b. “Privatized Fertilizer Importation January ­16. ­h ttp://www.newtimes.co.rw​ a nd D i s t r ibut ion Sy s t e m i n Rwa nd a : /­section/Printer/227457/. Contribution to Boosting P ­ roductivity.” New Policy ­A ssociates. ­2 017. Rwanda Agriculture Times, June 1 ­ 8. w­ ww.newtimes.co.rw/files​ Public Expenditure Review 2016, 2 ­ vols. /­promo/1435924746Minagri-July-3.pdf. Kigali: World Bank Group and ­ U KAid. ——— . ­ 2 017a. “Rwa nd a C omprehen sive Pritchard, ­ M. ­2013. “Land, Power, and Peace: Mechanization ­ Policy.” MINAGRI, Kigali, Tenure Formalization, Agricultural Reform, June ­15. ­ and Livelihood Insecurity in Rural ­ Rwanda.” ———. ­ 2017b. “Rwanda National Agricultural Land Use Policy 30 (1): ­ 186–96. Policy: A Productive, Green, and Market- R DB (Rwanda Development ­ B oard). ­2 017. Led Agricultural Sector towards 2 ­ 030.” “Investment Opportunities in the Irrigation MINAGRI, Kigali, ­ October. and Mechanization ­ S ectors.” RDB, ­ K igali. ———. ­ 2017c. “Strategic Plan for Agricultural ­h ttp://www.rdb.rw/fileadmin/user_upload​ Tr a n sfor m at io n 2 018 – 24 ­ ( ‘ P S TA4’).” / Docu ments /Ag ricu lt u re / I rrigation%20 MINAGRI, Kigali, ­ November. and%20Mechanization%20projects.pdf. MINICOM (Ministry of Trade and Industry), REMA (Rwanda Environment Management R wa n d a . ­ ­ 2 018 . “ N at io n a l Pol i c y o n Authority). ­ ­ 2011. “Green Growth and Climate Cooperatives in Rwanda: Toward Private Resilience National Strategy for Climate Cooperative Enterprises and Business Entities Change and Low Carbon ­ D evelopment.” for S o cio - E conom ic T ­ ra n sfor m at ion.” REMA, ­Kigali. MINICOM, Kigali, ­ January. ———. ­2015. Rwanda: State of Environment and MINIRENA (Ministry of the Environment), Outlook Report ­ 2015. Kigali: ­R EMA. ­ Rwa n d a . ­ 2012 . “S econd National S e i d l e r, E ­. 2­0 01. “ W hole s a le M a rke t ­Communication.” MINIRENA, ­Kigali. Development: FAO’s ­ E xperience.” Paper pre- Moreki, ­ J. ­C ., ­ N. ­ S . Ndubo, ­ T. Ditshupo, and pared for the 22nd Congress of the World ­ J. ­ B. ­ Ntesang. ­ 2012. “Cattle Identification and Union of Wholesale Markets, Food and Traceability in ­ Botswana.” Journal of Animal Agriculture Organization, Durban, South Science Advances 2 (12): ­ 925–33. Africa, ­September. T ransitioning A gricu l ture and F ood as an E ngine of G rowth 249 Tschirley, ­ T. Reardon, ­ D., ­ M . Dolislager, and J ­. and ­ O pportunities.” Policy Note ACS9719, S nyder. ­ ­ 2 015. “The Rise of a Middle Class World B a n k, Wash i ng ton DC , Ju ne ­ 7. in East and Southern Africa: Implications for ­h ttp://documents.worldbank.org /curated​ Food System ­ Transformation.” Journal of /e n / 8 8 3 4114 6 8 3 056 6 0 8 7 7/ p d f /AC S 9 7​ International Development 27 (5): ­ 628–46. 190W P0P140 0B ox385238B 0 0PU BL IC 0​ U.S. Department of ­ Agriculture. ­ 2017. International .pdf. Ag ricu lt u ral Produc tivit y ­ database. ———. ­ 2 015. “Rwanda Green L eaf Price Washington, DC: Economic Research ­ Service. Reform Impact Evaluation Baseline ­ Report.” ­h t tps: //w w w.ers.usda.gov/data-products​ Investment Climate Impact Program, World /­international-agricultural​-productivity/. Bank, Washington, ­ h ttp://pubdocs​ DC . ­ Vermeulen, ­ H., ­J. Kirsten, and K ­ .­ Sartorius. ­2008. .worldbank.org /en /858581473702690438​ “Contracting Arrangements in Agribusiness /­COMPEL-rwanda-baselinereport.pdf. Procurement Practices in South ­ A frica.” ———. ­ 2016. “Agricultural Finance ­ Diagnostic.” Agrekon 47 (2, June): ­ 198–221. Rwanda Country Support Program, Financial Weatherspoon, ­ D., and ­T. ­ Reardon. ­ 2003. “The Inclusion Support Framework, World Bank, Rise of Supermarkets in Africa: Implications Washington, ­DC. ­ for Agrifood Systems and the Rural ­ Poor.” ———. ­ 2 017. World Development Indicators Development Policy Review 21 (3): ­ 333–55. ­ d atabase. Washington, DC: World B ­ ank. World ­ B ank. ­2007a. “Restoring China’s Loess ­https://data.worldbank.org/products/wdi. Plateau.” Feature Story, World Bank, March ­ Xie, ­M., ­ N. Asquith, ­ J. Li, ­ J. Tyson, ­A. Kleine, and ­1 5. ­h ttp://www.worldbank​.org ​ /en /news​ Y. ­ ­ Huan. ­ 2010. “Rehabilitating a Degraded /­f eature​/2007/03/15/restoring-chinas-loess​ Watershed: A Case Study from China’s Loess -plateau. P lateau.” World Bank Institute, Climate ­ ———. ­2007b. World Development Report 2008: Change Unit, Washington, ­ DC. Agriculture for ­ D evelopment. Washington, Zeuli, ­ K ., ­R . Cropp, and ­ M. ­ S chaars. ­ 2 004. DC: World ­ Bank. Cooperatives: Principles and Practices in ———. ­ 2014. “Promoting Agricultural Growth the 21st ­ C entury. University of Wisconsin in Rwanda: Recent Performance, Challenges, Extension A1457, ­ Madison. ­ August. 6 Capable and Accountable State Institutions “ We never stopped learning because the key comparative advantage over most other situation kept on changing and we had to countries in the region. adjust our own policies.” (Lee 2000, 686) Nevertheless, achieving middle-income status will require further improvements in We are moving ahead, learning from what “ institutions to encourage greater creativity works and what doesn’t work and adjust- and initiative by central and local govern- ing without losing sight of our goals.” ment officials. As the economy gets bigger (Kagame 2015) and more complex, it will require far deeper coordination between institutions (for man- aging urbanization, industrialization, export Introduction promotion, and decentralization, for exam- Starting from very difficult initial conditions ple) and cooperation between the state and after the 1994 genocide against the Tutsi, the private sector (for public investments, for Rwanda has made remarkable progress in example), on the one hand, and between the strengthening governance. International state and the society (to ­ support innovation rankings place Rwanda at the average level at the local level and generate strong feedback of upper-middle-income countries, with loops), on the other. Achieving all of this will particularly strong performance on indica- require greater social capital of trust between tors of government effectiveness, control citizens, between civil society and the govern- of corruption, rule of law, and regulatory ment, and among local officials, civil servants, quality.1 The formerly bloated civil service and national political leadership to promote has been downsized, and the skills of gov- the innovation (in the bureaucracy, in the soci- ernment officials have improved markedly. ety) required for rapid economic growth. Reliance on homegrown approaches to per- This chapter addresses the following formance management has instilled a strong ­ question: How should Rwanda strengthen culture of results at all levels of government, its state efficiency and accountability greatly increasing government effectiveness. to best support successful transition to Rwanda’s strong governance is a driving becoming a modern, competitive, and glob- force behind its rapid growth and provides a ally connected ­ m iddle-income economy? 251 252 FUTURE DRIVERS OF GROWTH IN RWANDA The chapter applies the analytical frame- the entry of new firms remain low, reflect- work of the World Development Report ing gaps in market institutions. Creating 2017 (World Bank 2017) to discuss insti- special courts and fast-track procedures to tutional adaptation to induce the behavior adjudicate small claims; further promoting changes needed—in government, firms, and alternative means of resolving commercial ­ citizens—to promote higher levels of develop- disputes (such as arbitration, mediation, and ment. It describes Rwanda’s performance on conciliation); boosting reliance on technol- core measures of institutional development ogy; increasing training and specialization and highlights three pillars, or broad policy of judges, prosecutors, and investigators; objectives, to strengthen trust and improve and improving case management techniques government effectiveness. would enhance the effectiveness of the judiciary system for contract enforcement. ­ Further improvements in the efficiency of the Pillar 1: Enabling Innovative, land management system and implementa- Coordinated, Capable Bureaucracy, tion of the intellectual property law would and Local Governments strengthen property rights. Reviewing the Key challenges relate to incentives for inno- performance of fiscal incentives with a view vation among bureaucrats and local gov- to reducing Rwanda’s tax expenditures, ernments and to shortfalls of some core increasing local government revenue col- competencies in the civil service. Imihigo lection, ensuring that the state captures an (performance contracts) have made a strong appropriate share of the expected rise in min- contribution to instilling a culture of results, eral revenues, and raising Rwanda’s strong but these contracts can be improved to foster public financial management to international coordination between institutions, bottom- standards would strengthen the state’s finan- up initiatives, and innovation. To enhance ­ apacity. A better balance needs to be cial c innovation, Rwanda could consider adopt- achieved between public investments that ing multiyear targets in imihigo with annual are expected to generate high returns over benchmarks, making greater use of qualita- the long term and investments in areas with tive surveys, and focusing more on outcomes potential high short-term returns based on rather than processes. Another step forward market signals of scarcity. could be to increase the importance of the joint imihigo to at least 50 percent and to take bold measures to improve interagency Pillar 3: Enhancing the State’s coordination on key cross-cutting issues. Accountability to Citizens Further increases in training by the Rwanda Strengthening key watchdog agencies (the Management Institute (RMI) should be sup- Public Accounts Committee, the Office of the ported by skills audits and functional reviews Auditor General, the Ombudsman’s Office, to identify the missing skills in the public sec- and the Rwanda Governance Board) with tor. Regular compensation reviews, higher more operational capabilities and strengthen- compensation for individuals with scarce ing the media sector would further strengthen skills, and the development of nonfinancial government accountability. Increasing own incentives will become increasingly important revenue raising by local government and pro- to retain and attract highly skilled staff. viding greater fiscal decentralization would be a bold step toward making local officials and service providers more accountable. Pillar 2: Strengthening Market Deepening the initiatives that have sought Institutions and Building the State’s to involve citizens in local decision making Financial Capacity (for example, in setting imihigo objectives) Despite Rwanda’s impressive improvements and to provide greater local flexibility and in governance, private sector investment and reliance on more qualitative information in C a p a b l e and A ccounta b l e S tate I nstitutions 253 monitoring performance would strengthen Sub-Saharan Africa region and is ranked support for local government and improve 50 globally. T his positive assessment the quality of services. reflects Rwanda’s strong performance in process (low corruption, high efficiency) and compliance. Rwanda’s Record in Governance The Rwandan government has main- and Government Effectiveness tained high-profile anticorruption efforts. Institutional quality in Rwanda has improved significantly. Public sector reforms have led FIGURE 6.1  Relationship between government effectiveness and to an effective state with relatively low cor- level of development ruption and capable institutions. An almost across-the-board upward trajectory in institu- 2.5 tional quality is seen in the governance assess- Singapore ment conducted by the Rwanda Governance 2.0 Board (RGB) since 2010 (table 6.1). 2 This assessment is backed up by several cross- Government e ectiveness (Kaufmann index) 1.5 country measures of governance, where Chile ࡗMean 1.0 Rwanda emerges at par with middle-income Malaysia Korea, Rep. countries, with particularly strong per- 0.5 South Africa China Thailand formance on indicators of government Rwanda Vietnam Indonesia Turkey ­e ffectiveness, control of corruption, rule 0 India Morocco of law, and regulatory quality (figures 6.1 Tunisia ࡗMean Uganda and 6.2). According to the 2017 Ibrahim –0.5 Ethiopia TanzaniaࡗMean Index of African Governance, Rwanda is the Malawi Cambodia Bangladesh only country to have made consistent year- –1.0 ࡗMean Burundi on-year improvement on overall governance since the index was started. –1.5 Considering the grim conditions follow- –2.0 Somalia ing the genocide against the Tutsi, Rwanda has achieved high levels of s ­ tability and –2.5 security. According to the 2016 World 2.5 3.0 3.5 4.0 4.5 5.0 5.5 Internal Security and Police Index, which Log GDP per capita (constant 2011 international $) measures the ability of a country’s secu- Low income Upper-middle income rit y apparatus to respond to internal Lower-middle income High income security challenges, Rwanda (together Sources: Kaufmann, Kraay, and Mastruzzi 2010; World Development Indicators data from with Botswana) performs the best in the World Bank, various years. TABLE 6.1  The Rwanda Governance Scorecard: Scores for 2016 compared with previous scores First Second Third Fourth Change over Indicator edition edition edition edition five years Trend Rule of law 67.71 73.37 81.68 79.68 +11.97 È Political rights and civil liberties 71.43 73.62 77.05 81.83 +10.4 È Participation and inclusiveness 74.23 75.26 75.36 77.01 +2.87 È Safety and security 87.26 91.35 91.96 92.62 +5.36 È Investing in human and social development 82.41 78.8 81.54 74.88 −7.53 Í Control of corruption transparency and accountability 76.22 77.1 79.04 86.56 +10.34 È Quality of service delivery 66.21 70.44 72.00 72.93 +6.72 È Economic and corporate governance — 74.93 72.20 76.82 +1.89 È Source: RGB 2016b. Note: — = not available. The Rwanda Governance Scorecard is composed of eight indicators based on the guidelines of international context, recognized international indexes, and Rwandan homegrown indicators (RGB 2016a, 3). 254 FUTURE DRIVERS OF GROWTH IN RWANDA Institutions such as the Ombudsman’s Office has contributed to macroeconomic stability, have sought to root out corruption, and the while good-quality public financial man- courts have prosecuted cases at all levels of agement has improved service delivery. An government. Consistent with this, in 2017 important measure in this regard is revenue Rwanda was ranked 48 (out of 180 countries) generation. Government revenues (exclud- for control of corruption in Transparency ing grants) equal almost 17 percent of gross International’s Corruption Perception Index, domestic product (GDP) and in 2015 were a vast improvement over its 2006 ranking of above the average ratio for lower-middle- 121, placing it 3 (alongside Mauritius) on the income countries (figure 6.4). continent. The expenditure side of the budget is The 2003 Constitution called for poli- defined by a high ratio of public investment ticians and civil servants to declare their to GDP. At its peak in 2013–15, ­ p ublic assets, which has been followed by vari- investments reached 13 percent of GDP; ous enforcement measures (Freedom House quite high by international standards. They 2007), including prosecution of prominent have declined since then, stabilizing at about government officials. The 2016 Bertelsmann 10 percent of GDP.4 However, this number Stiftung’s Transformation Index, which does not capture the full extent of public assesses whether public servants and politi- investment, because it does not include net cians are held accountable, 3 found Rwanda lending (on average, 2 percent of GDP in to be at the level of upper-middle-income recent years).5 Transport, energy, and health countries (figure 6.3). have been the important priority sectors The country has also built relatively for public investment. The other large item strong fiscal institutions that have supported is governance and decentralization, which rapid growth. Prudent fiscal management includes capital transfers to the communities FIGURE 6.3  Correlation between prosecution of FIGURE 6.2  Benchmarking of Rwanda along the office abuse and GDP per capita various dimensions of governance 10 Control of Chile Singapore corruption 9 8 Prosecution of o ce abuse (index) Rule of law Mean 7 Korea, Rep. ࡗ South Africa Regulatory quality 6 India Malaysia ThailandTurkey Indonesia 5 Rwanda ࡗMean Government Morocco 4 Ethiopia Uganda ࡗ China e ectiveness Mean Mean ࡗ Tunisia 3 Vietnam Political stability Bangladesh 2 Cambodia Voice and 1 accountability 0 0 20 40 60 80 2.5 3.0 3.5 4.0 4.5 5.0 5.5 Percentile rank, 2016 Log GDP per capita (constant 2011 international $) Mean of UMICs Mean of LICs Low income Upper-middle income Mean of LMICs Rwanda Lower-middle income High income Source: 2016 Worldwide Governance Indicators from Kaufmann, Kraay, and Mastruzzi 2010. Sources: 2016 Bertelsmann Stiftung’s Transformation Index, Q3.3: Note: LIC = low-income country; LMIC = lower-middle-income Prosecution of office abuse; World Development Indicators 2016 country; UMIC = upper-middle-income country. data (World Bank, various years). C a p a b l e and A ccounta b l e S tate I nstitutions 255 for various infrastructure needs. On bal- international assessment noted favorably, ance, public investment has been managed “Despite the limited technical capacity well, and project implementation has been in some fields, Rwandan public adminis- particularly strong. 6 Rwanda scored high tration is relatively efficient, recruitment among low-income countries and came in ahead of 19 middle-income countries in a FIGURE 6.4  Revenue (tax and nontax) excluding grants 2012 International Monetary Fund assess- ment (Dabla-Norris et al. 2012).7 The 2016 Morocco public expenditure and financial account- Turkey ability assessment (PEFA Secretariat 2017) South Africa also shows strength in project selection and Tunisia ­ costing (table 6.2). Korea, Rep. Mean These achievements are due in large part Vietnam to a strong developmental vision driven by the Chile urgent need for recovery in the aftermath of Thailand the genocide against the Tutsi. Rwanda has Malaysia been characterized as a “developmental state” Singapore Rwanda (Booth and Golooba-Mutebi 2012; Crisafulli Cambodia and Redmond 2012), with the top leadership China playing a central role and showing the capac- Mean ity to commit credibly to coherent reform Mean policies and to ensure that the bureaucracy Indonesia and other relevant actors carry through on Tanzania India implementation. Strengthening administrative Ethiopia capacity, instituting a decentralized system, Uganda and building effective civil service monitoring Bangladesh were seen early on as being crucial for politi- 0 5 10 15 20 25 30 35 cal legitimacy.8 These aspects were comple- % of GDP mented by a series of homegrown initiatives Low income Lower-middle income Upper-middle income that sought to instill “Rwandan values” and mobilize collective action. Source: World Development Indicators (World Bank, various years); available years: 2011 for Ethiopia, Morocco, and the mean of low-income countries; 2012 for Tunisia; 2013 for India, Strong gains have been made in build- Vietnam, and the mean of low- and middle-income countries; 2014 for China and the mean ing a capable civil service (box 6.1). An of upper-middle-income countries; 2015 for all the others. TABLE 6.2  Components and ratings of public investment management framework P1–11 Dimension Score Justification Public investment B+ Scoring method M2 management 11.1 Economic analysis of B Ministries, departments, and agencies as well as district councils submit all proposed investment proposals capital investment projects to the Public Investment Committee for economic appraisal before approval in accordance with the public investment guidelines issued by National Development Planning and Research. Some results are published. 11.2 Investment project A All major investment projects are prioritized on the basis of the public investment selection guidelines before they are included in the budget. 11.2 Investment project A Public investment guidelines outline a comprehensive costing framework for both monitoring investment cost and forward-linked recurrent costs. The information is included in the budget documentation. 11.4 Investment project C The Project Management and Monitoring Unit of the Ministry of Finance and Economic monitoring Planning undertakes at least quarterly physical inspection of all approved government projects. It also prepares quarterly and annual project progress and financial reports, but these are not published. Source: PEFA Secretariat 2017. 256 FUTURE DRIVERS OF GROWTH IN RWANDA practices have improved over time, and Given the historical legacy of conflict, most technical positions, as well as most strengthening subnational institutions was low- to middle-ranking officials, seem considered a national priority. The country to be indeed recruited based on candi- embarked on a transformative process of date merit” (Hertie School of Governance decentralization to improve the efficiency 2014). A strong civil service, together with of the public sector, strengthen social cohe- a strong system of vertical accountability, sion, improve local governance, and pro- explains Rwanda’s strong reputation for mote access to basic infrastructure and implementation, control of corruption, and service delivery. In 2000, a comprehen- public financial management and its suc- sive National Decentralization Policy9 was cess in building basic infrastructure and adopted to “enhance the participation of improving outcomes. citizens in planning and [to] strengthen, BOX 6.1  A major turnaround in the Rwandan civil service The state inherited a heavily depleted civil service in 2005 (Hausman 2011). Today, Rwanda’s civil ser- after the genocide. In the absence of a well-organized vice is dominated by young, well-educated gradu- state apparatus, public sector organizations began ates, with a mean age below 35 years, making it one hiring at their own discretion. a Overstaffing soon of the youngest among comparators. Based on New became common, sometimes with several employ- Public Management approaches, reorganizations ees holding the same position. Moreover, many of since the early 2000s have sought to streamline the the new recruits lacked the necessary skills. By the number of institutions while separating the policy- end of 1997, there were about 40,600 civil servants, making function (left to ministries) from the imple- and Rwanda was facing problems that afflict many mentation function (assigned to boards, agencies, other countries in Sub-Saharan Africa: lack of uni- and districts). form procedures for recruitment, promotion, or per- Transparent and merit-based recruitment, an formance that contributed to an unsustainable wage area where many countries fail, was considered bill (given the government’s limited resources), even central to reconstruction of the country. Under the though the salaries offered were not competitive.b urugwiro dialogue (in 1998–99),c transparent and The Rwandan civil service had become too large, merit-based recruitment was increasingly empha- inefficient, and not sufficiently deconcentrated or sized, becoming a principle enshrined in the 2003 decentralized for service delivery (Hausman 2011). Constitution (Chemouni 2017),d replacing the prac- The first plan to reform the civil service, adopted tice of favoritism that dominated the pre-genocide in 1998, included downsizing, decentralization of era and contributed to social tensions (Hausman personnel, uniform recruitment policy and pro- 2011). The Public Service Commission, set up in cedures, and stronger selection of candidates. 2007,e was made responsible for the recruitment of Downsizing and voluntary departure programs central government employees.f The law provided were implemented in 1999, 2004, 2006, and 2009. both the commission and the ombudsman (who As a result, the size of the civil service and, more organized annual visits to different state organiza- important, the profile of civil servants changed dra- tions to detect cases of nepotism) with administra- matically: the share of civil servants with university tive autonomy, requiring them to report directly to degrees rose from just 6 percent in 1998 to 79 percent Parliament and the president. a. According to the interviews organized with the former head of the National Tender Board on February 18, 2015, some state organizations, such as the Ministry of Civil Service, were informally administering written exams and interviews when hiring for top positions (Chemouni 2017). b. In 1995 salaries accounted for 73.6 percent of total expenditures (Ministry of Public Service, Skills Development, and Labor 2005). c. Meetings at the Office of the President between politicians and civil society representatives to debate on the reconstruction of Rwanda (May 1998–March 1999). d. Articles 45 and 126. e. Article 181 of the 2003 Constitution. f. A series of measures have been introduced to improve transparency in the recruitment process, such as guidelines on timing and the modalities for publishing vacan- cies and job interviews, but also the provision that a candidate who feels that he or she has been treated unfairly can report the case to the Ombudsman’s Office or the Public Service Commission. The publication of the decisions taken at each stage of the recruitment process and the involvement of private companies to organize and evaluate recruitment tests also have supported transparent and merit-based recruitment. C a p a b l e and A ccounta b l e S tate I nstitutions 257 through various means, the ‘voice’ of citizens FIGURE 6.5  Rwanda’s decentralization framework for service in influencing service delivery providers.” In delivery the September 2005 National Conference on Decentralization, Accountability, and Service National policy makers Ubuyobozi Delivery, the “Rwanda model” of decen- tralization was laid out further, inspired by the latest international knowledge in the fields of citizens-centered governance and Local policy makers Inzego Z’Ibanze public service delivery, including the World Development Report 2004 (World Bank 2003b). The government confirmed this approach and further defined tools and Citizens' voice Compact mechanisms to incorporate the decentraliza- Ijambo Ry’ Umuturage Amasezerango Y’ Ibikorwa tion framework in subsequent policies, leg- islation,10 guidelines, and sectoral reforms (figure 6.5).11 Traditional Rwandan concepts and prac- Citizens Service providers tices have been strengthened and new ini- Abaturage Abatanga service tiatives have been put in place to improve Client power Ububasha Bw’ Umuguzi government accountability and enhance the voice and power of beneficiaries. Multiple Source: MINALOC 2006. platforms for citizen-state interaction have Note: MINALOC = (Rwanda) Ministry of Local Government. been institutionalized at central and local levels, such as Umushyikirano (National Dialogue Council), a forum that brings public ­s ervices.13 Nevertheless, the num- together the president of the republic and ber of access-to-­information requests seems citizens’ representatives to debate national low.14 In the health sector, after an encour- issues; the Joint Action Development Forum, aging pilot program, the government adopted a key participatory planning mechanism to performance-based financing at the national improve alignment between citizens’ and level. This approach allows the government to districts’ priorities and the Accountability contract both private and public health facili- Day (Inteko’ z’abaturage); the district bud- services. ties for the delivery of a package of ­ get hearings and Budget Day to present Health facilities are then paid according to and discuss the district budget; and the verifiable measures of quantity, quality, and post umuganda 12 meetings. Furthermore, equity of services.15 a bottom-up planning approach has been Imihigo (performance contracts), a genu- outlined in improved versions of the local ine Rwandan management tool, have been planning and budgeting guidelines issued adopted for boosting vertical accountabil- by the Ministry of Local Government and ity of civil servants and local governments the Ministry of Finance (see MINECOFIN toward central ministries and the presidency. 2007, 2009, 2013). The first imihigo were signed between the Concrete measures have sought to deepen president of the republic and district may- the public sector’s accountability further. ors in 2006. The imihigo have since become A law to strengthen access to informa- an annual undertaking (box 6.2). The origi- tion came into effect on March 11, 2013. nal goal was to enable local ­ governments Overseen by the Ombudsman’s Office, the to articulate their own objectives, which law has clear provisions on proactive disclo- reflect priorities of the local citizenry, sure of information and is especially valuable and to set strategies to achieve the objec- in its broad scope, covering not only public tives. According to the Ministry of Local bodies but also private bodies that provide Government’s decentralization assessment, 258 FUTURE DRIVERS OF GROWTH IN RWANDA imihigo became the most engaging and Remaining Governance credible planning instrument at the subna- Challenges tional level, with better-funded activities and the most effectively evaluated targets Rapid improvements in governance and insti- (MINALOC 2017). Imihigo have made a tutional structures have served Rwanda well, strong contribution to instilling a culture of contributing to its impressive broader devel- results in Rwanda, thereby increasing gov- opment outcomes. But the requirements of a ernment effectiveness. middle-income Rwanda will be different— The justice sector, too, has carried out more demanding—from those that have suf- several important reforms. In 2005, the ficed in the low-income period. The ambitious government reestablished the traditional aspirations being adopted are bound to rub community court system called Gacaca to against emerging institutional limitations. One mete out justice, achieve reconciliation, and of the most robust results of the development heal some of the festering wounds from the literature is that no non-resource-rich coun- time of the genocide. Under the Gacaca, a try has reached high income levels without fully homegrown system, local communities highly capable and accountable institutions. selected citizens as judges to hear trials of And building effective institutions is a long- suspects accused of all crimes (except their term endeavor. The improvements are neces- planning) during the genocide. By the time sarily incremental, building on global and local the Gacaca courts officially closed down on lessons and factoring in the specific context May 4, 2012, they had handled officially of the country. The process is also not linear. at that time a total of 1,237,356 cases. In Rwanda’s work in this regard will be critical. 2006, also with the purpose of promoting To assess remaining challenges and guide reconciliation and the rule of law and tack- the way forward, this chapter adopts an ling the huge court backlogs (this time for analytical framework that considers societal common-law offenses), the government cre- trust as a core underpinning of state effec- ated the abunzi (mediation committees) to tiveness and long-term development. A high provide free mediation services to resolve level of societal trust is necessary for the small conflicts. Abunzi have been effective, state to establish its legitimacy and to ensure handling more than 50,000 cases in 2017 c ooperative relationships between itself, ­ and earning high levels of citizen satisfac- the business sector, and citizens. Although tion (RGB 2016a). necessary for economic development in any BOX 6.2  Imihigo: A Rwandan management tool of accountability Performance contracts have played a major role in sanctioned or even fired. A fully homegrown initia- creating a strong vertical chain of accountability in tive, it was introduced in 2006 and has since become the civil service. Imihigo is the plural in Kinyarwanda an invaluable tool for planning and implementation of umuhigo, which means a “vow to deliver.” It also of development policies (RGB 2014). Derived from includes the concept of guhiganwa, which means “to the traditional practice where individuals voluntarily compete among one another.” It describes a practice set their own targets and publicly commit to achieve in which a set of targets or goals must be achieved them within time-bound periods, the revived concept by an official within a specific period, following cer- was intended to strengthen participatory priority set- tain guiding principles and overcoming all hurdles ting, bottom-up planning, and accelerated implemen- (RGB 2016b), short of which the official may be tation while focusing on time-bound results.a a. The imihigo are set yearly, but they are evaluated every six months. Citizens Report Cards, produced by the Rwanda Governance Board, are now part of the evaluation of imihigo (with a weight of 10 percent). C a p a b l e and A ccounta b l e S tate I nstitutions 259 BOX 6.3  Relationship between trust and long-term development Above all, strong growth requires building a deeper problems that impede economic growth (La Porta level of trust. A vast literature demonstrates how et al. 1997; Putnam 1993). Low- and middle- trust (between the people and the government and income countries that fail to overcome vested inter- among the people) plays a major role in growth ests through rules-based and transparent decision and economic development (Fukuyama 1995; making, efforts to improve inclusion, and strong Knack and Keefer 1997, for example). Trust low- accountability mechanisms often stagnate (for ers transaction costs (Fukuyama 1995); encourages many countries, referred to as the “middle-income the establishment of long-term arrangements—for trap”), because they are unable to induce the example, in the form of financial and employment behavioral changes needed to promote investment contracts—and fixed investments; shapes the rela- and innovation (World Bank 2017, spotlight 6). tionship between voters and politicians, which can Trust also facilitates adaptation to change with- lead to improved public policy, institutional reform, out excessive disruption. Low- and middle-income and stability; and strengthens the social compact countries experience violent transitions every eight (Knack and Keefer 1997). years on average (Cox, North, and Weingast 2015), Comparative experiences and a wealth of lit- underlining the difficulties involved in adapting erature point to the importance of deepened trust institutions to the management of emerging social and state legitimacy in resolving collective action tensions. FIGURE 6.6  A framework for assessing price signals, rests on a strong partnership effectiveness of state institutions in Rwanda between the private and public sectors, and is complemented by fiscal discipline State to match policies with the state’s financial e ectiveness capacity 3. The extent to which the state is account- able to its citizens, which ensures that Societal trust policies and programs are aligned with the needs and aspirations of the people Well-structured Strong market State and innovative government institutions accountability Building Trust Building trust and cooperation is an ongoing challenge in Rwanda. Interpersonal trust was situation (box 6.3), trust is especially mean- understandably very low in the immediate ingful in a postconflict situation. aftermath of the genocide against the Tutsi. State effectiveness and societal trust Impressive progress has been made since then both depend on three institutional pillars in the way of reconciliation, promoting eco- (figure 6.6): nomic and social development, and strength- ening the structures of governance; such 1. The level and quality of the state’s capac- progress has helped to close the trust gap. But ity, which are shaped by governance more work is needed in this regard. structures and the level of meritocracy, coordination, and rules-based authority in the country Structure and Capacity of Government 2. The presence of a well-functioning m­ arket-​ Despite the substantial progress that has based economic system, which ensures that been made, important challenges remain for resource allocation is guided by efficient building a well-structured, high-capacity 260 FUTURE DRIVERS OF GROWTH IN RWANDA government. The main ones emerge from agencies, some institutions (and staff) may the following: a centralized decision-­ have competing mandates, which increases making approach and lack of foresight and the coordination costs for businesses and innovation mind-set in local governments, citizens. An example may be the Rwanda ineffective coordination mechanisms across ­ Revenue Authority (RRA), which is man- different government entities, limited capac- dated to collect tax revenues, and the Rwanda ity, and insufficient training. Development Board, which is mandated to A key challenge is the issue of weak inno- attract and develop investment. Streamlining vation among bureaucrats and local govern- coordination between these two institutions ments. The identification and encouragement is essential to avoid any possible confusion to of local innovations in the current imihigo investors. evaluation framework are good examples Coordination challenges also may arise to emulate in order to mitigate the possible from some of the poorly perceived overlap- negative effects of centralized evaluation ping responsibilities of central and local and control. Strong performance monitoring authorities. Attempts to clarify the assign- can also create a culture of internal compe- ment of expenditure responsibilities have tition or a “hyper performance orientation” been successful to implement because of (MacArthur 2012).16 This culture can dis- coordination efficiencies between local gov- courage information sharing and coordi- ernments and sectoral ministries. Taking sec- nation among public officials. The imihigo toral decentralization to the next level will coordination unit should continue to monitor help to address these challenges. imihigo planning and implementation closely. In order to improve coordination between Moreover, even though transfers to local public institutions (such as between line min- governments have increased in the last istries and local governments), in 2015 the decade, they, too, can be overly prescrip- government adopted the joint imihigo, which tive. For example, 80 percent of transfers means that both institutions have to perform are earmarked funds, which involve com- tasks to achieve the joint performance con- pulsory spending by local governments, and tract. A weight is assigned to an individual 14 percent are discretionary block grants. institution’s imihigo and the joint imihigo. Worldwide, establishing an enabling and less Final lessons on the joint imihigo will have restrictive environment has proved to be more to be drawn later, taking into account, effective in encouraging local innovation among other things, the relationship between than strengthening the technical capacity of ­ c ooperation and competition during the the local governments in question (Campbell implementation of imihigo and the impact of and Fuhr 2004). Rwanda’s current strategy each on delivery. of granting more flexibility to districts over Rwanda’s public administration needs to their own generated revenues provides a improve some core competencies. For exam- good opportunity to spur creativity and effi- ple, public institutions in the agriculture and ciency in wealth creation at the local level, energy sectors suffer from inadequate staff which directly affects local and national eco- levels in central ministries and districts com- nomic development. The experience of the pared with agencies and boards and from a World Bank Municipal Innovations project in lack of long-term skills development plans, Indonesia (World Bank 2003a) confirms that including in the public sector (a 2009 skills successful innovations are more likely if cen- audit found that 60 ­ p ercent of skills were tral and local governments share a political in short supply in both sectors) (MacArthur commitment to create a supportive enabling 2012). Staffing requirements in some units environment. also need to be evaluated more closely. For Coordination between ministries and pub- example, some small units (with only two lic sector agencies is still a challenge. With to three staff) are in charge of several mega- the development of several semiautonomous projects and are subject to exceptionally C a p a b l e and A ccounta b l e S tate I nstitutions 261 heavy workloads and possible efficiency on the existence and implementation of anti- ­problems.17 Gaps are also emerging in public trust rules. The World Bank/Organisation for sector training. In the last five years, the RMI Economic Co-operation and Development only reached 9 percent of all civil servants product market regulation indicators simi- (World Bank 2016). larly show Rwanda’s mixed performance on various underpinnings of competition (­figure  6.7).18 The regulatory framework pro- Market Institutions and Fiscal tects incumbents in key input sectors, with Management especially high barriers to entry in network Rwanda’s future ambitions will require the industries such as telecommunications and government to strengthen key market institu- electricity. tions, while backing them up with financial Major progress has been made in strength- capacity generated by a prudent fiscal stance. ening private ownership rights, which are The evidence is compelling that well-­ enshrined in the 2003 Constitution (revised functioning markets require well-defined in 2015) and backed by several laws (for rules of the game that are enforced transpar- example, the 2013 land law and the 2007 ently and predictably. The counterfactual expropriation law), other legal instruments, yields an entrepreneurial landscape that is and land registration efforts. Article 34 of not conducive to private-sector-led growth. the Constitution notes, “Private property, As described by North (1991), “With inse- whether individually or collectively owned, cure property rights, poorly enforced laws, is inviolable,” while allowing for excep- barriers to entry, and monopolistic restric- tion to the inviolability for “public interest” tions, the profit-maximizing firms will tend purposes. Delays in compensation and their to have short time horizons and little fixed low levels relative to market value remain a capital and will tend to be small. The most challenge. profitable businesses may be in trade, redis- There is considerable scope to improve tributive activity, or the black market. Large efficiency of the commercial dispute reso- firms with substantial fixed capital will exist lution process further, including contract only under the umbrella of government pro- e nforcement—an important aspect of a ­ tection with substantial subsidies, tariff pro- functioning market economy that is essen- tection, and payoff to the polity—a mixture tial to economic development and sustained hardly conducive to productive efficiency.” growth (Esposito, Lanau, and Pompe 2014). This caution is pertinent to Rwanda, where Globally, Rwanda ranks 78 on the ease the private sector is small and informal for of enforcing contracts (among 190 econo- the most part (chapter 4 of this report). mies), which shows both the extent of prog- Rwanda has made good progress in ress made to date and the scope for further improving regulations that support compe- improvement that remains. According to the tition, but considerable scope remains for Rwanda Governance Scorecard 2016 (RGB further improvement. This is borne out by 2016b), with regard to the performance of the country’s performance on cross-country the judiciary, “Citizens are less satisfied by comparisons. The Global Competitiveness the level of backlog of courts cases, the inde- Report 2016–2017 ranks Rwanda 77 among pendence of courts, and implementation of 138 countries and 10 among 29 Sub-Saharan courts resolution.” There is widespread con- African countries on the intensity of local sensus among stakeholders that justice sec- competition, while giving it high rankings tor staff need continuous training in several on the extent of market dominance (34) and categories important for commercial mat- effectiveness of antimonopoly policy (26) ters, such as intellectual property, insurance (World Economic Forum 2016). According disputes, receivership cases, banking and to the Bertelsmann Stiftung’s Transformation finance, c ­ orporate governance, and arbitra- Index 2018, Rwanda scores 6 points out of 10 tion.19 A backlog of commercial cases, limited 262 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE 6.7  Rwanda’s performance on product market regulations Government involvement in network sectors 4.58 Barriers in network sectors 3.46 License and permit system 3.33 Barriers to trade facilitation 3.12 Tari barriers 3.0 Barriers in service sectors 2.64 Governance of state-owned enterprises 2.25 Antitrust exemptions 2.22 Scope of state-owned enterprises 2.22 Price controls 2.21 Communication and simpli cation of rules and procedures 2.11 Command and control regulation 2.1 Administrative burdens for sole-proprietor rms 1.8 Administrative burdens for corporations 1.8 Barriers to foreign direct investment 0.56 Di erential treatment of foreign suppliers 0.52 Legal barriers to entry 0.37 Government direct control over enterprises 0.22 0 1 2 3 4 5 Least restrictive Most restrictive State control subcomponents Barriers to entrepreneurship subcomponents Barriers to trade and investment Sources: For OECD (Organisation for Economic Co-operation and Development) countries, OECD Product Market Regulation database (http:// www.oecd.org/eco/growth/indicatorsofproductmarketregulationhomepage.htm); for non-OECD countries, product market regulation data from the World Bank Group’s Markets and Competition Policy database. Note: Based on a scale from 0 to 6, from least to most restrictive for competition. presence of commercial courts, and underde- by rationalizing spending and strengthening veloped case management techniques within revenue mobilization has become an impor- the judiciary are other important issues in tant medium-term priority. A comprehensive need of redress. The recent creation of the assessment of contingent fiscal liabilities and Court of Appeals within the Supreme Court close monitoring of emerging fiscal risks will could possibly help to reduce the backlogs. also be important to having realistic assess- With regard to maintaining a prudent ments of available fiscal space and maintain- and supportive fiscal policy, one challenge ing fiscal sustainability. The fiscal challenges will be responding to the financing needs extend to public investment management, of the economy while keeping a sustain- especially to weaknesses in ex post evalua- able level of public and publicly guaranteed tion arrangements and monitoring mecha- debt, which now equals about 48 percent of nisms for midcourse adjustments and the lack GDP (figure 6.8). The upward debt trajec- of a well-developed and clear cost manage- tory has been driven by a public investment ment system over the course of the project life push that has coincided with declining exter- cycle. nal grant financing. Fiscal space has shrunk considerably as a result, which may affect Public Sector Accountability Rwanda’s ability to implement countercycli- cal macro management in the future. Grant A final set of governance challenges has to financing is likely to continue declining as do with strengthening accountability mech- Rwanda approaches middle-income sta- anisms between the local government and tus. This decline will result in higher public citizens. Representation of the most vulner- debt if the government offsets it with com- able and poor has increased in the councils. mercial borrowing. Rebuilding fiscal buffers The list of remaining challenges runs along C a p a b l e and A ccounta b l e S tate I nstitutions 263 several dimensions. District councils, often FIGURE 6.8  Public and publicly guaranteed debt as a percentage constrained by limited budget and staff, are of GDP in Rwanda, 2010–17 not always effective in listening to citizens’ needs. In addition, district councils need 60 capacity building for effective delivery of their mandate (GIZ 2013; MINALOC 2017). 50 48.0 The role of civil society organizations (CSOs) 44.5 in public policy remains limited as a result 40 36.4 of various capacity gaps. Citizens them- Percentage selves could be better empowered and better 30.4 informed to ensure that their views are heard 30 27.5 by district authorities. 20.5 22.1 21.5 Even though participation in ubudehe 20 20 and umuganda meetings and forums as well as citizens’ engagement in finding solutions 10 are high, their capacity in influencing deci- sions is limited. Citizens’ participation and 0 district-level accountability have been found 2010 2011 2012 2013 2014 2015 2016 2017 to be limited (Ndahiro 2015). Consequently, External central government debt Domestic debt citizens’ participation in the imihigo process External public-guaranteed debt Public sector debt also needs to be improved. I n recent years, C SOs’ capacit y to Source: IMF and World Bank 2017. Note: IMF = International Monetary Fund. enhance public accountability has increased but needs to grow further. Their capacity to mainstream citizens’ input in policy plan- regard, which needs to be broadened. The ning has improved. Although CSOs engage neighborhood security (irondo) and hygiene in government programs, organizational (isuku) initiatives, 21 which help to improve and institutional capacity gaps continue to security and sanitation in many localities, hamper the growth and role of CSOs in pub- are interesting examples of citizens’ mobili- lic policy. The government and some devel- zation for the provision of public goods. In opment partners have jointly put in place the this regard, China’s experience of solidarity CSO empowerment program that enhances groups for informal accountability and local CSOs capacity to spur accountability and provision of public goods (Heilmann 2008; service delivery to citizens. In 2016, the Tsai 2007) shows how associations and inter- government of Rwanda adopted a joint pro- est groups could help citizens to voice their gram with CSOs in order to improve the needs more effectively and contribute to implementation of social protection pro- improved service delivery in Rwanda. grams within communities. The interim Citizens’ limited information about, and evaluation of this program in pilot districts power over, service providers (schools and found that CSOs have an important role to health facilities) impairs service delivery. play in improving delivery and enhancing Parents’ involvement in school manage- accountability in pro-poor programs. It also ment through parent-teacher committees or revealed that openness on the part of dis- as members of parent-teacher associations tricts to work with and receive inputs from is, in most of cases, still limited. According CSOs can be improved. to a study conducted in Gasabo District in In principle, associations and interest 2013, parents are interested in the education groups can mobilize citizens to find local of their children, but large families and par- solutions (Boix and Posner 1998; Edwards ents’ high illiteracy rates, among other fac- and Foley 1998; Ehrenberg 1999). Rwanda tors, limit parental involvement in schools has already had some early success in this (Kabarere et al. 2013). A 2014 Transparency 264 FUTURE DRIVERS OF GROWTH IN RWANDA International assessment reveals that par- as sanitation, potable water, electricity, and ents rarely attend general meetings to dis- road and telecommunications connectiv- cuss issues related to school management or ity, for example) would give everyone a fair their children’s education. As global experi- chance to enjoy the fruits of growth. ence shows, this is a missed opportunity. For The governance structures will have to example, the provision of information to par- become adept at encouraging local solu- ents had a strong positive impact on learning tions, taking informed risks, ensuring effec- by teachers and students in Chile (Bruns and tive coordination within the government, Luque 2015) and Mexico (Gertler, Patrinos, engendering the commitment and trust of and Rubio-Codina 2012). nonstate actors (including the private sector), Remaining gaps in the provision of health and instituting highly developed feedback care highlight some of the limitations of loops. Past ingredients of success, such as e xisting accountability mechanisms. For ­ strong implementation capacity and verti- example, performance-based financing has cal accountability, will need to be comple- had impressive results in health outcomes, mented with elements that enable economic but clients’ influence over health service pro- institutions to be more innovative, receptive viders has largely failed to improve. The per- to feedback, and solution driven. Vertical formance culture outstrips the verification accountability can contain corruption in the of quality of services and does not always short to medium term, but, as soon as the rely on citizen monitoring and feedback. pressure from the top diminishes or gets dif- In the 2016 Citizens Report Card, citizens fused in a more complex setting, corruption said that they highly appreciate health infra- and other forms of institutional malaise can structure, such as buildings and facilities, reemerge (Khan 2017). but their appreciation is substantially lower Strong and inclusive growth over the with regard to services provided at hospi- next decades will entail specific institu- tals, health centers, and health posts (RGB tional requirements, closely aligned with the 2016a). Inadequacy of health staff (doctors framework described in figure 6.6. It starts and nurses) and scarcity of medicines are with continuing to build trust between the the critical pitfalls in the health sector (RGB executive, the bureaucracy, the private sec- 2016a). The Rwandan experience is similar tor, and the society more broadly. Otherwise, to that of many low-income countries with transaction costs will be higher, long-term performance-based funding. investments will be lower, and innovation (in the bureaucracy, in the economy, and in the society) will not happen. Trust, above all, An Agenda for Reforms is about building social capital within local An upper-middle-income economy (which communities by generating more voice and Rwanda aspires to have by 2035) will nec- improving the quality of citizen participation essarily be more complex, innovative, and in decision making. sophisticated, requiring more adaptive insti- The task of continuing to build trust and tutions. It will also be more inclusive, with enhance state effectiveness will, therefore, strong emphasis on providing equality of have three core institutional pillars: (1) inno- opportunity for all citizens. With income vative, coordinated, capable, and empowered inequality already at elevated levels (the bureaucracy and local governments; (2) effi- Gini ­c oefficient for income equals 0.45), cient market signals to allocate resources and Rwanda will need to be keenly mindful of enable strong cooperation between the state its future trends. Building a strong system and the business sector, backed by fiscal disci- of social protection to provide basic safety pline to match policies with the state’s finan- nets to the most vulnerable and expanding cial capacity; and (3) popular participation to access to quality public services (education, ensure that policies and programs are aligned health care, and basic infrastructure such with the needs of citizens. C a p a b l e and A ccounta b l e S tate I nstitutions 265 The chapter argues that all four com- targets” (Shyaka et al. 2016), particularly ponents (trust and the three institutional in joint imihigo. pillars) are essential for an effective state, • Emphasize qualitative surveys and a which, in turn, is necessary for sustained stronger focus on outcomes as well as high growth: the absence of any of the four processes. components (and, therefore, of state effec- tiveness itself) will yield unsatisfactory Strengthen Coordination within outcomes. Government Coordination within government could possibly be improved through the strength- Pillar 1: Enabling Innovative, ening of the delivery unit, 22 based on Coordinated, Capable Bureaucracy agencies in high-performing East Asia and Local Governments e conom ie s — for ex a mple , M a laysia’s Strong growth over the next decades will Performance Management and Delivery require well-structured, capable, and empow- Unit. This unit would have the author- ered civil service and local governments, with ity to coordinate public policy setting resources and other incentives to be innovative and implementation for some key cross- (even entrepreneurial). Such institutions are cutting issues, such as industrial policy or needed to inform decision makers and chal- local development. 23 Delivery units focus lenge them with evidence and analysis when on producing better results more quickly needed; share information and ­ coordinate through a combination of change-manage- actions within and outside the administration; ment techniques and approaches to public and take informed risks, experiment, scale up service improvement. Having a delivery successes, and learn from failures. Creating unit near the nexus of political power is a this type of institutions will require increas- mechanism for government leadership to autonomy of civil servants and local ing the ­ signal its focus on results and improve the government, strengthening capacity and coor- unit’s management capacity. The approach dination within the government, and making seems to work best when it is embedded progress on decentralization to encourage in a robust ­ o rganizational performance local initiatives and enable local transforma- management framework (which is already tion, both in relation to human security and present in Rwanda); focuses on establish- accountability challenges, as well as boosting ing high-profile, well-publicized priorities; local economic development. and uses high-frequency data to support improvement processes (Shostak et al. Continually Adjust Imihigo to Promote 2014). In Rwanda, it will be important to More Flexible Approaches look into the existing institutional setup On the basis of worldwide experiences of and how it could be adjusted to play that performance contracting, imihigo will need role fully and to facilitate coordination and to be adjusted continually and enforced greater delivery. with a view to enhancing innovation within the ­government. The following main chal- Strengthen Civil Service Capacity lenges need to be considered: Further progress is necessary to attract, retain, and motivate public servants to deliver • Improve interagency coordination and high-quality services. The foundations of consider increasing the importance of the an effective civil service cover several critical joint imihigo. dimensions, including merit-based recruit- • Improve outcome-based monitoring of ment, the right incentives and compensation imihigo and adopt “multiyear targets, for staff, right training at entry and on the job, with annual benchmarking as opposed to career progression and performance moni- the current practice of setting only annual toring, efficient human resource information 266 FUTURE DRIVERS OF GROWTH IN RWANDA FIGURE 6.9  The main elements for an efficient and capable civil systems, and organizational structures that service encourage high performance (figure 6.9). Rwanda has to find its own way to build a modernized public administration, but it Recruitment could learn from the experience of successful Staff and training and East Asian economies (box 6.4). availability Capacity upgrading RMI could be expanded further to meet of staff current constraints and future training needs of the civil service. Although not all civil ser- Motivation, vants need to be trained, creating a critical Performance pay, benefits, and career mass requires a significant scale of training Incentives and retention management (MacArthur 2012). RMI is planning to train policy as many as 2,000 civil servants a year. Skills audits and functional reviews should be car- ried out to identify the missing skills in the Organizational Human resources Structures and structure and information public sector, and a new approach is needed institutional restructuring systems environment to financing RMI. A successful example is the use of both public and private fund- ing by the Singapore Civil Service College. The government of Singapore provides a Source: Based on MacArthur 2012. minimum level of funding, while requiring BOX 6.4  Lessons from East Asia on strengthening public administration The East Asia experience offers several lessons that agencies (Cheng, Haggard, and Kang 1998). Political would be useful for Rwanda. First, public admin- leadership strived to insulate some key parts of the istration can be modernized through a variety of economic bureaucracy from other branches of gov- means. For example, civil service recruitment in ernment and from interest group pressures. Japan and the Republic of Korea is through a unified Finally, successful East Asian countries have civil service exam, which has never been the case adopted two main approaches to implementing indus- in Singapore or in Taiwan, China. Nevertheless, trial policies and coordinating public interventions: all of these economies have achieved a certain level either by leaving line ministries in charge, as in Japan of meritocratic recruitment, “complemented by a or Korea, or by creating “pilot agencies” or “task career structure that produces rewards commensu- forces,” as in Taiwan, China, or “statutory boards,” rate with those that capable individuals could attain as in Singapore. Despite these different approaches, in the private sector” (Evans 1998). coherence was achieved by strong continuity in the Second, civil servants should keep learning, economic technocrats, who shifted positions but adapting, and innovating. For example, ­ management maintained a vision of the overall development goals of state-owned enterprises and institutional organi- (Evans 1998). Moreover, in order to improve public zation of the state have evolved c ­ onsiderably over policy coordination, Taiwan, China, has emphasized the years in Taiwan, China. Public administrators promoting within the civil service rather than hiring need to feel empowered to disrupt and try new managers from outside. This has enabled top bureau- options (even if they ultimately fail). crats to establish relationships across various struc- Third, political authorities need to delegate some tures and administrations and contributed to esprit de powers to bureaucrats and trust that they will make corps and loyalty in the civil service. Over the years, the best technical decisions. Even though political several task forces and agencies have been incorpo- power was concentrated in the executive for several rated gradually back into the bureaucracy, but some decades, some independent decision-making author- small-scale units outside ministerial structures have ity was delegated to relatively insulated technocratic remained (Cheng, Haggard, and Kang 1998). C a p a b l e and A ccounta b l e S tate I nstitutions 267 ministries or departments and state-owned institutions that convey prices, define prop- enterprises (SOEs) to devote at least a share erty rights, enforce contracts and competi- of their budget to the trainings offered. The tion policies, and close informational gaps college has developed joint programs and is between buyers and sellers (Commission on funded partly by its development cooperation Growth and Development 2008). As part or technical assistance budget. However, to of its long-term development, Rwanda also adopt this model, RMI will have to demon- has to develop these key market institutions, strate that it remains up-to-date, has methods backed by the strong financial capacity of the for systematically collecting data on train- state. ing needs and tailoring its programs to meet those needs, and is delivering top-­ quality Strengthen the Policy Framework for Free training and capacity building, including and Fair Competition “after-sales service.” If successful, the gov- Competitive markets enhance private sector ernment could encourage donors to use RMI dynamism through cost reductions, innova- in their support for capacity building. tion, and productivity growth. As noted in Career development instruments and World Bank (2014), “Firms operating in a regular compensation reviews will become competitive environment are more likely to increasingly important to establish adequate innovate and to increase their productivity. incentives for civil servants, particularly as Competition boosts investment, generates private companies increasingly compete for employment, speeds up economic growth, scarce skills. 24 Although intrinsic motiva- and improves overall welfare. Competition in tion seems to be important in the Rwandan input markets is a key driver of efficiency and public sector, compensation for some groups productivity growth in sectors that use these of civil servants, such as teachers in primary inputs. Empirical evidence strongly supports schools, is low. Better career development the positive effects of well-enforced compe- instruments and nonfinancial incentives tition policy on productivity growth. Tough could be developed, especially in sectors enforcement against the practices of cartels where retention of scarce skills is crucial is an effective tool in reducing the negative (technical line ministries and the judicial impact of anticompetitive behavior, ben- system, for example). A specific issue con- efitting consumers with lower prices, direct cerns the health sector, where success has savings, and improvements in the variety and ­ depended on community health workers. It quality of goods and services.” will be important to design and implement a An effective competition framework package of targeted interventions to improve addresses four broad areas: agency effec- the recruitment, retention, performance, and tiveness, effective enforcement, probity of motivation of health workers, particularly investigations, and effective advocacy. First, given the possibility of an increase in the rate the effectiveness of the competition author- of attrition and a consequent “generational ity in encouraging compliance with the law change” in the pool of prospective candidates depends on the institutional design, degree of (LSTM, RBC, and UNICEF 2016).25 independence of technical decisions, strategic planning, and technical capacity of its staff. Second, characteristics of the legal frame- Pillar 2: Strengthening Market work, including the enforcement powers Institutions and Maintaining Fiscal granted and provisions that regulate how the Prudence agency will assess restrictive trade practices The evidence is compelling that function- and mergers, help to determine the effective- ing markets require well-defined rules of the ness of enforcement. Third, the reputation game, enforced transparently and predict- of the agency, which is based on confidence ably. Advanced economies (almost as a rule) in the probity of its investigations, is essen- have a system of highly evolved economic tial for effectiveness. Finally, advocacy for 268 FUTURE DRIVERS OF GROWTH IN RWANDA pro-competition regulations and elimination The competition policy framework needs of anticompetitive practices is important to to be expanded to incorporate competi- generate a competition culture that can create tive neutrality—the principle that govern- demand for enforcement actions and regula- ment activities should not be accorded an tions that improve market functioning. advantage over competing private sector Recognizing the importance of competi- activities—and it would be useful to expand ­ tive markets, Rwandan authorities have sev- the mandate of RICA to monitor SOEs eral initiatives to strengthen the competition and assess competitive neutrality issues. policy framework. Parliament enacted the Similarly, rules that designate joint ventures competition and consumer law in 2012 and and select private sector partners for SOEs a law creating a competition authority in would benefit from provisions to enhance 2013, which was replaced in August 2017 transparency and equal treatment of poten- by Law no. 31/2017 creating the Rwanda tial partners. Inspectorate and Competition Authority There is also a strong need to support (RICA). Law no. 36/2012 tackles practices RICA to develop its operational framework, that undermine, prevent, or restrict competi- enforce the competition law, and strengthen a tion (such as abuse of dominance and car- culture of competition in Rwanda. Despite a tels) and mergers that are likely to prevent strong legal mandate, RICA is not fully oper- or lessen competition significantly. However, ational, making the allocation of budget and RICA is not operational, and the competi- staff with sufficient expertise a high priority. tion law has not yet been fully enforced. In Korea, the Korea Fair Trade Commission In addition, the government reformed the plays a major role relative to the nation’s investment code to level the playing field largest business conglomerates (chaebols), generated by firm-specific incentives, but including the monitoring of limits on cross investment incentives still need to be made shareholdings and cross-debt guarantees more efficient and less distortive of competi- (box 6.5). tion (see chapter 4 of this report). Effective Finally, ongoing initiatives to eliminate implementation of these important regula- distortions generated by differential incen- tory measures remains important to achieve tives for competitors in the same sector are desired outcomes. important and need to be scaled up. BOX 6.5  The role of competition policy in regulatory reform in the Republic of Korea Competition policy in Korea was developed and multiple purposes, including the promotion of bal- enforced alongside an overall program of reforms anced development and fairness as well as free com- in the 1980s and 1990s. These reforms prompted petition and efficiency. It prevents and punishes anti- greater reliance on markets than on central govern- competitive practices, protects consumers and small ment direction to drive growth and achieve greater and medium enterprises, and is a key member of the openness and transparency in public institutions and Regulatory Reform Committee in charge of eliminat- major private enterprises. The basic competition law, ing regulations that restrict competition and affect the the Monopoly Regulation and Fair Trade Act, covers business environment. An important policy issue in all of the principal problems with competition policy: Korea was establishing an environment in which the collusion, monopoly, mergers, unfair practices, con- chaebols could compete on equal terms with other tract fairness, and consumer protection. market actors; the act also gave the Korea Fair Trade The Korea Fair Trade Commission, Korea’s inde- Commission the power to regulate the chaebols’ cor- pendent competition agency, plays a central role in porate and investment structure, establishing new major reform efforts as well as enforcement. It has units to specialize in those subjects. Source: OECD 2000. C a p a b l e and A ccounta b l e S tate I nstitutions 269 In the telecommunications and electricity and Brazil found that firms located in prov- sectors, regulatory barriers could be mitigated inces with more effective courts have greater through vertical separation rules and regula- access to credit (World Bank 2004). A study tions ensuring third-party access to essential focusing on Mexico found that states with facilities. In other sectors (air transport and better court systems have larger and more professional services, for example), regula- efficient firms (World Bank 2004). Effective tions that result in differential treatment of courts reduce the risks faced by firms and foreign suppliers or providers can discour- increase their willingness to invest. Firms age competition and their impact needs to in Brazil, Peru, and the Philippines report be monitored constantly and reviewed in the that they would be willing to invest more future. The Bar Association’s practice of set- if they had greater confidence in the courts ting binding minimum and maximum legal (World Bank 2004). fees needs to be reviewed with time because it Specific recommendations for Rwanda may restrain competition on pricing and goes follow from these challenges and are moti- against antitrust principles. vated by the impact of contract enforcement on long-term growth. These recommen- Strengthen Enforcement of Contracts dations include (1) creating small-claims When U.S. businesses were asked, “How courts or a fast-track procedure for handling likely would you say it is that the litigation small claims (short term), (2) establishing environment in a state could affect an impor- strong c ­ ooperation between different actors tant business decision at your company such in the commercial process (short term), as where to locate or do business?” two- (3) ­determining clear rules for adjournments thirds of respondents stated “very likely” or and case-flow management (short term), “somewhat likely” (U.S. Chamber Institute (4) promoting the use of alternative means for Legal Reform 2010). Efficient contract to resolve commercial disputes, includ- enforcement is essential to economic develop- ing arbitration (medium to long term), and ment and sustained growth (Ball and Kesan (5) ­continuing to develop and implement the 2010; Dakolias 1999; Dam 2006; Esposito, e-court system (medium to long term). Lanau, and Pompe 2014; Rosales-López Small-claims courts are specialized courts 2008). Economic and social progress can- with specific duties and powers to adjudicate not be achieved without respect for the rule and resolve small-value disputes through a of law and effective protection of rights, both special, more relaxed proceeding. 26 Fast- of which require a well-functioning judiciary track procedures represent another option, that resolves cases in a reasonable time frame where small claims are processed by the first- and is predictable and accessible to the pub- instance courts applying a different set of lic (Dakolias 1999; Sherwood, Shepherd, and or simplified procedural rules. Small-claims de Souza 1994). Economies with a more effi- courts or fast-track procedures play a special cient judiciary, in which courts can effectively part in building public trust and confidence enforce contractual obligations, have more in the judicial system (Ramsay 1996). They developed credit markets and a higher level of help to meet the objectives of efficiency and development overall (Dam 2006). A stronger cost-effectiveness by providing a mechanism judiciary is also associated with more rapid for resolving legal disputes involving small growth of small firms (Islam 2003). Overall, sums of money quickly and inexpensively enhancing the efficiency of the judicial sys- (Axworthy 1976; Ramsay 1998). In addition, tem can improve the business climate, foster they tend to reduce backlogs and caseloads in innovation, attract foreign direct investment, higher or specialized courts. Faster and less and secure tax revenues (Esposito, Lanau, costly dispute resolution is extremely impor- and Pompe 2014). tant to small and medium enterprises, which A study examining the efficiency of may not have the resources for long, costly courts in different provinces of Argentina litigation. 270 FUTURE DRIVERS OF GROWTH IN RWANDA Recognizing these benefits, Rwanda is • Speedy resolution of disputes. Small- already considering establishing a fast-track claims courts or fast-track proceedings procedure for handling small claims at the generally apply short time limitations for primary courts level. Once the procedure is the completion of action by the court and instituted, regular reviews of the small-claims litigants. For example, in Korea (ranked caseload and an assessment of the effective- number 1 on enforcing contracts in the ness of this special proceeding should be Doing Business 2018 rankings), a judge undertaken to ensure that it is meeting the may take steps to expedite the process, needs of court users as well as to identify such as having the party concerned file areas for improvement. In addition, the fol- an application for evidence even before lowing good-practice elements need to be the legal date for pleading. Similarly, in woven in (Gramckow et al. 2016): Norway, once the case is filed and not resolved through mediation, the judgment • Coverage and monetary caps. Applicability must be issued no later than three months to all civil cases is important so as not to from filing.29 leave room for interpretation and confu- sion. Capping the case at limited monetary To improve predictability for decision value is also important, 27 with scope to making in the courts, a systematic approach review the monetary thresholds regularly to training judges and court staff should be to account for different types of cases, considered, along with increased specializa- inflation, economic growth, and local tion. Specialized judges who have experi- factors. ence in the subject matter of their case can • Affordability. Affordability is key for be expected to feel greater confidence in small-claims courts or fast-track proceed- their judgment than their generalist coun- ings to work. The costs should be kept terparts (Baum 2009). Further, building significantly lower in such courts than in cooperation between different actors in the regular first-instance courts. 28 Various commercial process could be considered an forms of legal aid also should be avail- interim step, as judges get trained and expe- able to provide greater accessibility and rienced. Commercial dispute resolution is affordability, including free legal help, rep- constantly changing, requiring regular dia- resentation, and advice; fee waivers and logue between politicians, members of the exemption schemes; and the acceptance of judiciary, civil society, and the private sector self-representation. to stay abreast of the changes. This initiative • Simplified rules and informality. Small- could result, for example, in the organiza- claims courts or proceedings are based tion of regular workshops on the issues that on the use of straightforward specific might require special substantive legal or procedures, which are separate from the procedural knowledge. Regular cooperation rules of regular civil procedures, are eas- on such issues will help judges to educate ier for litigants to understand and follow, themselves using the feedback and experi- and do not require the parties to rely on ence of the private sector and other involved the assistance of lawyers. In Norway, the entities, whereas the private sector can rules allow unrepresented parties to file obtain practical information on the proce- a case and submit their defense verbally. dures from judges in order to avoid common In Sweden, the entire process, from filing mistakes that might result in unnecessary to issuance of a judgment, is informal. adjournments and delays. Constructive According to World Bank (2018), only 12 cooperation would also generate ideas to of the 133 economies in which a small- improve the procedural and substantive claims court or fast-track procedure is in issues on the legislative level. place mandate that parties be represented To tackle the challenge of underdevel- by a lawyer during such disputes. oped case management techniques within C a p a b l e and A ccounta b l e S tate I nstitutions 271 the judiciary, the following actions should Provision of financial incentives for parties be targeted: (1) limit adjournments to to attempt mediation or conciliation could unforeseen and exceptional circumstances also be considered. Nonetheless, such mea- and (2) maintain realistic calendars of event sures should not seek to make mediation a hearings and trials. Well-performing courts mandatory process (at least not for all cases) manage cases according to clear rules that before a case is filed at the court. Alternative establish meaningful hearing schedules dispute resolution should be seen not as a and limit adjournment options (Gramckow replacement for traditional litigation but as a et al. 2016). tool for resolving disputes in a timely, cost-­ To verify that a court is meeting legal effective, and transparent way. Rwanda has and organizational performance standards, been commended for digitalizing access to practical and meaningful measurement sys- justice, through the Integrated Electronic tems need to be in place (Gramckow et al. Case Management System, and further con- 2016). Rwanda maintains certain internal solidation of efforts can only yield positive statistics—for example, to track the number ­ outcomes. of cases resolved versus the number of incom- ing cases. Currently, such statistics are avail- Strengthen Enforcement of Property Rights able to the public only on an annual basis.30 Strong property rights underpin invest- Rwanda may consider making them publicly ment in market-based societies, leading available on a quarterly basis in order to to economic growth and new and better enhance transparency of the judiciary’s per- jobs. In low- and middle-income coun- formance, while providing a solid basis—and tries, many people depend on land for their broadening support—for any other reform. livelihoods: land is an important natural Alternative dispute resolution (arbitration, resource for the survival and development mediation, conciliation) has proved a valuable of people and ecosystems globally and, pillar in enhancing access to justice, bring- therefore, plays a critical role in the econ- ing rapid, consent-based dispute resolution omy of every nation. to businesses in many economies. Effective Rwanda has done well by focusing atten- systems of domestic commercial arbitration tion on securing property rights, particu- and mediation or conciliation matter to pri- larly land property rights. Its strong efforts vate investors. Especially in smaller cases, in this space have contributed to the strong having a neutral mediator or arbitrator saves reputation for governance and its leadership businesses time and money in resolving com- in Africa in the areas of land demarcation, mercial disputes and provides greater con- registration, and transfer. Consolidating, trol over outcomes and confidentiality (Love strengthening, and deepening the institu- 2011; Pouget 2013; Stipanowich 2004). tions that support property rights is an Rwanda scores a credible 2.5 points out of 3 important long-term priority in Rwanda. on the Alternative Dispute Resolution Index The following measures are priorities for of Enforcing Contracts regarding the qual- attention: (1) ensure proper enforcement of ity of judicial processes.31 However, further legislation on land expropriation for pub- improvement, particularly with regard to lic purposes, (2) maintain a robust system enhancing trust in the enforceability of con- of land rights management, (3) maintain a ciliation agreements, would be another good comprehensive and accurate land register, step forward. (4) provide public information on land trans- As a first step for increasing trust and action values, and (5) strengthen implemen- confidence in the conciliation process, the tation of the law on intellectual property. legislation could establish a clearer rule Irregular implementation of expropriation regarding enforceable settlement agree- procedures risks undermining recently con- ments or any other agreement that is reached firmed land rights. Rwanda’s formal legisla- out of a conciliation or mediation process. tion governing the expropriation process is 272 FUTURE DRIVERS OF GROWTH IN RWANDA generally considered appropriate, but there property, Rwanda ranks second only to are growing concerns over cases where the New Zealand, which it surpasses on some legal process has not been adhered to during submeasures of the ranking (World Bank implementation. Expropriation complaints 2018). Maintaining a robust system is nec- are among the top land-related challenges essary to secure these remarkable achieve- expressed by citizens in governance assess- ments. Skilled personnel are needed in ments conducted by RGB. three areas related to managing the land Expropriation in the national interest is register effectively: database and software often used to support new land uses man- development, business and strategy devel- dated by the master plans, with land being opment, and cybersecurity. Benchmarking expropriated at its current use value (agricul- Rwanda’s resourcing against other leading tural, residential), before seeing considerable small economies (that is, Denmark, Estonia, appreciation as it gets developed for commer- Latvia, and New Zealand) would highlight cial or industrial purposes. The profits accrue the key issues that require closer attention. to the government or are used as a financial Maintaining the transparency and account- inducement for developers. This practice can ability of land administration institutions is lead to perceptions that land is being expro- also important. In many countries, citizens priated from poor landowners at low values lose trust in such institutions owing to rent- to transfer to well-connected developers. The seeking activities. perceived inequality is exacerbated when Keeping the land register comprehensive, landowners are paid out at low values that accurate, and current is critical. With land do not enable them to purchase similar prop- registration completed, it is now important erty on the open market and where the pro- to ensure that all secondary land transac- cedures set out in the expropriation law are tions are captured in the land register. There not followed. Going forward, Rwanda may have been recent reports (for example, GCC look at peer lessons in rapidly growing cit- Ltd. 2016) of informal secondary land trans- ies in the global South for alternative inno- actions. The underlying reasons—includ- vative approaches to spur people-centered ing a flat transaction fee of FRW 27,000 urbanization. (irrespective of size, location, and value), The issue can be best addressed by restrictions on the subdivision of agricul- reviewing ministerial orders on expropria- tural land smaller than 1 hectare, costs to tion and otherwise ensuring fair and timely subdivide land, and lack of awareness of compensation for land; ensuring consistent procedures—should be looked at system- application of current laws and regulations; atically and addressed. A system of cross- creating extensive public awareness, together subsidy could be c ­ onsidered—for example, with clarity and public consultations (as pro- to replace the current flat fee—so as not to vided for by the law) as to what constitutes deter the less well-off from registering land “national interest” for land expropriation; transactions. Sustained emphasis is needed establishing rapid intervention to ensure to build greater awareness of the importance that expropriation procedures are respected and benefits of recording land transactions, and to advocate for people whose property given that land historically has been trans- has been expropriated without following the acted informally. Further, as new types of established procedures; and building bet- land rights (such as ­ multilevel urban residen- ter public awareness and consultation on tial buildings) take shape, more awareness expropriation matters. Rwanda has made will be needed on how these buildings are remarkable strides toward establishing a owned and managed, using, for example, strong land rights management s ­ ystem—by the new condominium law in Rwanda as a far the best in Africa and on many measures reference point. If restrictions on the sub- among the best in the world. On the Doing division of agricultural land are to remain, Business 2018 ranking for registration of alternative approaches such as co-ownership C a p a b l e and A ccounta b l e S tate I nstitutions 273 and land use consolidation may be needed 2010. 32 Although this legislation has gone to prevent transactions from happening a long way toward establishing a strong informally. industry, limited information still results in Use of e-titles (pdf format) could be a wide range of valuations. The establish- offered to reduce paperwork and cost. This ment of “reference prices” to be used pre- option could be instituted incrementally, dominantly for land acquisition by the state starting with urban areas where access to the offers some guidance, but the wide range Internet is relatively high. Government could of published prices limits their u ­ tility. The also consider partnering with the private banking sector would also benefit from sector in the provision of some land-related better information, because it needs better services, including certification of land trans- price discovery to underpin its mortgage actions (private notary services). business. Financial institutions complain With a comprehensive land disputes data- that insufficient hard information makes base in place, it will be important to record it difficult to assess land values accurately, and expedite the resolution of outstanding l eading to a reluctance to extend further ­ land disputes. A web service currently being credit to this sector and restricting the developed to link the land register with the growth of an asset-backed wholesale lend- e-court system is an important step, although ing industry, which would inject further the register does not yet provide a clear cat- capital into the sector. egorization of disputes. The law on intellectual property draws As Rwanda rapidly urbanizes and under- heavily on model legislation. Although the takes structural reform in agriculture, better legislation is considered good practice, the information on market values of all catego- institutions supporting it are not yet well ries of land will support the transition and developed and awareness of the legisla- enable people to understand, value, and tion is low. At present, financial and human preserve their rights and lay stronger foun- resources to implement the law are limited. dations for a growing property industry. The government needs to review the current Rwanda has a world-class land management legal framework to support implementation system and has gathered information on and to increase awareness of the value of the sales price of private land transactions intellectual property among developing busi- for nearly five years. However, the price of nesses. It also needs to enhance the capacity land has been recorded in a database in only to carry out due diligence required for trade- about half of 112,000 land transactions. mark or brand registration to prevent avoid- Gathering the residual information should able disputes. be relatively simple, because it is recorded on the pdf files attached to each transaction. Tighten Fiscal Management for Stronger This information will be invaluable, because State Financial Capacity it will enable the establishment of market- Maintaining fiscal prudence over the long based valuation rolls to underpin property term requires careful planning. Grant financ- valuations. Following the example of other ing is likely to continue to decline, even countries with developed land markets, this as public investment remains a priority to information should be made public. Rwanda achieve the government’s ambitious devel- can again make use of the new technologies opment targets. Therefore, eliminating inef- to leapfrog legacy printing technologies and ficient expenditures and increasing revenues disclose this information online. are necessary to limit borrowing and to cre- The land valuation industry needs accu- ate some fiscal cushion for withstanding pos- rate transaction values. The valuation pro- sible shocks. A comprehensive assessment of fession in Rwanda is relatively new: the law contingent liabilities as a result of the gov- establishing and organizing the real prop- ernment’s proactive interventions and close erty valuation profession was enacted in monitoring of emerging fiscal risks also are 274 FUTURE DRIVERS OF GROWTH IN RWANDA important to assess available fiscal space fully integrated with procurement and pay- realistically. roll systems. However, the system could be Improving the allocation of resources and used more effectively in monitoring the allo- operational efficiency of budget institutions cation of funds to the service delivery units is critical. The government should focus on and in tracking analyses. Last, independent adjusting the composition of spending to performance evaluations of the efficiency and enhance the allocative efficiency of public effectiveness of service delivery programs by finance. Finding a better balance between RGB will have to be strengthened to ensure spending priorities is key for ensuring the greater coverage and more impact on national optimal use of scarce public resources, espe- transformation. cially given the expected decline in grant Various aspects of debt management financing. A better balance needs to be also need to improve further. Producing achieved between public investments that statistical bulletins and annual reports on are expected to generate high returns over government debt operations could help to the long term and those that have potential improve transparency in debt management. high economic or social returns in the short The final issue has to do with the man- term according to market signals regarding agement of mineral revenues, which are scarcity. expected to pick up sharply in the coming Even though revenue targets are usually years. Volatility in commodity prices has had exceeded, revenue collection remains below only a limited impact on fiscal policy so far potential. This may be due in part to the because of the artisanal nature of the sec- major role played by the informal sector in tor and the relatively low revenues that the the economy. However, according to the Tax government collects from mineral exports. Administration Diagnostic Assessment Tool However, recent discoveries are expected to (IMF 2015), revenue collection is impaired mineral exports, so establishing an opti- raise ­ by unreliable tax records, questionable on- mal fiscal regime to capture an efficient and time tax-filing data, inadequate generation fair share of mineral wealth is a policy pri- and archiving of internal documentation, ority. Government may envisage capturing a difficulty in obtaining data on tax arrears, larger share of mineral revenues through the and lack of an independent survey of the pub- involvement of SOEs in the sector. A portion lic’s confidence in the RRA. When Rwanda of mineral revenues could be deposited in a reaches a much higher level of development, sovereign wealth fund to limit short-term vol- RRA management will have to improve its atility in revenues due to swings in mineral documentation and archival practices and prices. to expand the taxpayer database. The low ratio of local government revenue to GDP Pillar 3: Strengthening the State’s indicates that there is still scope for improve- Accountability to Citizens ment in local government revenue collection in Rwanda (see chapter 3 of this report for a There is a risk of disconnect between citi- discussion of increasing local revenues from zens and government if citizens’ views are improved land use). not reflected adequately in local govern- Rwanda’s public financial management is ment services. A feedback loop is needed reasonably strong and improving, but more between civil society and local governments can be done to strengthen the operational and service providers. A lack of participa- efficiency of fiscal institutions. The fiscal tion may ultimately have a negative impact accounting system needs to be made compli- on the political settlement and in the short ant with international standards. The govern- term may undermine the effectiveness of ment has rolled out the Integrated Financial investments, for example, in the education Management Information System, which is sector. C a p a b l e and A ccounta b l e S tate I nstitutions 275 The Rwandan state has opted for a of government expenditures were audited ­ itizen- ​ c c entered governance strateg y. (in fiscal 2013/14), and 60 percent of the Different programs have been put in place recommendations in the OAG report were to ensure that this approach is effective, implemented (PEFA Secretariat 2017). although some of them may need further The legally independent Ombudsman’s consolidation. The government of Rwanda Office and RGB have an important man- can become more effective in the use of pub- date (among others) to carry out the fight lic resources, while containing any future against corruption and monitor service conflict through greater public involve- delivery. Strengthening checks and balances ment and feedback in decision making. among the branches of government would The capacity and motivation of citizens to help to increase accountability. aggregate and voice their concerns and to Strengthening the operational and out- participate with state actors in assessing the reach capacity of the media will be key to performance of service delivery are critical monitoring public policy decisions and (Brinkerhoff and Wetterberg 2016; World implementation on behalf of the popula- Bank 2003b). 33 However, participation tion. In 2013, the media sector was moved does not lead to more accountability with- from statutory to self-regulation. “Media out local political support for such involve- self-regulation is a joint endeavor by media ment and discipline imposed by higher levels professionals to set up voluntary editorial of government (Brinkerhoff and Wetterberg guidelines and abide by them in a learn- 2016). These two factors are essential to ing process open to the public. By doing so, ensure that decentralization can foster social the independent media accept their share of accountability. responsibility for the quality of public dis- course in the nation, while fully preserv- Improve Accountability over the Executive ing their editorial autonomy in shaping it” Reform measures can seek to strengthen (OSCE 2008). checks and balances further to improve The Rwanda media self-regulatory body accountability over the executive. Several was set up to regulate the daily functioning institutions created in recent years have of the media and the conduct of journal- made progress in strengthening over- ists (2013 media law, art. 4). The Rwanda sight over the executive branch, but their Media Barometer indicates that, over- independence and capacity need to be all, the level of media development in the strengthened. Parliamentary oversight of country stood at 69.6 percent in 2016 (up the executive is increasingly effective. For from 60.7 percent in 2013 when the first example, the Public Accounts Committee Media Barometer was published), reflect- was created in 2011 to scrutinize external ing the positive impact of the reforms audit reports and enforce audit recom- (RGB 2016c). Complaints to the Rwanda mendations. The committee has conducted M e d i a C o m m i s s i o n h ave g r adu a l l y in-depth hearings on audit findings, cov- increased over the last years, along with the ered by radio and national television and number of media houses. with attendance by senior officials, minis- Access to information is also essential to ters, and the like (PEFA Secretariat 2017). create more space and opportunity for the The Office of the Auditor General (OAG) public to participate meaningfully in the has built its own credibility as Rwanda’s affairs of state. Limited use of the access- supreme audit institution by contribut- to-information law shows that further out- ing to improved public financial man- reach is necessary; even more important, it agement in line with standards of the may reflect loopholes in the law. The big- International Organization of Supreme gest weaknesses are in the areas of appeals Audit Institutions. More than 80 percent and promotional measures. Therefore, the 276 FUTURE DRIVERS OF GROWTH IN RWANDA current law likely still has a limited impact. The revenue-raising powers of subna- Rwanda is ranked 56 out of 111 in the global tional governments should also be increased. right-to-information rating.34 This would allow districts to have a greater level of control regarding their own rev- Strengthen Accountability of Local Officials enues, which would allow them satisfy citi- and Service Providers zens’ needs and preferences. Decentralization should be deepened to Strengthening citizens’ participation in generate adaptive and innovative local decision making at the local level will bal- governance, to increase accountability, ance government power and spur ownership and to overcome inefficiencies in coordi- and localization of centrally driven devel- nation and in the use of public resources. opment. An essential step is to continue to One of the most important steps, in order adapt the imihigo system. It will be cru- to increase accountability, is to overcome cial to ensure that associations of citizens inefficiencies in the coordination and allo- participate in determining imihigo objec- cation of public resources. With this end in tives, setting targets, and selecting invest- view, a comprehensive package should be ment projects; to provide more flexibility in defined. A new framework to strengthen monitoring performance; and to use more sectoral decentralization should go hand qualitative information in performance in hand with corresponding fiscal decen- evaluations. In addition, it will be important tralization adjustments. Local government to enhance the role and capacity of the dis- should use free programmable resources to trict councils to channel input from citizens satisfy citizens’ needs and preferences and into decision making and feedback to local to invest in local economic development government and to articulate citizens’ needs on the basis of the locality’s comparative effectively in local development policies. advantages, while ensuring overall align- Improvements in the enabling environment ment with national priorities and goals. for citizens organizations, for deep commu- In the short term, it will be desirable to umuganda, and nity initiatives such as post-­ restructure the system of conditional or ear- for student-parent committees should be marked grants (see also the discussion under considered, and efforts should be expanded pillar 1). The conditional grant program to include feedback loops. These community should be conceived as supporting major initiatives also can be used for quality assur- thrusts in central government policies. ance. Indeed, community initiatives can work Performance-based grants also should be to uphold standards of service and inform the explored to drive central government poli- government of inconsistencies in the regula- cies while ensuring performance. tions or enforcement. Annex 6A  Rwanda’s Territorial Organization Rwanda is divided into two levels of admin- are integrated into districts, but these enti- istration: the central level and the local ties do not have administrative and finan- self-governments. Local self-government in cial autonomy (figure 6A.1). The 2005 Law Rwanda is formed at the district level. The on Territorial Organization consolidated provinces and the City of Kigali are decon- 12 provinces into 4 and 106 districts into 30. centrated administrative units of the cen- In addition, the City of Kigali serves as a spe- tral government. Sectors, cells, and villages cial city-province. C a p a b l e and A ccounta b l e S tate I nstitutions 277 FIGURE 6A.1  Territorial organization in Rwanda Central government Guverinoma Province Intara District Akarere Sectors Imirenge Cells Utugali Source: 2015 Law on Territorial Organization. Annex 6B  Investment as a Share of GDP FIGURE 6B.1  Gross fixed capital formation as a share of GDP in Rwanda, 1965–2015 30 25 20 % of GDP 15 10 5 0 65 67 69 71 73 75 77 79 81 83 85 87 89 91 93 95 97 99 01 03 05 07 09 11 13 15 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 20 Source: World Development Indicators data (World Bank, various years). Note: Gross fixed capital formation (formerly gross domestic fixed investment) includes land improvements (fences, ditches, drains, and so on); plant, machinery, and equipment purchases; and the construction of roads, railways, and the like, including schools, offices, hospitals, private residential dwellings, and commercial and industrial buildings. According to the 1993 System of National Accounts, net acquisitions of valuables are also considered capital formation. 278 FUTURE DRIVERS OF GROWTH IN RWANDA Annex 6C  Overview of the Public Investment Management Assessment Questionnaire Rwanda has a well-publicized strategic guid- guidelines in their implementation based on ance for preparing public investment project e-procurement. Costs have escalated in a proposals at central and subnational levels, few cases, mostly as a result of unforeseen including long-term national development events. strategies (Vision 2020 and the Economic In project adjustment and monitoring, Development and Pover t y Reduc tion implementing agencies are required to pre- Strategy, 2013–18) as well as sector strategy pare progress reports, which are compiled papers and district development strategies. by MINECOFIN for a quarterly report- The overall public investment management ing to the Council of Ministers on project is guided by the National Investment Policy. performance. In rare cases, projects may be All projects are submitted to the Public canceled because of changes in situation, Investment Committee for approval to be but no clear guidance is available for this considered in the budgeting process. This process. screening includes a multistage process, rank- In project evaluation, there is no practice ing, and prioritization and planning consul- of ex post evaluation, including formal insti- tations. On the basis of recent data, about 60 tutional arrangements. However, the new percent of project proposals are rejected at National Investment Policy provides for ex this stage. post evaluation. Guidelines on their conduct In project appraisal, the Ministry of Finance still need to be developed. and Economic Planning (MINECOFIN) plays the main role. This process is governed by the National Investment Policy and the Guidance Notes on National Feasibility Study, which provides 1. However, according to World Governance guidance on investment planning process and Indicators, Rwanda’s ranking on voice and procedures. For projects above a threshold of accountability remains low. FRW 750 million (less than US$1 m ­ illion), 2. T he Rwanda Governance Scorecard is a cost-effectiveness and benefit ­ a nalysis is composed of eight governance indicators required. There are no mechanisms for inde- developed based on international guide- lines, recognized international indexes, pendent project appraisal, however. and Rwandan homegrown indicators (RGB In project selection, MINECOFIN leads 2016a, 3). Apart from measuring perfor- the planning and budget consultations with mance over time, the scorecard also serves as all institutions. Project selection is based an advisory tool, with each edition concluding on three criteria: desirability, achievabil- with key recommendations for improving the ity, and viability. Projects not approved at worst-performing indicators. an earlier stage by the Public Investment 3. It also includes conflicts of interest and ethical Committee cannot be considered for selec- misconduct. tion by MINECOFIN. Financing of the 4. The Rwanda Social Security Board also has projects is usually adjusted downward in this invested in real estate as part of portfolio diver- process, with an average approved budget at sification. Taking this investment into account, it is possible that public investments through 60–65 percent of initial requests. public agencies reached 16 percent of GDP in In project implementation, no specific 2013–15. This figure does not include invest- guidance specifies institutional arrangement, ments made by state-owned enterprises, which roles, responsibilities, and requirements for have a prominent role in the economy. Public efficiency and accountability. All ministries investments play an important role in Rwanda, and other budget agencies, however, are where overall investments are about 25 percent required to prepare and publish their pro- of GDP. For the total trend of capital forma- curement plans and follow the procurement tion, see figure 6B.1 in annex 6B. C a p a b l e and A ccounta b l e S tate I nstitutions 279   5. Most of net lending has gone to SOEs, such charged by districts and determining their as Kigali Convention Center and Rwanda limits. Air, for investment. The government also 11. These guidelines include the 2006 revised used a special investment vehicle for Kigali Fiscal Decentralization Strategy, 20 07 Convention Center, which, in addition to Rw a n d a D e c e n t r a l i z a t i o n S t r a t e g i c receiving net lending from the budget, also Framework, and performance-based financ- borrowed from commercial banks with a gov- ing in the health sector. ernment guarantee. 12. Umuganda refers to the tradition, preva-   6. For the results of a recent public investment lent in the Great Lakes region, of obliga- management assessment, see annex 6C. tory communal labor on public projects.   7. Aspects such as open competition for award of In Rwanda, umuganda takes place on the contracts and complaints mechanisms relating last Saturday morning of every month. to procurement, funding flows during budget Following the physical work, a meeting is execution, effective internal controls (including held where, in theory, the communal work commitment controls), and internal audit were carried out that day is evaluated and plans the main strengths cited in the International for further works are discussed and agreed, Monetary Fund’s assessment (Dabla-Norris together with community issues and plans et al. 2012). The assessment found relatively more broadly. weak performance in ex post evaluations and 13. A website was created in January 2015 to external audits, the integration of recurrent facilitate requests for information (www​ and investment expenditures in the budget, and .sobanukirwa.rw). Since introduction of public access to key fiscal information. the law, 21 appeals have been lodged with   8. The first plan to reform the civil service, the Ombudsman’s Office for having been adopted in 1998, included, among others, refused access to information (15 of these downsizing, decentralization of personnel, were in 2015–16). The RGB carries out uniform recruitment policy, and stronger information campaigns regarding this leg- selection of candidates. islation through radio shows and meetings   9. The National Decentralization Policy was with citizens. revised in 2013. 14. According to the access-to-information web- 10. The Laws Determining the Organization and site, only 162 requests were made in the last the Functioning of the Provinces, Districts, two years (www.sobanukirwa.rw). Sectors, Cells, and Villages (adopted in 2006), 15. Health facilities submit invoices contain- defining the composition, organic structure, ing service delivery data on each of the per- and competencies of these entities and regu- formance indicators to the District Steering lating entities’ operations in a comprehen- C om m it tee i n cha rge of veri f ic at ion. sive manner; the Law on the Organisation Additional layers of verification also exist at and Functioning of the Different Layers of the hospital and national level. the Government and the Law Establishing 16. Bevan and Hood (2006) explain how the for- the Source of Revenue for Districts and Its mer Soviet Union achieved dramatic increases Management (2002), defining the financial in production from the 1930s to the 1960s sources at the disposal of the districts; the based mainly on increases in quantity but not Minister of Finance decrees; the Law 59/2011 in quality, which became increasingly prob- of 31/12/2011, establishing the sources of rev- lematic. They further explain, “In the Soviet enue and property of decentralized entities system, as all bodies responsible for supervis- and governing their management; Ministerial ing enterprises were interested in the same Order no. 005/12/10/ TC of 22/06/2012, success indicators, the supervisors … con- determining the modalities for implemen- nived at, or even encouraged, gaming.” tation of Law no. 59/2011 of 31/12/2011, 17. This statement is based on interviews. establishing the source of revenue and prop- 18. Of the 20 low- and middle-income countries erty of decentralized entities and governing in the product market regulation database their management; and Presidential Order with updated information as of 2013 or 2014, no. 59/01 of 15/09/2009 modifying and Rwanda ranks 11 (and 46 in the full sample of complementing Presidential Order no. 02/01 58 countries). of 31/03/2008, establishing the list of fees 19. This statement is based on interviews. 280 FUTURE DRIVERS OF GROWTH IN RWANDA 20. “Ubudehe provides a mechanism for the map- 29. See Article 7 of the Trial of Small Claims Act ping of poverty levels through a participative of the Republic of Korea. categorization of households in each village 30. Annual reports are published at the website by the community members themselves. It of the Judiciary of Rwanda (http://judiciary​ also allows community participation in select- .gov.rw/media_house/reports/judicial_reports​ ing the most critical poverty challenges and to .html). The website provides a combined over- agree on appropriate practices to meet these view of the whole judiciary, and information needs both at community and household on specific courts is limited to the clearance level. [It] is a traditional Rwandan practice rate reports, which are not easily accessible of mutual help or mutual assistance among (the annual report is lengthy and difficult to people in order to solve their problems” (RGB navigate). 2014). 31. See more at http://www.doingbusiness​.org​ 21. Other similar initiatives include community /data/exploreeconomies/rwanda#enforcing​ mediators (abunzi) and community agricul- -contracts. tural advisers (abajyanama). 32. Law no. 17/2010 of 12/05/2010. 22. Numerous countries have established such 33. The capacity and motivation of citizens to par- structures in the last two decades, one of ticipate and engage are called the “short route the first being the delivery unit in the United of accountability” in the World Development Kingdom. Report 2004 (World Bank 2003b). 23. For more details on the prerequisites for a 34. See http://www.rti-rating.org/country-data/. delivery unit to be successful, see Shostak et al. (2014). 24. The total public wage bill in Rwanda (mea- sured as percentage of GDP) is still low References (less than 5 percent of GDP and less than Axworthy, C. S. 1976. “Controlling the Abuse of 20 ­percent of total public expenditures) when Small Claims Courts.” McGill Law Journal benchmarked internationally, whereas some 22: 480–95. professions may be paid at a low level com- Ball, G. G., and J. P. Kesan. 2010. “Judges, Courts, pared with regional benchmarks. and Economic Development: The Impact of 25. A long-term strategy is needed for the future Judicial Human Capital on the Efficiency and role of community health workers, including Accuracy of the Court System.” Paper pre- efforts to increase their integration into the sented at the 15th “Annual Conference of the public service. International Society for New Institutional 26. Small-claims courts usually use informal Economics,” Stanford University, Stanford, hearings, simplified rules of evidence, and CA, June 16–18, 2011. more streamlined rules of civil procedure Baum, L. 2009. “Probing the Effects of Judicial and typically allow the parties to represent Specialization.” Ohio State University. themselves. Bevan, G., and C. Hood. 2006. “What’s Measured 27. For example, in the Small-Claims Court in Is What Matters: Targets and Gaming in the Washington, DC in the United States, the English Public Health Care System.” Public limit is currently set at US$5,000, while the Administration 84 (3): 517–38. Small-Claims Tribunal in Singapore has a Boix, C., and D. N. Posner. 1998. “Social cap of S$10,000, which can be increased to Capital: Explaining Its Origins and Effects on S$20,000 with the consent of both parties. Government Performance.” British Journal of However, establishing a cap on small-value Political Science 28 (4): 686–93. cases should not be final. Booth, D., and F. Golooba-Mutebi. 2012. 28. For example, the rules of the Small-Claims “Developmental Patrimonialism: The Case of Court in Washington, DC stipulate that, for Rwanda.” African Affairs 111 (444): 379–403. claims of US$500 or less, the cost for filing a Brinkerhoff, D., and A. Wetterberg. 2016. new claim is US$5; for claims of US$500 or “Gauging the Effects of Social Accountability more and up to US$2,500, the fee is US$10; on S er vices , G over na nce, a nd Citi zen and for claims over US$2,500 and up to Empowerment.” Public Administration US$10,000, the fee is US$45. Review 76 (2): 274–86. C a p a b l e and A ccounta b l e S tate I nstitutions 281 Bruns, B., and J. Luque. 2015. Great Teachers. of East Asian Economic Policies.” Journal of Washington, DC: World Bank. Development Studies 34 (6): 66–86. Campbell, T., and H. Fuhr. 2004. Leadership Freedom House. 20 07. C ountries at the and Innovation in Subnational Government: Crossroads. Washington, DC: Freedom Case Studies from Latin America. W BI House. https://freedomhouse.org /report​ Development Studies. Washington, DC: World /­c ountries-crossroads/countries-crossroads​ Bank. http://openknowledge.worldbank.org​ -2007?page=140&edition=8. /­handle/10986/15023. Fukuyama, F. 1995. Trust: The Social Virtues and the Chemouni, B. 2017. “The Politics of Core Creation of Prosperity. New York: Free Press. Public Sector Reform in Rwanda.” ESID GCC (Growth Consultants for Change) Ltd. Working Paper 88, University of Manchester, 2016. “A n I nvestigation into I nformal Manchester, U.K. www.effective-states.org. Transactions under the LTR Programme Cheng, T.- J., S. Haggard, and D. Kang. 1998. in Rwanda.” Report sponsored by Rwanda “Institutions and Growth in Korea and Natural Resources Authority and Department Taiwan: T he Bureaucracy.” Jour n al of for International Development. Development Studies 34 (6): 87–111. Gertler, P. J., H. Patrinos, and M. Rubio-Codina. Commission on Growth and Development. 2008. 2012. “Empowering Parents to Improve The Growth Report: Strategies for Sustained Education: Evidence from Rural Mexico.” G ro w t h a n d I n c l u s i ve D e ve l o p m e n t . Journal of Development Economics 99 (1): Washington, DC: World Bank. 68–79. Cox, G. W., D. C. North, and B. R. Weingast. G I Z (G er m a n A genc y for I nter n at ion a l 2015. “The Violence Trap: A Political- C o op erat ion). 2 013. A n a ly s i s of t he Economic Approach to the Problems of Involvement of the District Councils in Development.” Hoover Institution Working Ensuring Effective Financial Management of Paper, Stanford University, Stanford, CA. District Resources. Bonn: GIZ. Crisafulli, P., and A. L. Redmond. 2012. Rwanda, Gramckow, H., O. Ebeid, E. Bosio, and J. L. Silva Inc.: How a Devastated Nation Became an Mendez. 2016. Good Practices for Courts: Economic Model for the Developing World. Helpful Elements for Good Court Performance New York: St. Martin’s Press. and the World Bank’s Quality of Judicial Dabla-Norris, E., J. Brumby, A. Kyobe, Z. Process Indicators; Key Elements, Lessons Mills, and C. Papageorgiou. 2012. “Investing Learned, and Good Practice Examples. in Public Investment: An Index of Public Washington, DC: World Bank Group. Investment Efficiency.” Journal of Economic Hausman, D. 2011. “Rebuilding the Civil Service Growth 17 (3): 235–66. after War: Rwanda after the Genocide: 1998– Dakolias, M. 1999. “Court Performance around the 2009.” Innovations for Successful Societies, World: A Comparative Perspective.” Technical Princeton University, Princeton, NJ. https://­ Paper 430, World Bank, Washington, DC. successfulsocieties.princeton.edu/sites/success​ Dam, K. W. 2006. “The Judiciary and Economic fulsocieties/files/Policy_Note_ID163.pdf. Development.” John M. Olin Law and Heilmann, S. 2008. “Policy Experimentation in Economics Working Paper 287 (Second Series), China’s Economic Rise.” Studies in Comparative University of Chicago Law School, Chicago. International Development 43 (1): 1–26. Edwards, B., and M. W. Foley. 1998. “Civil Society Hertie School of Governance. 2014. Background and Social Capital beyond Putnam.” American paper on Rwanda for ANTICORR. Hertie Behavioral Scientist 42 (1): 124–39. https://doi​ School of Governance, Berlin. .org/10.1177/0002764298042001010. IMF (International Monetary Fund). 2015. Ehrenberg, J. 1999. Civil Society: The Critical “TADAT Rwanda—Report Assessment.” History of an Idea. New York: NYU Press. IMF, Washington, DC. Esposito, G., S. Lanau, and S. Pompe. 2014. IMF and World Bank. 2017. “Staff Report for “Judicial System Reform in Italy: A Key the 2017 Article Iv Consultation, Seventh to Growth.” IMF Working Paper 14/32, Review under the Policy Support Instrument, International Monetary Fund, Washington, DC. and Second Review under the Standby Credit Eva n s , P. 1998. “Tra n sferable L e s son s? Facility—Debt Sustainability Analysis Update.” Re-Examining the Institutional Prerequisites IMF and World Bank, Washington, DC. 282 FUTURE DRIVERS OF GROWTH IN RWANDA I sla m , R . 20 03. “Do More Tra nspa rent MINALOC (Ministry of Local Government), G over n m ent s G over n B e t ter?” Pol ic y Rwanda. 2006. Making Decentralized Service Research Working Paper 3077, World Bank, Delivery Work in Rwanda: Putting the People Washington, DC. at the Center of Service Provision. Kigali: K aba rere, V., L . N. M a kewa, T. Muchee , MINALOC. and E. Role. 2013. “Parental Involvement ———. 2017. Assessment of the Impact of in High- and Low-Performing Schools in Decentralization Policy Implementation in Gasabo District, Rwanda.” International Rwanda (2001–2017). Kigali: MINALOC. Journal about Parents in Education 7 (1): M I N E C OF I N (M i n i s t r y of Fi n a nc e a nd 30–42. Economic Planning), Rwanda. 2007. “Guide Kagame. P. 2015. “Homegrown Solutions Were to the District Development Planning and Born out of High Ambitions and Scarce Budgeting Process.” MINECOFIN, Kigali. Resources.” Speech before a delegation from ———. 2009. “Planning and Budgeting Guidelines St. Andrews University, Kigali, March 24. for L ocal G over n ment C onsolidated.” http://paulkagame.com/?p=4019. MINECOFIN, Kigali. Kaufmann, D., A. Kraay, and M. Mastruzzi. ———. 2013. “Planning and Budgeting Guidelines 2010. The Worldwide Governance Indicators: for Decentralized Development Projects.” M e t ho d o l og y a n d A n a l y t i c a l I s s u e s . MINECOFIN, Kigali. Washington, DC: Brookings Institution Press. Ministry of Public Service, Skills Development, Khan, M. 2017. “A New Approach to Tackle and Labor, Rwanda. 2005. “Reform Paper: Corruption: Lessons from a Decade of Anti- Salary Structure in the Public Sector.” Ministry Corruption Programmes.” Presentation at a of Public Service, Skills Development, and high-level conference on anti-corruption, Dar Labor, Kigali, September 5. es Salaam, June. Nd a h i ro, I . 2 015. “E x pl a i n i ng I m ig i ho Knack, S., and P. Keefer. 1997. “Does Social Performance in Gicumbi District, Rwanda: Capital Have an Economy Payoff? A Cross- The Role of Citizens Participation and Country Investigation.’’ Quarterly Journal of Accountability (2009–2014).” Master’s thesis, Economics 112 (4): 1251–88. Erasmus University, The Hague. https://thesis​ La Porta, R., F. Lopez-de-Silanes, A. Shleifer, .eur.nl/pub/32926/. and R. W. Vishny. 1997. “Trust in Large North, D. C. 1991. “Institutions.” Journal of Organizations.” American Economic Review Economic Perspectives 5 (1): 97–112. 87 (2): 333–38. OECD (Organisation for Economic Co-operation Lee, K. Y. 2000. From Third World to First: The and Development). 2000. “Korea—The Role Singapore Story, 1965–2000. New York: of Competition Policy in Regulatory Reform.” Harper. Country Studies, OECD, Paris. Love, I. 2011. “Settling out of Court.” Financial O S C E (O r g a n i z a t i o n f o r S e c u r i t y a n d a nd Pr ivate S e c tor D evelopment Vic e Cooperation in Europe). 2008. The Media Presidency, World Bank, Washington, DC. Self-Regulation Guidebook. Vienna: OSCE. LSTM (Centre for Maternal and Newborn PE FA (P ublic E xpendit u re and Financial Health), RBC (Rwanda Biomedical Centre), Accountability) Secretariat. 2017. Public and U NICEF (United Nations Children’s Expenditure and Financial Accountability Fund). 2016. Comprehensive Evaluation Assessment 2016: Final Report (February of the C o m m unit y He alth Prog ra m in 2 0 16 F ra m e w o rk ) . L ondon : A E C OM Rwanda. Kigali: Ministry of Health. International Development Europe for the MacArthur, T. 2012. Developing Sustainable Government of Rwanda. Public S ec to r C apa cit y fo r R wa n d a’s Pouget, S. 2013. “Arbitrating and Mediating Vision 2020: Enhancing Human Resource Disputes: Benchmarking Arbitration and M an age me nt . Washing ton, DC: World Mediation Regimes for Commercial Disputes B a n k . h t t p : / / d o c u m e n t s .w o r l d b a n k​ Related to Foreign Direct Investment.” Policy .or g /c u rat e d /e n / 96 0 8614 683 071376 41​ Research Working Paper 6632, World Bank, / D e velopi n g - su s t a i n able - publ ic - s e c tor​ Washington, DC. -c­apacity​-for-Rwandas-vision-2020-enhancing​ Putnam, R. D. 1993. “What Makes Democracy -human-resource-management. Work?” National Civic Review 82 (2): 101–07. C a p a b l e and A ccounta b l e S tate I nstitutions 283 Ramsay, I. 1996. “Small Claims Courts: A Review.” Stipanowich, T. 2004. “ADR and the ‘Vanishing In Rethinking Civil Justice: Research Studies Trial’: The Growth and Impact of Alternative for the Civil Justice Review, 491–541. Toronto: Dispute Resolution.” Journal of Empirical Ontario Law Reform Commission. Legal Studies 1 (3): 843–912. ———. 1998. “Review: Lowered Horizons and Tsai, L. 2007. “Solidarity Groups, Informal Broadened Vistas: The Small-Claims Court Accountability, and Local Public Goods and Access to Justice.” Justice of Law and Provision in Rural China.” American Political Society 25 (3): 438–42. Science Review 101 (2): 355–71. RGB (Rwanda Governance Board). 2014. The U.S. Chamber Institute for Legal Reform. 2010. Assessment of the Impact of Homegrown Lawsuit Climate 2010: Ranking the States. Initiatives. Kigali: RGB. Washington, DC: U.S. Chamber Institute for ———. 2016a. Rwanda Citizens Report Card. Legal Reform. Kigali: RGB. World Bank. 2003a. “Indonesia Municipal ———. 2016b. Rwanda Governance Scorecard I n n o v a t i o n s P r o j e c t .” Wo r l d B a n k , Report. Kigali: RGB. Washington, DC. ———. 2016c. Rwanda Media Barometer. Kigali: ———. 2003b. World Development Report 2004: RGB. Making Services Work for the Poor People. Rosales-López, V. 2008. “Economics of Court Washington, DC: World Bank. Performance: An Empirical Analysis.” European ———. 2004. World Development Report 2005: Journal of Law and Economics 25 (3): 231–51. A Better Investment Climate for Everyone. Sherwood, R. M., G. Shepherd, and C. M. Washington, DC: World Bank. de Souza. 1994. “Judicial Systems and ———. 2016. What Works for Strengthening Economic Performance.” Quarterly Review of Public Ser vice Training in Postconf lict Economics and Finance 34 (Suppl. 1): 101–16. Environments? Washington, DC: World Bank. Shostak, R., J. Watkins, A. Bellver, and I. John- ———. 2017. World Development Report 2017: Abraham. 2014. “When Might the Introduction Governance and the Law. Washington, DC: of a Delivery Unit Be the Right Intervention?” World Bank. Governance Practice Note, World Bank, ———. 2018. Doing Business 2018: Reforming Washington DC. http://siteresources.worldbank​ to Create Jobs. World Bank Group Flagship .org/PUBLICSECTORANDGOVERNANCE​ Report. Washington, DC: World Bank. /Resources/285741-1368636830774/ When​ ———. Various years. World Development _Might_the_Intro_of_a_DU_Be_the_Right​ Indicators database. Washington, DC: World _Intervention_FINAL.pdf. Bank. Shyaka, A., S. Kamikazi, J. Murwanashyaka, and Wo rld E c o no m i c Fo r u m . 2 016 . G l o b a l A. Mulindwa. 2016. “National Umushyikirano Competitiveness Report 2016–2017. Geneva: Council.” RGB, Kigali. World Economic Forum. ECO-AUDIT Environmental Benefits Statement The World Bank Group is committed to reducing its environmen- tal footprint. In support of this commitment, we leverage elec- tronic publishing options and print-on-demand technology, which is located in regional hubs worldwide. Together, these initiatives enable print runs to be lowered and shipping distances decreased, resulting in reduced paper consumption, chemical use, greenhouse gas emissions, and waste. We follow the recommended standards for paper use set by the Green Press Initiative. The majority of our books are printed on Forest Stewardship Council (FSC)–certified paper, with nearly all containing 50–100 percent recycled ­ content. The recycled fiber in our book paper is either unbleached or bleached using totally chlorine-free (TCF), processed chlorine–free (PCF), or enhanced elemental chlorine–free (EECF) processes. More information about the Bank’s environmental philosophy can be found at http://www.worldbank.org/corporateresponsibility. A strong and widely acknowledged record of economic success—including a three-and-a- half-fold increase in per capita income since 1994—places Rwanda among the world’s fastest-growing economies. Traumatic memories of the 1994 genocide are gradually fading, as associations begin to take a more positive form—of a nation on the rise, powered by human resilience, a sense of common purpose, and a purposeful government. Past successes and a sense of frailty have fueled aspirations for a secure, prosperous, and modern future. Sustaining high rates of economic growth is at the heart of these ambitions. Recent formulations of the nation’s Vision 2050 set a target of achieving upper-middle- income status by 2035 and high-income status by 2050. Future Drivers of Growth in Rwanda: Innovation, Integration, Agglomeration, and Competition, a joint undertaking by experts from Rwanda and the World Bank Group, evaluates the country’s possibilities and options in this endeavor. The report identifies four essential drivers of growth—innovation, integration, agglomeration, and competition—and reforms in six priority areas: human capital development, export dynamism and regional integration, well-managed urbanization, competitive domestic enterprises, agricultural modernization, and capable and accountable public institutions. ISBN 978-1-4648-1280-4 Government of Rwanda SKU 211280