84083 rev Afri cA D eve lopme nt f orum Youth Employment in Sub-Saharan Africa Youth Employment in Sub-Saharan Africa Youth Employment in Sub-Saharan Africa Deon Filmer and Louise Fox with Karen Brooks, Aparajita Goyal, Taye Mengistae, Patrick Premand, Dena Ringold, Siddharth Sharma, and Sergiy Zorya A copublication of the Agence Française de Développement and the World Bank © 2014 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 17 16 15 14 This work is a product of the staff of The World Bank with external contributions. Note that The World Bank and the Agence Française de Développement do not necessarily own each component of the content included in the work. The World Bank and the Agence Française de Développement therefore do not warrant that the use of the content contained in the work will not infringe on the rights of third parties. The risk of claims resulting from such infringe- ment rests solely with you. The findings, interpretations, 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, or the Agence Française de Développement. The World Bank does not guar- antee 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 Unported license (CC BY 3.0) http://creativecommons.org/licenses/by/3.0. 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: Filmer, Deon and Louise Fox. 2014. Youth Employment in Sub-Saharan Africa. Africa Development Series. Washington, DC: World Bank. doi:10.1596/978-1-4648-0107-5. License: Creative Commons Attribution CC BY 3.0 Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: 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. 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; fax: 202-522-2625; e-mail: pubrights@worldbank.org. ISBN (paper): 978-1-4648-0107-5 ISBN (electronic): 978-1-4648-0108-2 DOI:10.1596/978-1-4648-0107-5 Cover design: Critical Stages Cover image: © Corbis; permission required for further use. Library of Congress Cataloging-in-Publication Data has been applied for. Africa Development Forum Series The Africa Development Forum Series was created in 2009 to focus on issues of signifi- cant relevance to Sub-Saharan Africa’s social and economic development. Its aim is both to record the state of the art on a specific topic and to contribute to ongoing local, regional, and global policy debates. It is designed specifically to provide practitioners, scholars, and students with the most up-to-date research results while highlighting the promise, challenges, and opportunities that exist on the continent. The series is sponsored by the Agence Française de Développement and the World Bank. The manuscripts chosen for publication represent the highest quality in each institution and have been selected for their relevance to the development agenda. Working together with a shared sense of mission and interdisciplinary purpose, the two institutions are com- mitted to a common search for new insights and new ways of analyzing the development realities of the Sub-Saharan Africa region. Advisory Committee Members Agence Française de Développement Jean-Yves Grosclaude, Director of Strategy Alain Henry, Director of Research Philippe Cabin, Head of Research Publishing Division World Bank Francisco H. G. Ferreira, Chief Economist, Africa Region Richard Damania, Lead Economist, Africa Region Stephen McGroarty, Executive Editor, Publishing and Knowledge Division IBRD 39 February Sub-Saharan Africa CAPE MAURITANIA VERDE NIGER MALI SUDAN ERITREA SENEGAL CHAD THE GAMBIA BURKINA FASO GUINEA-BISSAU GUINEA BENIN NIGERIA ETHIOPIA CÔTE GHANA SIERRA LEONE CENTRAL SOUTH D’IVOIRE SUDAN AFRICAN REPUBLIC SOMALIA LIBERIA TOGO CAMEROON EQUATORIAL GUINEA UGANDA KENYA SÃO TOMÉ AND PRÍNCIPE GABON CONGO DEMOCRATIC RWANDA REPUBLIC OF CONGO BURUNDI TANZANIA SEYCHELLES COMOROS ANGOLA MALAWI Mayotte (Fr.) ZAMBIA MADAGASCAR ZIMBABWE MOZAMBIQUE MAURITIUS NAMIBIA BOTSWANA Réunion (Fr.) SWAZILAND LESOTHO SOUTH AFRICA Throughout this report, “Africa” is shorthand for “Sub-Saharan Africa.” In particular instances, “Sub-Saharan Africa” is retained to clarify comparisons across regions or to indicate a specific data set. Titles in the African Development Forum Series Africa’s Infrastructure: A Time for Transformation (2010) edited by Vivien Foster and Cecilia Briceño-Garmendia Gender Disparities in Africa’s Labor Market (2010) edited by Jorge Saba Arbache, Alexandre Kolev, and Ewa Filipiak Challenges for African Agriculture (2010) edited by Jean-Claude Deveze Contemporary Migration to South Africa: A Regional Development Issue (2011) edited by Aurelia Segatti and Loren Landau Light Manufacturing in Africa: Targeted Policies to Enhance Private Investment and Create Jobs (2012) by Hinh T. Dinh, Vincent Palmade, Vandana Chandra, and Frances Cossar Informal Sector in Francophone Africa: Firm Size, Productivity, and Institutions (2012) by Nancy Benjamin and Ahmadou Aly Mbaye Financing Africa’s Cities: The Imperative of Local Investment (2012) by Thierry Paulais Structural Transformation and Rural Change Revisited: Challenges for Late Developing Countries in a Globalizing World (2012) by Bruno Losch, Sandrine Fréguin-Gresh, and Eric Thomas White The Political Economy of Decentralization in Sub-Saharan Africa: A New Implementation Model (2013) edited by Bernard Dafflon and Thierry Madiès Empowering Women: Legal Rights and Economic Opportunities in Africa (2013) by Mary Hallward-Driemeier and Tazeen Hasan Enterprising Women: Expanding Economic Opportunities in Africa (2013) by Mary Hallward-Driemeier Urban Labor Markets in Sub-Saharan Africa (2013) edited by Philippe De Vreyer and François Roubaud Securing Africa’s Land for Shared Prosperity: A Program to Scale Up Reforms and Investments (2013) by Frank F. K. Byamugisha All books in the Africa Development Forum series are available for free at https://openknowledge.worldbank.org/handle/10986/2150 Contents Foreword  xix Acknowledgments  xxi About the Authors  xxiii Acronyms and Abbreviations  xxvii Overview  Youth Employment in Sub-Saharan Africa   1 Africa’s Large Youth Population   2 Growth, Jobs, and Africa’s Labor Force—Now and in the Future   3 Youth’s Transition to Productive Employment   6 Policy Priorities for Addressing Youth Employment   8 Human Capital: The Fundamental Role of Basic Education   10 Raising the Productivity of Smallholder Farmers   11 Increasing the Productivity of Nonfarm Household Enterprises   15 Improving Competitiveness to Boost Modern Sector Wage Jobs   18 Building an Effective Youth Employment Policy    21 Notes  22 References  22 1 Opportunities and Challenges for Youth Employment in Africa   25 Africa’s Working-Age Population: Very Young and Growing Rapidly   25 Can Africa’s Youth Bring an Economic Advantage?    26 What Is a Job, and Where Do Most Africans Find One?    28 Growth, Jobs, and Africa’s Labor Force—Now and in the Future   30 Framework of This Report   41 Notes  41 References  42 Focus Note 1  Jobs: More Than Just Income   43 The Value of Jobs   43 Jobs and Aspirations   44 ix x Contents Jobs and Development   45 References  47 2 Youth: A Time of Transitions   49 Youth’s Transition from School   50 Youth’s Transition to Work    53 The Parallel Transitions: Choices Influencing Health, Family Formation, and Civic Engagement    60 Challenges for Females Transitioning to Work   62 Facilitating Transitions from School, to Work, and across Sectors of Employment   64 Notes  64 References  64 3 Skills for Productive Employment   67 Schooling, Educational Attainment, and Work   68 Building a Foundation: Cognitive, Socioemotional, and Behavioral Skills   75 Building Skills through Post-School Training    89 Government Interventions and the Post-school Training Market   94 Conclusion: A Skills Agenda for Youth    102 Notes  103 References  104 4 Agriculture as a Sector of Opportunity for Young Africans   113 Agriculture: Potential Opportunity, Room to Grow   114 Recognizing the Opportunity in Agriculture for Young People   117 Agricultural Career Paths for the Future   120 Lifting Key Constraints on Capital, Land, and Skills   123 Land Policies That Benefit the Young   127 Enhancing Skills and Building a Better Educational Foundation   130 Current Agricultural Programs Deliver Too Little, Too Slowly, to Meet the Needs of Africa’s Young People   135 Harnessing Agriculture’s Youth Dividend   137 Notes  137 References  138 Focus Note 2  Safety Nets and Pathways to Productive Employment   142 Short-Term Benefits of Safety Net Programs   142 Safety Nets Plus Explicit Productive Components    143 Do Complementary Interventions Open Pathways toward Productive Employment Youth?  144 for  Contents xi Safety Nets as Vehicles to Deliver Interventions Aimed at Improving Youth Employment Outcomes  145 Notes  146 References  146 5 Creating Productive Employment for Youth in the Household Enterprise Sector  149 The Household Enterprise Sector Today   152 The Business: Constraints and Opportunities    153 Constraints for Young People to Enter the HE Sector   157 Creating and Sustaining Productive Employment in Household Enterprises   159 What Should Governments Do to Help Equip Youth with Skills for the HE Sector?   167 Market Access and Voice   174 Conclusion  177 Notes  178 References  180 Focus Note 3  Financial Inclusion and the Transition to Sustainable Livelihoods for Young People   183 Gaps in Access to Savings and Credit   183 Formal Savings Services: Issues and Options   184 Formal Credit Services: Challenges in Expanding Options for Smallholders, Household Enterprises, and Young People   188 Notes  194 References  195 6 Raising Productivity in Africa’s Modern Wage Enterprises to Foster Job Growth for Youth  197 Africa’s Modern Enterprise Sector: An Overview   198 Modern Manufacturing Delivers Little Employment, Few Exports   199 How Competitive Is Modern Manufacturing in Africa?   200 Sources of Productivity Gaps in African Manufacturing   207 Making African Firms Competitive: Priorities for Improving the Business Climate and Workforce Skills  212 Building Skills for the Modern Wage Sector   224 Many Options to Promote Competitive Modern Enterprises   231 Notes  232 References  233 xii Contents Focus Note 4  Youth Unemployment in South Africa: Different Configuration, Different Approaches  238 Conventional Perceptions of the Causes of Unemployment   238 The Real Unemployment Problem: Slow Growth    239 Inefficient Job Search and Matching Processes   240 Policies to Reduce Unemployment   240 References  241 7 Conclusion: Building an Effective Youth Employment Policy   243 A Programmatic Approach   243 “Do Now for Now”: Address the Constraints Facing Households and Firms   244 “Do Now for Later”: Take Action for Medium-Term Payoffs   246 Appendix  A Note on Data   249 Standardized and Harmonized Household and Labor Force Data   249 Demographic and Health Survey Data   249 Boxes O.1 What is a “job”?   4 O.2 The youth employment challenge in resource-rich and some middle-income countries Africa  6 in  O.3 What will happen to employment if light manufacturing increases dramatically Africa?  7 in  O.4 Youth employment versus overall employment   8 O.5 Framework for analyzing youth employment   9 O.6 Does labor need to move out of agriculture as productivity grows?   13 1.1 Youth versus overall employment   29 1.2 How are our employment estimates done?   32 1.3 The challenge of diversifying output and employment in resource-rich countries: Examples from Central Africa   34 1.4 How are our employment projections done?   36 F1.1 Employment, conflict, and violence: Is there a link?   46 2.1 Child labor in Sub-Saharan Africa   51 2.2 Intergenerational transmission of occupations in Africa   54 3.1 Estimating the returns to education   74 3.2 A second chance at education for African youth    81 3.3 Northern Ghana’s School for Life second-chance education program   82 3.4 Returns to behavioral skills in wage employment: Evidence from Peru   86 3.5 Developing socioemotional and behavioral skills in postconflict settings   87 3.6 Promoting socioemotional and behavioral skills   88 3.7 On-the-job training varies by country and type of firm, and it is not for everyone    95 Contents xiii 3.8 Kenya’s Jua Kali voucher program    98 3.9 Can incentives improve training quality and participation?   100 3.10 Impact evaluation to build the evidence base on youth employment programs   101 4.1 Women and girls: A major force within Africa’s agricultural labor force   115 4.2 When agriculture is more productive, economies can grow   116 4.3 Compromising growth and job creation without improving long-term food security  117 4.4 Producer organizations and the transition to modern supply chains    119 4.5 Options for establishing or leaving a farm in Kenya   119 4.6 High-value agriculture and opportunities for employment off the farm   123 4.7 Red Fox Ethiopia: More technically sophisticated wage work   124 4.8 Documenting land rights: Encouraging investment and reducing the cost of land transfers  128 4.9 Mexico’s program to speed intergenerational land transfers   129 4.10 Information and communication technologies: Altering the flow of agricultural information  131 4.11 Vocational agricultural education: A poor substitute for general education   132 4.12 Innovations in agricultural extension: Relying on farmers to improve service delivery  134 4.13 Overview of the Comprehensive African Agriculture Development Programme (CAADP)  136 5.1 What is a household enterprise?   150 5.2 Why do females earn less? Occupational segregation in the household enterprise sector  154 5.3 Small and medium enterprises are not the same as household enterprises   156 5.4 Informal training is normal for youths seeking to operate a household enterprise   158 5.5 Rwanda: Many programs to support household enterprises, but little information results  159 on  5.6 Need for a comprehensive approach   161 5.7 Ghana’s integrated approach to HE development   162 5.8 Weakening the local economy by perpetuating the machinga cycle in Dar es Salaam    164 5.9 Clustering household enterprises for the benefit of all   165 5.10 Sources of credit used by households to start a business in Tanzania   166 5.11 Rural enterprise projects bolster skills and business development in Ghana and Senegal  172 5.12 Taking a household business to the international market: Gahaya Links and Rwanda’s peace baskets   176 5.13 Reducing isolation and exploitation through self-help associations   177 F3.1 Rural banks in Ghana: Reaching clients who are underserved by other banks   185 F3.2 Use of technology: Bringing secure financial services to new markets in Kenya and India   186 F3.3 Linking smallholders to supply chains to improve their access to financial services   189 xiv Contents F3.4 Shifting from group to individual liability lending as a positive microfinance innovation: Evidence from the Philippines   190 F3.5 ROSCAs, VSLAs, SHGs, and SACCOs: Examples of informal savings and credit systems  191 F3.6 Village women in Mali: Achieving food and financial security through savings and credit  groups  194 6.1 Where do young people figure in Africa’s wage employment picture?   201 6.2 Measurement issues and other limitations of unit labor costs   203 6.3 Does Africa really have a labor cost advantage?   207 6.4 The high price of inefficiency at the port of Dar es Salaam   215 6.5 Improving land transportation through increased international cooperation and comprehensive procedural reforms: The Malaba border-crossing pilot   217 6.6 Who must pay bribes? How much? And does it matter?   221 6.7 Knowledge, technology, and the emergence of successful enterprise clusters in Africa  223 6.8 Why have special economic zones failed to thrive in Africa, and what are the lessons for the future?   224 6.9 Reforming TVET systems in Africa   226 6.10 Does government have a role in on-the-job training?   228 Figures O.1 The structure of Sub-Saharan Africa’s population is different than that in other regions  3 O.2 Over the past two decades, agriculture’s share in GDP contracted in Africa, but manufacturing did not replace it   4 O.3 Where are Africans working?   5 O.4 Informal will be normal in 2020   6 O.5 The transition from school to work in Sub-Saharan Africa is slow   8 O.6 Family formation starts earlier for young women than for young men   9 O.7 Education shapes opportunities   10 O.8 Young people are unlikely to own land   14 O.9 At the same level of income per capita, national policies can produce very different levels of financial inclusion   17 O.10 The share of youth in wage employment tracks the share in the general population  18 1.1 Africa’s population is young and will remain so   26 1.2 The structure of Sub-Saharan Africa’s population is different than that in other regions  27 1.3 Unlike in other regions, the number of young people in Sub-Saharan Africa will increase dramatically in the near future   27 1.4 In East Asia, the dependency ratio changed quickly; in Sub-Saharan Africa, it is changing, but slowly   28 1.5 The reduction in the total fertility rate has stalled in several African countries   29 Contents xv 1.6 Africa’s growth miracle, 2005–12    30 1.7 Over the past two decades, agriculture’s share in GDP contracted in Africa, but manufacturing did not replace it   31 1.8 Exports are a smaller share of GDP in Africa than in East Asia and a larger share than in South Asia, but African countries, even richer ones, export commodities, not manufactured goods   31 1.9 The majority of Sub-Saharan Africa’s workers in low- and lower-middle-income countries work in agriculture and in nonfarm household enterprises   32 1.10 Growth has moved the structure of employment away from agriculture in some countries, but not others    33 1.11 Most wage jobs are in the private sector (but not in resource-rich countries)    33 1.12 Industrial wage employment has not risen with GDP in Africa the way it has in manufacturing exporters at similar income levels   35 1.13 Where will Africa’s 125 million new jobs be created?   39 1.14 Informal will remain normal in much of Sub-Saharan Africa   39 1.15 Even game-changing growth will have limited effects on the distribution of employment in the near term   40 F1.1 Jobs and life satisfaction across regions   44 F1.2 Jobs drive development   45 2.1 Many 18-year-olds in Africa are still in school, but half of them are still in primary school  50 B2.1.1 Percentage of children working   51 2.2 School and work are often combined   52 2.3 Most Sub-Saharan African youths start out working for their families and then become  self-employed  53 B2.2.1 Intergenerational transmission of employment sector in five African countries    54 2.4 Youths in urban Tanzania are unlikely to move between employment sectors during their working lives   55 2.5 Youth transitions to sector of employment vary across urban and rural areas and between male and female youths   56 2.6 Sectoral mobility among urban youth in Uganda   57 2.7 In Tanzania, many work in two or more activities   57 2.8 Younger people are most often engaged in casual wage employment   58 2.9 Personal networks are key to finding a job   58 2.10 Migration increases mobility across sectors of employment   60 2.11 Family formation starts earlier for young women than for young men   61 2.12 As they get older, young people increasingly engage as citizens   62 3.1 Primary school completion rates have risen substantially in Sub-Saharan Africa, 1990–2011  68 3.2 Africa’s young people have more education than ever before, but average education attainment is still low   69 3.3 Educational attainment in Sub-Saharan Africa is projected to rise substantially   69 3.4 Schooling maps to sector of employment   70 xvi Contents 3.5 Wage workers with more schooling are more likely to work under contract   70 3.6 Relationships between schooling and work vary depending on gender and urbanization  71 3.7 Education is associated with higher earnings in household enterprises   72 3.8 Education is associated with higher wages   73 3.9 The convex relationship between earnings and education: Schooling and earnings in urban West Africa   75 3.10 The ability to read in early grades is alarmingly low   77 3.11 By the end of primary school, many students have not mastered even basic competencies: 2007 SACMEQ results for math and reading proficiency   77 3.12 Secondary school students in Sub-Saharan Africa perform poorly on internationally comparable assessments: TIMSS results for math proficiency, 2011   78 3.13 Learning trajectories by age are flat in Guinea-Bissau   78 3.14 Learning trajectories by grade are flat: Performance on test scores in selected African countries, by educational achievement   79 B3.2.1 Proportion of 15- to 19-year-olds who have completed each grade   81 3.15 Cognitive skills increase slowly, especially for the poorest   83 3.16 Behavioral skills take longer to develop than cognitive skills   84 3.17 The returns to schooling are higher for those with greater ability: Returns of an additional year of schooling by quantile   88 3.18 Many young people, especially in West Africa, have been an apprentice, whereas experience with TVET is less prevalent   90 3.19 Apprenticeships are geared toward youths with lower levels of education   91 3.20 TVET is geared toward wage work, whereas apprenticeships lead to working in either a household enterprise or wage work   92 3.21 Africa has a wide range of informal private training providers, 2009   93 B3.7.1 On-the-job training in African firms varies by country   95 B3.7.2 On-the-job training in African firms varies by firm size and export orientation, 2006  95 3.22 There are stark differences between rich and poor in experience with apprenticeships and TVET  97 3.23 Men and women take up different types of vocational training in Kenya   97 3.24 Training costs and the price paid by participants vary by type of training in Kenya    99 B4.2.1 Higher total factor productivity helped U.S. farmers to compensate for declining terms of trade, 1975–2009   116 4.1 The young do not typically own land   118 F2.1 Impacts of conditional cash transfers plus business grants or conditional cash transfers plus skills training in Nicaragua    145 B5.1.1 Most people working in the HE sector are owners   150 B.5.1.2 Most HEs are family operations   150 5.1 Household enterprises are an important share of nonfarm employment in low- and middle-income countries  151 5.2 In Africa, as labor moved out of agriculture, more people moved into the HE sector than into private wage employment   151 Contents xvii 5.3 Most HEs are in the trading sector   152 B5.2.1 Earnings are higher in male-dominated subsectors than in female-dominated subsectors  154 5.4 Younger HE owners tend to have more education than older ones   155 5.5 Household enterprise owners tend to be over 25 years of age   158 F3.1 Percentage of population ages 15 and older who reported saving any amount in the last 12 months, by savings mechanism   184 F3.2 Percentage of individuals with an account at a formal financial institution   188 6.1 Africa has less nonfarm wage employment than other regions   199 6.2 Wage employment is growing inconsistently across Africa   200 B6.1.1 The share of youth in wage employment tracks the share in the general population   201 6.3 Services form the largest share of nonfarm wage employment; within manufacturing, the food and textile industries dominate   201 6.4 Africa’s modern wage manufacturing sector continues to have a small share in employment  202 6.5 African firms export relatively little   203 6.6 China’s unit labor costs have remained below those of other emerging economies   204 6.7 Growth in manufacturing wage employment has been inconsistent across African countries   205 6.8 Based on unit labor costs, some African countries could become competitive with other emerging economies  205 6.9 Africa does not have a uniform wage advantage over other developing regions   206 6.10 Labor productivity is particularly low in low-income African countries    208 6.11 African workers are not as well equipped with fixed capital as their counterparts elsewhere in the world   208 6.12 African countries have lower productivity than other regions   209 6.13 Africa fares poorly in firm-reported indicators of transport, water, and electricity supply  210 6.14 Decomposing textile industry productivity   211 6.15 Trading across borders is expensive and slow in Africa   212 6.16 African firms use relatively little bank financing   218 6.17 Coverage of credit bureaus is generally low in Africa   218 6.18 Africa has the highest formal costs of setting up a business   220 6.19 Business setup costs are higher in resource-rich African countries   220 F4.1 Patterns in economic and employment growth in South Africa    239 Tables BO.6.1 Increase in crop area harvested, agricultural labor force, and output per worker in Sub-Saharan Africa, 1960–2008 and 1990–2008   13 O.1 Priority actions to take now to address the youth employment challenge   22 1.1 African countries have less wage employment than high-growth comparator countries  35 B1.4.1 Growth elasticities of employment   37 xviii Contents 1.2 Average annual growth, by sector and country income level, 2005–20   38 1.3 Framework of this report    41 B3.2.1 Costs per pupil in second-chance education programs and formal public schools    82 3.1 Developmental tasks, by age    84 4.1 Wholesale prices of unprocessed maize and rice in selected countries (average, January–April 2012)    115 B4.2.1 Percentage change in prices of selected food products in the United States, 1960–2010   116 4.2 Pathways for agricultural employment and their requirements   120 B4.6.1 Employment in horticultural supply chains in Sub-Saharan Africa    123 5.1 Rural versus urban household enterprises    153 5.2 Sources of capital for household enterprises: Start-up and credit for operations    155 7.1 Priority actions to take now to address the youth employment challenge   245 A.1 Standardized and Harmonized Household and Labor Force Surveys used in report analysis  250 A.2 Demographic and Health Survey Data Sets used in report analysis   251 Foreword Africa is now the world’s youngest continent. With its comprehensive analysis of youth While national populations in most parts employment in Africa, this report updates how of the world are aging, young people are now and where countries are creating jobs and in the majority in many African countries. looks to the future of promising new direc- Whether they live in the cities and towns of a tions for the continent. It focuses on how pol- rapidly urbanizing Africa, or in rural villages icy makers can help young people walk and settlements; whether they come from through the front door of the labor market and middle-class backgrounds or from vulnerable into both fast-growing modern wage jobs as families that are living in poverty, one thing is well as into other sustainable economic oppor- certain—these young people have high expec- tunities—improving income by raising pro- tations, and African policy makers are increas- ductivity in the informal sector, where many ingly concerned about how to meet them. young people will work for the foreseeable Today, jobs and opportunity for young peo- future. ple are consistently at the top of the develop- Youth employment is not a one-dimen- ment agenda in virtually every country on the sional challenge. Addressing it will require continent. attention to the quality of basic education and Although data on employment is patchy training to improve young people’s produc- and not generally comparable across African tivity, while also removing current obstacles countries, we can nonetheless learn much from that hinder progress in agriculture, household the available evidence. Contrary to popular enterprises, and the modern wage sector. perceptions, unemployment is but one of the If children do not know how to read or do main problems facing low-income countries in basic mathematics at the end of their years in Africa. Beyond simple unemployment, Africa is primary school, this basic skills gap has a pro- challenged to address underemployment: in the found impact on productivity in adulthood. absence of adequate social safety nets, young If young people have weak access to land and people are compelled to take low-productivity, to credit, their dreams of becoming entrepre- low-wage jobs for their very survival. neurs or working in high-paying, rewarding While urban youth tend to be much more jobs will be dashed. If university graduates vocal about their challenges and aspirations, acquire degrees and knowledge that have little and highly visible in their job-seeking, millions practical application in Africa’s fast-changing of young men and women in rural and semi- labor force, then their investment of time and urban areas struggle to find better-paying jobs money will have been largely in vain, with and opportunities to escape poverty for them- few prospects for strong careers in the private selves and their families. sector. xix xx Foreword I hope that this report will help African advantage of the opportunities that are countries develop customized strategies to steadily growing across the continent whether address youth employment and to invest in in export-oriented manufacturing, high- better information on how well these strate- productivity agriculture and agribusinesses, or gies are working. profitable small enterprises. Collectively we must redouble our efforts Africa’s future will be profoundly shaped by to meet the paramount challenge of youth how well it manages this challenge. employment so that young people can take Makhtar Diop Vice President for Africa The World Bank Acknowledgments This report was prepared by a core team led by Kadzamira. The report was edited by Kelly Deon Filmer and Louise Fox and comprising Cassaday (overall) and Amy Gautam (who Karen Brooks, Aparajita Goyal, Taye Mengis- had primary editing responsibility for chapter tae, Patrick Premand, Dena Ringold, Siddharth 4). Several World Bank staff, as well as policy Sharma, and Sergiy Zorya. Additional contri- makers, academics, and other stakeholders butions were made by Shubha Chakravarty, provided comments at various stages of the Florence Kondylis, Obert Pimhidzai, Raju development of this report. Richard Damania, Singh, and Erik von Uexkull. The team was Mary Hallward-Driemeier, Margaret Grosh, directed by Shantanayan Devarajan and Ritva and Ravi Kanbur provided careful peer review Reinikka. Research assistance was provided by comments. Any errors or omissions are the Jorge Munoz, Lena Nguyen, and Thokozani responsibility of the team. xxi About the Authors Deon Filmer Deon Filmer is a Lead Economist in the World Bank’s research department and the HD Econom- ics Unit of the Human Development department of the World Bank’s Africa region. He works on issues of youth employment and skills, service delivery, and impact evaluation of policies and pro- grams to improve human development outcomes. He has spent many years in the World Bank’s research department where his research has covered the areas of education, health, social protec- tion, and poverty analysis. He has published numerous articles in top-ranked economics journals. Recent work includes studies of the impact of demand-side programs on schooling outcomes; of the roles of poverty, orphanhood, and disability in explaining education inequalities; of the determinants of effective service delivery and the evaluation of interventions aimed at improving it; of the determinants of fertility behavior, and on trends in adult mortality around the world. He was a core team member of the World Development Report 1995: Workers in an Integrating World and World Development Report 2004: Making Services Work for Poor People, and a contributor to the World Development Report 2007: Development and the Next Generation. He holds PhD and MA degrees from Brown University. Louise Fox Louise Fox was, until her recent retirement from the World Bank, a Lead Economist in the Africa region of the World Bank. During her long career at the Bank, she worked on a wide range of topics; her specialties included the analysis of employment and labor markets, poverty and inequality, and the economics of social service delivery, with the overarching theme of the links among policies, outcomes, and poverty reduction. Prior to her position in the Africa region, Dr. Fox spent 13 years working on issues of labor market adjustment, poverty, and social protection in transition economies, including China and Mongolia, the Baltic States, and Eastern Europe. Before that she researched poverty, inequality, and macroeconomic adjustment in Latin America. Her most recent published work has been on the topics of poverty reduction and inclusive growth, the political economy of poverty reduction, and on employment, labor markets, and labor regula- tion, all with respect to Sub-Saharan Africa. She has also published in the areas of pension reform, reform of child welfare systems, social protection, public expenditures in the social sectors and poverty reduction, female-headed households and child welfare, stabilization policies and poverty reduction, the social costs of adjustment, and the economic history of poverty and inequality in Brazil. Dr. Fox received a PhD from Vanderbilt University. Karen Brooks Karen Brooks is Director of the CGIAR research program on Policies, Institutions and Markets. Brooks worked for the World Bank for 20 years. For the last 10 years of her tenure there, she xxiii xxiv About the Authors managed analytical and operational programs in agriculture and rural development for the Africa region. Brooks also worked extensively in both the Europe and Central Asia and Africa regions; in the former on issues related to the transition from central planning, and in the latter on the invest- ment and policy agenda associated with the renewed commitment to agricultural growth. Prior to joining the World Bank, she was Associate Professor in the Department of Applied Economics at the University of Minnesota. Brooks received both her PhD and MA degrees in Economics from the University of Chicago, where she was a National Science Foundation Graduate Fellow. Aparajita Goyal Aparajita Goyal is an Economist with the Agriculture team of the Africa region of the World Bank. Previously, she worked in the Bank’s Agriculture and Environment Services department, as well as the Bank’s Development Research Group. Her work has focused on microeconomic issues of development, with a particular emphasis on technological innovation in agriculture, access to markets, and intellectual property rights. Her work has been published in leading academic economics journals, and has also been featured in the popular press, including Frontline, The Economist, and the Wall Street Journal. She has previously worked with ActionAid in London and the Right to Food Campaign in India. She obtained a PhD in Economics from the University of Maryland and an MSc degree in Development Studies from the London School of Economics. Taye Mengistae Taye Mengistae is a Senior Economist with the Finance and Private Sector Development, Eastern and Southern Africa team of the World Bank’s Africa region. Previously he was with the Bank’s Development Research Group. He has published numerous articles in leading academic economic journals. His country work experience includes South Africa, Kenya, Ghana, Ethiopia, Zambia, India, China, and Pakistan. His main areas of research interest include globalization, export com- petitiveness, capital flows, innovation, and education. He holds a PhD from Oxford University and an MA degree from the University of Leeds. Patrick Premand Patrick Premand is a Senior Economist in the HD Economics Unit of the Human Development department of the World Bank’s Africa region. He works on issues of youth employment and skills, early child nutrition and development, and the impact evaluation of policies and programs to improve human development outcomes. He has previously held positions in the Office of the Chief Economist for Human Development Department and in the Bank’s Latin America and Caribbean region’s Poverty team. He holds PhD and MSc degrees from Oxford University. Dena Ringold Dena Ringold is a Lead Economist in the Social Protection team of the World Bank’s Africa region, with fifteen years experience working in Human Development at the World Bank. Her research interests include social inclusion of minorities, safety nets, and governance. Prior to joining the Africa region, she was a core team member of the World Development Report 2013: Jobs, and she previously worked in the Office of the Chief Economist for Human Development. Dena began her career at the Bank in the Europe and Central Asia (ECA) region, where she worked on operations and analytical programs with a focus on social protection and local service delivery in Central and Southeastern Europe. While in ECA she helped to initiate the Bank’s first qualitative and quantita- tive analyses of the Roma minority and helped to set up the Roma Education Fund. She has also worked on social protection in the Latin American and Caribbean (LAC) region. In 2005, Dena was an Ian Axford Fellow in public policy based at the Ministry of Maori Development in Wel- lington, New Zealand. She holds an MSc in Economics and Government from the London School of Economics and a BA in History and Political Science from Swarthmore College. About the Authors xxv Siddharth Sharma Siddharth Sharma is an Economist in the Finance and Private Sector team of the World Bank’s Europe and Central Asia Region. He previously worked in the Social Protection team of the Africa region and in the Poverty team of the Middle East and North Africa region. His research focus has been on labor markets, productivity, and firms. He has published academic papers in leading economics journals. He holds a PhD from Yale University, and an MA and BA from the University of Delhi Sergiy Zorya Segiy Zorya is a Senior Economist in the Sustainable Development team of the World Bank’s East Asia and Pacific Region. He was previously with the Bank’s Agriculture and Rural Development (ARD) Department. He works on issues pertaining rural policies and strategies, food prices, man- agement of public expenditure, and global food security. Prior to ARD, he worked in three Bank regions: Europe and Central Asia, East Asia and Pacific, and Africa. His most recent work was in South Sudan, Tanzania, Uganda, and Zambia. Sergiy holds an MA degree in Agricultural Econom- ics from the Agricultural University of Zhytomyr, Ukraine, and a PhD in Agricultural Economics from the Georg-August University of Goettingen, Germany. Acronyms and Abbreviations AfDB African Development Bank FAO Food and Agriculture ALPs accelerated learning programs Organization (United Nations) BARA Bureau of Applied Research in FET further education and training Anthropology G2P government to person CAADP Comprehensive African GDP gross domestic product Agriculture Development GNI gross national income Programme GRADE Grupo de Análisis para el CECAM Caisses d’Epargne et de Desarrollo Crédit Agricole Mutuelles HE household enterprise (Madagascar) HIV/AIDS human immunodeficiency CEMAC Commission de la virus/acquired Communauté Economique immunodeficiency syndrome et Monétaire de l’Afrique ICT information and Centrale (Economic and communication technology Monetary Community of IDS Institute of Development Central Africa) Studies CFA Communauté Financière IFAD International Fund for d’Afrique Agricultural Development COTVET Council for TVET (Ghana) IFC International Finance CREATE Consortium for Research on Corporation (World Bank Educational Access, Transitions, Group) and Equity IFPRI International Food Policy CREO Comprehensive Review of Research Institute Education Outcomes IIEP International Institute for CSAE Centre for the Study of African Educational Planning Economies ILO International Labour DFID Department for International Organization Development (United Kingdom) IMF International Monetary Fund DHS Demographic and Health IPA Innovations for Poverty Action Surveys IPAR Institute of Policy Analysis and EPAG Economic Empowerment of Research Rwanda Adolescent Girls and Young IPPG Institutions and Pro-Poor Women (Liberia) Growth xxvii xxviii Acronyms and Abbreviations KEPSA Kenya Private Sector Alliance SEZ special economic zone KYEP Kenya Youth Empowerment SfC Saving for Change Project SfL School for Life LISGIS Liberia Institute of Statistics SHG self-help group and Geo-Information Services SHIP Survey-Based Harmonized MFI microfinance institution Indicators Program MICS Multiple Indicator Cluster SIEF Spanish Impact Evaluation Survey Fund MIJARC Mouvement International de SMEs small and medium enterprises la Jeunesse Agricole et Rurale SPV special-purpose vehicle Catholique SSATP Sub-Saharan Africa Transport MIX Microfinance Information Policy Program Exchange SUMATRA Surface and Marine Transport MSME micro, small, and medium Authority (Tanzania) enterprise TESDA Technical Education and Skills NBSSI National Board for Small Scale Development Authority (the Industries (Ghana) Philippines) NGO nongovernmental organization TEVETA Technical Education and OECD Organisation for Economic Vocational Education and Co-operation and Training Association (Malawi) Development TFP total factor productivity OHADA Organisation pour TICTS Tanzania International l’Harmonisation en Afrique Container Services du Droit des Affaires TIMSS Trends in Math and Science (Organization for the Study Harmonization of Business Law in Africa) TPA Tanzania Port Authority OSBP one-stop border post TVET technical vocational education and training PMC project management consultant TVVP Technical and Vocational PROMER Promotion of Rural Vouchers Program (Kenya) Entrepreneurship (Senegal) UEMOA Union Economique et R&D research and development Monétaire Ouest Africaine RCB rural and community bank (West African Economic and RCT randomized controlled trial Monetary Union) REP Rural Enterprise Project UNESCO United Nations Educational, (Ghana) Scientific, and Cultural ROSCA rotating savings and credit Organization association UNICEF United Nations Children’s Fund SACCO savings and credit cooperative UNIDO United Nations Industrial SACMEQ Southern Africa Consortium Development Organization for Measuring Educational USAID United States Agency for Quality International Development SEWA Self-Employed Women’s VSLA village savings and loan Association (India) association Overview Youth Employment in Sub-Saharan Africa S ub-Saharan Africa has just experienced The incidence of poverty has fallen as Sub- one of the best decades of growth since Saharan economies have expanded, yet overall the 1960s. Between 2000 and 2012, gross growth in Sub-Saharan Africa has not been domestic product (GDP) grew more than as pro-poor as growth in other regions. Each 4.5 percent a year on average, compared to 1 percent increase in average per capita con- around 2 percent in the prior 20 years (World sumption has been associated with a reduction Bank various years). In 2012, the region’s GDP in poverty of 0.69 percent; elsewhere in the growth was estimated at 4.7 percent—5.8 per- world, that reduction has averaged just over cent if South Africa is excluded (World Bank 2 percent (World Bank 2013). In part, Afri- 2013). About one-quarter of countries in the ca’s poverty reduction has been less marked region grew at 7 percent or better, and several because in many countries the source of African countries are among the fastest grow- growth is primarily oil, gas, and mineral extrac- ing in the world. Medium-term growth pros- tion, not labor-intensive sectors such as agri- pects remain strong and should be supported culture or manufacturing. Young people, who by a rebounding global economy. have weaker links to the world of work than At the same time, many Africans are dissat- the general population, are therefore doubly isfied with this economic progress. According disadvantaged. to the latest Afrobarometer data, 65 percent of Although the current generation of Africans the surveyed population consider economic entering the labor force is the most educated conditions in their country to be the same or ever, many are finding that their prospects for worse than the year prior, 53 percent rate their employment and earnings differ very little national economic condition as “very bad” or from those of their parents. In a few countries, “fairly bad,” and 48 percent say the same about they are worse. their personal economic condition (Afro- Youth in urban areas have been vocal about barometer 2011–12. www.afrobarometer.org). their dissatisfaction. Urban demonstrations 1 2 Youth EmploYmEnt in Sub-Saharan africa consisting primarily of politically active and markets and can, in the longer term, stimu- disaffected youth have become more common late economic growth and development in African capitals. The causes of urban violence • The role of mineral exports—which have undoubtedly include factors much broader shaped the structure of recent economic than employment status (such as inequality and growth but failed to sufficiently increase exclusion), yet dissatisfaction with opportuni- the number of wage jobs most desired by ties, especially in relation to expectations, can youth—and the prospects for reversing this be a contributing factor. Understandably con- trend in the future cerned, especially in light of the Arab Spring, policy makers in Sub-Saharan Africa are mak- • The largely untapped reservoir of opportu- nities in farming, at a time of high global ing youth employment a high priority. Focusing prices for agricultural commodities and ris- on urban youth, they are seeking policies and ing local and regional demand for food programs that can ameliorate the dissatisfac- tion of young people and ease their transition • The massive expansion in access to educa- tion, which is adding many years of school- into adulthood by encouraging the creation of ing, but much less learning, during child- sustainable, productive employment. hood and youth But urban youth are only the most visible and audible part of the employment problem. • The aspirations of youth and policy mak- The majority of young people still live in Afri- ers, which focus on the wage employment ca’s rural areas and small towns. Poorer and less sector at the expense of more immediate educated than their urban counterparts, they opportunities in family farming and house- too struggle to find pathways to adulthood, hold enterprises. especially to stable, remunerative employment Recognizing that it is the private sector that that allows them to support a family. For young creates jobs, the report examines obstacles women, the pathway can be especially treacher- faced by households and firms in meeting ous. As they navigate the school-to-work tran- the youth employment challenge. It focuses sition, their control over their own destiny and primarily on productivity—in agriculture, in their employment choices may be limited by nonfarm household enterprises (HEs), and in social norms. the modern wage sector—because productivity The challenge of youth employment in is the key to higher earnings as well as to more Africa may appear daunting, yet Africa’s vibrant stable, less vulnerable, livelihoods. To respond youth represent an enormous opportunity, to the policy makers’ dilemma, the report iden- particularly now, when populations in much tifies specific areas where government interven- of the world are aging rapidly. Youth not only tion can reduce those obstacles to productivity need jobs, but also create them. Africa’s grow- for households and firms, leading to brighter ing labor force can be an asset in the global employment prospects for youth, their parents, marketplace. Realizing this brighter vision for and their own children. Africa’s future, however, will require a clearer understanding of how to benefit from this asset. Meeting the youth employment challenge in all Africa’s Large Youth Population its dimensions—demographic, economic, and social—and understanding the forces that cre- Sub-Saharan Africa today faces an unprec- ated the challenge, can open potential pathways edented opportunity. Half of the population is toward a better life for young people and better under 25 years of age. Each year between 2015 prospects for the countries where they live. and 2035, there will be half a million more This report begins by laying out the dynam- 15-year-olds than the year before. Meanwhile, ics of the youth employment challenge in Sub- the population in the rest of the world is, or will Saharan Africa today: soon be, aging (figure O.1). • The demographic transition, which created Africa’s youth bulge offers a range of oppor- the youth bulge that is entering African labor tunities. First, the world’s goods and services overview 3 Figure O.1 The structure of Sub-Saharan Africa’s population is different than that in other regions a. Sub-Saharan Africa b. South Asia c. East Asia and Pacific 80+ 80+ 80+ 70–74 70–74 70–74 60–64 60–64 60–64 Age group (years) Age group (years) Age group (years) 50–54 50–54 50–54 40–44 40–44 40–44 30–34 30–34 30–34 20–24 20–24 20–24 10–14 10–14 10–14 0–4 0–4 0–4 100 80 60 40 20 0 20 40 60 80 100 100 80 60 40 20 0 20 40 60 80 100 160 120 80 40 0 40 80 120 160 Population (millions) Population (millions) Population (millions) Male 2035 Male 2015 Female 2015 Female 2035 Source: Based on United Nations 2011. cannot be produced without working- primary commodities for their exports. The age labor. Sub-Saharan Africa, along with failure of this growth to reduce poverty is stark South Asia, can be the main supplier of the in several countries, including oil-rich Angola, world’s workforce, either by producing goods Gabon, and Nigeria, and noticeable in oth- and services in the region or by sending work- ers, such as Mozambique and Zambia. Labor- ers to regions with a shortage of workers. Sec- intensive manufactured exports—the force ond, manufacturing wages in other regions are behind employment and economic transfor- rising. Africa’s labor force should compete for mation in East Asia—are far from taking off in these jobs. Third, increasing concentrations of Africa. In fact, manufacturing’s share in GDP is workers in urban areas can be a source of inno- lower in Sub-Saharan Africa today than it was vation and rapid economic growth (World Bank in 1980; over the same period in Asia, it rose 2008). Young people will be at the forefront of in both lower- and middle-income countries these developments. Finally, if fertility contin- (figure O.2). ues to decline, rapid growth in Africa’s work- To understand the challenge of youth force will mean that the number of working- employment in this context, we start with where age adults relative to “dependents” will rise from Africans work today (see box O.1 for an expla- just around 1 in 1985 to close to 1.7 in 2050, nation of how this report defines employment). providing the space for savings, investment, Contrary to popular perceptions, measured and sustained economic growth. Yet the demo- unemployment in low-income Africa is only graphic transition is not automatic. A critical 3 percent.1 Even in lower-middle-income coun- concern is that the decline in fertility rates has tries, unemployment is quite low (figure O.3). stalled—or not even started—in many African These low unemployment rates may seem countries. But it is also a critical concern that counterintuitive, given widespread concern those of working age are able to be productive. about “unemployed youth,” but most Africans simply cannot afford to be idle. Very few fami- lies can fully support a recent graduate while Growth, Jobs, and Africa’s Labor he or she seeks a job. That the unemployment Force—Now and in the Future rate is highest among university graduates— who mostly come from the top end of the Despite 15 years of relatively rapid economic income distribution—is no coincidence. Only growth averaging more than 4.5 percent a year, in upper-income countries, with broader safety almost all African countries still depend on nets, does substantial unemployment persist, 4 Youth EmploYmEnt in Sub-Saharan africa Figure O.2 Over the past two decades, agriculture’s share in GDP contracted in Africa, but manufacturing did not replace it a. Low-income countries b. Lower-middle-income countries 100 100 90 90 80 80 70 70 60 60 Percent Percent 50 50 40 40 30 30 20 20 10 10 0 0 1990 2010 1990 2010 1990 2010 1990 2010 1990 2010 1990 2010 Sub-Saharan Africa South Asia East Asia and Pacific Sub-Saharan Africa South Asia East Asia and Pacific Agriculture Industry (excluding manufacturing) Manufacturing Services Source: World Bank various years. Box O.1 What is a “job”? In assessing the challenges of youth employment, it is impor- self-confidence; they can contribute to an individual’s overall tant to take stock of what it means to have a job and to have life satisfaction. Not all jobs contribute to these dimensions employment. To some, having a job is synonymous with of well-being. The type of job, working conditions, contract, having a wage or salaried position with an employer. Most benefits, and safety and security at work all matter. Beyond work in Africa is not structured that way, however. This study personal life satisfaction, jobs also contribute to social cohe- follows the approach adopted in the World Development sion through various channels: jobs can shape identities and Report 2013: Jobs, which defines jobs as “activities that gen- how individuals relate to one another, jobs can connect peo- erate actual or imputed income, monetary or in kind, for- ple to one another through networks, and the distribution mal or informal.” That report also notes that not all forms of jobs within society and the perceptions about who has or work can be considered jobs—for example, activities that access to opportunities, and why, can shape people’s expec- are performed against the will of the worker or that violate tations and aspirations for the future, their sense of having a basic human rights. stake in society, and their perceptions of fairness. All of these Across countries, including those in Sub-Saharan Africa, intrinsic aspects of jobs are particularly important for youth. people report that jobs have a broader importance than the income they provide. Jobs can convey identity, status, and Source: World Bank 2012. including among youth. Because the youth regular wage, sometimes with associated ben- employment challenge is configured somewhat efits. In low-income countries, these jobs are differently in Africa’s resource-rich countries divided roughly equally between the public and some middle-income countries, they will and private sectors, although the private share need to approach the challenge in somewhat grows with per capita income. The industrial different ways (see box O.2). sector (mining, manufacturing, and construc- Where do most Sub-Saharan Africans find tion) accounts for less than 20 percent of wage employment? About 16 percent of those in the jobs (about 3 percent of total employment). labor force have “wage jobs”—jobs that pay a The remaining jobs are either on family farms overview 5 (62 percent) or in household enterprises (22 Figure O.3 Where are Africans working? percent), which may be collectively described Estimated structure of employment in Sub-Saharan Africa by country type, 2010 as the informal sector (see box O.5 for a defi- 100 nition of household enterprises). These kinds 90 of jobs—working a small plot of land, selling 80 vegetables on the street, sewing clothes in one’s home—often generate low earnings, partly 70 because the “enterprises” tend to be tiny, typi- 60 % of total cally involving only the family. 50 Will this pattern of employment persist? 40 After all, countries that are not resource rich 30 are creating private wage jobs at a rapid clip— often faster than GDP is growing. The kinds of 20 jobs that are created will depend partly on the 10 structure of growth that Africa attains. Growth 0 and employment projections for this report Low-income Lower-middle- Resource-rich Upper-middle- Total countries income countries income countries assume that growth will remain strong (5–6 countries countries percent a year) and will be fueled by Africa’s 183 40 150 21 395a natural resources—minerals and agriculture. Labor force (millions) The mineral sector is not expected to create Agriculture Household enterprises Wage services Wage industry Unemployed very many jobs. Increases in wage employment (as a share of total employment) will come Source: Fox et al. 2013. from continued diversification of output and Note: On the horizontal axis, numbers show size of the labor force, ages 15–64, in each group. Resource-rich countries included are Angola, Chad, the Democratic Republic of Congo, Guinea, exports and from increased domestic demand Nigeria, the Republic of Congo, Sudan and Republic of South Sudan, and Zambia. a. Numbers do not add to total because of rounding. for services as incomes grow. Since service employment is projected to grow faster than employment in industrial sectors, as it has in tivity of the almost 80 percent of the workforce the past, most nonfarm employment will be who will be in the informal sector—thereby created there. Based on these assumptions, the addressing the underemployment associated number of industrial sector wage and salary with work in this sector. The size of the youth jobs is projected to increase 55 percent over the bulge in Africa and the current structure of the next 10 years. The problem is that this growth economy mean that the majority of this gen- starts from such a small base that it does not eration’s workers will remain in the informal even come close to absorbing the millions of sector for the duration of their working lives. young people entering the labor force each To be sure, in the long run these workers (or year. Because of the low base, the share of their children) will move to the formal sector, industrial wage jobs in total employment will like their counterparts in East Asia and Latin rise only from 3 to 4.5 percent in Sub-Saharan America. Africa, still below the share in other developing This focus on raising productivity in the regions. The share of wage jobs in the service informal sector may seem unusual, given the sector is projected to rise from 13 to 22 percent. publicity around high unemployment among In other words, over the next 10 years, at best university graduates and the recent emphasis only one in four of Sub-Saharan Africa’s youth on creating jobs in the formal manufactur- will find a wage job, and only a small fraction ing sector (Dinh et al. 2012). But university of those jobs will be “formal” jobs in modern graduates still represent only a tiny fraction enterprises. Most young people will end up (about 3–4 percent) of the labor force, come working where their parents do—in family from the richest households, and have the best farms and household enterprises (figure O.4). job prospects. Creating jobs in the formal sec- The employment challenge is therefore not tor is important and should be encouraged, just to create jobs in the formal sector, impor- but the reality is that even if African countries tant as that may be, but to increase the produc- were able to attract an extraordinary infusion 6 Youth EmploYmEnt in Sub-Saharan africa Box O.2 The youth employment challenge in resource-rich and some middle-income countries in Africa Resource-rich countries present particular challenges when it and Malaysia) or to non-resource-based primary commodities comes to employment. Natural resource rents, if poorly man- (Chile)—through sound macroeconomic policies, open trade aged, lead to overvalued exchange rates and uncompetitive and investment policies, strategies to build human capital, and real wages. Such conditions severely hamper job creation in a good business environment (Gelb and Grasmann 2010). export-oriented sectors. At the same time, the few but highly Youth employment problems in some middle-income paid employment opportunities in the natural resource sec- countries, such as Mauritius and South Africa, resemble the tor encourage young people to “wait for a job”—behavior crisis in youth unemployment occurring outside Sub-Saha- that can distort educational choices and aggravate skill mis- ran Africa. Especially in South Africa, unemployment is high matches in the labor market. Resource rents can also engen- (25–40 percent, depending on the definition used), youth der substantial governance problems that stifle growth in unemployment is even higher, and the informal sector is employment. very small. The challenge is to reduce unemployment, which If countries manage their natural resource endowments involves understanding its determinants. While the symp- well, however, they can amass the financial resources to toms and some of the causes may differ (for instance, labor support smart investments in human development, infra- regulations may play a significantly larger role in these more structure, and the promotion of new sources of economic developed countries), some of the solutions proposed in this growth. Resource-rich countries noted for successfully grow- report for other countries are quite similar to those needed ing their economies, such as Chile, Indonesia, and Malaysia, in South Africa—for example, steps to increase agricultural have diversified economically—to manufacturing (Indonesia productivity or improve workers’ foundational skills. Figure O.4 Informal will be normal in 2020 pessimistic about Africa’s future? That we are Jobs by sector denying African workers the hope of emerging from informal employment? On the contrary, 300 raising the productivity of smallholder farms 250 and household enterprises is precisely what will enable the formal sector to develop and thrive. Jobs (millions) 200 It was the key to structural transformation in 150 Asia and Latin America, and it is the key to 100 Africa’s future as well. 50 0 Agriculture Household Wage Wage Youth’s Transition to Productive enterprise services industry Employment Jobs in 2010 New jobs in 2020 Source: Fox et al. 2013. Transition is the defining feature of youth (see Note: The projected number of new jobs added by 2020 will be World Bank 2006). The young leave school 125 million. for work, eventually settling on a career. They grapple with the many decisions that influ- of private investment in very labor-intensive ence when they start a family and how health- enterprises, the formal sector would draw only ily they live (including decisions to engage in a small number of workers from the informal risky behaviors such as tobacco and alcohol sector in the near future (see box O.3). use). Most young people also begin to engage Even if it is realistic to emphasize the role of in the rights and duties of citizens, such as the informal sector, does this mean that we are voting in elections. The question, therefore, is overview 7 Box O.3 What will happen to employment if light manufacturing increases dramatically in Africa? Some observers contend that the structure of employment ca’s low- and lower-middle-income countries. The average could change more rapidly if Africa experienced a radical, annual growth of industrial wage employment would double “game-changing” departure from its current growth path over the decade to 12 percent a year, and total wage employ- (see Lin and Monga 2012; Dinh et al. 2012). What employ- ment would grow 6 percent a year. ment prospects would open to youth by 2020 if African Ultimately, however, the structure of employment would countries picked up manufacturing industries and jobs from change very little compared to today. Low-income countries East Asia beginning in 2015, just as Asian countries such as could expect about 5 million more wage jobs a year, and Bangladesh, Cambodia, and Vietnam picked up industries lower-middle-income countries could expect about 2 million and jobs from Japan and the Republic of Korea in the 1980s new wage jobs—a shift of 10 percent of total new jobs in and 1990s? these countries, representing a small change in the pros- To test the possibilities, we simulated this recent Asian pects for new entrants. These modest gains partly reflect the experience in low-income and lower-middle-income coun- short period used for the projection (which covers 5 years, tries of Africa.a For the simulation, the wage employment whereas the structural change in Vietnam unfolded over 20 elasticity rises to 1.2 to match the historical wage employ- years). They also reflect Africa’s larger labor force and the ment elasticity estimated for Bangladesh, Cambodia, and lower base from which industrial development must start. Vietnam—meaning that employment in the industrial sec- Africa will need at least two decades to change the structure tor would grow 20 percent faster than value added, which of employment sufficiently to offer dramatically different implies very labor-intensive growth. The industrial growth prospects to its youth, which underscores the importance of projection for low- and lower-middle-income countries is starting the change process now. also revised upward to 10 percent a year over 2015–20. This figure is slightly above the median and average indus- try growth rate experienced by Bangladesh, Cambodia, and Source: Fox et al. 2013. Vietnam during the most recent decade (9.3 percent a year). a. Africa’s resource-rich countries are excluded from this simulation, If this “game-changing” scenario could be realized, indus- because even resource-rich countries in East Asia did not achieve the trial wage employment would grow much faster across Afri- transformation of employment simulated here. how to help youth to make these transitions similar arrangements, but others do not. Evi- in a way that puts them on a pathway to pro- dence from urban Tanzania (Bridges et al. ductive employment. The particular challenges 2013) shows that some young people do odd that young people—especially young women jobs and are supported by their families for as or poorer youth—encounter during these long as five years before they settle into wage transitions increase the difficulty of finding a jobs or (mostly) self-employment. Moreover, pathway to productive employment (although first-generation school leavers aspiring to be securing productive employment is important wage workers lack a family history in formal for all members of society, as box O.4 explains). employment. They may not have networks to The transition from school to work as well help them to find jobs. as between sectors of employment (between Young women may be particularly disad- farming and a wage job, for example) is par- vantaged by other dimensions of the transi- ticularly difficult for young Africans. Many tion, such as family formation, compared with lack the means, skills, knowledge, or connec- young men (figure O.6). Social norms tend to tions to translate their education into produc- enforce job segregation by gender. For instance, tive employment. Nor is there a structured young women in the household enterprise sec- path to follow. Many young people combine tor work mostly in narrowly defined fields such school with work for many years (figure O.5). as dressmaking, even though a range of other Some move straight into apprenticeships and occupations could be more lucrative. 8 Youth EmploYmEnt in Sub-Saharan africa Box O.4 Youth employment versus overall employment A key part of improving employment opportunities for Afri- to help young people to enter agriculture or start household ca’s youth is to understand and address Africa’s overall eco- enterprises do not generally reduce such opportunities for nomic challenges. For that reason, much of the discussion in adults. But efforts to facilitate access of youth to the modern this report focuses on identifying policies that can increase wage sector could potentially displace adults working in the the productivity of all employment—in agriculture, nonfarm sector. On the one hand, assistance for young people might household enterprises, and the modern wage sector. The have long-term benefits if it sets them on a productive path. report also identifies promising policies that focus specifically On the other hand, older workers may have families and on helping youth to transition more successfully to higher- other economic dependents who rely on their income. There productivity work in each of those three sectors. is no clear argument that social welfare will be improved by The focus on youth raises the question of whether it might favoring younger workers over older ones. This report advo- be socially beneficial to support employment policies that cates seeking to increase the opportunities for all workers, favor youth over other members of society. Special efforts while helping youth to overcome their particular constraints. Figure O.5 The transition from school to work in Sub-Saharan Africa is slow a. Rural b. Urban 100 100 80 80 60 60 Percent Percent 40 40 20 20 0 0 15 18 21 24 27 30 33 15 18 21 24 27 30 33 Age (years) Age (years) Working, not at school Both working and at school At school, not working Not at school, not working, and looking for work Source: Based on standardized and harmonized household and labor force surveys (see appendix). Policy Priorities for Addressing that shape young people’s potential to find a Youth Employment pathway to productive work in the three sec- tors: human capital and the business environ- To understand the challenges constraining ment. Box O.5 describes this framework more productivity and earnings for youth and to fully. orient how policies should be targeted, this On the human capital side, the role of basic report considers the three main sectors where education dominates interventions in all three people work: agriculture, nonfarm household sectors. As a complement, different approaches enterprises, and modern wage employment. are needed to build post-school skills in agri- It then distinguishes between two dimensions culture, household enterprises, and the modern overview 9 Figure O.6 Family formation starts earlier for young women than for young men a. Females b. Males 100 100 80 80 % of age cohort % of age cohort 60 60 40 40 20 20 0 0 15 20 25 30 35 15 20 25 30 35 Age (years) Age (years) Ever had a child Ever married Ever had sexual intercourse Source: Based on demographic and health survey data in 28 countries (see appendix). Box O.5 Framework for analyzing youth employment Three main sectors of employment: The modern wage sector includes small, medium, and large firms that employ five or more workers on a continu- Agriculture is where the vast majority of work takes place ous basis. It also includes the public sector, which in some in Sub-Saharan Africa. Agriculture occupies more than 70 countries is a large share of the modern wage sector. In the percent of the labor force in Africa’s low-income countries low- and lower-middle-income countries of Sub-Saharan and more than 50 percent in its lower-middle-income coun- Africa, roughly half of wage employment is in the public sec- tries. African farmers are predominantly smallholders who tor. In this report, we focus only on the private sector, where consume a large share of what they produce. One recent the potential for job growth is the greatest. collection of household surveys estimates that the share of own-consumption is around 50 percent (versus 20–30 per- Two dimensions that shape the pathways to productive work: cent outside of Sub-Saharan Africa). • Human capital—the supply side, meaning the abilities, edu- Household enterprises are unincorporated, nonfarm busi- cation, skills, family connections, networks, and other char- nesses owned by households. They include self-employed acteristics that are embedded in an individual and allow people running businesses that may employ family mem- that person to find opportunities to be productive, increase bers without pay, but may also employ less than five non- earnings, and achieve income security family workers on a continuous basis. The vast majority • Business environment—the factors outside the worker’s (70 percent) of nonfarm enterprises today are pure self- immediate control that affect productivity (access to land, employment: just the owner operating the HE. About 20 capital, and finance; infrastructure; technology; and mar- percent of these enterprises include a family member in the kets), as well as the government policies, regulations, and operation, and only 10 percent have hired someone outside programs that may affect the choice of economic activity of the family. and how the activity is conducted. wage sector. On the business environment side, structure services and access to finance as well raising farm productivity requires enabling as a space to operate will play critical roles; to farmers to gain access to finance and secure boost modern sector wage jobs, business regu- land tenure; in the nonfarm HE sector, infra- lations and infrastructure will be important. 10 Youth EmploYmEnt in Sub-Saharan africa Human Capital: The Fundamental complete primary school have low levels of Role of Basic Education basic skills: among sixth-graders 43 percent in Tanzania and 74 percent in Mozambique are Sub-Saharan Africa has seen a rapid increase at or below the “basic numeracy” level, while in the number of children who complete pri- 44 percent in Mozambique cannot “read for mary school, from about 50 percent in 1991 to meaning” (Hungi et al. 2010). A few years of 70 percent in 2011. The average young Gha- low-quality basic education will not confer naian or Zambian today has more schooling much of a gain in productivity if students never than the average French or Italian citizen had master even basic literacy and numeracy— in 1960 (Pritchett 2013). The current cohort of although so-called “second-chance education” youth in Sub-Saharan Africa will be the most approaches offer some potential for catching schooled ever. up. For many, however, primary schooling is Educational attainment shapes employment the highest level of schooling they will achieve. opportunities, as reflected in the substantial These results are especially troubling because variation in the educational profiles of young they suggest that school leavers have a fragile workers in each sector (figure O.7). Interna- foundation on which to build more specialized tionally benchmarked learning assessments skills. suggest that many young people nevertheless Even students who make it to the second- lack the skills to compete in a global market- ary level—those who will most likely head place. Schooling is not learning. Deep deficien- to the modern wage sector—are not globally cies in the quality of education mean that the competitive. In the most recent internationally effect of schooling on productivity is far below benchmarked assessment of eighth- and ninth- its potential. The poor quality of education grade students, 79 percent of Ghanaians and 76 directly constrains productivity and hinders percent of South Africans do not surpass the individuals from acquiring new skills. lowest measured level of math proficiency. For Learning in primary school is often mini- comparison, the international mean is 25 per- mal: 80 percent of Malian third-graders and cent, and the corresponding scores are 67 per- more than 70 percent of Ugandan third-graders cent for Indonesian students and 45 percent cannot read a single word (Cloutier, Rein- for Jordanian students.2 Beyond these cogni- stadtler, and Beltran 2011). Even students who tive skills, many youth lack the behavioral and Figure O.7 Education shapes opportunities a. Ages 15–24 b. Ages 25–34 100 100 80 80 60 60 Percent Percent 40 40 20 20 0 0 Agriculture Nonfarm Wage Wage Agriculture Nonfarm Wage Wage household without with household without with enterprise contract contract enterprise contract contract No education Primary not completed Primary completed Lower secondary completed Upper secondary completed Post-secondary Source: Based on standardized and harmonized household and labor force surveys (see appendix). overview 11 socioemotional skills, sometimes called “soft It has plentiful supplies of land and often of skills,” that are needed to get, keep, and be pro- water, unlike other regions. If young people can ductive in a job. gain access to these resources and use them in Addressing this lack of learning is not easy. conjunction with strategies to make agriculture Surveys of schools reveal substantial failures more productive, the results could be transfor- in service delivery. For example, absenteeism mative for livelihoods and economic growth. among teachers is between 16 and 20 per- More than two-thirds of the young people who cent on a given day in Kenya, Senegal, and work in rural areas already work in agriculture, Tanzania, and primary school students in and most will remain there, even if the non- those countries experience only about two farm sector develops extremely rapidly. to three hours of learning a day.3 Reform- Although agriculture is the most immediate ing the accountability framework that allows means of generating income and employment such poor performance to persist is key. Bet- for large numbers of young people, efforts to ter information on performance must be accelerate agricultural growth and improve complemented by targeted approaches that food security in Africa have been conceptually increase oversight by the people who are most separated from efforts to create jobs for young affected: students and their parents. Steps to people. Yet these goals are highly complemen- ensure that teachers are well prepared for tary. Increasing young people’s opportunities teaching and supported in their tasks are crit- for productive work in rural areas is arguably ical for creating a cadre of high-performing the most important catalyst for Africa to reap professionals. The rise of private schools in its demographic dividend. Africa—schools that often deliver superior Low agricultural productivity is the primary performance at lower cost—should not be impediment to overcome. Agricultural produc- stifled but should be encouraged and chan- tivity remains lower in Africa than in any other neled to give larger numbers of students the region of the world, and agriculture is the least opportunity to learn. productive sector in African economies. This is Improvements in basic education will lay true, despite the fact that total factor productiv- the foundation for improvements in produc- ity, as well as land and labor productivity, have tivity. At the same time, to maximize young been increasing in African agriculture since the people’s chances of transitioning successfully 1990s (Fuglie and Rada 2013; Nin-Pratt, John- to remunerative employment, complemen- son, and Yu 2012). Productivity may increase tary actions are required to improve the busi- further as food prices continue to rise, because ness environment and develop human capital. the value of output for the same amount of Those actions will vary by sector of employ- inputs will increase, but these productivity ment, as discussed next. indicators are far below levels achieved in other regions during their phases of rapid economic growth. Indeed, African countries are not fol- Raising the Productivity of lowing the trajectory of other regions in which Smallholder Farmers productivity gains on farms, combined with higher productivity and more opportunities off Agriculture can and should be a sector of the farm, shifted labor rapidly out of agriculture. opportunity for Sub-Saharan Africa’s youth. The effects of low agricultural productivity The growing demand for food means that extend well beyond rural areas and farm house- there is scope for supply to take advantage of a holds. An underappreciated result of low agri- growing market. And growth in demand is not cultural productivity in Africa is high domestic limited to Africa’s expanding domestic markets. food prices. Local prices are poorly correlated Global food prices are at their highest point with global prices, especially in the interior. in several decades, and they are expected to When domestic prices are high, this increases remain high for at least the rest of the decade. the cost of food and pushes up wages, contrib- Africa is well positioned to produce for this uting to Africa’s overall lack of competitiveness. large and potentially lucrative global market. High domestic prices undermine real earnings 12 Youth EmploYmEnt in Sub-Saharan africa for everyone except those farmers who are net on regulatory and policy reform, and some food sellers and whose production costs are attention to including young people, agricul- relatively low. Greater agricultural productivity ture could absorb the large numbers of new job will reduce the domestic cost of food and cre- seekers and offer meaningful work with large ate more and better-paying nonfarm jobs for public and private benefits. everyone, especially youth. Broadly speaking, three pathways are avail- Understanding why the least productive able for rural youth in agriculture: (1) full-time sector is also the largest may provide insight work on family farms, (2) part-time farm work, into how farmworkers can increase their pro- combined with running a household enterprise, ductivity. Small-scale farmers may be caught which can include the sale of farm services or in a trap that prevents them from generating inputs, and (3) wage work. To increase the sufficient earnings to invest in expanding pro- productivity of these pathways to agricultural duction and productivity. The vast majority employment, constraints need to be relaxed in of African farmers work on tiny plots of land, at least four areas: credit and financial services, often under uncertain tenure arrangements. land policies, infrastructure, and skills. They cannot take advantage of economies of scale (where they exist), modern agricultural Credit and Financial Services inputs, and mechanization. Poor rural infra- Because they work in a risky environment and structure (transport, electricity, and irrigation) lack collateral, rural households face major frustrates farmers’ efforts to obtain afford- constraints in obtaining capital and credit. able inputs such as seed and fertilizer, market Traditional financial institutions do not find their output profitably, or harness new land it profitable to provide agricultural credit. for cultivation. The lack of irrigation makes Instead, various nongovernmental organiza- agriculture more vulnerable to the vagaries of tions (NGOs) and banks have been innovating weather. with new instruments and institutions, some Low levels of education and pervasive health of which look promising. First started in Niger problems (two outcomes of the poor delivery in 1991, village savings and loan associations of services) prevent farmers from increasing (VSLAs)—where members save on a regular their own productivity, much less migrating to basis and lend money on terms determined by areas where agriculture or some other occupa- the group—have spread to 39, mostly African, tion might be more productive. Rural youth countries. They hold great potential for assist- have significantly lower levels of education ing young people to save the funds to invest in than their urban counterparts. Many endemic a farm and to get access to credit, while also diseases are not difficult to manage or cure, but benefiting from the mentoring and access to these areas of health policy often receive little information that come from being a member. attention despite their high cost to the rural VSLAs could help youth in rural areas to estab- economy. Illness, apart from reducing the labor lish themselves in agriculture as well as in non- available to farm households, can deplete sav- agriculture sectors. ings through costly treatments and cause dis- Various institutional arrangements and tress sales of assets. incentive schemes can also widen access to For Africa to raise agricultural productiv- credit. Examples include different forms of ity sufficiently to support overall growth and collateral (chattel mortgages and others), leas- provide a remunerative livelihood for those ing (which requires no collateral, such as the working in the sector, farming must shift rap- DFCU Leasing Company in Uganda), linking idly from being an occupation of last resort and credit to extension services (thereby addressing low productivity to one of technical dynamism multiple constraints simultaneously, because and recognized opportunity (see box O.6 for a young people also need information), and con- discussion of the link between productivity and tract farming (in which the wholesaler provides jobs in agriculture). With much higher priority credit for inputs). accorded to well-designed programs of public None of these innovations in rural finance investment in agriculture, continued progress is exclusively for young people. Nor should overview 13 Box O.6 Does labor need to move out of agriculture as productivity grows? If Africa’s farms get bigger as productivity grows, will work- in the developed world in favor of technology-intensive ser- ers be pushed out of the sector and have nowhere to go? vices and products. The cost of capital in Africa remains high, Some observers have raised this concern, and it is under- reflecting low saving rates, high costs of doing business, standable. In other parts of the world, growth in produc- and rising demand for capital-intensive infrastructure invest- tivity has been accompanied by an increase in farm size, ments. Agriculture’s share in African GDP (and ability to a reduction in labor intensity, and the exit of labor from employ labor) therefore could remain steady or even grow, agriculture. rather than shrink, with development. Reserves of good Africa’s endowment of land and labor and its recent his- tory offer perspectives to quell this concern. If land is avail- Table BO.6.1 Increase in crop area harvested, agricultural labor able and crop area is still expanding, increased farm size force, and output per worker in Sub-Saharan Africa, 1960–2008 need not displace labor, especially where the cost of capital and 1990–2008 to invest in mechanization is high. As shown in table BO.6.1, % increase between 1960 and 2008, crop area in Africa expanded, Indicator 1960–2008 1990–2008 the workforce in agriculture grew, and output per worker crop area harvested 42 20 increased despite the larger number of workers per hect- are. During the later years, from 1990 to 2008, the pace of agricultural labor force 125 21 growth in the labor force slowed relative to the expansion of output per worker 21 9 land, and output per worker continued to increase. Source: Fuglie and Rada 2013. Africa’s farms can grow in number, size, and productiv- ity without displacing labor. The subcontinent’s experience land, abundant water, and an energetic young labor force with structural change in the twenty-first century is not likely are tremendous assets at a time when the global economy to replicate the experience in other regions in earlier times. urgently needs more food and fiber. Whether Africa can use The current context is one of high global food prices, huge these assets productively depends on finding ways to ease potential for growth in area and yield, few nontradable the constraints on access to land and address the barriers to manufactured goods, and shifts in comparative advantage agricultural growth highlighted in this report. young people be segregated as a group and The problem of insecure and unclear land offered financial services designed specifically rights can be addressed by developing an inven- for them. The risks of working with this cli- tory of registered land and improving land ent base are high, and separating young people registration. Geographic information systems from a larger pool for sharing risks would make have made this process increasingly practical. them even less attractive to financial institu- Tenure security is also reinforced by improving tions. A better approach is to support a range land titling procedures. When farmers know of innovations in finance that facilitate out- that their land rights are secure, they are more reach to small farmers and rural entrepreneurs. likely to invest in improving their land. A recent When necessary, additional features should be impact evaluation of a land registration pilot added to enhance the ability of these programs in Rwanda shows that more secure land tenure to serve young people. increased investments in soil improvement by 9 percent among male farmers and 18 percent Land Policies among female farmers (Ayalew Ali, Deininger, Insecure and unclear land rights, as well as and Goldstein 2011). constraints on renting or otherwise using land, Once land is registered and titled, land rental pose problems for young people in agriculture. markets can develop. Land rental markets have Some young people own land (albeit small been shown to promote commercial farming in plots), but ownership is strongly concentrated Ghana and to encourage the transfer of land among older adults (figure O.8). to smaller-scale farmers in Sudan. By contrast, 14 Youth EmploYmEnt in Sub-Saharan africa Figure O.8 Young people are unlikely to own land 2008), but even here, investments should be evaluated case by case and not accepted across 80 the board. For example, paved rural roads are 70 not always a good investment, especially in areas % who own at least one plot of land where motorized traffic is low. Paved roads are 60 more expensive to maintain, so maintenance is 50 often not funded; another issue is that roads are prone to becoming tools of political patronage 40 (Raballand and Macchi 2009). 30 Skills 20 Given their low levels of formal education (fig- 10 ure O.7), youth in rural areas can increase their productivity with more and better schooling. 0 In turn, higher productivity increases demand 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60+ for schooling, triggering a virtuous cycle. The Age (years) evidence used to identify externalities in edu- Tanzania Uganda Malawi cation usually comes from agriculture: farm- ers learn from neighboring farmers, especially Source: Based on data from the World Bank’s Living Standards Measurement Study, Integrated Surveys for Agriculture (2011 in Malawi, 2010 in Tanzania and Uganda). more educated ones, who are also more likely to adopt new technologies (Conley and Udry 2010; Rosenzweig 2012). All of these effects apply more strongly to young farmers. restrictions on land rental markets have inhib- In addition to basic education, high- ited both the farm and nonfarm sectors, dis- productivity farming requires specific skills, couraging people with land from taking jobs in such as skills in processing, marketing, machin- the nonfarm sector for fear of losing their land. ery operation and repair, transport, logistics, Agribusiness can increase productivity, and and quality control. In some countries, agri- parts of Africa, especially the Guinea savannah, cultural vocational training institutes (some of have huge potential for commercial agricul- which are associated with universities) tradi- ture on both large and small farms (Morris, tionally have provided these skills. These insti- Binswanger-Mkhize, and Byerlee 2009). Large tutes have a mixed track record, mostly owing commercial landholdings and agribusiness to the disconnect between academic, lecture- may prove politically contentious, however. style teaching and the need for on-the-ground, How large farms are integrated into a diverse practical training. farm structure and how smallholders are com- Agricultural extension programs have a dis- pensated for land that they make available to appointing history in Africa, mainly because large commercial operators are critical issues to of poor incentives and accountability. Better resolve. Other ways of aggregating smallhold- results are coming from new programs that ers, such as producer associations or contract empower farmers by giving them a choice of farming, could also be practical approaches for providers and services from among a range of improving productivity when increasing the public, private, and nongovernmental agen- scale of production will contribute to lower cies. Another approach, farmer field schools, costs. involves participatory methods of learning, technology development, and dissemination Infrastructure and appears to be especially successful in build- In some—but not all—cases, investments in ing women’s skills. Business incubators and rural infrastructure can have huge rates of rural alliances that bring together commercial return. Typically such investments occur in buyers with producer organizations are further more densely populated areas (World Bank ways of boosting agricultural incomes. In all overview 15 such efforts, the use of information and com- and imports. As a result, HEs that are involved munication technology can benefit, and benefit in manufacturing do not persist as long as HEs from, the participation of young people. that provide services. Historically, HEs have tended to remain tiny or to disappear; very few grow into even small Increasing the Productivity of or medium enterprises. Data from West Africa Nonfarm Household Enterprises show that, even after 10 years of operation, the capital stock of these businesses remains The majority of people who work outside the the same. Most enterprises never hire another farm sector are engaged in informal, household worker. So the employment they provide, enterprises. Often such an enterprise is one pil- including the employment for young people, lar of a diversified livelihood strategy: Many comes from seizing a business opportunity and households are engaged in the farm and non- starting a new enterprise. farm sectors at the same time (30–50 percent Despite their small scale, HEs are an instru- of rural households have a nonfarm HE). In ment for reducing poverty in Africa, with the urban areas, many households with an enter- potential for becoming an even more power- prise have a family member earning a wage ful one. They tend to be found in richer areas. income, a pattern that is likely to increase over Households with enterprises are less likely to time (Fox and Sohnesen 2012). be poor and are clustered in the middle quin- Although their productivity is relatively tiles of the income distribution. Rural house- low, HEs provide earnings that are usually holds gain a higher hourly income from HEs higher than anything their owners could than from agricultural work. In urban areas, obtain in the agriculture sector. Most HEs have some HE owners make more money than they no employees and are pure self-employment. would from a wage job. In fact, controlling for Few include a family member as additional education and skills, the reported consumption labor, and only 10 percent hire outside the of rural and urban households with an HE in family. Some are started in response to a Africa is no different from that of households local business opportunity (such as increased in wage employment (Fox and Sohnesen 2012). demand for a service), whereas others are Indeed, they also report being “happier” (Falco started because the household lacks alterna- et al. 2013). tive means to earn a living. Most governments continue to ignore, What do HEs do? They mostly sell services neglect, or undermine the potential of this (hairdressing, repairs) and internationally and sector. Hawkers and sellers are regarded as an locally produced consumer goods (used cloth- ing, household supplies, vegetables, eggs). They unattractive annoyance to be chased out of the also contribute to the industrial sector, trans- business districts in capital cities. Advocates forming agricultural goods or natural resources of formal employment criticize HEs for not into charcoal, bricks, iron work, or processed offering the benefits and security of a wage job. grain. Some pursue artisanal activities such as Thinking that HEs can be transformed into woodworking, dressmaking and tailoring, and small and medium enterprises—for which they construction. have a strategy—governments try to formalize HEs sell low-cost goods and services mainly these informal enterprises. This transformation for the local market, which lacks a modern rarely takes place, however, because this is not service sector. In urban areas, the traders and the intention of the owners. Lacking support, hawkers substitute for convenience stores and HEs just try to survive. In Tanzania, the law shopping malls. The low-quality manufac- prohibits businesses from operating without tured goods these enterprises produce, such fixed premises, but it does not stop the govern- as homemade bricks and furniture, eventually ment from collecting taxes and fees from those are replaced by higher-quality locally produced same enterprises. (This describes 80 percent goods (made in brick and furniture factories) of HEs.) 16 Youth EmploYmEnt in Sub-Saharan africa By contrast, Ghana, a country with a rich cult to implement the law—if there is one. For trading history dating to precolonial times, example, in Dar es Salaam, Tanzania, local gov- explicitly incorporates HEs in its strategy doc- ernments are not empowered to decide whether uments and the institutions that implement hawkers can use the land next to national roads. them. In 2006 the government established an Just as land rental markets can facilitate access objective to “enhance productivity and income/ to agricultural land, they can facilitate access to wages … in all sectors of the economy, includ- space for HEs to do business. ing the informal economy” (Republic of Ghana In addition to secure space, HEs need ser- 2006). The Trade Union Congress supports the vices, such as security, sanitation, electricity, development of HE organizations and their transport, and water supply. For the most part, integration into the consultative mechanisms they are willing to pay for these services (and between government and the private sector. do pay for them) through fees and taxes. In fact, To realize the potential of the HE sector for HEs pay local business taxes at a higher rate productive youth employment, national strat- than large businesses but often fail to receive egies that recognize the sector’s potential and any services. HEs have little leverage to improve propose a supportive policy framework need this situation (Fox and Sohnesen 2012). Local to be developed. Such strategies should endorse authorities are not accountable to HEs, because the creation of independent HE associations to ineffective political decentralization and weak reduce the costs of reaching individual enter- legal status deprive HEs of a voice in local prises and to give this sector a voice in govern- governance. ment decision making. Local interventions need to address three key areas: the local busi- Credit and Financial Services ness climate in urban areas, access to credit and Lacking access to formal sources of finance, financial services, and skills. young people struggle to raise capital to start and operate a business. The problem is exacer- Urban Policy bated by the fact that business and household One of the most frequently cited constraints finances are linked, so that lumpy household on the productivity of urban HEs is the lack of expenditures (school fees, repairs, and so on) access to space and sometimes outright harass- and external shocks (family illness) spill over ment, legal and extralegal, by local authorities. to the business. Virtually all HEs in Africa Insecurity of premises discourages entry (the today report that their own funds or loans main form of growth in this sector) as well as from friends and relatives enabled them to investment in the enterprise. Governments can start their businesses, and the overwhelming help rather than hurt this sector by incorporat- majority of existing businesses report that they ing the growth of HEs in planning processes. did not obtain any type of loan over the last 12 Without planning, traders and vendors crowd months. An expansion of credit cannot make sidewalks and roads, leading to massive conges- up this gap and may even make it worse, as tion. Usually the situation escalates to a crisis, recent experiences in India have shown. To use and authorities use police or security forces credit effectively, a borrower has to be able to to “decongest” the city. Because the eviction put aside money regularly to service the loan. policy is rarely permanent, the cycle usually For this purpose, the establishment of a savings resumes. If governments had planned for the mechanism is critical. growth of HEs and provided adequate space for At the root of the credit problem for HEs is them in the key areas of foot traffic in the city, the lack of financial inclusion among house- employment growth and social stability would holds in Africa (World Bank 2014). Households have followed. not only lack sources of credit but also have dif- Alongside planning, national policy makers ficulty finding reliable places to safeguard their can clarify land rights in urban jurisdictions, savings. As a result, it is challenging for them giving local governments scope to provide HEs to accumulate the funds to start or expand a with locations to operate. Overlapping land business. This problem is especially acute for regulations and responsibilities make it diffi- youth and for females. Research has shown that overview 17 access to a secure place for savings is particularly develop low-cost products for small savers and important for women’s ability to build up sav- for households to use banks by channeling pay- ings for a business (Dupas and Robinson 2013). ments to households through banks, including The challenge of providing access to finan- microfinance banks with a client base among cial services for poor people is a common one. lower-income households.4 Ghana developed Because of economies of scale, the spread of a system of rural banks to process payments formal banking services (banks, postal savings, to cocoa farmers; today these banks provide formal savings institutions other than banks) low-cost accounts in rural areas. Ghana and usually rises with income and urbanization— Rwanda are both encouraging the spread of banks go where the money is. But this is not mobile banking to further widen access to the whole story in Africa. FINDEX data show financial services. Countries in the Western that even at the same level of income per cap- and Central African monetary unions using ita, national policies can produce very different the CFA currency (UEMOA and CEMAC) are results (figure O.9). Low- and lower-middle- also changing their regulations to foster more income countries such as Ghana, Kenya, and inclusion. Benin, which encouraged the estab- Rwanda have achieved greater financial access lishment of microfinance institutions, was an than other African and non-African countries early mover in this group. at similar income levels. They have done so pri- marily by reducing the costs of serving small Skills savers and those in remote areas. Training can help to structure a pathway to In Kenya, banking access increased through youth employment in HEs. Training pro- the pioneering use of mobile banking technol- grams—both for entry and for improving ogy. By increasing branchless banking, mobile incomes and sustainability—are the most com- banking allows accounts to be maintained monly provided government and donor inter- at relatively little cost to savers and borrow- vention to support HEs, whether or not they ers. Today, about half of the adult population target youth. Programs provide (1) technical of Kenya uses mobile banking. In Rwanda, training in a specific sector (such as tailoring, government provided incentives for banks to metalworking, operating a bakery); (2) business Figure O.9 At the same level of income per capita, national policies can produce very different levels of financial inclusion Individuals ages 25 and older with a bank account (%) 100 90 80 MUS MNG 70 THA CHN 60 ZWE RSA 50 AGO 40 RWA KEN IND EGY NGA BWA 30 UGA SWZ GHA COM BFA GAB 20 10 COG SEN NER MLI TKM 0 350 3,500 GDP per capita 2011 (current US$) CFA countries Non-CFA Sub-Saharan African countries Rest of the world Trend of CFA countries Trend of non-CFA Sub-Saharan African countries Trend of rest of the world Source: Based on Global Financial Inclusion (Global Findex) Database, World Bank, Washington, DC, http://www.worldbank.org /globalfinder Note: The x axis of this figure is on a log scale. 18 Youth EmploYmEnt in Sub-Saharan africa skills or financial literacy (such as basic account- the engine for employment and growth in the ing or money management); (3) behavioral and medium to long term for Africa. No country life skills; or (4) a mix of these skills. Programs has developed without this sector coming to specifically intended for youth focus primarily dominate employment. This is the sector that on providing the skills needed to enter the HE can exploit economies of scale and produce sector and have included all four types of train- for export. The multiplier effects to the house- ing. Programs targeted at existing HEs tend to hold economy from the creation of wage jobs focus more on business skills to strengthen or are strong. Most secondary-school graduates expand HEs. aspire to work in this sector. When these aspi- The good news for youth employment is rations are not fulfilled and graduates must that programs designed to facilitate entry (ver- resort to working in the household enterprise sus those focusing on productivity) appear to sector, for example, the risks for social insta- have had some success, so positive models are bility and political violence are high. While emerging. Apprenticeships and on-the-job Africa’s young people seem to have no spe- training can help young people, provided that cial advantage or disadvantage when it comes these programs are closely tied to market sig- to modern wage employment, they remain a nals. For this reason, private providers, includ- consistent share of that employment when it ing existing businesses, are often the best source grows (figure O.10). for this training. Youth often face multiple The modern wage sector has been creat- constraints in entering the HE sector, and the ing jobs at a fairly rapid pace in Sub-Saharan most promising pilots are delivering interven- Africa—usually faster than GDP growth. The tions that tackle multiple constraints (offering problem is that the sector has grown from such behavioral, business, or technical skills training a small base that it still cannot absorb the mil- together, or combining training with measures lions of young people entering the labor force to tackle credit constraints through savings every year. To generate jobs at a rate that is groups, grants, and other means). Many of these commensurate with growth in the labor force, “bundled” interventions have been expensive, the export-oriented enterprises—with their however, and they are yet to scale up in Africa. potential to sell to global markets—will have Overall, despite the large number of train- ing programs, evidence of their effectiveness Figure O.10 The share of youth in wage employment in the HE sector in Africa remains thin. More tracks the share in the general population systematic use of careful evaluations is clearly required, including impact evaluations that 100 measure outcomes among program partici- Share of wage employment for population 90 pants and compare them to a relevant group 80 of nonparticipants. At the very minimum, gov- 70 ernments should encourage all programs to ages 15–34 (%) track and report outcomes. At the same time, 60 governments should not attempt to deliver 50 training directly but rather focus their efforts 40 on market-enhancing programs that dissemi- 30 nate information about training opportunities and enable disadvantaged youth to access train- 20 ing that is already available. 10 0 0 10 20 30 40 50 60 70 80 90 100 Improving Competitiveness to Share of wage employment for population Boost Modern Sector Wage Jobs ages 15–64 (%) 45 degree line Although small (16 percent of the labor force), Source: Based on household and labor force surveys (most recent the formal wage employment sector represents data available). overview 19 to be the engine for job creation in this sector. tion. Infrastructure policies and regulations Since most African economies are small, access block firms’ access to infrastructure services, to external markets is the key to unleashing and building infrastructure without improving the modern wage sector’s full potential. The policies and institutions is unlikely to solve the scope for trade is wide. Even services are traded problem. Conversely, improving policies such internationally, although these tend to call for as electricity and water tariffs or trucking regu- relatively high-level skills still lacking in much lations can go a long way to improving services of Sub-Saharan Africa. Moreover, the employ- and attracting the private sector. ment effects of trade go beyond exporting Expensive or limited financing prevents firms, as rising demand from the export sec- firms from investing to improve productivity tor—for inputs, consumer goods, and ser- and also keeps productive firms from grow- vices—increases opportunities in other parts ing. Banks in Africa set high collateral require- of the economy. ments and high risk premiums partly because What factors are constraining export- they lack the credit information systems that oriented enterprises in Africa today? The main help lenders to evaluate prospective custom- constraint on the growth of an export-oriented ers. This problem can be addressed by policy sector in Africa is low productivity. and regulatory reform. Africa also needs bet- The underlying causes are not identical ter, well-enforced creditor rights, which will across the continent, even if they have similar ease lending by providing protection for lend- effects. In some countries, the cost of comple- ers when borrowers default. Improved creditor mentary inputs to labor (electricity, overland rights and contract enforcement will also allow transport, and so on) is too high. Clearly, the borrowers to use a broader range of assets as cost of transporting goods across borders is collateral. prohibitively high in the region, and the need In developed economies, innovative new for better transport infrastructure, simpler cus- firms are constantly emerging and growing, toms procedures, and expedited inland border while unproductive incumbents are leaving. crossings is acute. In other countries, bureau- This churning is a major source of aggregate cratic red tape delays investors’ access to land productivity and employment growth in the or permits. The high costs of financial interme- modern wage sector. In Africa, this process diation are starving firms and entrepreneurs of is held back not only by financial constraints the capital needed to implement good ideas. on entrepreneurs but also by difficult formal In many countries, poor connectivity has frag- requirements for business entry or expansion. mented local markets, suppressing competition Governance issues, such as corruption in the and reducing the pressure to improve produc- granting of business licenses and other permits, tivity. These business climate issues would be hamper this process as well. a problem for productivity (and hence, youth In short, the constraints to Africa’s pro- employment) even if firms only produced for ductivity are a combination of market and domestic markets. government failures. Government failures in What can be done to increase Africa’s low particular either have been difficult to rectify— labor productivity? The entire business climate especially when they require economywide pol- comes into play here, with some exceptions. For icies such as deregulation or tariff reform—or, example, labor market regulations—consid- when implemented, have not delivered results. ered an important determinant of productivity Vested interests inside and outside government elsewhere—do not play a major role in Africa, can prevent reforms or their implementation. except for South Africa. In many settings, reg- Analysis of Ethiopia’s light manufacturing sec- ulations may exist on the books, but they are tor shows that the leather goods industry could rarely enforced. create 90,000 jobs (it currently employs 5,000). To unleash this potential, value-chain analy- Infrastructure and the Business Climate sis shows that the government has to remove Infrastructure is a major problem, but build- a series of policy-induced constraints, such as ing new infrastructure may not be the solu- trade restrictions, anticompetitive practices, 20 Youth EmploYmEnt in Sub-Saharan africa and financial regulations that affect the sector important for all types of wage workers, partly (Dinh et al. 2012). The scale of such a reform because they facilitate further training and on- means that it could take some time. the-job learning in firms. A complementary approach is to create a To avoid creating more unemployable uni- localized environment—say, a special economic versity graduates, higher education policies and zone (SEZ), with the requisite infrastructure curriculums also need to be geared toward pri- and deregulation—in which industries that vate demand. Financing in public universities could benefit from proximity can cluster and should include a private component (especially flourish. The government provides land and for those who can pay), so that signals from the functioning infrastructure services for the SEZ. private sector are received (Devarajan, Monga, Although in China this approach attracted for- and Zongo 2011). Such a shift would improve eign investment and know-how and helped equity in access to higher education as tuition China to become an export manufacturing payments become means-tested rather than powerhouse, it has not yet been implemented free across the board. Also, those who pay will successfully in Africa. demand value for money. Besides giving careful thought to the The Sub-Saharan African experience with design and implementation of these interven- technical vocational education and training tions, policy makers must evaluate each one (TVET) has been disappointing. Secondary in terms of its susceptibility to political cap- and post-secondary vocational training costs ture, which plagued industrial policy when it at least three times more than basic second- was last attempted on a large scale in Africa. ary education, yet often provides no better Any subsidy creates a rent. Politically powerful foundation for private sector jobs. Training in interests, if they capture those rents, will resist government-run programs has not been geared efforts to reduce them, even though that step to private sector needs. is necessary for industries to compete in world Governments in Sub-Saharan Africa should markets. focus on support for public goods in TVET such as quality assurance and information. To Skills promote access to training for poor and disad- Is demand for secondary-school graduates vantaged youth, governments should provide simply insufficient in Africa’s modern wage financial support for training in either the pub- sector, or is there a skills mismatch? In fact, lic or the private sector. Information about the both of these problems appear to be present. returns to alternative training options can help Secondary and post-secondary graduates say to align young people’s training choices with the they have trouble finding a job because of lack realities of the labor market. To the extent that of demand. There is a much larger supply of governments support specific training options, labor for unskilled (factory floor) jobs than for those options should emphasize portable skills skilled jobs as mechanics and factory engineers rather than the firm- or job-specific skills that or for office jobs as accountants and managers. employers should already have an incentive to Meanwhile, employers are requesting permits provide. They should ensure that programs are to import experienced skilled labor. Graduates closely linked to the private sector, potentially at all levels without technical training and some through public-private partnerships. Programs work experience (where they can acquire and for disadvantaged youth that integrate training demonstrate the equally valuable “soft” skills) with internships show promise. face an especially crowded job market, reflect- The visibly poor management practices ing an “aspirational” mismatch as much as a in African firms suggest considerable scope skills mismatch. for improving productivity by investing in As with the farm and HE sectors, perhaps business and management skills training and the most important step toward resolving these perhaps even in individualized management problems in the modern wage sector is to get consulting. The evidence for such programs is basic schooling right. Foundational skills are mixed but promising, and governments should overview 21 consider testing and refining them through It is equally urgent to undertake other careful piloting. actions to address constraints that will only yield payoffs in the medium term. For exam- ple, improvements in basic education, the Building an Effective Youth foundation for developing all other skills, will Employment Policy take time to translate into higher productiv- ity and better youth employment outcomes. The challenge of youth employment in Africa Reforms to the business climate will require is not amenable to simple solutions. It reflects sustained effort, and it may take time for the challenges and opportunities of countries investors to respond. But policy makers must themselves in a globalized world. The key rise to these urgent challenges. Failure to act employment issue is that productivity, and now means that future cohorts of young peo- therefore earnings, are low, while aspirations, ple may also lack clear pathways to productive especially those of youth, are high—and per- work. haps higher than those of their parents. Despite These priorities are, of course, only a general progress in many countries, most youth in guide. Addressing the challenge in any given Africa today will not have an easy or structured country will require a country-specific analy- path to a sustainable livelihood, one of the sis. The framework put forward in this report, core aspects of adulthood. All stakeholders— the general diagnostics that it provides, and the governments, private firms, private and non- evidence that it marshals to illustrate successful governmental training providers, and young or promising approaches provide a foundation people themselves—have a role to play in sup- for such an analysis and indicate policy direc- porting this transition. tions to pursue. But a country-based analysis Progress requires a comprehensive approach is required to address the specific local issues to relieve the constraints on human capital and surrounding the following questions: What is the business environment that prevent the constraining earnings growth in agriculture, private sector from seizing opportunities and HEs, and the modern wage sector? Why is it increasing productivity in agriculture, HEs, and hard for youth to start HEs and work produc- the modern wage sector. Governments need to tively in them? Why isn’t private investment in take a holistic view of how to address the situ- labor-intensive firms flowing in to increase the ation—they need to “own” the whole problem. number of modern wage jobs on offer? What A common tendency is to perceive can and should government, NGOs, and other government-provided technical and vocational private sector actors do to ease the constraints skills training as the key. But such action, by that youth face in making the transition to pro- itself, will not address the more fundamental ductive employment? problems. Government intervention should The basis for a successful country-specific focus on public goods—those things that will analysis will be more and better data on support higher productivity in the economic employment; such data are currently sparse activities of households and enterprises. Spe- and often of low quality (see the discussion in cific actions can relieve the most pressing con- Fox and Pimhidzai 2013). The basis for build- straints in the short term, such as increasing ing a much better evidence base of which access to finance for HEs as well as modern approaches are potentially effective and cost- firms; improving access to land and technol- effective will be more experimentation with ogy for young people to expand earnings in promising interventions—and careful evalua- agriculture; building supportive infrastructure tion of their impacts. that enables all enterprises to be more produc- At its core, the youth employment challenge tive; and opening access to regional markets is closely aligned with the challenge of pro- so that firms can broaden the reach of their moting inclusive growth, defined not only as products. Table O.1 summarizes the priority growth in which the poorest segments of soci- actions. ety share but also as growth in which young 22 Youth EmploYmEnt in Sub-Saharan africa Table O.1 Priority actions to take now to address the youth employment challenge “Do now for now”: “Do now for later”: Area for intervention Actions to affect the current cohort of youth Actions to affect future cohorts of youth agriculture 1. Enable rental markets for land 1. Establish effective land registration and transaction systems 2. pilot intergenerational land transfer programs 2. Scale up intergenerational land transfer programs, based on lessons from pilot programs 3. Support high-quality, demand-driven extension services (covering information as well as skills) 3. mainstream youth into smart interventions aimed at increasing productivity (producer organizations, livestock development, 4. link agricultural credit to extension services irrigation, and others) 4. build skills through rapid improvements in education systems in rural areas agriculture and household 5. promote rural village savings and loan associations and self-help enterprises groups 6. Enable financial inclusion for households 7. use safety net programs as a platform to deliver interventions to disadvantaged youth household enterprises 8. Develop a national strategy for household enterprises that reflects 5. build foundational skills through rapid improvements in education the voice of their owners and youth systems 9. Ensure access to workspace and infrastructure for household 6. address infrastructure needs of household enterprises in urban enterprises through improved urban policy development planning 10. leverage nongovernmental organizations to deliver interventions that support disadvantaged youth to enter the sector by addressing multiple constraints (building a range of skills together; building skills along with providing access to finance) modern wage sector 11. reduce the cost of infrastructure services by addressing quality 7. increase the quantity of infrastructure services and efficiency 8. Expand regional markets for products 12. address logistics bottlenecks 9. build foundational skills through rapid improvements in education 13. reduce corruption and the cost of business start-up systems 14. reform technical and vocational education and training and pursue 10. improve access to credit through financial sector reform public-private partnerships for delivering demand-driven training cross-cutting areas 15. increase awareness of opportunities and pathways to self- 11. promote early child development and nutrition to build a stronger employment, especially for young women foundation for skills development 16. consider second-chance education for basic skills 12. build behavioral skills (consider reforms within the school system) 13. reduce fertility rates to lower the size of future youth cohorts (through more girls’ education, improved maternal and child health, increased access to family planning) 14. build better employment data and a stronger evidence base to identify country constraints, priorities, and opportunities people’s vitality is harnessed and rewarded. For References African countries that meet this challenge, the Ayalew Ali, Daniel, Klaus Deininger, and Markus P. benefits will build on each other. The demo- Goldstein. 2011. “Environmental and Gender graphic dividend will yield returns, and Africa’s Impacts of Land Tenure Regularization in Africa: prosperity will grow and be shared. Pilot Evidence from Rwanda.” Policy Research Working Paper 5765, World Bank, Washington, DC, August. Notes 1. Defined as having no work at all in the last seven Bridges, Sarah, Louise Fox, Alessio Gaggero, and days and actively looking for work (ILO 1982). Trudy Owens. 2013. “Labour Market Entry and 2. In Botswana form-two and in South Africa Earnings: Evidence from Tanzanian Retrospec- grade-nine students were tested, correspond- tive Data.” Background paper presented at the ing to nine years of schooling; in Ghana junior CSAE Conference on Economic Development in high school form-two students were tested, Africa, Oxford University, March. corresponding to eight years of schooling. Cloutier, Marie-Hélène, C. Reinstadtler, and Isa- 3. See www.sdindicators.org. bel Beltran. 2011. “Making the Grade: Assessing 4. A similar approach was used in Mongolia. Literacy and Numeracy in African Countries.” overview 23 DIME Brief, Washington, DC, World Bank. Commission. IMF Country Report 06/225, IMF, http://go.worldbank.org/15Y7VXO7B0. Washington, DC, June. Conley, Timothy, and Christopher Udry. 2010. Hungi, Njora, Demus Makuwa, Kenneth Ross, “Learning about a New Technology: Pineapple Mioko Saito, Stéphanie Dolata, Frank van in Ghana.” American Economic Review 100 (1): Cappelle, Laura Paviot, and Jocelyne Vellien. 35–69. 2010. “SACMEQ III Project Results: Pupil Devarajan, Shantayanan, Célestin Monga, and Achievement Levels in Reading and Mathe- Tertius Zongo. 2011. “Making Higher Educa- matics.” Working Document 1, Southern tion Finance Work for Africa.” Journal of African Africa Consortium for Measuring Educational Economies 20 (supplement 3): iii133–54. Quality (SACMEQ), Harare. http://www. sacmeq.org/downloads/sacmeqIII/WD01_ Dinh, Hinh T., Vincent Palmade, Vandana Chandra, SACMEQ_III_Results_Pupil_Achievement. and Frances Cossar. 2012. Light Manufacturing in pdf. Africa: Targeted Policies to Enhance Private Invest- ment and Create Jobs. Washington, DC: World ILO (International Labour Organization). 1982. Bank. “Resolution Concerning Statistics of the Eco- nomically Active Population, Employment, Dupas, Pascaline, and Jonathan Robinson. 2013. Unemployment, and Underemployment.” “Savings Constraints and Microenterprise Devel- Adopted by the Thirteenth International opment: Evidence from a Field Experiment in Conference of Labour Statisticians, Geneva. Kenya.” American Economic Journal: Applied Economics 5 (1): 163–92. Lin, Justin Yifu, and Célestin Monga. 2012. “Solving the Mystery of African Governance.” New Politi- Falco, Paolo, William F. Maloney, Bob Rijkers, and cal Economy 17 (5): 659–66. Mauricio Sarrias. 2013. “Heterogeneity in Sub- jective Well-Being: An Application to Occupa- Morris, Michael, Hans Binswanger-Mkhize, and tional Allocation in Africa.” Documento CEDE Derek Byerlee. 2009. Awakening Africa’s Sleeping 2013–01, Universidad de los Andes, Department Giant: Prospects for Commercial Agriculture in the of Economics, Bogotá. http://dx.doi.org/10.2139/ Guinea Savannah Zone and Beyond. Washington, ssrn.2229328. DC: World Bank. Fox, Louise, and Obert Pimhidzai. 2013. “Different Nin-Pratt, Alejandro, Michael Johnson, and Dreams, Same Bed: Collecting, Using, and Inter- Bingxin Yu. 2012. Improved Performance of preting Employment Statistics in Sub-Saharan Agriculture in Africa South of the Sahara: Taking Africa: The Case of Uganda.” Policy Research Off or Bouncing Back? IFPRI Discussion Paper Working Paper 6436, World Bank, Washington, 01224. Washington, DC: International Food Pol- DC. icy Research Institute. Fox, Louise, and Thomas Pave Sohnesen. 2012. Pritchett, Lant. 2013. The Rebirth of Educa- “Household Enterprises in Sub-Saharan Africa: tion. Washington, DC: Center for Global Why They Matter for Growth, Jobs, and Liveli- Development. hoods.” Policy Research Working Paper 6184, Raballand, Gaël, and Patricia Macchi. 2009. “Trans- World Bank, Washington, DC. port Prices and Costs: The Need to Revisit Fox, Louise, Alun Thomas, Cleary Haines, and Donors’ Policies in Transport in Africa.” Working Jorge Huerta Munoz. 2013. “Africa’s Got Work to Paper 190, Bureau for Research and Economic Do: Employment Prospects in the New Century.” Analysis of Development, Duke University. IMF Working Paper 13–201, International Mon- http://ipl.econ.duke.edu/bread/papers/ etary Fund, Washington, DC. working/190.pdf. Fuglie, Keith O., and Nicholas E. Rada. 2013. Rosenzweig, Mark R. 2012. “Microeconomic “Resources, Policies, and Agricultural Productiv- Approaches to Development: Schooling, Learn- ity in Sub-Saharan Africa.” Economic Research ing, and Growth.” Journal of Economic Perspec- Report 145368, U.S. Department of Agriculture, tives 24 (3): 81–96. Economic Research Service, Washington, DC. United Nations. 2011. World Population Prospects: Gelb, Alan, and Sina Grasmann. 2010. “How The 2010 Revision. New York: Department Should Oil Exporters Spend Their Rents?” Work- of Economic and Social Affairs, Population ing Paper 221, Center for Global Development, Division. http://esa.un.org/wpp/Excel-Data/ Washington, DC. http://dx.doi.org/10.2139/ population.htm. ssrn.1660570. World Bank. 2006. World Development Report 2007: Ghana, Republic of. 2006. “Poverty Reduction Development and the Next Generation. New York: Strategy Paper.” National Development Planning Oxford University Press. 24 Youth EmploYmEnt in Sub-Saharan africa ———. 2008. World Development Report 2009: ———. 2014. Global Financial Development Report Reshaping Economic Geography. New York: 2014: Financial Inclusion. Washington, DC: Oxford University Press. World Bank. ———. 2012. World Development Report 2013: ———. Various years. World Development Indica- Jobs. New York: Oxford University Press. tors. Washington, DC: World Bank. http://wdi ———. 2013. Africa’s Pulse: An Analysis of Issues .worldbank.org. Shaping Africa’s Economic Future 7 (April). Chapter 1 Opportunities and Challenges for Youth Employment in Africa Two trends are converging in Africa, with immediate attention, but it can be met. Indus- potentially profound effects on how Africa’s try—especially export manufacturing—has been economy will grow and where it will create jobs. a vibrant source of wage jobs in other regions, First, Africa’s economies, spurred by high prices most notably East Asia. In Africa, industry is in for primary export commodities, are growing the very early stages of development. It needs time again after a hiatus of many decades. Output is and the right policy environment to grow. Gov- shifting out of agriculture and largely into ser- ernments must also focus on tapping the more vices, and employment is shifting slowly toward immediate potential for productive employ- services. The second trend is that Africa’s rapidly ment in agriculture and household enterprises. growing population will constitute the world’s This report assesses the specific challenges and largest reservoir of working-age individuals for opportunities related to youth employment on generations to come, and the majority of this farms, in nonfarm household enterprises, and population will be young. These trends mean in modern wage jobs. It examines these issues that the structure of employment will continue and possible interventions in light of two types to change in Africa, but the transformation will of binding constraints to higher productivity for be slow. young people in those sectors: constraints related Africa stands to gain economically, socially, to human capital and constraints related to the and substantially from channeling the energy business environment. of its young labor force into more productive employment. The challenge is large and demands Africa’s Working-Age Population: Throughout this report, “Africa” is shorthand for “Sub- Very Young and Growing Rapidly Saharan Africa.” In particular instances, “Sub-Saharan Africa” is retained to clarify comparisons across regions or The median person in Africa is 18 years old— to indicate a specific data set. 7 years younger than the median age in South 25 26 Youth EmploYmEnt in Sub-Saharan africa Asia, which is the next youngest region (figure lion (figure 1.3). Each year between 2015 and 1.1). In other words, fully half of Africa’s popu- 2035, there will be half a million more 15-year- lation is under 18 years old. Africa will remain olds than the year before. This rapid increase the youngest region in the world in the decades contrasts starkly with the Middle East and to come, and the age gap with other regions North Africa, where increases in the size of this will increase. According to projections by the cohort have steadied, and even with East Asia, United Nations, the median age in Africa will where numbers are dominated by China and increase only to 21 in 2035 and to 24 in 2050. In the size of this cohort is expected to fall from other regions of the world, the median person 350 million in 2010 to 225 million by 2050.1 In in 2050 will be more than 35 years old—almost South Asia, the size of the cohort is expected 45 in East Asia and the Pacific. to increase, and then begin to decline after In Africa, the population structure resem- 2030. bles an Egyptian pyramid, with a narrow top and a wide base (figure 1.2). The wide base is Africa’s “youth bulge”—Africa has twice as Can Africa’s Youth Bring an many 15-year-olds as 35-year-olds. In other Economic Advantage? regions of the world, the structure is elongated, reflecting a more even distribution across age Africa’s young and growing working-age popu- groups. South Asia’s demographic profile is lation compels attention for many reasons, not closest to Africa’s, whereas in East Asia and the least because a rapidly expanding working- Pacific, the pyramid is inverted, with a greater age population spearheaded the economic number of older than younger people. Projec- transformation that occurred in East Asia and tions suggest that the shape of the pyramid in the Pacific between 1965 and 1990. Over that Africa will remain as it is in the near future— period, the working-age population in East just with more people at each age. Asia and the Pacific rose by almost 500 million These population trends suggest that the (from 541 million to 1.039 billion). More cru- number of young people entering Africa’s cially, the number of “dependents” (ages 0–14 working-age population will be rising for years and 65 and over) increased by only 143 million to come. The United Nations estimates that in (from 437 million to 580 million). The region 2015 Sub-Saharan Africa will have 193 million had just over one working-age adult for each people between the ages of 15 and 24; by 2035, dependent in 1965, but by 1990 almost two it will have 295 million, and by 2050, 362 mil- working-age adults supported each dependent (figure 1.4). During those years, gross domes- Figure 1.1 Africa’s population is young and will remain so tic product (GDP) per capita in East Asia and the Pacific increased from around US$1,300 to 50 US$3,300. Based on cross-country analysis of 45 the relationship between growth rates and the 40 changing structure of the population, analysts 35 have attributed one-third to half of East Asia’s Median age (years) 30 economic growth to changes in demography 25 (Bloom and Williamson 1998; Bloom, Can- 20 ning, and Malaney 2000). 15 East Asia’s “demographic dividend” is argued to have transformed the economy 10 through two main channels. The first was the 5 increased availability of workers. More work- 0 Sub-Saharan South Middle East Latin America East Asia Europe and ers mean more output, and if there are more Africa Asia and and the and Pacific Central Asia workers relative to the population, then out- North Africa Caribbean put per capita will rise. The second channel 2015 2035 2050 was the continued expansion of the working- Source: Based on United Nations 2011. age population relative to the population as a opportunities and challenges for Youth Employment in africa 27 Figure 1.2 The structure of Sub-Saharan Africa’s population is different than that in other regions a. Sub-Saharan Africa b. South Asia c. East Asia and Pacific 80+ 80+ 80+ 70–74 70–74 70–74 60–64 60–64 60–64 Age group (years) Age group (years) Age group (years) 50–54 50–54 50–54 40–44 40–44 40–44 30–34 30–34 30–34 20–24 20–24 20–24 10–14 10–14 10–14 0–4 0–4 0–4 100 80 60 40 20 0 20 40 60 80 100 100 80 60 40 20 0 20 40 60 80 100 160 120 80 40 0 40 80 120 160 Population (millions) Population (millions) Population (millions) Male 2035 Male 2015 Female 2015 Female 2035 Source: Based on United Nations 2011. Figure 1.3 Unlike in other regions, the number of young people in Sub-Saharan Africa will increase dramatically in the near future 400 350 Population ages 15–24 (millions) 300 250 200 150 100 50 0 1950 2010 2050 1950 2010 2050 1950 2010 2050 1950 2010 2050 1950 2010 2050 1950 2010 2050 East Asia and Pacific Europe and Central Asia Latin America Middle East and South Asia Sub-Saharan Africa and the Caribbean North Africa Source: Based on United Nations 2011. Estimates Projections Note: Each bar shows an estimate or a projection of the number of 15- to 24-year-olds for one year at five-year intervals. whole, a result of the rapid decline in fertility. will not materialize. The rapid change in the With fewer dependents, working-age adults can ratio between the two groups (illustrated in increase their savings—which can be converted figure 1.4) was critical for the boom in produc- into productive investments, which further tivity in East Asia and the Pacific (Bloom and stimulate economic growth. Williamson 1998). For Africa to realize a demographic divi- Historically, declines in child mortality have dend of its own, it is not enough to have a large preceded declines in fertility—a sequence that working-age population; fertility rates need creates a true youth bulge (a bulge that is fol- to fall. Unless the number of dependents per lowed by a decline in the youth population working-age adult decreases rapidly, the poten- once the number of children born falls). In tial benefits of a changing population structure Africa, child mortality has declined dramati- 28 Youth EmploYmEnt in Sub-Saharan africa Figure 1.4 In East Asia, the dependency ratio changed quickly; in Sub-Saharan Two important differences help to explain Africa, it is changing, but slowly why the changing population structure in South Asia and Latin America has not trans- 2.5 lated into a demographic dividend. First, the rate of increase has been much slower in those non-working-age population 2.0 regions than in East Asia (where the ratio Ratio of working-age to increased from 1.1 in 1970 to 1.8 by 1990). Second, as noted in the seminal analysis of the 1.5 demographic dividend, East Asians had “the social, economic, and political institutions and policies that allowed them to realize the growth 1.0 potential created by the transition” (Bloom and Williamson 1998). 0.5 Recent analysis has pointed to the economic 1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 threat posed in Africa by a growing population, low savings rate, and low productivity, which Sub-Saharan Africa Middle East and North Africa could mean a limited demographic dividend Latin America and the Caribbean South Asia (Eastwood and Lipton 2011). An economic East Asia and Pacific environment that is conducive to investment Source: Based on United Nations 2011. and growth—into which the population, with Note: Working-age population is defined as individuals ages 15–64; non-working-age population is defined as individuals ages 0–14 and 65 and older. its large cohorts of young people, will arrive and find productive employment—is vital for Africa’s growing labor force to have a positive cally in the past two decades, although some effect on economic and social development. African countries still have the highest rates in The next section discusses in some detail the the world. In the first decade or so of this cen- types of jobs that exist and are likely to exist in tury, the under-5 mortality rate fell from 160 to Africa, because overall employment is a precon- 70 in Benin, 219 to 129 in Burkina Faso, 166 to dition of youth employment (see box 1.1). As 88 in Ethiopia, 189 to 112 in Malawi, 274 to 127 chapter 2 discusses, young people have unclear in Niger, and 151 to 90 in Uganda.2 and constrained pathways to productive work, Africa’s youth bulge will not taper off, how- and as the subsequent chapters in this report ever, unless fertility rates decline much more show, a range of policies will be necessary to rapidly and systematically. In the 1970s the fer- enhance those pathways. tility rate in Asia and Latin America was iden- tical to the rate in Africa today, but fertility is falling much more slowly in Africa today than What Is a Job, and Where Do Most in those regions at that time (Bongaarts and Africans Find One? Casterline 2013). A more important consider- ation may be that fertility rates have declined In assessing the challenges of youth employ- inconsistently across Africa; in fact, the decline ment in Africa, it is important to take stock of has stalled in several countries (figure 1.5). what it means to have a job and to have employ- “My father has Even with a rapidly changing population ment. To many, having a job is synonymous never worked. structure, the demographic dividend is not with having a wage or salaried position with He has spent automatic. After all, as figure 1.4 shows, East an employer (see quote at left). The majority his whole life Asia is not the only region where the population of work in Africa is not structured in such a here in this structure has changed. In Latin America, the way, however. This study follows the approach plantation which ratio of working-age population to dependents adopted in the World Development Report 2013: he inherited.” rose from 1.1 in 1970 to 1.9 in 2010. In South Jobs, which defines jobs as “activities that gen- Tanzania Asia, it rose from 1.2 in 1975 to 1.8 in 2010. erate actual or imputed income, monetary or in opportunities and challenges for Youth Employment in africa 29 kind, formal or informal” (World Bank 2012e). Figure 1.5 The reduction in the total fertility rate has stalled in several African This includes part- or full-time in-household countries economic activities, such as subsistence farm- 9 ing, regardless of whether anything is ever sold. That report also notes that not all forms of 8 work can be considered jobs. Examples include 7 activities that are performed against the will of the worker and activities that violate basic 6 Total fertility rate human rights. 5 In many countries, including in Africa, people report that jobs have a broader impor- 4 tance beyond the income they provide (see 3 focus note 1). Jobs can convey identity, status, and self-confidence; they can contribute to an 2 individual’s overall life satisfaction. Some jobs 1 contribute to these dimensions of well-being more than others. The type of job, working 0 1950–55 1960–65 1970–75 1980–85 1990–95 2000–05 conditions, contract, benefits, and safety and security at work all matter. Jobs also influence Nigeria South Africa Botswana Sub-Saharan Africa social cohesion by shaping individuals’ identi- Tanzania Niger Kenya ties and relations to one another and bringing Source: United Nations 2011. them together in networks. The distribution of jobs within society and perceptions about who has access to opportunities, and why, can shape Box 1.1 people’s expectations and aspirations for the future, their sense of having a stake in society, and their perceptions of fairness. All of these Youth versus overall employment intrinsic aspects of jobs are particularly impor- A key part of improving employment opportunities for Africa’s youth tant for youth. is to understand and address Africa’s overall economic challenges. For In Africa, the vast majority of work takes that reason, much of the discussion in this report focuses on identify- place in agriculture. Agriculture occupies more ing policies that can increase the productivity of all employment—in than 70 percent of the labor force in Africa’s agriculture, nonfarm household enterprises, and the modern wage low-income countries and more than 50 per- sector. The report also identifies promising policies that focus specifi- cent in its lower-middle-income countries. cally on helping young people to make a more successful transition to higher-productivity work in each of those three sectors. African farmers are predominantly smallhold- The focus on youth raises the question of whether it might be ers who consume a large share of what they socially beneficial to support employment policies that favor youth over produce. Data from recently collected house- other members of society. Special efforts to help young people to enter hold surveys indicate that the share consumed agriculture or start a household enterprise do not generally reduce such is around 50 percent, compared with 20–30 opportunities for adults, but targeted programs in the modern wage percent outside of Africa (Losch, Freguin- sector could have a detrimental effect on adult employment in the sec- Gresh, and White 2013). tor. On the one hand, assistance for young people might have long- Others find employment in household term benefits if it sets them on a productive path. On the other hand, enterprises (HEs), which are unincorporated, older workers may have families and other economic dependents who nonfarm businesses owned by households. rely on their income. There is no clear argument that social welfare will be improved by displacing these older workers to the benefit of They include self-employed people who run younger workers. Increasing the opportunities for all workers, while a business that may employ family members helping youth to overcome their particular constraints, is the strategy without pay but also self-employed people that this report advocates. who run a business that employs a small num- ber of nonfamily workers on a casual basis. 30 Youth EmploYmEnt in Sub-Saharan africa “We work mostly The vast majority (70 percent) of nonfarm (Chuhan-Pole and Angwafo 2011). The flow of as subcontractors HEs today are pure self-employment—just the private capital to Africa now exceeds the flow owner operating the HE. About 20 percent of of foreign aid. As a result, the structure of out- for the textile these enterprises include a family member in put has changed: the share of GDP generated factories. . . . the operation, and only 10 percent have hired by agriculture is falling, and the share gener- The entire someone outside of the family. ated by industry and services is rising. By 2010, village lives on The modern wage sector includes small, agriculture’s share in GDP had fallen to 30 per- the income from medium, and large firms that employ five or cent in low-income countries and 16 percent embroidery, more workers on a continuous basis. It also in lower-middle-income countries, while the includes the public sector, which in some coun- share of the industrial and services sectors had smocking, and tries is a large share of the modern wage sector. increased. agriculture, In the low- and lower-middle-income coun- The drivers of this growth were economic which we do tries of Africa, roughly half of wage employ- policy reforms, which were necessitated by mis- when we don’t ment is in the public sector. This report focuses guided steps taken in the past, and higher com- have work.” only on the private sector, where the potential modity prices, which produced better terms of Madagascar for job growth is greatest. trade (Devarajan and Fengler 2013). These two factors allowed domestic demand, especially for private sector services and construction, to Growth, Jobs, and Africa’s Labor power growth. Africa’s impressive growth tra- Force—Now and in the Future jectory has largely followed commodity prices, however, and African exports are still concen- Since 2000, Africa has seen more than a decade trated in primary commodities. In contrast, of economic growth, the longest continuous it was the rapid rise in export manufacturing expansion in more than 50 years. Until the that allowed East Asia to capture its demo- 2008–09 global economic crisis, Africa’s GDP graphic dividend. Over the past two decades, grew relatively rapidly, averaging 5 percent a manufacturing’s share of GDP actually fell in year, and growth had resumed by 2010 (figure Africa, while rising in Asia’s lower-income and 1.6). Between 1998 and 2008, mineral-exporting middle-income countries (figure 1.7). Today, countries experienced an exceptionally steep the share of manufactured goods in merchan- rise in GDP; 22 countries that are not oil pro- dise exports is 30 percent in Africa, compared ducers averaged 4 percent growth or higher with 50 percent in Latin America (another resource-rich region) and 60 percent in lower- middle-income countries on average. Manu- Figure 1.6 Africa’s growth miracle, 2005–12 factured exports have led growth in only one country in Africa: Mauritius. Indeed, Africa is 8.0 so far behind East Asia that it will take some 7.0 time to catch up. 6.0 5.0 The Structure of Employment in Real GDP growth (%) 4.0 African Countries 3.0 Africa’s dependence on commodity exports, 2.0 aid, and domestic demand as sources of 1.0 growth did not lead to a major transformation 0.0 in employment (figure 1.9; see box 1.2 for an –1.0 explanation of how employment estimates are –2.0 derived and countries are classified). Although 2005 2006 2007 2008 2009 2010 2011 2012 agriculture’s share in GDP fell substantially, Developing (excluding China) Sub-Saharan Africa (excluding South Africa) almost 60 percent of Africa’s labor force in 2010 Sub-Saharan Africa still reported that agriculture was their main Source: World Bank 2012a. economic activity. opportunities and challenges for Youth Employment in africa 31 Figure 1.7 Over the past two decades, agriculture’s share in GDP contracted in Africa, but manufacturing did not replace it a. Low-income countries b. Lower-middle-income countries 100 100 90 90 80 80 70 70 60 60 Percent Percent 50 50 40 40 30 30 20 20 10 10 0 0 1990 2010 1990 2010 1990 2010 1990 2010 1990 2010 1990 2010 Sub-Saharan Africa South Asia East Asia and Pacific Sub-Saharan Africa South Asia East Asia and Pacific Agriculture Industry (excluding manufacturing) Manufacturing Services Source: World Bank various years. Figure 1.8 Exports are a smaller share of GDP in Africa than in East Asia and a larger share than in South Asia, but African countries, even richer ones, export commodities, not manufactured goods Sub-Saharan Africa East Asia and Pacific South Asia Exports of goods and services (% of GDP) Exports of goods and services (% of GDP) Exports of goods and services (% of GDP) 100 100 100 90 90 90 80 80 80 70 70 70 60 60 60 50 50 50 40 40 40 30 30 30 20 20 20 10 10 10 0 0 0 1990 1995 2001 2005 2011 1990 1995 2001 2005 2011 1990 1995 2001 2005 2011 Sub-Saharan Africa East Asia and Pacific South Asia 100 100 100 90 90 90 (% of merchandise exports) (% of merchandise exports) (% of merchandise exports) 80 80 80 Manufactures exports Manufactures exports Manufactures exports 70 70 70 60 60 60 50 50 50 40 40 40 30 30 30 20 20 20 10 10 10 0 0 0 1990 1995 2001 2005 2011 1990 1995 2001 2005 2011 1990 1995 2001 2005 2011 Low-income countries Lower-middle-income countries Upper-middle-income countries Source: World Bank various years. 32 Youth EmploYmEnt in Sub-Saharan africa Figure 1.9 The majority of Sub-Saharan Africa’s workers in low- and lower-middle- in agriculture combined with the persistently income countries work in agriculture and in nonfarm household enterprises low growth in agricultural productivity help to Estimated structure of employment in Sub-Saharan Africa by country type, 2010 explain the continued low incomes and poverty 100 in Africa. The data confirm that unemployment 90 remains low, except in upper-middle-income countries and especially in South Africa, where 80 unemployment is stubbornly high (see focus 70 note 4). 60 Africa’s decade of growth was not “jobless,” % of total 50 but patterns of employment growth differed 40 across countries. Growth moved away from agriculture in some countries: for example, 30 Ghana, Nigeria, Rwanda, Tanzania, and Uganda 20 all experienced high growth in both wage 10 and HE employment (figure 1.10). Mineral- 0 exporting countries such as Cameroon and Low-income Lower-middle- Resource-rich Upper-middle- Total Mozambique experienced job growth but saw countries income countries income countries countries countries little or no transformation of the employment 183 40 150 21 395a structure, although the creation of nonfarm Labor force (millions) employment in Nigeria over the last decade shows that diversification is possible in an Afri- Agriculture Household enterprises Wage services Wage industry Unemployed can oil-exporting country. Source: Fox et al. 2013. Note: On horizontal axis, numbers show size of labor force in each group. Country classification is In rapidly growing countries where growth based on per capita gross national income (GNI) in World Bank (various years). See box 1.2 for an did not depend significantly on mineral explanation of how employment estimates are derived and countries are classified. a. Numbers do not add to total because of rounding. resources, employment followed output in shift- ing out of agriculture—the sector with the low- est productivity—and into higher-productivity This result is not wholly unexpected: the industry and services. Those countries saw transformation in labor always lags the trans- nonfarm wage and salary employment in the formation in output (more capital per worker private sector advance at a rapid pace that is needed to employ people in more produc- often surpassed growth in GDP. Most of the tive jobs). Yet the large share of the labor force increase in wage employment (public and pri- Box 1.2 How are our employment estimates done? Data on the structure of employment in Africa are dif- of the labor force for each country was calculated by taking ficult to obtain. Many countries do not collect these data the United Nations projection for the working-age popula- frequently; often the data are not released until long after tion in each country and applying a labor force participation they have been collected; and in some cases, only pub- rate. Countries were grouped according to income, with lished tabulations are released. Regardless of how they low income set at gross national income (GNI) per capita are released, the quality is often poor (Fox and Pimhidzai under US$1,000, lower-middle income at GNI per capita of 2013). For these estimates, all possible data sources were US$1,000–US$4,000, and upper-middle income at GNI per consulted to get at least one observation on the structure capita above US$3,000. Resource-rich countries were esti- of employment for each country over the period 2000–10 mated as a separate group. Figure 1.9 shows the baseline to form the baseline. We were able to do this for 28 of 47 distribution estimated using this method by country group countries in Africa, covering 73 percent of the labor force. for 2010. For the other countries, we imputed a distribution of the labor force from the averages of similar countries. The size Source: Fox et al. 2013. opportunities and challenges for Youth Employment in africa 33 vate) occurred in the expanding services sector; Figure 1.10 Growth has moved the structure of employment away from agriculture growth in labor-intensive industry was slow, in some countries, but not others owing to the limited development of export 20 manufacturing. 15 Percentage point change in By 2010, the private sector was creating 10 employment share most of the wage jobs found in Africa (figure 5 1.11). Much of the increase in wage employ- 0 ment was in noncontract employment such as –5 day labor or temporary jobs. These jobs, often –10 called “informal wage employment,” represent –15 about half of all nonfarm wage jobs in Africa. –20 Despite the recent creation of private wage –25 jobs, the share of wage employment remains Nigeria Rwanda Ghana Uganda Tanzania Senegal Côte d’Ivoire low throughout Africa, partly because the pub- Agriculture Private wage Household and medium-size enterprises lic sector shed so many jobs in the 1980s and Source: Based on standardized and harmonized household and labor force surveys. See appendix A. 1990s (Fox and Gaal 2008). Although most nonfarm wage employment was found in the public sector in the 1960s and 1970s, this pat- Figure 1.11 Most wage jobs are in the private sector (but not in resource-rich tern was reversed in non–resource-rich coun- countries) tries during their recent growth spurt. In con- 100 trast, economies that grew through increased 90 mineral extraction maintained the pattern: 80 the state distributed part of the resource rents Public wage as a share of all wage employment (%) in the form of public sector jobs. The pub- 70 lic sector continues to provide most of the 60 wage employment in resource-rich countries, 50 because incentives for the private sector to cre- 40 ate employment remain weak (box 1.3). 30 Private nonfarm employment grew the 20 most rapidly in the HE sector, as rural and 10 urban households used their extra income to 0 start businesses (figure 1.10). The data pre- Low-income Lower-middle- Resource-rich Upper-middle- Weighted sented here are likely to understate growth in countries income countries countries income countries average (including South Africa) the HE sector, because they show only pri- Source: Fox et al. 2013. mary employment, when in fact a sizable share of the labor force—up to 50 percent in some countries—undertakes two or more economic agriculture, which is about the average for activities (jobs) over a 12-month period. Par- Africa’s lower-middle-income countries (table ticularly in rural areas, where the majority of 1.1). The difference is that these East Asian Africa’s population still lives, a household is countries have higher agricultural productiv- likely to report its HE as a secondary activity, ity, which has helped to reduce rural poverty after farming (Fox and Sohnesen 2012). to levels well below those in Africa. Africa and these East Asian countries may have the same The Structure of Employment in Africa share of the labor force in agriculture, but the Compared with Other Regions persistently low growth in agricultural produc- The share of the labor force in agriculture is tivity in Africa prevents the African labor force not unusually high in Africa, especially in its from reducing poverty to the same extent as lower-income countries. In Vietnam and the their Asian counterparts (IMF 2012). The HE Lao People’s Democratic Republic, where per sector is also large in all of the lower-income capita income just exceeds US$1,000, 50 per- comparator countries, especially in Bangla- cent or more of the labor force still works in desh, the poorest of the comparators. 34 Youth EmploYmEnt in Sub-Saharan africa Box 1.3 The challenge of diversifying output and employment in resource-rich countries: Examples from Central Africa Debate remains lively on the existence and nature of the so- structure, and the promotion of new sources of economic called “resource curse,” but few doubt that youth employ- growth. A central concern is to ensure coherence across ment faces special challenges in countries with large endow- policies that promote human capital development, includ- ments of mineral resources. Resource rents, if not carefully ing specific programs for youth employment, and those that managed, lead to overvalued exchange rates and contribute promote public investment in infrastructure and productive to noncompetitive wages. Both of these outcomes cripple sectors. In Central Africa, for example, public expenditure the development of sectors that trade in goods other than reviews in Equatorial Guinea and Gabon found that pub- mineral resources, particularly if a country’s human capital or lic spending has overemphasized investments in physical productive infrastructure has not improved as a result of its infrastructure relative to human capital (World Bank 2010, resource wealth. The traditional tradable sectors of the econ- 2012c). Despite high unemployment among well-educated omy, including export-oriented agriculture, can shrink rap- youths, firms in newly emerging economic sectors have dif- idly; at the same time, the industries based on extracting nat- ficulty finding qualified personnel, indicating that the skills ural resources create few new wage-earning opportunities. taught by the education system are not adequately aligned Most workers are left with few choices. They either revert with efforts to promote new economic activities (World to low-productivity subsistence farming or migrate to urban Bank 2012d, 2013). Arbitrary limits on the employment of areas, where they often remain underemployed. The few foreigners can also be a significant business constraint, and opportunities for highly paid employment—in the natural even when expatriates are employed, all too little is done to resource sector and the public sector (where growth is fed transfer their expertise to locals. Often foreign contractors by resource rents)—contribute to a pattern of youth unem- implement large-scale public investment projects that use ployment in which young people simply opt to wait for a very little domestic labor, even for low-skilled activities that job in those sectors and develop a motivational bias against could employ local youths in the short run. A review of youth other forms of employment. The expectation of a public sec- employment programs in Cameroon found that they address tor “desk job” distorts young people’s educational choices a wide range of specific problems but lack coherence and and creates a large group of graduates whose skills are ill are not clearly integrated with broader policies to remove matched to the needs of the private sector. binding constraints to growth and employment (World Bank Often natural resource wealth goes hand in hand with 2012b). significant deficits in the institutional environment that Successful resource-rich countries outside Africa, such hamper growth in private sector employment. When natu- as Chile and Malaysia, have put skills and education at the ral resource revenue is available, incentives for business cli- center of their diversification strategies (Gelb 2010). Rather mate reforms are reduced, and the public sector assumes an than “picking winners” and subsidizing specific sectors, such excessive role in economic activity. countries have used natural resource revenues strategically The positive side of this story is that countries with exten- to identify and address constraints on skills in close coop- sive natural resources have substantial financial resources eration with the private sector. This approach has effectively at their disposal to invest in human development, infra- promoted economic diversification and youth employment. The low share of the labor force in private manufacturing jobs. Clearly the importance of industry is what makes the employment struc- mineral rents in raising per capita incomes in ture so different in low-income and lower- Africa’s lower-middle-income countries con- middle-income countries of Africa compared tributes to this discrepancy. Resource extrac- to the rapidly growing countries of Asia or tion does not generate many jobs, and high Latin America (table 1.1). All the comparator resource rents can create an economic structure countries except Lao PDR and Mongolia have unfriendly to private, labor-intensive industry a larger share of employment in industrial (Gelb 2010). The economies of Lao PDR and wage jobs, because they have a high number of Mongolia, both mineral exporters in East Asia, opportunities and challenges for Youth Employment in africa 35 Table 1.1 African countries have less wage employment than high-growth comparator countries % of employment Wage job Income level and region or country All Industry Services HE Agriculture Total Low income Sub-Saharan africa 12.3 2.3 10.0 18.3 69.4 100.0 bangladesh 25.7 10.8 14.9 27.7 46.6 100.0 cambodia 23.3 11.1 12.2 21.0 55.7 100.0 Low-middle income Sub-Saharan africa 13.9 2.0 11.9 31.1 55.1 100.0 bolivia 43.0 12.6 30.4 28.1 28.9 100.0 lao pDr 13.5 5.4 8.1 19.0 67.5 100.0 mongolia 39.3 5.9 33.4 16.0 44.7 100.0 nicaragua 43.9 13.3 30.6 22.9 33.2 100.0 philippines 48.7 12.6 36.1 19.5 31.8 100.0 Vietnam 31.8 14.3 17.5 19.1 49.1 100.0 Source: Fox et al. 2013. have structures similar to those in resource- stantial productivity growth seen in rapidly rich African countries (figure 1.12). Even in this growing economies outside Africa. To create case, African countries with high levels of min- more productive employment, Africa still faces eral exports stand out, because they have even the dual challenge of increasing productivity in less wage employment in industry than Lao agriculture and diversifying employment out of PDR or Mongolia. Resources are not destiny, agriculture. however, as Bolivia’s successful performance in export-oriented manufacturing shows. The lack of jobs in export-oriented manu- Figure 1.12 Industrial wage employment has not risen with GDP in Africa the way it has in manufacturing exporters at similar income levels facturing is not the only factor setting Africa apart. As discussed, the labor force is grow- 20 Wage industry employment as a share of employment (%) ing much faster in Africa than in Asia or 18 SWZ Latin America, making it that much harder to transform the structure of employment. For 16 VNM example, because Vietnam’s labor force grew NIC 14 at only two-thirds of the pace of Senegal’s over BOL the last decade (2.1 versus 3.1 percent a year), 12 KHM PHI BGD every dollar invested in creating labor-intensive 10 STP manufacturing jobs will have a stronger effect on the structure of employment (measured 8 as a share of the labor force) in Vietnam than 6 LAO KEN LSO MNG in Senegal. In other words, Senegal needs 50 MWI 4 RWA COM GHA LBR percent more investment in manufacturing SEN CMR than Vietnam needed, just to bring its share of 2 NGA COG AGO COD TCD CIV employment in industry to the level in Vietnam NER SLE BFA 0 in 2008. 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 In sum, after a decade of growth and job GDP per capita, 2010 (current US$) creation, the majority of Africa’s labor force Sub-Saharan Africa, low- and low-middle-income countries Comparators still works in its least productive sector—agri- culture—which has yet to experience the sub- Source: World Bank various years for GDP; Fox et al. (2013) for employment structure in 2010. 36 Youth EmploYmEnt in Sub-Saharan africa “We both started What Can Africa Expect in the Future? Basic Scenario: Strong, Steady Growth this profession Africa’s strong growth early in this decade Continues [silk production has fueled expectations that by the end of the In this scenario, the main sources of growth decade most of its population will live in coun- in Africa over the next decade are the same and weaving] tries with a per capita income (GNI) above as those in the last decade, resulting in steady when we were US$1,000—what the World Bank defines as growth across country groups (table 1.2). Agri- eight years old. “middle income” (World Bank 2012a). Cur- cultural growth in lower-middle-income and There was no rently almost 60 percent of the labor force low-income countries is projected to remain other option for works in low-income countries, including strong at about 5 percent a year, driven by us; this is our those classified as resource rich, so this shift increased factors of production (land and heritage, and we would be significant. labor) and by increased labor productiv- How might this continued strong growth ity (from increased input use and improved proudly continue influence the number and types of jobs avail- access to markets). This growth will cause real this tradition.” able to Africans by 2020? And what would incomes to rise for farmers, as demand for food Madagascar happen to employment if growth were even crops from Africa is projected to remain high stronger? We developed two sets of employ- regionally and internationally, keeping prices at ment projections to explore those questions. current real levels. Projected industrial growth First, we project what might happen if the in lower-middle-income and low-income growth remains robust, following its present countries reflects a combination of new mining course. Next, we project what might happen projects and higher manufacturing output to to employment if Africa experiences a “game- serve the domestic market, but no major gains changing” growth scenario, arising from a in manufacturing exports. Some countries with surge in export-oriented manufacturing. (For mining projects are projected to have very high background on the data and methods used industrial growth: Ghana at 19 percent a year, to develop the employment projections, see as oil production comes on stream; Liberia at box 1.4.) The sections that follow outline the 17 percent a year, from iron ore mining; and assumptions behind these scenarios and the Sierra Leone, with an average industrial growth resulting picture of employment. rate exceeding 50 percent, from iron ore min- Box 1.4 How are our employment projections done? For projecting the distribution of employment across sectors, the early 1990s through the late 2000s, and (3) computed the first step was to develop economic growth projections estimates for African countries for which at least two high- by sector for each country. These were made based on pro- quality employment and national accounts data points could jections of area and yield for agricultural products and on be obtained. Using that data, we developed a set of median projections of underlying sectoral benchmarks, such as elec- sector elasticities for each country grouping, sector, and tricity usage, cement usage, road and rail transport, telecom- type of nonagricultural job (see table B1.4.1). The industry munications, and hotel stays, for the other sectors. and services elasticities for the middle-income countries are Then sectoral elasticities of employment with respect to comparable to the other estimates. For low-income coun- growth were developed and applied. In developing the esti- tries, the industry elasticities are considerably lower in Africa mates for employment growth between 2010 and 2020, we than in Asia (Bangladesh, Cambodia, and Vietnam), because consulted (1) computed sectoral elasticity estimates for some industry growth has been much more labor intensive in the African countries for the previous decade, (2) computed Asian countries. In contrast, the services employment elas- estimates for selected Asian comparator countries from ticities are estimated to be slightly higher in Africa than in opportunities and challenges for Youth Employment in africa 37 Box 1.4 (continued) Asia. The employment elasticity is much lower in industry income countries, because the explosive growth anticipated for resource-rich countries than for low-income countries in this sector will come from increased extraction of iron ore, because of the importance of mining in the former group which is not labor intensive. To close the model, the unem- and the prevalence of labor-intensive production in the lat- ployment rate for low-income and lower-middle-income ter group. The agricultural sector elasticities are negative countries was held constant, and all projected employment for South Africa and the middle-income countries because not allocated to industry and services was assigned to agri- they have been losing agricultural employment over the past culture. This feature of the model is consistent with agricul- decade, in contrast to the earlier period (Kapsos 2005) and ture’s current function as the fallback economic activity for to middle-income countries in Asia. most households, but it means that employment estimates Finally, the elasticities of employment were applied to the for agriculture in low-income and lower-middle-income projected growth rates to project employment by sector for countries are not based on growth elasticities. In middle- each country to 2020. Since elasticity estimates vary consid- income countries, labor was allocated to each sector, with erably across countries, the median estimate was adjusted in unemployment as the residual. The resulting baseline sce- some countries based on the economic structure and pro- nario shows only a modest decline in unemployment in jected future performance of the country. For example, the these countries. The detailed country employment estimates employment elasticity of industrial sector growth in Sierra were aggregated back into the country groupings. The final Leone was adjusted downward from the median for low- result is shown in figure 1.14. Table B1.4.1 Growth elasticities of employment a. Our elasticity parameters Upper-middle Lower-middle income (except Sector Low income income Resource rich South Africa) South Africa a agriculture −0.8 −1.0 Wage industry 0.9 0.8 0.6 0.6 0.5 nonwage industry 0.7 0.6 0.7 0.3 0.3 Wage services 0.8 0.8 0.8 0.7 0.5 nonwage services 0.8 0.9 0.7 0.6 0.5 b. Comparators Asia 1990–2010 ILO, SSA, 1990–2003b Vietnam, Cambodia, Low- and lower- Upper-middle Sector Bangladesh Indonesia, Philippines middle income income agriculture 0.3 0.3 0.7 0.1 Wage industry 1.2 0.4 0.6 0.8 nonwage industry 1.1 0.4 0.6 0.8 Wage services 0.7 0.7 0.8 0.7 nonwage services 0.7 0.7 0.8 0.7 a. Agricultural employment closes the model for low-income, low-middle income, and resource-rich countries. b. Data from Kapsos 2005. Source: Fox et al. 2013. Note: Resource-rich countries are defined here as Angola, Chad, the Democratic Republic of Congo, Guinea, Nigeria, the Republic of Congo, Sudan and Republic of South Sudan, and Zambia. 38 Youth EmploYmEnt in Sub-Saharan africa Table 1.2 Average annual growth, by sector and country income level, 2005–20 percent Actual Projected 2005–10 2010–15 2015–20 Country Real Agri- Real Agri- Real Agri- group GDP culture Industry Services GDP culture Industry Services GDP culture Industry Services low income 6.5 4.8 6.6 7.7 6.2 5.0 7.9 6.2 6.4 5.8 7.3 6.2 lower-middle 4.4 4.1 4.9 4.5 5.5 4.1 9.9 4.0 5.9 4.7 6.9 5.8 income lower-middle income, except 3.3 3.9 3.4 3.2 4.9 4.3 6.6 4.4 5.8 4.9 7.1 5.4 Ghana resource rich 6.8 7.5 3.1 11.9 6.8 6.5 4.6 9.5 6.4 6.4 3.7 8.5 South africa 3.6 2.5 2.3 4.1 3.1 2.1 2.2 3.5 3.2 3.2 1.3 3.8 other upper- 3.0 1.5 1.5 4.8 4.6 3.2 4.7 5.1 4.5 4.7 4.1 4.9 middle income Source: Fox et al. 2013. Note: See box 1.4 for an explanation of how employment projections are derived and countries are classified. ing. When those countries are excluded from the mining sector will not directly create very their respective groups, a smoother pattern many jobs. By 2020, wage and salary jobs will emerges for growth rates. Resource-rich coun- account for 29 percent of the net new jobs but tries, whose main export is expected to remain only 25 percent of the total jobs taken by new unprocessed minerals, will be distinguished by entrants (some new entrants will replace work- higher growth in the services sector (public sec- ers leaving the labor force). In other words, at tor growth funded by mineral rents). Upper- best one in four African youth will find a wage middle-income countries are projected to job, and only a small fraction of such jobs will continue recovering from the slump that fol- be “formal” jobs in modern enterprises. Unem- lowed the financial crisis of 2007–08. They are ployment is assumed to remain low in the low- expected to diversify their exports into services, income and lower-middle-income countries, giving a boost to that sector. and we project that it will fall slightly in the A key assumption behind this optimism is upper-middle-income countries if high growth that Africa will not experience another major rates are realized. economic shock from external or internal The HE sector is projected to create even sources, such as another major global reces- more jobs than the wage sector, accounting for sion that shatters demand for African exports, 45 percent of the net new jobs and employing the outbreak of a regional conflict, or a pro- 37 percent of new entrants through the start-up longed climatic disaster in the region. Without of new businesses. This sector feeds off demand such shocks, African countries could realize for goods and services created by employment 10 years of growth averaging 4.5–6.0 percent a and earnings growth in the wage and agricul- year, slightly above what was achieved during tural sectors, so balanced growth is necessary to 2005–10, which included the shock to world realize this part of the projection. financial markets. Because the majority of new jobs will need Figures 1.13 and 1.14 present initial answers to be created in countries currently classified as to the questions about the number and types low income (such as the Democratic Republic of jobs that this kind of growth might create. of Congo and Ethiopia), the agricultural sec- Although industrial wage employment is pro- tor will remain important for creating employ- jected to increase through continued modest ment. In agriculture—unlike other sectors— diversification of output and exports, service the projection of new jobs (in low-income and employment is projected to keep growing lower-middle-income countries) is not based on faster than industrial employment because demand for labor in the sector. Instead it repre- opportunities and challenges for Youth Employment in africa 39 Figure 1.13 Where will Africa’s 125 million new jobs be created? a. Jobs by sector b. Projected distribution of new entrants by sector, 2010–20 300 4% Wage industry 21% 250 Wage services 37% Agriculture Jobs (millions) 200 150 100 50 38% 0 Household Agriculture Household Wage Wage enterprise enterprise services industry Jobs in 2010 New jobs in 2020 Source: Fox et al. 2013. Note: See box 1.4 for an explanation of how employment projections are derived and countries are classified. sents the labor force that does not find a wage Figure 1.14 Informal will remain normal in much of Sub-Saharan Africa job or start a business (box 1.4). The absolute Projected structure of employment in Africa, by country type, 2020 number of people working in agriculture is pro- 100 jected to rise by about 33 million. Since about 70 90 percent of those who will exit the labor force are 80 now working in agriculture, the actual number 70 of new entrants that agriculture needs to absorb is over 62 million, or about 38 percent of the 60 % of total new entrants. Stronger growth in other sectors 50 could push this number down slightly, but it is 40 unlikely that the labor force in agriculture will 30 shrink over the next decade—young people 20 seeking jobs will simply have no other place to go. If African agriculture realizes its potential, 10 however, agricultural jobs will be more produc- 0 Low-income Lower-middle- Resource-rich Upper-middle- All tive, higher-earning jobs than they are today. countries income countries income countries Although the projection implies high countries countries growth in nonagricultural employment, the 248 52 200 23 523 projected structure of employment shown Labor force (millions) in figure 1.14 is not much different from the Agriculture Household enterprise Wage industry Wage services Unemployed current structure shown in figure 1.9. Agricul- Source: Fox et al. 2013. tural employment will decline in all country Note: See box 1.4 for an explanation of how employment projections are derived and countries are groups, yet the share of industrial wage jobs in classified. total employment will rise from 2.3 to only 3.2, below other developing regions, because those jobs are growing from such a small base. provided here—30 million new wage jobs over How do these estimates compare with others the coming decade? The main difference is that in the literature? A recent McKinsey study con- we use Africa-specific data and projections to tends that Africa will create about 122 million forecast the employment profile, whereas the new jobs over the next 10 years and that almost McKinsey team uses data from fast-growing half of them will be wage-paying jobs (Fine et developing and emerging market countries on al. 2012). Why is McKinsey’s employment fore- other continents.3 That method clearly imparts cast considerably more optimistic than the one an upward bias to McKinsey’s results. 40 Youth EmploYmEnt in Sub-Saharan africa “Game-Changer” Scenario: A Surge in To test the possible implications of such a Labor-Intensive Export Manufacturing “game change” for employment in Africa, we Is there a better scenario for African jobs? The simulated this recent Asian experience for low- scenario just presented is already optimistic in income and lower-middle-income countries of the sense that it is based on sustained growth Africa.4 In this simulation, we raised the wage across Africa, but it is not based on a radical, employment elasticity to 1.2 to match the his- “game-changing” departure from Africa’s cur- torical wage employment elasticity estimated rent growth path. Some observers contend that for Bangladesh, Cambodia, and Vietnam— if Africa increases output in export-oriented, meaning that employment in the industrial labor-intensive light manufacturing as dramat- sector would grow 20 percent faster than value ically as high-growth comparator countries in added, which implies very labor-intensive Asia managed to do, the structure of employ- growth. The industrial growth projection for ment could change more rapidly (Lin and low-income and lower-middle-income coun- Monga 2012; Dinh et al. 2012). When growth tries was also revised upward to 10 percent a in manufacturing employment was at its peak year over 2015–20. This figure is slightly above in Bangladesh, Cambodia, and Vietnam, those the median and average industry growth rate countries had annual growth rates of 10 per- experienced by Bangladesh, Cambodia, and cent or more in their industrial sectors and Vietnam over the most recent decade (9.3 per- were creating industrial wage jobs at an even cent a year). Figure 1.15 compares the structure faster pace. of employment for the original and the alterna- How would the structure of employment tive, “game-changing” scenario. differ if low-income and lower-middle-income If the alternative scenario could be realized, African countries changed their policies and industrial wage employment would grow much investments to achieve a comparable perfor- faster across Africa’s low- and lower-middle- mance? What employment prospects would be income countries, which would account for open to youth in those countries in 2015–20 if almost 60 percent of Africa’s labor force in Africa picks up manufacturing industries and 2020. The average annual growth of indus- jobs from East Asia beginning in 2015, in much trial wage employment would double over the same way that other East Asian countries the decade to 12 percent a year, and total wage picked up industries and jobs from Japan and employment would grow 6 percent a year. Yet the Republic of Korea in the 1980s and 1990s? even then the structure of employment would Figure 1.15 Even game-changing growth will have limited effects on the distribution of employment in the near term Original Alternative Original Alternative scenario scenario scenario scenario 16 100 14 90 Wage industry jobs, 2020 80 12 70 10 % of total (millions) 60 8 50 6 40 30 4 20 2 10 0 0 Low-income Lower-middle- 248 248 52 52 countries income countries Low-income Lower-middle- countries income countries Original Additional jobs under scenario alternative scenario Labor force (millions) Agriculture Household enterprise Wage services Wage industry Unemployed Source: Fox et al. forthcoming. Note: See box 1.4 for an explanation of how employment projections are derived and countries are classified. opportunities and challenges for Youth Employment in africa 41 look about the same. Low-income countries Table 1.3 Framework of this report could expect about 5 million more wage jobs Three sectors Two dimensions a year, and lower-middle-income countries • Agriculture • Human capital could expect about 2 million new wage jobs— • Nonfarm household • Business environment a shift of 10 percent of total new jobs in these enterprises countries, representing a small change in the • Modern wage employment prospects for new entrants. These modest gains partly reflect the short period used for the projection, which covers productivity (access to land, capital, and only five years, whereas the structural change finance; infrastructure; technology; and in Vietnam took 20 years to unfold. They also markets), as well as the government policies, reflect the larger labor force and the lower base regulations, and programs that may affect from which industrial development would the choice of economic activity and how the start. Africa will need at least two decades to activity is conducted. change the structure of employment suffi- ciently to offer dramatically different prospects Chapters 4, 5, and 6 address each sector to its youth, which underscores the importance in turn (agriculture, HEs, and modern wage of starting the change process now. employment in the private sector). They delve into the particular ways that binding constraints related to the business environment and human Framework of This Report capital influence young people’s potential for productive employment, and they describe how Starting from the reality of youth in Africa— those constraints might be relieved. the size of the cohort and the events that shape To provide the context for those chapters, their lives during the many transitions that chapter 2 discusses the transitions that char- youth entails—this report analyzes young peo- acterize youth, particularly the overlapping ple’s employment prospects and experiences transitions from school and to work. Chapter and examines how to create pathways leading 3 looks at skills, an issue that cuts across all to productive work. Using a simple analytical sectors of employment, focusing on the skills framework, the report develops a systematic that are critical for productive employment and and detailed understanding of the challenges how they are acquired. The chapter assesses the involved in improving the productivity, earn- role of schooling in producing education and ings, and efficiency of the transitions of youth— learning and describes the wide landscape of as well as the various interventions that show apprenticeships and other forms of training promise in addressing those challenges (table that develop skills outside of school. 1.3). We focus on the three sectors in which pro- ductivity increases will be most critical: agricul- Notes 1. Excluding China, the estimates for East Asia and ture, nonfarm HEs, and the modern wage sector. the Pacific are 115 million youth ages 15–24 in We distinguish two types of constraints that 2015 and 101 million in 2050. The decline in the limit young people’s potential for finding path- cohort size begins in 2010 (as it does for China). ways to productive work in the three sectors: 2. Figures are based on Demographic and Health 1. Human capital. The supply side, or the abili- Surveys (DHS) final reports (www.measuredhs .com). ties, education, skills, family connections, 3. McKinsey based its estimate on data for the Arab networks, beliefs, and other character traits Republic of Egypt, Indonesia, Korea, Malaysia, that are embedded in an individual and Mexico, Pakistan, the Philippines, Thailand, and allow that person to find opportunities to be Turkey. productive, increase earnings, and achieve 4. Africa’s resource-rich countries were excluded income security from this simulation, because even the resource- 2. Business environment. The factors outside rich countries in East Asia did not achieve the type a worker’s immediate control that affect of employment transformation simulated here. 42 Youth EmploYmEnt in Sub-Saharan africa References Policy Research Working Paper 6184, World Bloom, David E., David Canning, and Pia N. Bank, Washington, DC. Malaney. 2000. “Population Dynamic and Eco- Fox, Louise, Alun Thomas, Cleary Haines, and nomic Growth in Asia.” Population and Develop- Jorge Huerta Munoz. 2013. “Africa’s Got Work to ment Review 26 (supplement): 257–90. Do: Employment Prospects in the New Century.” Bloom, David E., and Jeffrey Williamson. 1998. IMF Working Paper, International Monetary “Demographic Transitions and Economic Mir- Fund, Washington, DC. acles in Emerging Asia.” World Bank Economic Gelb, A. 2010. “Economic Diversification in Review 12 (3): 419–55. Resource-Rich Countries.” Paper prepared for Bongaarts, John, and John Casterline. 2013. “Fertil- the conference on Natural Resources, Finance, ity Transition: Is Sub-Saharan Africa Different?” and Development: Confronting Old and New Population and Development Review 38 (supple- Challenges, Central Bank of Algeria and the IMF ment 1): 153–68. Institute, Algiers, November 4–5. Chuhan-Pole, Punam, and Manka Angwafo, eds. IMF (International Monetary Fund). 2012. 2011. Yes Africa Can: Success Stories from a “Regional Economic Outlook: Sub-Saharan Dynamic Continent. Washington, DC: World Africa Sustaining Growth amid Global Uncer- Bank. tainty.” IMF, Washington, DC. Devarajan, Shantayanan, and Wolfgang Fengler. Kapsos, Stephen. 2005. “The Employment Intensity 2013. “Africa’s Economic Boom: Why the Pessi- of Growth: Trends and Macroeconomic Deter- mists and the Optimists Are Both Right.” Foreign minants.” Economic Strategy Paper 2005/12, Affairs (May–June): 68–81. International Labour Organization, Geneva. Lin, Justin Yifu, and Célestin Monga. 2012. “Solving De Wolf, Stefan, Yves Rolland Rakotoarisoa, Lau- the Mystery of African Governance.” New Politi- rence Vanpaeschen, and Honoré Rabekoto. 2008. cal Economy 17 (5): 659–66. Madagascar: Le Grand Livre des Petits Métiers: Portraits of Daily Life Professions. Belgium: Snoek Losch, Bruno, Sandrine Freguin-Gresh, and Eric Publishers. Thomas White. 2013. Structural Transformation and Rural Change Revisited. Washington, DC: Dinh, Hinh T., Vincent Palmade, Vandana Chandra, World Bank. and Frances Cossar. 2012. “Light Manufacturing in Africa: Targeted Policies to Enhance Private United Nations. 2011. World Population Prospects: Investment and Create Jobs.” Agence Française The 2010 Revision. New York: United Nations, de Développement and World Bank, Washing- Department of Economic and Social Affairs, ton, DC. Population Division. http://esa.un.org/wpp/ Excel-Data/population.htm. Eastwood, Robert, and Michael Lipton. 2011. World Bank. 2010. “Equatorial Guinea Public “Demographic Transition in Sub-Saharan Africa: Expenditure Review.” World Bank, Washington, How Big Will the Economic Dividend Be?” Popu- DC. lation Studies 65 (1): 9–35. ———. 2012a. Africa’s Pulse 6 (October). World Fine, David, Arend van Wamelen, Susan Lund, Bank, Washington, DC. Armando Cabral, Mourad Taoufiki, Nor- bert Dörr, Acha Leke, Charles Roxburgh, Jörg ———. 2012b. “Employment in Cameroon: Stock Schubert, and Paul Cook. 2012. “Africa at Work: Take of Programs, Assessment of Existing Gaps Job Creation and Inclusive Growth.” McKinsey and Opportunities, and Proposed Next Steps.” Global Institute, Washington, DC. World Bank, Washington, DC. Fox, Louise, and Melissa S. Gaal. 2008. “Working ———. 2012c. “Gabon Public Expenditure Review: out of Poverty: Job Creation and the Quality Better Management of Public Finance to Achieve of Growth in Africa.” World Bank, Washington, Millennium Development Goals.” World Bank, DC. Washington, DC. ———. 2012d. “Gabon: Rapport sur la croissance Fox, Louise, and Obert Pimhidzai. 2013. “Different et l’emploi.” World Bank, Washington, DC. Dreams, Same Bed: Collecting, Using, and Inter- preting Employment Statistics in Sub-Saharan ———. 2012e. World Development Report 2013: Africa: The Case of Uganda.” Policy Research Jobs. New York: Oxford University Press. Working Paper 6436, World Bank, Washington, ———. 2013. “Economic Growth in Equatorial DC. Guinea: Paths to Inclusiveness and Sustainabil- Fox, Louise, and Thomas Sohnesen. 2012. “House- ity.” World Bank, Washington, DC. hold Enterprises in Sub-Saharan Africa: Why ———. Various years. World Development Indica- They Matter for Growth, Jobs, and Livelihoods.” tors. Washington, DC: World Bank. Jobs: More Than Just Income FOCUS NOTE 1 a job—be it for a wage or not—is almost nic backgrounds in Rwanda’s coffee industry always more than just an income. A job found that workplace interactions are associ- affects a person’s core sense of identity ated with better attitudes toward collabora- and at the same time establishes how a person tion across ethnic boundaries and less distrust is perceived by society. The kind of job that (Tobias and Boudreaux 2011). people do exerts a powerful influence on their Jobs in agriculture, too, can connect peo- social well-being and economic development. ple through networks. Studies in Ghana and The World Development Report 2013: Jobs cov- Uganda illustrate how farmers connected ers these themes in detail and serves as the basis through networks can obtain information “[In my job] I for the brief discussion that follows. and increase productivity. In Ghana, pine- meet a lot of peo- apple farmers adjusted their use of fertilizer in response to the successful or unsuccessful ple, learn how to The Value of Jobs experiences of their neighbors. Farmers just express myself starting to cultivate pineapples are more likely and how to go People throughout the world consider jobs to make changes based on information received about personal to be more than a task or an income. Jobs say from other farmers, showing the potential of communication.” something about an individual’s place and on-the-job interactions and learning from oth- Ghana identity in society and contribute to an individ- ers (Udry and Conley 2004). In rural Uganda, ual’s satisfaction with life (World Bank 2012). a recent randomized experiment studied the In a 2012 survey in Sierra Leone, 90 percent of productivity effect of networks by pairing cot- respondents judged their job to be somewhat ton farmers to stimulate the exchange of infor- or absolutely meaningful (Hatløy et al. 2012). mation. The pairs were encouraged to discuss Similarly, youths interviewed in a qualitative farming activities, problems, and solutions and study in Ghana reported that they value jobs to set a target for increased cultivation. Farm- that allow them to acquire new knowledge ers who participated in the project, especially and skills or to connect with other people women, significantly increased their produc- through social networks (Anarfi, Anyidoho, tivity. Connecting farmers with other farmers and Verschoor 2008). The type of job, contract, outside their established social circles helped to benefits, and safety and security at work all spread information that otherwise would not influence such perceptions of well-being. This have been shared (Vasilaky 2010). may explain why measures of job satisfaction While jobs can connect people through net- are lower in Sub-Saharan Africa than in other works, they can also exclude. Across countries, regions across all types of jobs—agricultural, most people find their jobs through connec- household enterprise, and modern wage (see tions with friends, relatives, and other acquain- figure F1.1). tances. In the 2012 jobs survey in Sierra Leone, In addition to their contribution to status, 75 percent of respondents reported that their empowerment, identity, and well-being, jobs job is important for establishing contacts with (including nonwage jobs) connect people others (Hatløy et al. 2012). Yet networks may through networks. Jobs connect people with have negative social consequences when they others—of different backgrounds, ethnicities, exclude people and groups who lack such con- and gender—with whom they would not oth- nections. A case study of a weaving cluster erwise interact and with information, includ- in Ilorin, a city in the Yoruba Muslim part of ing information about job opportunities. For western Nigeria, and a shoe and garment clus- example, a study of workers from different eth- ter in Aba, a city in the Igbo Christian area of 43 44 Youth EmploYmEnt in Sub-Saharan africa Figure F1.1 Jobs and life satisfaction across regions 50 45 40 35 30 Percent 25 20 15 High-income countries Latin America and the Caribbean 10 Europe and Central Asia 5 East Asia and Pacific Middle East and North Africa 0 South Asia Wage employed Self- Sub-Saharan Africa employed, Self- employed, Unemployed non-farm farm Source: World Bank 2012, based on Gallup 2009, 2010. Note: Figure shows the share of respondents who rate their present life at 7 or higher and who anticipate that their life in five years will be at 8 or higher on a scale of 10. eastern Nigeria, found that the greater inflow only consider “jobs” to be salaried employment, of producers increased the reliance on connec- whereas other forms of livelihood and work tions and social tensions. The poorest produc- activities do not qualify (World Bank 2013). ers had no regular suppliers or credit networks, In a similar study in Sierra Leone, youth con- but depended mainly on customers from their sidered “employment” to mean having a stable own village (Meagher 2011). and salaried position. They held office jobs in particular esteem. Such jobs are often referred to as “Englishman” jobs. Positions in teaching Jobs and Aspirations and nursing are also often treated with respect. Young people refer to casual, informal jobs that The distribution of jobs within a society can provide low levels of daily income as “dishonor- affect expectations and aspirations. As children able jobs.” and teenagers form goals for the future, their Wage jobs are not always the most coveted, aspirations may be influenced by whether their however. The reality is more complex, based on parents have jobs and the types of jobs their the context, available opportunities, and char- parents have. Frustration and even social unrest acteristics of youth. In Ghana, wage work is not may develop when education and effort are not necessarily preferred among different types of rewarded or when people perceive the distribu- employment (Falco et al. 2012). In fact, owners tion of jobs to be unfair. In many countries in of informal firms that employ others are sig- Africa, the conventional wisdom suggests that nificantly happier than people working in the people, and youth in particular, prefer wage formal private sector. Young people explained employment to other types of jobs, including that status, autonomy, and income cause them jobs in farming and household enterprises, to prefer self-employment. A 22-year-old stu- and that the lack of wage jobs can provoke dent explained, “Here in Ghana, you don’t earn social tensions. Wage employment, especially much by working for somebody. You are able in urban areas, is perceived to be more lucra- to make more from your own business than tive and secure and to have a higher status. from working for someone” (Anarfi, Any- For example, young people interviewed for a idoho, and Verschoor 2008). An unemployed qualitative study in Liberia explained that they youth echoed that message: “There is nothing focus note 1 45 like doing [your own work], and it gives you Figure F1.2 Jobs drive development the idea that, one day, I have to work hard and, Jobs if possible, establish my own company, employ connected to people.” Youth surveyed in Zambia expressed Jobs in global markets Jobs that are functional environmentally similar sentiments; those employed in wage cities benign jobs noted that they have to supplement their income with informal activities because of low DEVELOPMENT pay as well as limited job security (PREM Pov- erty Reduction Group 2008). Jobs for Jobs that give the poor a sense LIVING SOCIAL of fairness PRODUCTIVITY STANDARDS COHESION Jobs and Development Jobs that empower Jobs that link to women networks Jobs have three transformational dimen- sions for individuals and society (World Bank Jobs that do not Jobs that shift burden shape social 2012). The first one is living standards: poverty to others JOBS identity falls as people work their way out of hardship, Source: World Bank 2012. especially in countries where the scope for redistribution is limited. The second is produc- tivity: efficiency increases as workers get better at what they do and move from less productive individual may have different effects on society. jobs to more productive ones. The third is social Good jobs for development are those with the cohesion: societies flourish as jobs create a sense highest value for society, taking into account of opportunity and get people from different not only the value they have to the people who ethnic and social backgrounds to work together. hold them, but also the potential spillovers on When jobs are examined in light of their others—positive or negative. Jobs that reduce potential to contribute to those outcomes, poverty, connect the economy to global mar- it becomes clear that some jobs do more for kets, or foster trust and civic engagement can development than others (figure F1.2). For do more for development than others. policy makers, therefore, it is not only the num- The particular jobs that are good for devel- ber of jobs that matters, but their quality and opment will vary with each country’s level of contribution to a country’s development. As development, demography, endowments, and discussed, individuals value jobs for the earn- institutions. For example, ings and benefits they provide, along with their contributions to self-esteem and happiness. • In agrarian countries, most people live in But some jobs have broader effects on society. rural areas and their jobs are in agricul- Jobs for women can change the way households ture. Making smallholder farming viable invest in the education and health of children. is critical because poverty rates are high. Jobs in cities support greater specialization and Higher agricultural productivity can help the exchange of ideas, making other jobs more the development of off-farm employment. productive. And in turbulent environments, At the same time, urban jobs connected to jobs can contribute to peace and social cohe- world markets set the foundation for cities sion (see box F1.1). to become dynamic. Often the individual and social perspectives • In conflict-affected countries, the most on jobs coincide, but not always. Jobs with high immediate challenge is to support social pay and benefits may be coveted by individuals, cohesion. Employment for former com- but they may be less valuable to society if they batants or youths vulnerable to participa- are supported through government transfers or tion in violence is particularly important. restrictive regulations, undermining the earn- Construction can help, as it is labor inten- ings or job opportunities of others. Because of sive and can thrive even in a poor business gaps like these, jobs that look equivalent to an environment. 46 Youth EmploYmEnt in Sub-Saharan africa Box F1.1 Employment, conflict, and violence: Is there a link? Dissatisfaction with the quality and availability of jobs among banis 2005; Urdal 2004). Other work fails to find empirical youth in Africa has raised concerns in the media and in pub- evidence for the relationship between youth unemployment lic debate about the risks of violence and social tensions. and armed conflict, however, so the evidence remains mixed Across countries, however, the connections between jobs, (Cramer 2010). conflict, and violence are not straightforward. Only limited Overall, where unemployment is high or employment and contradictory evidence on those connections is available opportunities are poor, violence and tensions probably result from developing countries. Generally, the literature suggests from accumulated risk factors, such as exclusion, perceptions that relationships between conflict, violence, and employ- of opportunities, and family dynamics, rather than employ- ment status are indirect and may operate through channels ment status alone. For example, young people may turn to such as identity and social dynamics (Cramer 2010). gangs or other violent groups to compensate for the lack of The literature linking crime to unemployment comes ties they have in economic and social life. A longitudinal study mostly from developed countries, and it finds no consistent of youth in Ecuador found that members of gangs involved link between unemployment and violent behavior. Studies with drugs and guns joined “because they were searching for from the United Kingdom and the United States have linked the support, trust, and cohesion—social capital—that they youth unemployment to property crime, including burglaries maintained their families did not provide, as well as because and vehicle break-ins (Bell and Blanchflower 2010). The evi- of the lack of opportunities in the local context” (Moser dence is weaker for violent crimes. The literature on unem- 2009). Similarly, analysis in the United States found that ployment and conflict presents more consistent results, and gangs provide youth with the income, respect, and social ties although causality is difficult to establish, there is evidence that they are unable to find in jobs, particularly given the lim- that poor economic performance, including youth unem- ited opportunities available in cities such as Chicago and New ployment, can be associated with conflict (Collier and Sam- York that have lost stable manufacturing jobs (Padilla 1992). • In urbanizing countries, productivity growth • Fundamentals. Because jobs improve with in agriculture frees people to work in cities. development, it is necessary to create a pol- Jobs for women, typically in light manufac- icy framework that is conducive to growth. turing, can benefit households. Avoiding That task requires attending to macroeco- urban congestion and allowing the country nomic stability, an enabling business envi- to move up the value-added ladder are top ronment, human capital accumulation, priorities. and the rule of law—including respect for • In resource-rich countries foreign exchange rights. earnings may be substantial, but the abun- • Labor policies. Labor policy should avoid dance may undermine the competitiveness the distortionary interventions that clog of other activities and encourage the cre- the creation of jobs in cities and in global ation of jobs supported through transfers. value chains and that lack mechanisms for Jobs that lead to a diversification of exports giving voice and protection to the most vul- can have large development payoffs. nerable workers, regardless of whether they Ultimately the role of government is to are wage earners or not. ensure that the conditions are in place for • Priorities. Because some jobs do more for strong private sector–led growth, to under- development than others, it is necessary to stand why there are not more good jobs for understand where good jobs for develop- development in a particular country, and to ment lie, given the country context. Policies remove or mitigate the constraints that prevent should remove or offset the market imper- the creation of more of these jobs. The World fections and institutional failures that pre- Development Report 2013: Jobs outlines a three- vent the private sector from creating more layered policy approach: good jobs for development. focus note 1 47 References Moser, Caroline O. N. 2009. Ordinary Families, Anarfi, John Kwasi, Nana Akua Anyidoho, and Extraordinary Lives: Assets and Poverty Reduc- Arjan Verschoor. 2008. “The Economic Empow- tion in Guayaquil, 1978–2004. Washington, DC: erment of Young People in Ghana.” Report pre- Brookings Institution Press. pared for the World Bank, Washington, DC. Padilla, Felix M. 1992. The Gang as an American Bell, David, and David Blanchflower. 2010. “Youth Enterprise. Piscataway: Rutgers University Press. Unemployment: Déjà Vu?” Discussion Paper PREM Poverty Reduction Group. 2008. “The Eco- Series 4705, Institute for the Study of Labor, nomic Empowerment of Young People in Zam- Bonn. bia.” Report 51431, World Bank, Washington, Collier, Paul, and Nicholas Sambanis, eds. 2005. DC. Understanding Civil War. Vol. 1: Africa. Washing- Tobias, Jutta M., and Karol C. Boudreaux. 2011. ton, DC: World Bank. “Entrepreneurship and Conflict Reduction in the Cramer, Christopher. 2010. “Unemployment and Post-Genocide Rwandan Coffee Industry.” Jour- Participation in Violence.” Background paper for nal of Small Business & Entrepreneurship 24 (2). World Development Report 2011: Conflict, Secu- Udry, Christopher, and Timothy G. Conley. 2004. rity, and Development, World Bank, Washington, “Social Networks in Ghana.” Discussion Paper DC. 888, Economic Growth Center, Yale University, Falco, Paolo, William F. Maloney, Bob Rijkers, New Haven, CT. and Mauricio Sarrias. 2012. “Heterogeneity in Urdal, Henrik. 2004. “The Devil in the Demograph- Subjective Well-Being: An Application to Occu- ics: The Effect of Youth Bulges on Domestic pational Allocation in Africa.” Policy Research Armed Conflict, 1950–2000.” Social Develop- Working Paper 6244, World Bank, Washington, ment Paper 29740. World Bank, Washington, DC. doi: 10.1596/1813-9450-6244. DC. Hatløy, Anne, Tewodros Kebede, Huafeng Zhang, Vasilaky, Kathryn. 2010. “As Good as the Networks and Ingunn Bjørkhaug. 2012. “Perceptions of They Keep? Expanding Farmer’s Social Net- Good Jobs: Port Loko and Freetown, Sierra works Using Randomized Encouragement in Leone.” Background paper for the World Devel- Rural Uganda.” Yale University, New Haven, CT. opment Report 2013, World Bank, Washington, Processed. DC. World Bank. 2012. World Development Report 2013: Meagher, Kate. 2011. “Informal Economies and Jobs. Washington, DC: World Bank. Urban Governance in Nigeria: Popular Empow- erment or Political Exclusion?” African Studies ———. 2013. Understanding Youth Violence: Cases Review 54 (2): 47–72. from Liberia and Sierra Leone. Washington, DC: World Bank. Chapter 2 Youth: A Time of Transitions By definition, youth is a time of many tran- Often women’s opportunities are circumscribed sitions—from school, to work, through risky by social norms related to family responsibility, behaviors, to founding a family, and to exercising agency, and acceptable economic activities and citizenship. These transitions, which are shaped by fears of harassment. Given these challenges, by the opportunities available to each individual, the pertinent policy question is how to make the as well as by social norms and aspirations devel- paths to employment easier for young people to oped during childhood and adolescence, have find and navigate. long-lasting consequences. For youths, the path from school to a livelihood in agriculture, house- Different countries and different institu- hold enterprise (HE), or wage employment can tions define youth differently. Broadly speaking, be particularly long. Young people in rural areas the United Nations defines youth as ranging may work on their parents’ farms for some time from ages 15 to 24, the African Union defines prior to establishing their own farm or HE. In youth as ranging between ages 15 and 35, and urban areas, they may exit the labor force for a many Sub-Saharan countries apply their own year or two and then spend several years seek- definitions (such as ages 15–40 in Mali or ages ing but not finding a wage job. They may also 15–30 in Kenya). For this report, we do not work in the family business to gain experience define youth as a specific age range. Rather, we and save money to start their own HE or invest focus on the fact that youth is a period of tran- in a farm. Once young people start working in a sition—from school, to work, through risky sector—whether in agriculture, an HE, or a wage behaviors, to founding a family, and to exercis- job—they tend to stay there, although mobility ing citizenship (World Bank 2006). We present across sectors is slightly higher in urban than data for the various age ranges during which in rural areas. Migration can offer an opportu- these transitions occur (most commonly ages nity to change these patterns and substantially 15–24 and 25–34). Generally we refer to the increase earnings, but migration for employ- older group as “young adults” to distinguish ment is (perhaps surprisingly) low in Africa. For them from the younger group. young women, the transition from school to high- The transitions of youth have long-lasting productivity work entails additional challenges. consequences. Clearly, the stage at which a young 49 50 Youth EmploYmEnt in Sub-Saharan africa person leaves the education system and starts are still in school, but half of them are still in working is a key determinant of that person’s primary school (figure 2.1). Of the 16 percent skill level and career path, but an individual’s of 24-year-olds who remain in school, nearly stock of human capital is also formed through three-quarters are still in secondary or lower early work experience. A prolonged period of levels. Thus, in addition to their many peers initial unemployment or an initial job that offers who never attended school or left it at a pre- no opportunities for learning and growth can mature age, many Africans leave school only potentially hurt an individual’s future produc- in their late teens or 20s—without having tivity. Similarly, decisions regarding health and acquired much education. family made during adolescence and early adult- For far too many young Africans, this mis- hood can have long-term effects on an individ- match between age and academic level means ual’s health and career. that years of childhood and adolescence that Because adolescence is a critical period of should be spent learning in the classroom are development, including for socioemotional being used inefficiently. A large part of early skills, economic or health shocks occur- adulthood that should be invested in learning ring during that time can have long-lasting on the job and gaining experience is instead consequences. being used to finish school. Work experience is important not only Aside from slow advancement through because it is needed to master skills but because school, an unusually late start to school and it affects adolescents’ expectations, goals, and interruptions to schooling combine to ensure confidence. In some settings (primarily upper- much the same result. Young people graduate income countries where a wage job is the norm at a late age, even though they have not spent for first employment), prolonged unemploy- enough years in school. Understanding why ment during youth can affect mental and these failures are endemic is essential to devis- physical health, and it has been linked to stress, ing a supportive youth employment policy. depression, and illness later in life (Lundberg Lengthy yet insufficient schooling may arise and Wuermli 2012). from the same factors that prevent many chil- Many types of market or government fail- dren from attending school in the first place or ures could potentially constrain these transi- that force them to leave school at an early age, tions. Examples include labor market rigidities that lengthen the school-to-work transition, inadequate information on the risks of certain choices, and lack of access to finance for pursu- Figure 2.1 Many 18-year-olds in Africa are still in ing higher education or starting a business. The school, but half of them are still in primary school remainder of this report examines such failures 100 in depth. Enabling youth to manage these tran- 90 sitions better—in particular, to develop path- % of age cohort still in school 80 ways to higher productivity and higher-earning 70 jobs—is the central challenge for promoting 60 youth employment in Sub-Saharan Africa. 50 40 30 Youth’s Transition from School 20 10 More than 80 percent of 12-year-olds in the 0 developing world are going to school. As 12 15 20 25 30 34 expected, this number declines with age until, Age (years) by the age of 24, few people remain in school. Primary or less Secondary Post-secondary In Sub-Saharan Africa this transition is slow. Source: Based on standardized and harmonized household and About 50 percent of 18-year-olds in Africa labor force surveys (see appendix). Youth: a time of transitions 51 such as the poor quality of schooling, difficulty common in Africa (box 2.1), and strong evi- of access, and lack of financial resources. dence indicates that it is associated with lower Another key question is the extent to which school attendance and educational attainment. lengthy yet insufficient schooling is caused by For some children, work marks a permanent children or youths starting to work at an age end to schooling when they are still very young. when they should be in school. Child labor is For other children or young people, the effects Box 2.1 Child labor in Sub-Saharan Africa Many African children ages 5–14 engage in some type of finance, land, or other resources require more work from work. Compared to other regions, Sub-Saharan Africa has household members (World Bank 2012b). the largest share of child labor (World Bank 2012b). Child labor is typically associated with delays in the accu- The worst types of child labor are unambiguously harm- mulation of human capital, and it has been shown to reduce ful to children and include all forms of slavery, bondage, school attainment and school attendance (Beegle, Dehejia, military conscription, trafficking, and using, procuring, or and Gatti 2009). More limited evidence indicates that child offering children for prostitution, pornography, or other labor leads to long-term “scarring” over and above its nega- illicit activities. Extreme child labor leads to psychological tive impact on schooling. Yet the long-term implications and physical scarring, which in turn affects learning capacity of child labor for young people’s ability to engage in more and future earnings. For instance, in Uganda, child soldiering productive activities or for their welfare in general remain had a corrosive effect on human capital by keeping children unclear. In Tanzania, working in childhood increased the away from school and creating high levels of psychological probability that boys would be involved in farming 10–13 distress, especially among those who experienced the most years later, and it also pushed girls toward domestic work violence (Blattman and Annan 2010). and early marriage (Beegle et al. 2008). However, child Most children are involved in other forms of labor, such labor was also associated with an increased likelihood of as working on the farm while going to school (particularly wage work, which could be linked to higher living standards in agricultural households) as well as performing domes- (Beegle, Dehejia, and Gatti 2009). tic work, selling, begging, and engaging in some type of manufacturing. The most recent round of the Multiple Indi- Figure B2.1.1 Percentage of children working cator Cluster Survey (MICS) collected data on child work. It defines child labor among 5- to 11-year-olds as being Cameroon engaged in economic activities (working, paid or unpaid, for Chad someone who is not a member of the household, or work- ing in a family business, on the farm, or selling goods in the Congo, Dem. Rep. street) for at least 1 hour in the 7 days prior to the survey Ethiopia or being engaged in domestic activities (also described as Guinea-Bissau household chores) for at least 28 hours a week. Among 12- to 14-year-olds, only those who engage in economic activ- Nigeria ity for 14 hours or more or in domestic work for 28 hours Sierra Leone or more in the 7 days prior to the survey were counted as Swaziland engaged in child labor. According to these definitions, child labor among 5- to 11-year-olds is typically around 40–60 Togo percent; child labor among 12- to 14-year-olds is typically 0 20 40 60 80 100 closer to 20 percent (figure B2.1.1). In both age groups, the Percent results from these surveys show that the majority of children Ages 5–11 Ages 12–14 work “for the family business” in either the farm or non- Sources: Multiple Indicator Cluster Survey Reports (MICS), various farm sector. years (circa 2010). Available at www.childinfo.org. Child labor tends to be more prevalent among poor Note: Child labor is defined as follows: (a) for ages 5–11, as eco- nomic activity for at least 1 hour a week or domestic work for at households. Children in wealthier households may, in some least 28 hours a week; (b) for ages 12–14, as economic activity for at settings, engage in labor if household assets and access to least 14 hours a week or domestic work for at least 28 hours a week. 52 Youth EmploYmEnt in Sub-Saharan africa Figure 2.2 School and work are often combined are already working exclusively (not counting 100 those who are at work and in school). The inci- 90 dence of exclusive work then rises rapidly. By 80 age 20, it is well above 50 percent, and by age 70 25, it is at 75 percent. % of age cohort 60 Because a sizable number of those who 50 reach working age have worked while attend- 40 ing school, as many as 70 percent of African 30 youths complete the transition to work by the 20 time they are 18. With age, the ranks of work- 10 ing youth increasingly comprise those who are 0 exclusively at work. 15 20 25 30 34 The incidence of concurrent schooling and Age (years) work is especially high in agriculture, where Working and not in school Working and in school almost 90 percent of such work involves youth working on the family farm. In this context, Source: Based on standardized and harmonized household and labor force surveys (see appendix). young people may have fewer opportunities to leave the farm sector. At the other end of the spectrum, the transi- of premature work may be less extreme but tion from school to work appears to be exces- nonetheless limiting. Some children interrupt sively long for some young people. About 17 their schooling to perform intermittent work. percent of 26-year-olds and 10 percent of Others engage in work that is “light” enough 34-year-olds are not working. This pattern does to allow them to attend school but that delays not vary by gender, so lower female labor force their progression through grades. Indeed, participation is an unlikely explanation. Some household surveys indicate that a sizable per- of these people may be in a temporary phase centage of young people are both at work and of unemployment between jobs. That, too, is in school; as many as 45 percent of 15-year-olds unlikely to be the main explanation for their may fall in this category (figure 2.2). This work- delayed transition to work, however, given the school overlap becomes less common with age, low share of wage employment among youth. declining to about 10 percent among 24-year- Identifying the groups most likely to experi- olds, presumably because they switch to full- ence such difficulty in starting work should be time work. a priority. As this discussion shows, for many indi- Traditional policies related to the school- viduals the transition from school to work is to-work transition often use informational not a clearly demarcated progression from interventions and other strategies to link job schooling to graduation and work. Often the seekers to employers, but their relevance in transition is quite blurry—school and work Africa is limited. As the next section discusses overlap, or schooling is interrupted for work in greater detail, even though most young peo- and then resumed—so policies that target ple are already working by the age of 20, most only those who have just left school might have not yet left the family enterprise or farm. not be enough. At 18 years of age, nearly 70 percent of work- For many young people in Sub-Saharan ing youths are occupied in farming. At 24 years Africa, the transition from school begins of age, nearly 60 percent of those working are relatively early. Although in most developing still on the farm, and about 25 percent are in countries this transition can begin at any time a nonfarm family enterprise or some form of between the ages of 12 and 24, it picks up after self-employment. Only for cohorts in their 18 years of age, crossing the 50 percent mark mid-20s and onward does wage employment only among those in their early 20s. In contrast, represent a notable share (more than 15 per- more than 20 percent of 15-year-olds in Africa cent) of total employment. Youth: a time of transitions 53 Traditional school-to-work transition poli- more than 60 percent of men and 70 percent cies may, however, be more relevant to those in of women are working for themselves. Working their mid-20s and early 30s who are yet to start for someone else increases modestly as people working and are possibly looking for a wage age, reaching about 30 percent for men and job. Evidence from firm surveys shows that the about 20 percent for women. majority of such jobs are found through social Given the slow transformation of the struc- networks, leaving those with weak networks at ture of employment, even after they become a strong disadvantage (see the discussion of fig- economically independent from their parents ure 2.9 later in this chapter). most people end up working in the same type of economic activity as their parents (box 2.2). In part because of the slow spread of educa- Youth’s Transition to Work tion into rural areas and limited opportunities to enter nonfarm sectors, males born to house- As young people begin to work less for their holds where the father was a farmer tended to families and more for themselves (figure 2.3), become farmers themselves. Those born into they increasingly leave the family activity households where the father had nonfarm work and strike out on their own. The vast major- tended to end up in the nonfarm sector—usu- ity become self-employed. By their early 30s, ally because they grew up in an urban area and Figure 2.3 Most Sub-Saharan African youths start out working for their families and then become self-employed a. Male, rural b. Female, rural 100 100 80 80 % of age cohort % of age cohort 60 60 40 40 20 20 0 0 15 20 25 30 34 15 20 25 30 34 c. Male, urban d. Female, urban 100 100 80 80 % of age cohort % of age cohort 60 60 40 40 20 20 0 0 15 20 25 30 34 15 20 25 30 34 Worked for family (past 12 months) Worked for someone else (past 12 months) Worked for self (past 12 months) Source: Based on Demographic and Health Survey (DHS) data in 35 countries for females and 28 countries for males (see appendix). 54 Youth EmploYmEnt in Sub-Saharan africa Box 2.2 Intergenerational transmission of occupations in Africa The trend for young people to end up in the same activity in Côte d’Ivoire. In spite of the structural decline of agricul- as their parents has deep roots. A study covering five Afri- tural employment, those numbers are not much lower than can countries reveals strong patterns in the intergenerational for previous generations. The intergenerational transmission transmission of occupations from fathers to sons. A farmer’s of nonfarm employment is also substantial (figure B2.2.1, son is much more likely to become a farmer than to exit agri- panel b). In Côte d’Ivoire, 93 percent of the sons born in the culture (figure B2.2.1, panel a). In Madagascar, 87 percent 1960s of nonfarmers end up being nonfarmers, compared of farmers’ sons born in the 1960s are also farmers at the to 89 percent in Guinea, 73 percent in Ghana, 69 percent in beginning of their working life, as are 72 percent in Uganda, Madagascar, and 61 percent in Uganda. Overall, these pat- 70 percent in Guinea, 69 percent in Ghana, and 64 percent terns reveal limited intergenerational mobility. Figure B2.2.1 Intergenerational transmission of employment sector in five African countries a. Probability that farmers’ sons stay farmers b. Probability that nonfarmers’ sons stay nonfarmers 100 100 80 80 60 60 Percent Percent 40 40 20 20 0 0 Côte d’Ivoire Ghana Guinea Madagascar Uganda Côte d’Ivoire Ghana Guinea Madagascar Uganda Son born in 1930–39 Son born in 1940–49 Son born in 1950–59 Son born in 1960–69 Source: Based on Bossuroy and Cogneau 2013. encountered better opportunities for education over the two time periods—mainly because and nonfarm employment. Data from urban they were more likely to stay in agriculture.1 areas in West Africa collected in 2001–02 show In urban Tanzania, labor force histories col- that more than 60 percent of those who were lected from respondents who were 20–39 in self-employed had fathers who were also self- 2005 showed that among those whose first job employed (Pasquier-Doumer 2013). was wage employment, only 20 percent ever The limited longitudinal data available— shifted into an HE (figure 2.4). The same sur- from Uganda and urban Tanzania—suggest vey reveals that those whose first job was in the that once individuals enter a type of job—agri- HE sector (self-employment or family worker) culture, household enterprise (HE), nonfarm also tended to stay in that sector. The transition wage—few of them switch to another type of that determines what type of job a young per- job. Nationally representative panel data from son will hold therefore occurs primarily when Uganda suggest that, overall, 61 percent of that person exits school, not through subse- youth ages 20–29 in 2005 who were economi- quent mobility across employment sectors.2 cally active had the same type of employment five years later, with an especially high persis- Moving into the Three Sectors of tence in agriculture (73 percent). Persistence is Employment lower in HEs and wage work, but nevertheless Given the relatively small amount of switching high (at around 50 percent). Women were more between sectors, the transition to the first type likely to stay in the same type of employment of work is key. But little is known about exactly Youth: a time of transitions 55 how young people transition into the three main Figure 2.4 Youths in urban Tanzania are unlikely to move between employment sectors of employment. The cross-sectional data sectors during their working lives used in this report do not record individual his- tories and thus hide variation between cohorts as well as within cohorts on the timing and effi- 100 ciency of the transition. Longitudinal data from 90 Tanzania and Uganda provide some insights, as 80 they track individuals over time. Focus group 70 data collected from African youths help to fill % of youths out the picture. As noted, however, the under- 60 standing of how young people make key choices 50 leading to a transition to stable employment is 40 Wage unclear. What is clear is that the pathways and 30 employed constraints are different depending on the type Self-employed 20 of activity concerned and an individual’s loca- Family worker 10 tion and gender. 0 Unemployed As discussed, the transition into agriculture Entry as begins early. The vast majority of teenagers unemployed Entry as Out of labor force who work are working in agriculture (figure family worker Entry as self-employed Entry as 2.5). At age 15, of the 60 percent of those who wage employed are working, almost 90 percent are working in agriculture. The share working outside agricul- ture increases steadily with age, largely because Source: Bridges et al. 2013. Note: Sample includes youth ages 20–39 and in the labor force in 2005. young people who leave school at higher grades enter other sectors. In rural areas, where lim- ited educational opportunities prevent youths Like their adult counterparts, the young are from staying in school for very long, agricul- more likely to work in an HE than in a wage job. ture employs more than 90 percent of 15- and At age 19, just over 5 percent work in an HE, 16-year-olds, and about 80 percent of young while few work for a wage; by age 24, almost 20 people ages 24 and older remain in agriculture percent work in an HE, and just over 10 per- (although some who report agriculture as their cent are in a wage job; and by age 34, almost 30 primary activity also have a nonfarm activity percent work in an HE, and the share in wage as well). Women who work are more likely to work remains around 20 percent (figure 2.5). work in agriculture than men—and unlike As they leave school, women, like men, who are men their probability of working in agriculture not in agriculture are substantially more likely does not decrease much with age. One reason to work in an HE than to obtain wage employ- why so many women remain in agriculture is ment. Longitudinal data from urban Tanzania that they leave school sooner, so employment suggest that many women enter this sector by opportunities are set much earlier for females helping out in a family business and then tran- than for males. sitioning to their own HE. The same data also Although agriculture is primarily a rural show that other women enter the HE sector activity, young people and adults in second- only after a long period of searching for a wage ary cities and periurban areas report working job or failing to find one to their liking or with in agriculture as well. For example, almost 50 remuneration as high as they could earn in the percent of youth ages 20–29 in urban Uganda HE sector (Falco et al. 2012). in 2010 reported working in agriculture as their Because it is based on primary occupation, primary job, and at least 60 percent of those which shows substantial persistence, this gen- had entered the sector five years before. Young eral picture conceals changes happening below people in urban areas are less likely to stay in the surface. Within the HE sector and agricul- agriculture, however; it is often a stepping stone ture alike, young people may start working with to starting a nonfarm business (figure 2.6). their parents and then acquire their own plot of 56 Youth EmploYmEnt in Sub-Saharan africa Figure 2.5 Youth transitions to sector of employment vary across urban and rural areas and between male and female youths a. Overall 100 90 80 70 % of age cohort 60 50 40 30 20 10 0 15 18 21 24 27 30 33 b. Rural c. Urban 100 100 90 90 80 80 70 70 % of age cohort % of age cohort 60 60 50 50 40 40 30 30 20 20 10 10 0 0 15 18 21 24 27 30 33 15 18 21 24 27 30 33 d. Female e. Male 100 100 90 90 80 80 70 70 % of age cohort % of age cohort 60 60 50 50 40 40 30 30 20 20 10 10 0 0 15 18 21 24 27 30 33 15 18 21 24 27 30 33 Agriculture Household enterprise Wage Source: Based on standardized and harmonized household and labor force surveys (see appendix). land or start their own business. In both rural save enough money to start their own busi- and urban areas, multiple jobs are common. A ness. Labor force survey data from Tanzania young person may start out in agriculture but indicate that by age 25 about half of those who add nonfarm work as a second job. Even young had employment also pursued a secondary people with a wage job may start out in casual economic activity (figure 2.7). The probability labor but move to a more stable wage job or of having multiple sources of employment is Youth: a time of transitions 57 Figure 2.6 Sectoral mobility among urban youth in Uganda Sector of employment in 2010 by sector of employment in 2005 70 60 50 40 Percent 30 20 10 0 Agriculture, Household enterprise owner, Family worker Wage employed, Unemployed or out of labor force, 2005 (33%) 2005 (17%) (household enterprise), 2005 (22%) 2005 (13%) 2005 (15%) Agriculture, Household enterprise owner, Family worker (household enterprise), Wage employed, Unemployed or out of labor force, 2010 2010 2010 2010 2010 Source: Based on standardized and harmonized Uganda panel survey (see appendix). much higher in rural areas. Having multiple Figure 2.7 In Tanzania, many work in two or more jobs helps the worker to minimize the risks activities associated with operating a farm or HE and to 60 % of age cohort holding multiple jobs cope with seasonal variation in employment opportunities and income. The higher prob- 50 ability of having multiple sources of employ- 40 ment in rural areas also suggests that mobility 30 may be higher in rural areas than the data on primary economic activity indicate. 20 Young people find it particularly challeng- 10 ing to enter into modern wage employment. Employment histories of urban Tanzanian 0 15 20 25 30 34 youths suggest that most did not enter directly Age (years) into a wage job (20 percent; figure 2.4). How- ever, 69 percent of those who entered a wage Urban All Rural job remained in wage employment—26 per- Source: Based on standardized and harmonized Tanzania labor force survey, 2005 (see appendix). cent became self-employed, and 4 percent became unemployed. Among the few young people who work for wages, less than half have a contract. Only around 20 percent of African ing experienced long periods out of the labor teenagers with wage employment have a con- force or in unemployment before entering into tract; this rate increases to around 50 percent stable employment. In the longitudinal data, for those in their mid-20s and around 60 per- the average age of leaving school was two years cent for 30-year-olds (figure 2.8). Young people less than the average age of entering the labor are most often engaged in casual wage employ- force, suggesting that even those who did not ment, which can take the form of part-time or report entering the labor force as unemployed seasonal wage work. These arrangements are (for example, searching for a wage job) spent often verbal and of very short duration. time idle or doing odd jobs of very short dura- The desire for a wage job can lead urban tion (Bridges et al. 2013). Most of those who youth to search a long time, without success. reported a significant period of unemploy- In urban Tanzania, young people reported hav- ment (not working and looking for a job) did 58 Youth EmploYmEnt in Sub-Saharan africa Figure 2.8 Younger people are most often engaged getting information on the work-related char- in casual wage employment acteristics of youth and the need to have a trust- 100 worthy workforce, employers prefer to rely on contacts to attract new entrants. Urban youths % of wage-employed youth 80 who participated in focus group discussions in Kenya complained that either a personal con- with contract 60 tact from the same ethnic group or a bribe or 40 both were required simply to get anyone in a private firm to look at their résumé, even for 20 those with postsecondary education (World Bank 2012a). As long as firms have an ample 0 supply of entry-level candidates, this behavior 15 20 25 30 34 is not likely to change. Age (years) Networks are important beyond the wage Source: Based on standardized and harmonized household and sector. In a survey in Ghana, most youths cited labor force surveys (see appendix). networks as important for getting any type of work. Young men mentioned that friends would tell them about work opportunities, not end up getting a wage job. Only 25 percent especially in petty trading, street vending, of those who were no longer unemployed at farming and fishing, and construction. Young the time of the survey had found a wage job; women tended to get referrals from family the rest went into the HE sector (family or self- members for jobs in an HE, in many cases with employment). The average duration of unem- other family members. ployment before becoming self-employed was close to 4.5 years. Moving into Employment through If a young person is hoping for a wage job Migration but ultimately settles for self-employment, 4.5 Urbanization is fundamentally altering Africa. years is a long time to search for work before Even though the share of the population liv- realizing that “informal is normal.” The aver- ing in urban areas remains well below the share age duration of unemployment for those who in the rest of the world, in Africa urbanization entered the labor force as unemployed but has accelerated over the past couple of decades. managed to find a wage job in the end was 5.5 years. Youth entering as unemployed generally Figure 2.9 Personal networks are key to finding a job have higher education and come from wealth- ier families. Individuals who went through long Family/friends spells of unemployment did not suffer an earn- Public announcement/ ings penalty. If one makes it into wage employ- advertisement ment, even a long job search can pay off, but Private placement office the odds of finding a wage job are low (Bridges Public placement office et al. 2013). School-related network The importance of networks and contacts is one reason why the search for a wage job can Applicant walked in be frustrating for youth. Across countries, most Other people get their jobs through contacts from family and friends, especially in the case of 0 10 20 30 40 50 60 70 % of employers who filled most modern wage jobs. Almost 60 percent of enter- recent position with each method prises surveyed in 14 countries report that their Source: Based on World Bank enterprise surveys (aggregated most recent position was filled through con- from surveys conducted in Angola, Botswana, Burundi, Camer- tacts with “family or friends” (figure 2.9). For oon, the Democratic Republic of Congo, The Gambia, Guinea, Mauritania, Namibia, Niger, Rwanda, Swaziland, Tanzania, and a variety of reasons, including the difficulty of Uganda in 2006–07). Youth: a time of transitions 59 Census data from 42 Sub-Saharan African place between urban areas or between rural countries show that by 2010 two-fifths of areas. Seasonal migration between rural areas Africa’s urban population already lived in a big is common in some countries, Niger for exam- city (with a population of 1 million or more), ple, where it can involve up to one-third of the while two-fifths lived in a small town (with population (Simkins 2013). less than 250,000 people; Dorosh and Thurlow A snapshot of migration patterns and their 2013). The share of the population living in an implications for employment mobility is pro- urban area is expected to rise from 39 percent vided by a recent study from Kegera, Tanzania. in 2010 to 57 percent in 2060 (Simkins 2013). The study followed individuals from the pre- Contrary to popular belief, fertility—not dominantly rural region of Kagera in Tanza- migration—still drives most urban population nia who were 0–11 years old in 1991–94 (De growth in Africa. After slowing from the 1960s Weerdt and Kutka 2013). Two decades later, in to the 1990s, migration to urban areas has risen 2010, 56 percent of those individuals lived in in recent years, but it is expected to constitute the same or a neighboring village, 18 percent only about 30 percent of urban population had moved to a village in the same region, 9 growth between 2010 and 2060 (Simkins 2013). percent had moved to an urban area in the same Urbanization can be beneficial if it creates region, and 8 percent had moved to another agglomeration effects that can be a source of region. The main reason that young people had long-term growth and structural transforma- migrated was marriage (35 percent), followed tion. Africa’s urbanization has not yet delivered by attending school (16 percent), looking for many of these beneficial effects. One reason is work (15 percent), and working (10 percent). discussed in chapter 1: the lack of a manufac- Migration was associated with a substantial turing sector. The beneficial effects of agglom- amount of employment mobility. Although eration are more often seen in lower-income 57 percent of youths ages 18–32 in 2010 who countries with the development of industy, not stayed in the village remained in agriculture, services, but the economy of African cities is only 30 percent of those moving to a nearby dominated by services. Where agglomeration town, and 3 percent of those moving to a city, benefits occur in service clusters, the sectors, were still working in agriculture (figure 2.10). such as information technology, tend to require Young people in cities were over three times higher levels of education than most countries more likely to be working for a wage and nearly in Africa have. As a result, agglomeration forces twice as likely to be in nonfarm self-employ- are weak, and Africa remains a low-density ment. These moves were associated with large continent (see World Bank 2008). At the same increases in measured consumption. Young time, urban growth rates in some capital cities people who stayed in the village or moved to a are very high and difficult to manage. local town saw incomes increase by 70 and 91 Africans tend to migrate within their percent, respectively, between 1991 and 2010; country of origin (Sander and Maimbo 2003; those moving to a regional town had incomes McKenzie 2007; Simkins 2013). Few migrants 150 percent higher; and those who moved to a come to Sub-Saharan Africa from outside the city had income gains of around 300 percent region, and international migration within (De Weerdt and Kutka 2013). Most of these the region is limited to a handful of “mag- gains were associated with the move outside of net” countries, such as Kenya, Nigeria, and agriculture. South Africa (Simkins 2013). Most migration Clearly, urban growth and migration from within countries occurs for purposes that are rural to urban areas can be a powerful engine not directly related to employment, includ- for transforming the structure of employment. ing education, marriage, or avoiding conflict To ensure that the economic infrastructure can or insecurity. Local moves in particular are support both indigenous population growth often motivated by marriage. Longer moves, and migrants, however, forward-looking urban out of district boundaries, are more likely to be planning and investments are required as part related to work. Much of the migration takes of an overall development strategy. 60 Youth EmploYmEnt in Sub-Saharan africa Figure 2.10 Migration increases mobility across sectors of employment Human immunodeficiency virus/acquired immune deficiency syndrome (HIV/AIDS) and Employment status in 2010, by location 60 teenage pregnancy are the two biggest dangers % of individuals who were age 0–11 in 1991–94 in Kagera, Tanzania 50 associated with teenage sexual activity, and both loom large in Africa. Sub-Saharan Africa 40 is the region most severely affected by HIV. 30 According to survey data collected between 2006 and 2011, the age-specific fertility rate 20 for young women ages 15–19 (births per 1,000 10 women) was above 100 in all Sub-Saharan countries except Rwanda, and it surpassed 150 0 in Chad, Guinea, Lesotho, Madagascar, Malawi, School Agriculture Self-employed Casual wage Employed Not employed work Mali, Mozambique, and Zambia—rates that are at the higher end compared to rates in other Village Town City developing regions (United Nations 2007). Source: De Weerdt and Kutka 2013. In thinking about policies to prevent risky teenage behavior, it is important to consider how such behaviors are connected with the other youth transitions. For example, given the The Parallel Transitions: Choices evidence that stress and mental health prob- Influencing Health, Family lems are associated with risky sexual activity Formation, and Civic Engagement and substance abuse among young people, the stress from idleness or job dissatisfaction could Certain decisions taken and habits formed dur- increase the chances of engaging in risky behav- ing adolescence have lifelong consequences. ior. Another example is the rising evidence that Adolescence is when most individuals begin school attendance lowers teenage sexual activ- to have control over their own decisions and ity and pregnancy. A conditional cash transfer behaviors. It is also a period of experimenta- targeting young Malawian women (13–22 years tion, when youth start engaging in risky activi- of age) provided incentives to current stu- ties such as smoking and consuming alcohol. dents and dropouts to attend school and led to Youth is also the period in which sexual activ- large increases in school enrollment as well as ity is initiated, family formation begins (figure declines in early marriage, teenage pregnancy, 2.11) (see de Walque 2014). This is also the sexual activity, and risky sexual behavior (Baird time when people begin to engage with society et al. 2010). Similarly, a study in Kenya finds that as citizens. reducing the cost of schooling (by paying for uniforms) reduced dropout rates, teen marriage, Taking Health Risks and childbearing (Duflo et al. 2006). Schooling Smoking is one of the leading preventable is not the only determinant, however. In the causes of death worldwide, and youth is when it Malawi study, girls who received an uncondi- is most likely to develop into a habit. Less than tional cash transfer—a transfer not linked to 10 percent of boys ages 13–15 smoke cigarettes school participation and performance—also in most of Africa, except the southern tip— had large reductions in sexual activity, teenage Botswana, Namibia, and South Africa (Warren pregnancy, and marriage (Baird et al. 2011). et al. 2009). This rate of tobacco use is one of the lowest in the developing world, but even so, Starting a Family the potential for damage should not be under- Along with starting work, starting a fam- estimated. As countries grow richer, they tend ily is one of the biggest decisions taken dur- to pass through a cycle of tobacco prevalence ing youth. By the age of 25, nearly 80 percent (Eriksen, Mackay, and Ross 2012). The second of women in Africa have married and given stage of this cycle is when smoking prevalence birth (figure 2.11).3 This transition happens tends to rise sharply. later among men. While more than half of Youth: a time of transitions 61 Figure 2.11 Family formation starts earlier for young women than for young men a. Females b. Males 100 100 80 80 % of age cohort % of age cohort 60 60 40 40 20 20 0 0 15 20 25 30 35 15 20 25 30 35 Age (years) Age (years) Ever had a child Ever married Ever had sexual intercourse Source: Analysis of DHS data for 28 countries (see appendix). all women are married by the age of 20, the health of both child and mother, should be majority of men are likely to remain unmar- of particular concern to policy makers. How ried before the age of 25 and to marry only in young parents invest in their children will have their late 20s or early 30s. a large impact on the well-being of future gen- Decisions about employment and family for- erations. Evidence is growing on the range of mation are linked. The decision to start a family is interventions that could prepare youth for this often taken after securing the means to support a transition (summarized in World Bank 2003). family, which could be the reason why most men For instance, micronutrient supplementation marry late. The linkages are particularly salient and food fortification for children and for for women. Economically active women may young women before and during pregnancy decide to postpone marriage, schedule births are known to improve birth weight and sub- later in life, and have fewer children on average sequent child development, as are conditional than women who are not economically active. cash transfers for the use of preventive health For example, in Asia and Latin America, wom- services. Other interventions, such as informa- en’s employment in the garment sector has been tion campaigns on reproductive health, safe linked to later age at marriage and lower fertility motherhood, and child health, look promising. (Amin et al. 1998; McLeod et al. 2005). In the other direction, there is evidence—albeit mostly Exercising Citizenship from developed countries—that childbearing Along with participating in the civic discourse lowers female labor force participation (Bloom on important social, economic, and other et al. 2009). In Africa, where most employment issues, citizenship means having certain rights is with the family farm or enterprise, the nature and obligations, such as the right to vote, which of the relationship between work and family African youths exercise in large numbers (fig- formation may be different. Early marriage or ure 2.12). Youth is also when individuals begin childbearing may not prevent work so much as to express themselves as citizens in other ways, trap young women in low-productivity activities such as attending community meetings, join- performed in the household. ing with others to raise issues, or taking part While the main drivers of fertility are largely in demonstrations. Survey data from 2008 economic, public policy has a role in support- indicate that in the past year between 60 and ing healthier family formation, as financial or 70 percent of African youth ages 20–30 had informational constraints may lead young par- engaged in at least one of those activities.4 ents in Africa to underinvest in family planning Civic engagement is necessary for govern- or maternal health services. Early pregnancy, ments to take young people’s ideas and aspira- which can have a large adverse impact on the tions into account effectively when formulating 62 Youth EmploYmEnt in Sub-Saharan africa Figure 2.12 As they get older, young people agriculture, and they are also less likely to gain increasingly engage as citizens wage employment. Women’s employment 100 opportunities are constrained not only by the fact that they tend to exit school earlier, but % of cohort who voted in 80 most recent election also by gender-specific constraints associated 60 with marriage and fertility choices. Women’s employment opportunities are also more likely 40 to be constrained by occupational segregation, 20 social norms, or the fear of sexual harassment. 0 Many young women in Africa have to decide 18 20 22 24 26 28 30 32 34 at the same time whether and how to enter the Age (years) labor market and whether and when to have Source: Afrobarometer 2008. children. For some women, marrying and hav- ing children (or marrying early) can be a means to escape from poverty by relying on men for policy. Such engagement is particularly relevant economic support. Although recent analy- for policy to reflect an understanding of the sis suggests that young women in Africa are types of work to which the young aspire and delaying having children in comparison with perhaps even to inform those aspirations. The previous generations, fertility rates for young common perception that the young are simply women remain much higher in Africa than in waiting for prized jobs in firms may have some other regions (Sneeringer 2009). Across Africa, truth, but it is unlikely to be relevant to most 25 percent of females ages 15–19 have already young people in Africa, who are predominantly had their first child or are pregnant.5 By the rural and have little education. Some studies time they reach age 20, the average age at first find that young people in low-income countries job, they are already caring for young children. are concerned mainly about jobs that guarantee Once faced with responsibility for child care, at least basic incomes and job security (Wietzke young women are constrained in their employ- and McLeod 2012). But others suggest that the ment choices. They may choose not to work, labor market aspirations of African youth are they may work fewer hours than they otherwise more heterogeneous. This complexity makes would, or they may choose occupations that it all the more important to facilitate the civic offer flexible schedules and home-based work. channels through which the young can express In Liberia, 41 percent of young women, com- themselves. If the young cannot engage as citi- pared to 31 percent of adult women and just 11 zens, their frustrations about not being heard percent of young males, reported family respon- could foment economic and social instability. sibilities as a reason for “inactivity” (Ruiz Abril Civic organizations can also broaden access 2008). A time-use study in Guinea concluded to economic opportunities for youth by pro- that women, in particular, cannot increase the viding better information and facilitating net- number of hours devoted to paid work because works. This kind of action may be particularly of their care duties (Bardasi and Wodon 2009). important to previously excluded groups (such Where care options are limited, the need for as women) and to youth from disadvantaged flexible work arrangements pushes women backgrounds. from formal into informal work that can be combined with child care, such as HEs. The experience from developed countries suggests Challenges for Females that while part-time work is often a good option Transitioning to Work for women seeking flexibility, it can often trap them in low-quality jobs even after that flexibil- Employment opportunities are more con- ity is no longer needed (World Bank 2011). strained for young women than for young Gender segregation in jobs is evident men. Women are less likely to be able to leave throughout Sub-Saharan Africa, as in other Youth: a time of transitions 63 developing regions. Men are concentrated in South Africa. In that village, another explains, “For women, manufacturing, construction, mining, and “My parents would never allow me to go to distance becomes transportation, while women are concentrated other places. They think I will fall pregnant and a barrier because in retail trade, education, health, and social bring shame upon the family name” (Petesch work. Gender segregation has potentially large and Rodríguez Caillava 2012). A young woman of lack of bicycles effects on overall productivity and economic who is well educated might be able to take a and a bad eye growth by allocating labor in a less than opti- distant job—for instance, as a teacher. on women who mal way. Labor force surveys in Africa con- Sexual harassment may also affect women’s ride bicycles.” sistently find that women are concentrated in employment options. Nationally represen- Tanzania sectors with low wages and low productivity. tative statistics on the prevalence of sexual For example, the 2010 Liberia labor force sur- harassment are hard to find, but smaller sur- vey found that women are underrepresented in veys have found that the perceived risk of the growth sectors of mining, manufacturing, harassment from bosses and other authority construction, and services (LISGIS 2011). Even figures factors highly into a woman’s choice of though women have made strides in entering work. In qualitative assessments conducted in the services sector in the past 20 years, they Liberia and Rwanda in advance of job training continue to lag in industry (ILO 2012). When programs, young women reported a preference “When I was women strike out on their own as entrepre- for self-employment or female-dominated looking for neurs, they still tend to work in less productive, industries because of the threat of sexual female-dominated sectors. In agriculture, plots harassment.6 Surveys have found similar sen- (secretarial) controlled by women are much more likely to timents among female entrepreneurs in Tanza- work here in be used for less profitable staple food crops, nia (Holla, Leonard, and Wilson 2010). Nega- Mombasa, I while male-controlled plots are used for cash tive experiences with teachers in school (for went to a certain crops (see World Bank 2011, ch. 5). These pat- example, teachers trading sex for grades) can company, and terns are reinforced by the fact that training reduce the willingness of prospective trainees when I showed for women is often concentrated in less lucra- to participate in training programs. Risks of tive female-dominated trades, such as arts and harassment are particularly high for women that boss my crafts, tailoring, and hairdressing (discussed in who work in other people’s homes, where the papers he said it chapter 3; Fox et al. 2011). isolation of the work and the dependence on was OK, but then Social norms strongly shape women’s the employer for housing and food can make I had to give employment opportunities. Expectations with workers especially vulnerable to physical and him a date first.” respect to child care duties, mobility and trans- sexual violence, nonpayment of wages, limi- Kenya port, and occupational choice all play a role. tations on mobility, and (in extreme cases) When women do choose to work outside of the forced labor. home, they often spend more time traveling on Cross-border traders are particularly vulner- slower modes of transport; as a result, they are able. A recent survey of traders on the border limited to employment options closer to home between Rwanda and the Democratic Repub- (Uteng 2011). Lack of agency also restricts their lic of Congo documented high rates of threats, options. In a focus group in Bombouaka, Togo, harassment, and violence such as beatings, young women said that a husband would likely stripping, and even rape among the predomi- veto a job that requires a woman to travel out- nantly female traders (Brenton et al. 2011). A side the village (Petesch and Rodríguez Caillava survey of Liberian female cross-border traders 2012). in 2007 revealed that, in addition to traditional Domestic and child care responsibilities and challenges such as high duties, lack of market concerns about their reputations and safety information, and lack of capital, women also make it much harder for women to travel: “If reported high rates of sexual harassment and a woman is gone for a few days, people would rape. The survey also found that women were start making up all sorts of bad stories about more likely than men to join or to form groups her. The norm is that a woman should stay at to jointly transport goods to reduce costs, home and raise her children and look after the improve safety, and collectively resolve disputes home,” says a young woman from Ezinyathini, with border authorities (UNWomen 2012). 64 Youth EmploYmEnt in Sub-Saharan africa Facilitating Transitions from Nigeria, Senegal, South Africa, Tanzania, Uganda, Zambia, and Zimbabwe. School, to Work, and across 5. Average across the most recent demographic Sectors of Employment and health surveys in Sub-Saharan Africa for which this indicator is available. Based on cal- In all likelihood, most young people who enter culations using DHS StatCompiler. into employment over the next 10 years in Sub- 6. In Liberia, 20 percent of those who reported Saharan Africa will continue to work in the a preference for self-employment mentioned same sectors as the previous generation, with avoiding sexual harassment as a reason for their some expansion into the HE sector, especially preference (Ruiz Abril 2008). in market towns and rural areas near popula- tion centers. The transition from school to work References will continue to be unstructured and often slow Afrobarometer. 2008. Abrobarmeter Surveys Round for these youths. Agriculture is the gateway to 4. http://www.afrobarometer.org/. work and, for many, will become the sector of Amin, Sajeda, Ian Diamond, Ruchira T. Naved, and lifetime employment. If productivity and earn- Margaret Newby. 1998. “Transition to Adult- hood of Female Garment-Factory Workers in ings in agriculture do not improve, the wider Bangladesh.” Studies in Family Planning 29 (2): economy in rural areas, where the majority of 185–200. the population still lives, will be depressed, and Baird, Sarah, Ephraim Chirwa, Craig McIntosh, the options for youth will shrink. and Berk Özler. 2010. “The Short-Term Impacts Once youth start working—whether in of a Schooling Conditional Cash Transfer Pro- agriculture, HEs, or wage jobs—they tend to gram on the Sexual Behavior of Young Women.” stay in the same sector, with mobility across Health Economics 19 (1): 55–68. doi: 10.1002/ sectors being slightly higher in urban than in hec.1569. rural areas. Migration offers an opportunity to ———. 2011. “Cash or Condition? Evidence from change these patterns and substantially increase a Cash Transfer Experiment.” Quarterly Journal of Economics 126 (4): 1709–53. earnings. For young women, the transition from school to high-productivity work entails Bardasi, Elena, and Quentin Wodon. 2009. “Work- ing Long Hours and Having No Choice: Time additional challenges. The remainder of this Poverty in Guinea.” Policy Research Working report identifies ways to support young people Paper 4961, World Bank, Washington, DC. in making these transitions, by equipping them Beegle, Kathleen, Rajeev H. Dehejia, and Roberta with the skills that will be needed wherever they Gatti. 2009. “Why Should We Care about Child work (chapter 3) and by ensuring that the busi- Labor? The Education, Labor Market, and Health ness environment and their human capital are Consequences of Child Labor.” Journal of Human conducive to high productivity in agriculture Resources 44 (4): 871–89. (chapter 4), HEs (chapter 5), and the modern Beegle, Kathleen, Rajeev H. Dehejia, Roberta Gatti, wage sector (chapter 6). and Sofya Krutikova. 2008. “The Consequences of Child Labor: Evidence from Longitudinal Data in Rural Tanzania.” Policy Research Work- Notes ing Paper 4677, World Bank, Washington, DC. 1. Based on analysis of Uganda national panel sur- doi: 10.1596/1813-9450-4677. vey data. Blattman, Chistopher, and Jeannie Annan. 2010. 2. The study is based on the Tanzania household “The Consequences of Child Soldiering.” Review urban panel survey, which collects retrospec- of Economics and Statistics 92 (4): 882–98. tive information on previous jobs (Bridges et al. Bloom, David E., David Canning, Günther Fink, 2013). and Jocelyn E. Finlay. 2009. “Fertility, Female 3. The percentage of women who have given birth Labor Force Participation, and the Demographic can include unmarried women. Dividend.” Journal of Economic Growth 14 (2): 4. Based on analysis of Afrobarometer round 4 79–101. surveys (Afrobarometer 2008) in 20 African Bossuroy, Thomas, and Denis Cogneau.2013. countries: Benin, Botswana, Burkina Faso, Cape “Social Mobility in Five African Countries.” Verde, Ghana, Kenya, Lesotho, Liberia, Mada- Review of Income and Wealth 59: s84–110. doi: gascar, Malawi, Mali, Mozambique, Namibia, 10.1111/roiw.12037 Youth: a time of transitions 65 Brenton, Paul, Celestin Bashinge Bucekuderhwa, Lundberg, Mattias, and Alice Wuermli, eds. 2012. Caroline Hossein, Shiho Nagaki, and Jean Bap- Children and Youth in Crisis: Protecting and Pro- tiste Ntagoma. 2011. “Risky Business: Poor moting Human Development in Times of Eco- Women Cross-Border Traders in the Great Lakes nomic Shocks. Washington, DC: World Bank. Region of Africa.” Africa Trade Policy Notes 11, McKenzie, David. 2007. “A Profile of the World’s World Bank, Washington, DC. Young Developing Country Migrants.” IZA Bridges, Sarah, Louise Fox, Alessio Gaggero, and Discussion Paper 2948, Institute for the Study of Trudy Owens. 2013. “Labour Market Entry and Labor, Bonn. Earnings: Evidence from Tanzanian Retrospec- McLeod, D., R. Ramirez, M. Davalos, and W. Gru- tive Data.” Background paper presented at the ben. 2005. “Apparel Jobs: Ladder Up or Poverty CSAE Conference on Economic Development in Trap?” Paper presented at the 2005 meeting of Africa, Oxford University, March. the Latin America and Caribbean Economics de Walque, Damien, ed. 2014. Risking Your Health: Association, Paris, October 27–29. Causes, Consequences, and Interventions to Pre- Pasquier-Doumer, Laure. 2013. “Intergenerational vent Risky Behaviors. Washington, DC: World Transmission of Self-Employed Status in the Bank. Informal Sector: A Constrained Choice or Bet- De Weerdt, Joachim, and Andreas Kutka. 2013. ter Income Prospects? Evidence from Seven West “Urbanisation and Youth Employment in Tan- African Countries.” Journal of African Economies zania: Preliminary Analysis.” Background paper, 22 (1): 73–111. doi: http://dx.doi.org/10.1093/ World Bank, Washington, DC. jae/ejs017. Dorosh, Paul, and James Thurlow. 2013. “Agri- Petesch, P., and Rodríguez Caillava, I. 2012. “Voices culture and Small Towns in Africa.” Agricul- of Young Villagers in Sub-Saharan Africa.” Back- tural Economics 44 (4-5): 449–59. doi: 10.1111/ ground paper, World Bank, Washington, DC. agec.12027. Ruiz Abril, Maria Elena. 2008. “Liberia: Girls Vul- Duflo, Esther, Pascaline Dupas, Michael Kremer, nerability Assessment.” Government of Liberia, and Samuel Sinei. 2006. “Education and HIV/ Nike Foundation, and World Bank, Washington, AIDS Prevention: Evidence from a Randomized DC. https://docs.google.com/document/d/1FvEg Evaluation in Western Kenya.” Policy Research QNMsZHVfMLYxfp62DUly8nY1850lh2v Working Paper 4024, World Bank, Washington, GmiYHk_0/edit?hl=en_US&pli=1. DC. Sander, Cerstin, and Samuel Munzele Maimbo. Eriksen, Michael, Judith Mackay, and Hana Ross. 2003. “Migrant Labor Remittances in Africa: 2012. The Tobacco Atlas. 4th ed. Atlanta: Ameri- Reducing Obstacles to Developmental Contribu- can Cancer Society. tions.” Africa Region Working Paper 64, World Falco, Paolo, William F. Maloney, Bob Rijkers, Bank, Washington, DC. and Mauricio Sarrias. 2012. “Heterogeneity in Simkins, Charles. 2013. “Urbanization in Africa Subjective Well-Being: An Application to Occu- and Its Relation to the Demographic Dividend.” pational Allocation in Africa.” Policy Research Background paper for the Africa Regional Study Working Paper 6244, World Bank, Washington, on the Demographic Dividend, World Bank, DC. doi: 10.1596/1813-9450-6244. Washington, DC. Fox, Louise, Katie Kibbuka, Jorge Huerta Munoz, Sneeringer, Stacy E. 2009. “Fertility Transition in and Thomas Pave Sohnesen. 2011. “Small Is Big: Sub-Saharan Africa: A Comparative Analysis Development of the Household Enterprise Sec- of Trends in 30 Countries.” DHS Comparative tor in Ghana; a Quantitative Look at Household Reports 23, ICR Macro, Calverton, MD. Enterprises.” World Bank, Washington, DC. United Nations. 2007. “Age-Specific Fertility Rate.” Holla, Alaka, Sibomana Leonard, and Wema Wil- United Nations, Statistics Division, New York. son. 2010. “The Secret Lives of Economically http://data.un.org/Data.aspx?d=GenderStat& Empowered Women: Qualitative Evidence on f=inID%3A13. Sensitive Issues from Female Entrepreneurs in UNWomen. 2012. “Breaking the Cycle of Violence Dar es Salaam.” World Bank, Washington, DC. and HIV in Liberia.” http://www.unwomen.org/ ILO (International Labour Organization). 2012. en/news/stories/2012/7/breaking-the-cycle-of- Global Employment Trends for Women, 2012. violence-and-hiv-in-liberia/. Geneva: ILO. Uteng, Tanu Priya. 2011. “Gender and Mobility LISGIS (Liberia Institute of Statistics and Geo- in the Developing World.” Background paper Information Services). 2011. “Report on the for the World Development Report 2012: Gender Liberia Labor Force Survey 2010.” LISGIS, Equality and Development, World Bank, Wash- Monrovia. ington, DC. 66 Youth EmploYmEnt in Sub-Saharan africa Warren, Charles W., Samira Asma, Juliette Lee, ———. 2006. World Development Report 2007: Veronica Lea, and Judith Mackay. 2009. Global Development and the Next Generation. New York: Tobacco Surveillance System: The GTSS Atlas. Oxford University Press. Atlanta, GA: CDC Foundation. ———. 2008. World Development Report 2009: Wietzke, Frank-Borge, and Catriona McLeod. 2012. Reshaping Economic Geography. New York: “Jobs, Well-Being, and Social Cohesion: Evidence Oxford University Press. from Value and Perception Surveys.” Background ———. 2011. World Development Report 2012: paper for World Development Report 2013, World Gender Equality and Development. New York: Bank, Washington, DC. https://openknowledge. Oxford University Press. worldbank.org/bitstream/handle/10986/12138/ WDR2013_bp_Jobs_Wellbeing_SocialCohesion. ———. 2012a. “Kenya at Work: Energizing the pdf?sequence=1. Economy and Creating Jobs.” Kenya Economic Update 7 (December). World Bank. 2003. World Development Report 2004: Making Services Work for Poor People. New York: ———. 2012b. World Development Report 2013: Oxford University Press. Jobs. New York: Oxford University Press. Chapter 3 Skills for Productive Employment A critical part of the policy agenda for youth the skills demanded by employers, and to build employment in Africa is to strengthen human the evidence needed to guide specific programs capital. Although training alone will not deliver to improve skills for youth employment. Govern- more and better jobs for youth—much also ments must also bear in mind that a dynamic depends on policies to strengthen the business private market exists for training and avoid environment, as discussed elsewhere in this introducing undue distortions in this market. report—the education that young people receive Priority areas for government intervention are to and the skills they acquire can expand the spec- provide information and facilitate access to exist- trum of employment opportunities they can ing training for disadvantaged youths (such as access and the earnings they are likely to com- young women or the poorest) as well as to ensure mand. Skills strongly influence where people the availability of better-quality training options. work and how much they earn. A key problem is In the presence of active training markets, public that across Africa rapid increases in school par- interventions need to be selective, performance ticipation and educational attainment have come driven, and evidence based. at the cost of quality, contributing to a serious shortfall in the skills for productive employment. Low educational achievement and limited Those skills include basic and higher-order cogni- skills contribute to difficult transitions into tive skills, behavioral and socioemotional skills, work and limited employment mobility among technical or vocational skills, and business skills. African youth. Education and skills open path- To facilitate entry, improve productivity, and ways into productive employment. Schooling raise earnings across the range of employment is a good predictor of an individual’s eventual in agriculture, household enterprises (HEs), and occupation, but the quality of education also the modern wage sector, the most pressing pri- matters for productivity. In Africa more chil- ority is to increase the quality of schooling and dren attend school than ever before. Yet the ensure that it delivers actual learning and skills. generally poor quality of the education they Other important priorities are to identify and receive means that schooling has relatively small directly build the socioemotional and behavioral effects on productivity, earnings, and poverty skills that contribute to productivity, including reduction. Rapid improvements in the quality 67 68 Youth EmploYmEnt in Sub-Saharan africa Figure 3.1 Primary school completion rates have risen substantially in Sub-Saharan of learning provided in school are critical Africa, 1990–2011 to increase the productivity and earnings of Africa’s youth. Guinea-Bissau Mali Guinea Schooling, Educational Chad Niger Attainment, and Work Burkina Faso Measured against the objective set by the Mil- Benin lennium Development Goal for education— Ethiopia “ensure that all boys and girls complete a full Mozambique course of primary schooling”—Sub-Saharan Malawi Africa has made remarkable progress. Across Mauritania the region, the share of children completing Central African Republic primary school rose from 51 percent in 1990 Angola to 70 percent in 2011. Some countries made Madagascar exceptional progress (figure 3.1). Burkina Faso, Togo Chad, Ethiopia, Madagascar, Malawi, Mau- Côte d’Ivoire ritania, Mozambique, and Niger more than Gambia, The doubled their primary completion rates; Benin, Burundi Guinea, Guinea-Bissau, and Mali more than Senegal tripled theirs. These achievements rival those Uganda of countries with the best historical increases, Rwanda such as the Republic of Korea, and far exceed Congo, Dem. Rep. those typical of most developing countries Tanzania since 1960 (Clemens 2004). An average young Equatorial Guinea Ghanaian or Zambian today has more school- Cameroon ing than an average French or Italian person in 1960 (Pritchett 2013). Cape Verde Nevertheless, these improvements started Lesotho from a very low base, and Sub-Saharan Africa Congo, Rep. lags behind other world regions in primary Ghana school completion. The young people enter- Swaziland ing Africa’s labor force right now have more Namibia schooling than any previous generation, but South Africa they still have little overall schooling. Nearly Zambia 60 percent of 15- to 24-year-olds have com- Kenya pleted only primary school (figure 3.2). Among Botswana the group that is most likely to have exited school for good—those who are 25–34 years Sub-Saharan Africa old—30 percent have no education, 21 percent South Asia have completed only some primary schooling, Middle East and North Africa and 14 percent have completed only primary Latin America and the Caribbean education. About 35 percent have continued East Asia and Pacific beyond primary school, and 19 percent have gone beyond lower-secondary school. Edu- 0 20 40 60 80 100 cational attainment is higher in urban areas, Percent although more than 40 percent of 25- to 1990 2011 34-year-olds in those areas have not completed Source: World Bank various years. lower-secondary school. Skills for productive Employment 69 Figure 3.2 Africa’s young people have more education than ever before, but average education attainment is still low a. Ages 15–24 b. Ages 25–34 100 100 80 80 60 60 Percent Percent 40 40 20 20 0 0 All Female Male Rural Urban All Female Male Rural Urban No education Primary incomplete Primary complete Secondary incomplete Secondary complete Higher education Source: Based on demographic and health surveys in 30 countries (see appendix). Figure 3.3 Educational attainment in Sub-Saharan Africa is projected to rise substantially a. Male b. Female 100 100 80 80 60 60 Percent Percent 40 40 20 20 0 0 2010 2015 2020 2025 2030 2010 2015 2020 2025 2030 No education Primary education Secondary education Tertiary education Source: Based on International Institute for Applied Systems Analysis data. Younger generations are likely to do better. ing has come at the cost of learning outcomes, Thanks to recent progress in getting children as discussed later. into school, incoming cohorts should have sub- stantially higher levels of education than past Schooling and Sector of Employment cohorts—in fact, they will be the most highly The links between educational attainment and educated cohort ever in Africa. By 2020, more sector of employment are very clear. Most peo- than half of men ages 15–24 are projected to ple who never finish primary school work in have attained secondary education; women agriculture. Those with a primary or lower-sec- will cross that threshold in 2025 (figure 3.3). Of ondary education work in nonfarm household course these projections depend on policies to enterprises (HEs), whereas those with higher come, and they are only a rough guide to actual levels of education are more likely to enter the outcomes. Moreover, greater access to school- modern wage sector. 70 Youth EmploYmEnt in Sub-Saharan africa Figure 3.4 Schooling maps to sector of employment a. Ages 15–24 b. Ages 25–34 100 100 80 80 60 60 Percent Percent 40 40 20 20 0 0 Agriculture Nonfarm Wage Agriculture Nonfarm Wage household household enterprise enterprise No education Primary incomplete Primary complete Lower secondary complete Upper secondary complete Post-secondary Source: Based on standardized and harmonized household and labor force surveys (see appendix). Since schooling strongly influences the sector of youths and young adults who work in HEs is in which people work, the educational profile of only somewhat better, but about half have not wage earners is very different from that of work- completed even primary schooling. ers in the agricultural and HE sectors in Africa. Completion of primary or additional edu- The great majority of youths and young adults cation becomes the norm only among young working in agriculture never completed pri- adults in the wage sector. Almost 65 percent of mary school. Close to 80 percent of those ages 15- to 24-year-olds who work for wages have 25–34 who work in agriculture have primary completed primary school. Because youths who schooling or less; 40 percent have no education transition out of school later are more likely to at all (figure 3.4). The educational attainment enter the wage sector, the share of young adults (25–34) who finished primary school and are employed in that sector increases to over 80 Figure 3.5 Wage workers with more schooling are more likely to work under percent. Wage employees are significantly more contract likely to have a secondary or a tertiary education: 100 among 25- to 34-year-olds, wage employees are more than twice as likely as the self-employed to 80 have completed secondary schooling. Wage workers with more schooling are also 60 more likely to work under contract—an indi- Percent cator that they perform more formal modern 40 wage work (figure 3.5). Close to 70 percent of wage employees without primary schooling 20 lack a contract, compared to less than 40 per- cent of those with upper-secondary schooling. 0 The relationships between schooling and No education Primary Primary Secondary Secondary Post- work in each sector are broadly similar across incomplete complete incomplete complete secondary various subgroups of the population, with two No contract Contract main distinctions. First, women are substan- Source: Based on standardized and harmonized household and labor force surveys (see appendix). tially more likely to work in the nonfarm HE Skills for productive Employment 71 Figure 3.6 Relationships between schooling and work vary depending on gender and urbanization a. By gender b. By urban/rural 70 70 60 60 50 50 40 40 Percent Percent 30 30 20 20 10 10 0 0 No Primary Primary Low Upper Post- No Primary Primary Low Upper Post- education incomplete complete secondary secondary secondary education incomplete complete secondary secondary secondary complete complete complete complete Female wage Rural wage Male wage Urban wage Male nonfarm household enterprise Rural nonfarm household enterprise Female nonfarm household enterprise Urban nonfarm household enterprise Source: Based on standardized and harmonized household and labor force surveys (see appendix). sector as their educational attainment rises, modern wage sector cannot absorb the incom- suggesting that education facilitates women’s ing cohorts of youth. The majority of young transition out of agriculture (figure 3.6). Sec- people will still create their own livelihoods. ond, participation in the wage sector begins at As a result, a growing share of workers in agri- lower levels of educational attainment in urban culture and nonfarm HEs will probably have areas, reflecting the larger pool of educated higher levels of schooling. youth in urban areas as well as the greater avail- Shifts in employment patterns will not occur ability of jobs for relatively unskilled workers in automatically. Additional years of schooling “Education the modern wage sector. that impart actual competencies will be needed for young people to develop the skills that can guarantees a Will Workers’ Increasing Educational promote such a shift. Research has suggested better job and a Attainment Change Employment that higher levels of educational achievement stable life.” Patterns? and cognitive skills are associated with overall Togo There is potential for employment patterns to economic growth (Hanushek and Woessmann shift, if current patterns in educational attain- 2012) as well as with the employment of a ment and sector of work persist in Africa and larger share of youth in modern wage jobs out- if educational attainment improves as pro- side of agriculture (Lee and Newhouse 2012). jected. Africa is positioned for a large shift of If increases in educational attainment are not workers out of the farm sector and into HEs, associated with a greater accumulation of skills, along with an increase in modern wage sector schooling will have a limited effect on overall employment. growth and composition of employment. Then Analysis of past trends in four countries there is the demand side of the labor market (Al-Samarrai and Bennell 2007) suggests that to consider. The economy will need to generate increasing shares of secondary school gradu- new employment opportunities to make use of ates are working in self-employment (Malawi, the learning and skills acquired by young peo- Tanzania, and Zimbabwe, for example), except ple. As discussed throughout this report, poli- where economic growth has led to the creation cies to address human capital are not sufficient of substantially more self-employment and in themselves to address the youth employment wage jobs (Uganda). The employment projec- challenge; improvements in the business envi- tions in chapter 1 suggest that even a growing ronment are needed as well. 72 Youth EmploYmEnt in Sub-Saharan africa “I have always Schooling Increases Productivity and farmers are more likely to adopt new agricul- been limited Earnings tural technologies first, either because they ini- Schooling is associated not only with the sec- tially have more information about the tech- by not being tor in which youth—and older workers—work, nology or because they are able to learn more schooled . . . . than less educated farmers exposed to the same but also with greater productivity and earnings. I don’t want Therefore, in addition to facilitating entry into information. Schooling enhances the capacity my children to nonfarm HEs and modern wage jobs, educa- to learn throughout life. The introduction of have the same tion is also essential to tackle the policy chal- new technologies is a learning opportunity experience lenge of increasing workers’ productivity across that reveals the complementarity between the spectrum of employment. schooling and the adoption of profitable new I had.” The education level of household members technology. Madagascar Outside of agriculture, in the HE and wage is a strong determinant of standards of living. Households where adults have attained higher sectors, the link between education and earn- levels of schooling have substantially higher ings is strong. Figure 3.7 shows that higher levels consumption levels per capita than households of education among HE owners are associated where members have less schooling. This dif- with higher earnings in Ghana, Rwanda, Tan- ference partly reflects the selection of workers zania, and Uganda. Figure 3.8 shows the aver- into sectors described previously. age wage increment associated with increasing Considerable research in the agricultural levels of education by comparing wage workers sector indicates that farmers with primary at each education level to uneducated workers schooling tend to have higher profits than in a sample of nine African countries. In both farmers without schooling, even for those figures, earnings differentials are adjusted for holding similar assets. Educated farmers gen- the age, gender, and rural or urban location of erally are the first to adopt new seeds, tillage the worker. The figures should not be under- practices, fertilizer, and animal breeds (see, for stood as indicating “returns to education.” They example, Welch 1970; Huffman 1977; Besley merely capture an association between educa- and Case 1993; Foster and Rosenzweig 1995; tion and earnings and may not fully control Abdulai and Huffman 2005). The benefit of for factors that confound the analysis (box 3.1 education is particularly pronounced in envi- discusses the challenges in estimating returns ronments undergoing rapid technical change to education). (Foster and Rosenzweig 2010). More educated Earnings Tend to Increase Little with Only a Few Years of Primary Schooling Figure 3.7 Education is associated with higher earnings in household enterprises A review of rates of return to education across several developing countries finds that the 180 mean rate of return to an additional year of 160 education is almost 9 percent (Banerjee and 140 Duflo 2005). In other words, six years of pri- Percentage increment 120 mary schooling yield an increment of almost 100 70 percent. Although figures 3.7 and 3.8 do 80 not necessarily represent estimates of rates of 60 return to education (box 3.1), they illustrate an 40 association between earnings and education that is substantially lower than the link docu- 20 mented in other developing countries. A few 0 years of basic education appear to yield much Ghana Rwanda Tanzania Uganda lower productivity gains in Africa than in other Primary incomplete Primary Lower secondary Upper secondary Post-secondary regions. Source: Based on standardized and harmonized household and labor force surveys (see appendix). For instance, owners of HEs who have not Note: Figure displays net average earnings (sales minus expenditures per month) for a sample of self- employed owners of HEs by level of education (relative to HE owners without education). completed primary education do not earn sig- Skills for productive Employment 73 nificantly more than owners of HEs without Figure 3.8 Education is associated with higher wages any education—suggesting that minimal pro- 250 ductivity gains are associated with a few years of primary school (figure 3.7).1 200 In contrast, HE owners who have com- Percentage increment 150 pleted primary education earn significantly more than HE owners who have no education. 100 The earnings increment associated with pri- mary school completion ranges from 25 per- 50 cent in Ghana to 46 percent in Rwanda. Low 0 earnings differentials at low levels of education Cameroon Côte Ghana Kenya Malawi Mozam- Rwanda Sierra Uganda are a serious issue, because most young people d’Ivoire bique Leone in the HE sector have not completed primary Primary incomplete Primary complete school.2 Lower secondary complete, upper secondary incomplete Upper secondary complete In the wage sector, workers with incom- Post-secondary plete primary schooling earn a modest pre- Source: Based on standardized and harmonized household and labor force surveys (see appendix). mium (10–25 percent on average) compared Note: Based on a regression of hourly wages (in logs) by education level, adjusted for age, gender, and whether the workplace is urban or rural. Figure displays hourly wages among wage workers by level to workers with no education, and often the of education (relative to workers without education). The dependent variable is ln (wage/hours) using difference is not even statistically significant information from the last seven days. Wages were adjusted for local consumer price index (2005 = 100) and purchasing power parity. This dependent variable was run against the dummy level of educa- (Ghana, Mozambique, and Sierra Leone).3 In tion variables listed above, and the “no education” dummy was excluded. As the mean wage among contrast, wage workers who have completed those with no education is normalized to zero, the wages shown are not absolute but are depicted in comparison to workers with no education. Workers are ages 20–64. primary education earn consistently more than uneducated wage workers. The wages of primary school completers typically are more than 20 percent higher than the wages of work- analysis of seven cities in West Africa found ers without education; the gap reaches as much evidence of strong associations between edu- as 52 percent in Cameroon. Wage workers with cation and earnings in HEs as well as evidence upper-secondary education typically earn of an increase in the earnings premium across wages that are 100–150 percent higher than education levels (figure 3.9; Kuepie, Nordman, those without education. and Roubaud 2009). In particular, the study found that marginal returns for an additional Earnings Increase Faster with Education year of postprimary education are higher than beyond Primary School marginal returns associated with an additional In contrast with the limited earnings gains asso- year of primary education. ciated with a few years of basic education, earn- As seen in figure 3.8, individuals in the wage ings and wages increase faster when education sector who have completed upper-secondary extends beyond primary school. For owners of school earn 30–155 percent more than individ- HEs, the increment in earnings associated with uals without any schooling. Postsecondary edu- completing lower-secondary education ranges cation is associated with a premium of 70–200 from around 35 percent in Ghana and Tanzania percent over no schooling. As in the HE sec- to around 60 percent in Rwanda and Uganda. tor, the increment in earnings is substantially Upper-secondary and postsecondary education higher at higher levels of education (Teal 2010; are associated with even higher increments to Söderbom, Teal, and Harding 2006; Bigsten et earnings, although few individuals in the HE al. 2000). Each additional year of schooling in sector acquire that much education.4 Kenya, for example, is estimated to increase These descriptive patterns are consistent earnings by 3 percent among those with only with more robust econometric evidence tack- 1–7 years of schooling, 9 percent among those ling the selection issues involved in estimating with 8–11 years, and 24 percent among those returns to education (Fasih et al. 2012), par- with 12 years or more (Söderbom, Teal, and ticularly in Africa and in the HE sector.5 An Harding 2006). A review of studies concludes 74 Youth EmploYmEnt in Sub-Saharan africa Box 3.1 Estimating the returns to education When the more educated differ inherently from the less supply-side factors have influenced levels of schooling inde- educated, the relationship observed between earnings pendently of confounding factors such as ability. A surpris- and education levels ceases to be a reliable guide to the ing finding is that the estimates of the returns to school- causal impact of schooling on earnings and productivity. ing from such natural experiments are not much lower than For instance, when the more educated are of higher-than- unadjusted estimates. Although quasi-experimental meth- average innate ability, the earnings gap observed between ods relying on “natural experiments” or other econometric individuals with high and low levels of education partly techniques cannot eliminate all bias resulting from omitted reflects the higher-than-average ability of the more edu- variables, the typically high estimates emerging from these cated group. An individual of average ability and an indi- careful studies make it much harder to reject the idea that vidual of above-average ability should not expect the same schooling leads to substantial earnings gains. outcome from the same education. If those who are more Developed countries have produced a vast body of evi- educated are also more likely to come from advantaged dence on the adjusted returns to schooling. The studies use a backgrounds, then the measured returns to education range of econometric techniques to correct for the problems would be overstated, since they include returns to family just mentioned. Estimates from studies in developing coun- background. In the wage sector, they can also reflect con- tries tend to be higher than those from developed coun- nections in labor markets, particularly because job referrals tries, but developing countries rarely provide opportunities from family members appear to be particularly important in to use panel data or natural experiments to identify causal developing countries. effects. One of the best-cited examples concerns the impact Concerns about bias arising from omitted factors or of a program to build schools in Indonesia (Duflo 2001). The selection effects have motivated an entire field of research study benefited from the fact that the program, launched focused on correctly measuring the causal impact of educa- in 1974, differentially increased school opportunities among tion on earnings. One approach is to adjust the estimates children from different cohorts and different regions. Look- for omitted factors by including their proxy measures in a ing at wages in 1995, the study found returns to education regression analysis (for example, by using measures of cogni- in the range of 6.8–10.6 percent, approximately similar to tive skills to proxy for innate ability). However, this approach returns found in previous studies. A recent review of rates of may not correct for all omitted factors. For that reason, some return to education across multiple countries found that the of the best-regarded studies have chosen to take advantage mean rate of return to an additional year of education was of “natural experiments”—instances in which policies or almost 9 percent (Banerjee and Duflo 2005). that while the wage gains associated with an often accompanied by a decline in the earnings additional year of secondary or higher edu- increment associated with being a primary or cation are in the 10–15 percent range, gains secondary school graduate. In India in 1994, associated with an additional year of primary for example, the wage increment for a primary schooling are only 3–10 percent (Bigsten et school graduate versus someone with no edu- al. 2000; Schultz 2004). Here again, patterns cation was 50 percent, but by 2010 it had fallen are consistent with more robust economet- to 30 percent. The wage increment for a univer- ric results of increasing returns to education, sity graduate versus an upper-secondary gradu- including evidence from urban West Africa ate was 10 percent in 1994, but had risen to 36 (figure 3.9). percent by 2010. Similar changes occurred in other South Asian countries (Nayar et al. 2012). Returns to Education Change with In Latin America between the mid-1990s and Educational Attainment the late 2000s, the wage increment associated Returns to education change over time as the with completing secondary relative to primary relative shares of the population with no, pri- schooling declined systematically—for exam- mary, and secondary schooling change. In ple, from around 40 percent to about 30 per- other regions of the world, these shifts were cent in Brazil and Peru—while the increment Skills for productive Employment 75 Figure 3.9 The convex relationship between earnings and education: Schooling and earnings in urban West Africa a. Household enterprises b. Private wage employment 1,200 2,500 1,000 2,000 Hourly earnings (CFA) Hourly earnings (CFA) 800 1,500 600 1,000 400 500 200 0 0 0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18 Years of schooling Years of schooling Lomé Ouagadougou Cotonou Niamey Bamako Abidjan Dakar Source: Kuepie, Nordman, and Roubaud 2009, based on 1-2-3 surveys on employment and earnings in urban West Africa. Reproduced with permission of the authors. Note: In the study, household enterprises are defined as “production units with no fiscal or statistical identity or without any formal accountancy” and labeled as “informal sector” by the authors. We use the term “household enterprises” for consistency with terminology used in this report. associated with being a tertiary graduate rose emotional and behavioral skills. These short- (Aedo and Walker 2012). falls can prevent youth from reaching their All else being equal, the rising share of full potential. Cognitive, socioemotional, and Africa’s population with completed primary behavioral skills create the foundation for schooling should put downward pressure on acquiring higher-order cognitive and technical the earnings increment to schooling. However, skills, whether through more formal education, all else is not necessarily equal. The returns to training, or on-the-job learning. Improvements schooling depend on the quality of schooling in the quality of basic education are urgently and the skills it supplies, as discussed in the required to ensure that incoming youth acquire remainder of this chapter. Returns also depend the necessary foundational skills. on the demand side of the labor market. The Schooling aims to develop an assortment of productivity of the work done by individuals skills, including literacy and numeracy as well with a given level of schooling, in the particu- as higher-order cognitive, socioemotional, and lar economic environment where they put their behavioral skills. The skills shaped by the edu- skills to work, will determine how effectively cation system explain part of the gains in earn- their schooling turns into earnings. Chapters 4, ings that are associated with more schooling. 5, and 6 elaborate on how to improve employ- The fact that a few years of education confer ment opportunities and the business environ- only a small increase in earnings is related to ment on the demand side of the labor market. the low quality of basic education in Africa as well as other economywide factors beyond the education system. Moreover, the pattern Building a Foundation: Cognitive, of increasing marginal returns to education Socioemotional, and Behavioral in Africa suggests that productivity increases Skills faster at higher levels of education—consistent with the notion that skills can complement one Many types of skills are needed for produc- another and that “skills produced at one stage tive employment. Weak education systems are raise the productivity of investment at subse- contributing to a critical shortfall in two key quent stages” (Cunha and Heckman 2007). sets of skills—cognitive skills as well as socio- The fundamental issue is that many African 76 Youth EmploYmEnt in Sub-Saharan africa children never acquire the skills that are the requirement for professors and scientists, but foundation for a productive life because of the less so for entry-level or “semi-skilled” workers. low quality of basic education systems. Technical and vocational skills matter in tech- This section discusses how skills are built nical occupations. Other characteristics of per- through the education system, in early child- sonality, such as “openness to experience” and hood, and through other avenues for learning. “agreeableness,” have been shown to matter for Productivity-enhancing skills can be catego- particular career paths (Cobb-Clark and Tan rized as follows:6 2010). Still, the first step is a basic education of sufficient quality that lays the foundations • Basic cognitive skills, such as numeracy and for young people to acquire a range of relevant literacy skills later in life—through additional formal • Higher-order cognitive skills, such as prob- education, training, on-the-job learning, or lem solving and critical analysis other means. • Behavioral and socioemotional skills (also called soft skills or life skills), includ- Education Systems in Africa Are Failing ing a broad range of skills, such as social to Produce Critical Foundational Skills skills, self-regulation, self-confidence, and Rapid increases in school participation and conscientiousness educational attainment have undoubtedly • Technical or vocational skills, often specific come at the cost of quality. Empirical evidence to each occupation confirms that poor learning outcomes—a few • Business skills, such as entrepreneurship years of low-quality education, producing indi- skills, managerial skills, and financial viduals who are barely literate and numerate— literacy. are hampering the potential for education to increase productivity, even in African econo- The track record of the education system in mies that have been performing well. Learning producing these skills, including basic cognitive assessments in Africa show that most primary skills, has been abysmal in Africa. If the quality students still lack basic proficiency in reading of education does not improve rapidly, produc- at the end of second or third grade. In several tivity and earnings are likely to remain low—a countries, a very large proportion of primary problem that will only become more acute as school students are illiterate. For example, new generations of Africans enter and graduate more than 80 percent of Malian third-graders from school in larger numbers. Without sub- and more than 70 percent of Ugandan third- stantial improvements in quality, the rewards graders cannot read a single word (figure 3.10). that graduates of primary and lower-secondary Household surveys that measure numeracy school can expect to reap in the labor market will and literacy are consistent with these trou- surely fall as more graduates enter the workforce. bling results. In Tanzania, for example, a 2011 To some extent, such outcomes can be mitigated assessment of children’s abilities revealed that by urgent improvements in the education sys- 70 percent of students complete standard two tem that enable young people to develop more without meeting the numeracy standards of productivity-enhancing skills. Of course, educa- that level (Uwezo Tanzania 2011). Assessments tion is only a starting point; graduates will also in Kenya and Uganda revealed similar shortfalls need an economic and business environment in students’ cognitive skills. where they can use their skills productively. Even children who complete primary school The relevance of the range of productivity- have low levels of basic skills. In the regionally enhancing skills can vary by sector of employ- benchmarked Southern Africa Consortium for ment as well as across occupations. For Measuring Educational Quality (SACMEQ) in instance, the mix of skills required for pro- 2007, 43 percent of sixth-graders in Tanzania ductivity in wage employment will depend on and 74 percent in Mozambique did not get the complexity of the job. Higher-order cog- beyond the “basic numeracy” level, while 44 nitive skills matter more when jobs are more percent in Mozambique could not “read for complex. Analytical thinking is an absolute meaning” (figure 3.11). The results from an Skills for productive Employment 77 Figure 3.10 The ability to read in early grades is alarmingly low a. Students who cannot read a single b. Students who cannot read at least word of a simple paragraph 45 words per minute 90 90 80 80 70 70 60 60 Percent Percent 50 50 40 40 30 30 20 20 10 10 0 0 Ghana Kenya Liberia Mali Senegal Senegal The Uganda Ghana Liberia Senegal The Gambia Gambia Grade 3 End grade 2 End grade 3 Grade 5 Grade 6 Sources: Cloutier, Reinstadtler, and Beltran 2011; Gove and Cvelich 2010. Note: These findings refer only to students who have stayed in school. Those who have dropped out—who would have scored even lower on these assessments—are excluded from these calculations. The many primary school dropouts who lack the most basic skills, including simple numeracy or literacy, are largely young people who will find work in agriculture and HEs. Figure 3.11 By the end of primary school, many students have not mastered even basic competencies: 2007 SACMEQ results for math and reading proficiency a. Math b. Reading All All Zambia Malawi Malawi Zambia Namibia Lesotho Lesotho South Africa Uganda Uganda Mozambique Mozambique Zanzibar Namibia South Africa Zimbabwe Zimbabwe Botswana Botswana Seychelles Swaziland Zanzibar Tanzania Mauritius Seychelles Kenya Kenya Tanzania Mauritius Swaziland 100 80 60 40 20 0 20 40 60 80 100 100 80 60 40 20 0 20 40 60 80 100 Percent Percent Pre-, emergent, and basic numeracy (levels 1, 2, and 3) Pre-, emergent, and basic reading (levels 1, 2, and 3) Beginning numeracy (level 4) Reading for meaning (level 4) Competent and above (levels 5, 6, 7, and 8) Interpretive and above (levels 5, 6, 7, and 8) Source: Hungi et al. 2010. earlier round of SACMEQ (in 2003) were gen- est round (in 2011) of globally benchmarked erally similar. learning assessments, the Trends in Math and Beyond primary school, learning outcomes Science Study (TIMSS). Among the eighth- remain a cause for concern. Botswana, Ghana, grade (Ghana) and ninth-grade (Botswana and South Africa all participated in the lat- and South Africa) students tested, 79 percent 78 Youth EmploYmEnt in Sub-Saharan africa Figure 3.12 Secondary school students in Sub-Saharan especially given the grade in which this test is Africa perform poorly on internationally comparable administered. (The international mean was assessments: TIMSS results for math proficiency, 2011 25 percent; the result was 67 percent for Indo- nesian students and 45 percent for Jordanian Korea, Rep. students.) International average School-age children are building skills too Malaysia slowly. The amount of learning that takes place in schools from one year to the next is very Chile low. Household surveys that include the same Jordan measures of skills for all children, regardless Botswana of the grade they are attending, illustrate this phenomenon. Learning trajectories demon- Indonesia strate just how slowly school-age children are South Africa acquiring skills. A study from Guinea-Bissau Ghana that mapped performance in basic literacy and numeracy by children’s ages (figure 3.13) 100 80 60 40 20 0 20 40 60 80 100 shows that only half of 8-year-olds recognize % of students single digits, while less than half of 9-year-olds < 400 (low) 400–625 (intermediate high) recognize letters of the alphabet (Boone et al. > 625 (advanced) 2013). Literacy and numeracy performance Source: Mullis et al. 2012. improves only very slowly with age. An average Note: Students tested in grade eight in Ghana and grade nine in 16-year-old still cannot read a paragraph and Botswana and South Africa. cannot subtract a single-digit number from a two-digit number. These low learning trajecto- of Ghanaians and 76 percent of South Afri- ries are widespread: 351 schools were visited as cans did not surpass the lowest benchmarked part of the Guinea-Bissau study, but in only 6 level of mathematics proficiency (figure 3.12).7 of them could an average student read a para- In other words, all of those students failed to graph. In none of the schools could an average meet the proficiency criterion for this level— student multiply or divide. for example, “students have some knowledge Learning trajectories are flat not only for of whole numbers and decimals, operations, basic literacy or numeracy but for a range of and basic graphs”—which is a low threshold, other cognitive skills, such as language, math- Figure 3.13 Learning trajectories by age are flat in Guinea-Bissau a. Literacy b. Numeracy 8 An average An average 12 An average An average An average 14-year-old sixteen-year-old An average 14-year-old can sixteen-year-old 7 11-year-old can recognize can recognize and 10 11-year-old add two single can’t subtract a 6 can recognize words, but comprehend words, can’t add digits but cannot single digit from Numeracy test score letters but not has trouble but cannot read a two single a two-digit Literacy test score 8 add two-digit 5 words comprehending paragraph digits numbers to a number them single digit 4 Less than 6 Only half of eight-year- number half of nine- 3 year-olds can olds can 4 recognize letters recognize 2 of the alphabet single digits 2 1 0 0 7 8 9 10 11 12 13 14 15 16 17 7 8 9 10 11 12 13 14 15 16 17 Age (years) Age (years) Source: Boone et al. 2013. Reproduced with permission of the authors. Skills for productive Employment 79 Figure 3.14 Learning trajectories by grade are flat: Performance on test scores in selected African countries, by educational achievement a. Ghana b. Malawi Normalized test score (standard deviations) Normalized test score (standard deviations) 2.0 2.0 1.5 1.5 1.0 1.0 0.5 0.5 0.0 0.0 –0.5 –0.5 –1.0 –1.0 –1.5 –1.5 None Some Primary Lower Upper Tertiary None Primary Lower Upper primary secondary secondary secondary secondary 25th percentile 50th percentile 75th percentile c. South Africa 2.0 Normalized test score (standard deviations) 1.5 1.0 0.5 0.0 –0.5 –1.0 –1.5 5 6 7 8 9 10 11 12 Grade Level 2/3 Level 3/4 (anchored to grade 8) Sources: Calculations from Ghana 2004 living standards survey, Malawi Zomba Pilot control group, and South Africa 2008 national income dynamics study survey. Note: Standardized score tests are different and cannot be compared across countries. In South Africa, different tests were administered at different grades; scores on the two tests have been anchored at grade eight. ematics, and problem solving. Figure 3.14 another analysis using TIMMS data to infer illustrates how problem-solving skills are asso- learning trajectories, which shows that they ciated with education in samples from Ghana, are very flat in Ghana (Pritchett 2013). The Malawi, and South Africa. Learning trajecto- learning trajectories for problem-solving skills ries from Ghana show that the median level of in Malawi or South Africa are likewise very flat problem-solving skills is higher for individuals (figure 3.14). Overall, students at each level who have completed more schooling. Never- vary greatly in their problem-solving skills, but theless, the increases are very slow. The poorest- across levels of schooling, the accumulation of performing students in lower-secondary skills varies little around a generally low level. school (the twenty-fifth percentile) perform Africa’s schools are not effectively imparting roughly the same as the median performer basic numeracy or literacy skills, nor are they among those with incomplete primary school. imparting other cognitive skills such as prob- This finding is consistent with the results of lem solving or critical analysis. 80 Youth EmploYmEnt in Sub-Saharan africa Poor Cognitive Skills Contribute to Poor learning a day.10 Reforming the accountability Employment Outcomes framework that allows such poor performance Because returns to schooling depend consider- to persist is key (World Bank 2003; Bruns, ably on the quality of education (Card 1999), Filmer, and Patrinos 2011). Better informa- returns to basic education will remain low as tion on performance must be complemented long as new cohorts of young people enter the by targeted approaches that increase oversight labor force with more schooling but limited by the people who are most affected: students skills. Studies seeking to disentangle the returns and their parents. Steps to ensure that teachers to cognitive skills from other effects that school- are well prepared for teaching and supported ing might have on earnings have demonstrated in their tasks are critical for creating a cadre of a strong association—including in Africa.8 For high-performing professionals. What teachers example, one careful study of wage earners in know and what they do is the cornerstone of Ghana found that cognitive skills have signifi- good teaching and learning. The rise of private cant positive effects on wages. An earlier study schools in Africa—schools that deliver supe- of urban wage earners in Kenya and Tanzania rior performance often at lower costs—should produced similar findings (Glewwe 2002). not be stifled; rather they should be encour- Recent evidence from Kenya documents the aged and channeled to give more students the chain of causation from schooling to cogni- opportunity to learn. Such strategies would tive skills to employment outcomes.9 Second- need to be undertaken in tandem with efforts ary school graduates were shown to perform to ensure that children are prepared for learn- significantly better than comparable non- ing when they attend school. graduates on vocabulary and reasoning tests in adulthood. Test scores were higher by 0.6 Children Acquire Limited Cognitive and standard deviation, which is a large effect. By Socioemotional Skills before Entering their mid-20s, school completers had a roughly School 50 percent lower probability of low-skilled There are strong signs of cognitive delays self-employment compared to noncompleters. among African children even before they enter They were about 30 percentage points more school. Data from the Democratic Republic of likely to be in formal employment. Congo, Nigeria, Sierra Leone, and Togo show How can African countries overcome the that young children acquire cognitive skills problem of limited accumulation of cognitive more slowly than expected (figure 3.15). Signs skills? For those who have already dropped of delays are already apparent at age 36 months out, especially younger workers, one option is and get worse as children grow older (Naudeau to expand access to second-chance education. et al. 2010). Some children acquire skills faster Second-chance education is an option to assist than others, however. For instance, the accu- the large stock of young workers who have mulation of cognitive skills in early childhood already dropped out of school, but it is poten- is associated with household socioeconomic tially expensive and has had mixed success (see status. Children from poorer households score boxes 3.2 and 3.3). worse on measures of cognitive skills such as For current and future schoolchildren to receptive language or memory than children in become productively employed, however, an better-off households. In addition, the gap in immediate priority is to address the lack of cognitive skills between the poor and nonpoor learning in basic education. Improving the widens with age. These results are consistent quality of basic education will not be easy. Sur- across the region, including detailed studies in veys of schools reveal substantial failures in ser- Madagascar and Mozambique as well as other vice delivery. For example, absenteeism among parts of the world.11 Delays in early childhood teachers is on the order of 16 to 20 percent on development can affect children throughout a given day in Kenya, Senegal, and Tanzania; their lives, limit school readiness, and contrib- primary school students in those countries ute to the slow accumulation of skills in school experience only about two to three hours of (Shonkoff and Phillips 2000). Skills for productive Employment 81 Box 3.2 A second chance at education for African youth Many young people in Africa have not completed basic educa- Accelerated learning programs (ALPs) give children and tion and lack numeracy and literacy skills. In the school year youth an opportunity to catch up on missed education in ending in 2010, it is estimated that 23 percent (31 million) of a short period (Mattero 2010). They use intensive, flexible 132 million primary school–age children (5–14 years) in Sub- methods or schedules to complete the curriculum faster than Saharan Africa were out of school or not enrolled.a A large pro- in traditional education and help youth to reenter the for- portion of children never attend school (28 percent of young mal primary or secondary school system. Such programs are people ages 25–34 never started school). Many who manage common in countries where children’s schooling was inter- to start school drop out after just a few years (figure B3.2.1). rupted by armed conflict or other kinds of social upheaval: As learning achievement tests demonstrate, when the quality • Republic of South Sudan. ALPs provide basic education to of education is poor, even formal schooling is no guarantee those who missed the opportunity during the civil war. The that students will master basic literacy and numeracy skills. eight primary grades are reduced to four years of learning. The programs are part of a broader alternative education Figure B3.2.1 Proportion of 15- to 19-year-olds who have system that reaches more than 165,000 students (mostly completed each grade ages 12–18), roughly equal to the number of fourth-grade 100 students in primary schools nationwide (Government of South Sudan 2011). • Sierra Leone. The Complementary Rapid Education Pro- 80 gramme for Schools, implemented in three districts, reduced the traditional curriculum and introduced teach- Percentage 60 ing strategies, thereby shortening the six-year primary cycle to three years. 40 • Northern Ghana. The School for Life Program, described in box 3.3, was implemented to address a shortage of 20 teachers in areas where literacy among 12- to 18-year- olds was only 5 percent. The program recruited volunteer 0 teachers with some secondary education and gave them 0 2 4 6 8 10 modest incentives along with a short induction training to Grade teach out-of-school youth. Burkina Faso Senegal Ethiopia Nonformal education programs provide youth with instruc- Congo, Dem. Rep. Malawi Uganda tion equivalent to formal education, focusing on essential learning needs and basic skills such as literacy, oral expres- Source: Based on demographic and health survey data, except for the Demo- cratic Republic of Congo, which is from the Multiple Indicator Cluster Survey. sion, numeracy, and problem-solving skills (Delors 1996). Data are for 2010 or 2011 (http://econ.worldbank.org/projects/edattain). Most nonformal education courses range from a few months to years and can be offered on a part-time or full- Second-chance education programs have the potential to time basis. They are normally delivered face-to-face in formal bring school-age children back to school as well as to equip school facilities and learning centers but can also be pro- youth with the basic literacy and numeracy skills needed for vided through e-learning and radio. Community schools are productive employment. They enable individuals to complete a well-known example of nonformal education programs in general primary or secondary education, either by substitut- Africa. In Zambia, community schools enroll individuals who ing for formal education or by offering “bridges” to return to are 14 and older and are vulnerable, orphaned, or unable the formal education system (Mattero 2010). The main types to meet the costs associated with formal schooling. In Mali, of second-chance programs are (1) accelerated learning pro- community schools originally designed for adults have been grams, (2) nonformal education programs, and (3) education extended to rural children, offering primary school educa- equivalency programs. Although adult nonformal education tion up to sixth grade. In Eritrea, to ensure basic education as well as programs providing technical, vocational, and life for all (especially girls), a UNICEF-supported Complementary skills are sometimes referred to as second-chance education Elementary Education Program provided nonformal educa- programs, this report considers them separately. tion to children and young adults (ages 10–14 years old) (continued) 82 Youth EmploYmEnt in Sub-Saharan africa Box 3.2 (continued) who missed an opportunity to complete a five-year primary proficiency tests compared with students in public schools. school curriculum. Students who attended community schools in Mali outper- Equivalency degree programs are nonformal education formed their public school counterparts in school completion programs leading to qualifications equivalent to those gained examinations (DeStefano et al. 2006). through formal education programs. Equivalency degree Studies suggest that the costs per learner can be higher programs target primary or secondary school dropouts and in second-chance programs than in traditional public schools provide corresponding degrees, signaling that the recipient (table B3.2.1), but the limited information on the ultimate has demonstrated the ability to read, write, think, and com- outcomes of interest makes it hard to evaluate their cost- pute at the level for which the degree was offered (Boesel, effectiveness. Much more data are needed to understand Alsalam, and Smith 1998). Equivalency programs vary in how these programs can cost-effectively build skills and terms of admission, age, place, and pace, and they are deliv- increase youth productivity and earnings. ered either via face-to-face learning or distance education. Systematic evidence on the effectiveness of second- chance programs is quite limited and mixed. In Sierra Leone, Table B3.2.1 Costs per pupil in second-chance education programs and formal public schools students in ALPs performed consistently better than stu- dents in traditional schools, with a 91 percent passing rate Annual cost on primary school achievement tests. Students progressed Country and type of education per pupil (US$) to secondary schooling at similar rates (Nicholson 2006). For Ghana students in Ghana’s School for Life, the transition back into School for life 39 the formal system is close to 90 percent (DeStefano et al. public 27 2006). However, in the Republic of South Sudan the dropout rates for ALPs during the 2010–11 school year were 52 per- Mali cent for females compared to around 20 percent in formal community schools 47 schools (Government of South Sudan 2011). Results are bet- public 30 ter for males; only 5 percent dropped out of the ALP versus Zambia around 25 percent in formal schools. In Zambia students in community schools performed well on math and English community schools 39 public 67 a. UNESCO 2012. Source: DeStefano et al. 2006. Box 3.3 Northern Ghana’s School for Life second-chance education program School for Life (SfL) offers a nine-month literacy cycle in the children to maintain daily duties; and training takes place in mother tongue for children ages 8–14. The curriculum aims primary school classrooms after school, which provides a link to meet the first three years of the formal school system’s with formal education. requirements and transition participants into the formal sys- The program’s reach is broad and growing. Approxi- tem on graduation. Facilitators—volunteers selected by the mately 800 facilitators served 20,000 learners in 17 districts community—receive an initial three-week intensive training, in 2012–13, up from 10,000 in 9 districts the previous year. supplemented with follow-up workshops and training. Com- The vast majority complete the program—the graduation munities provide their own teachers or facilitators, who are rate was 97 percent in 2011–12—and most join the formal literate in the community’s language; the mother tongue is schooling system. As many as 90 percent of the 2011–12 both the language of literacy and the medium of instruction; cohort entered primary school on leaving SfL. the community is actively involved; the school calendar and hours are flexible and adapted to local conditions, allowing Sources: CREATE 2010; DFID 2012. Skills for productive Employment 83 Figure 3.15 Cognitive skills increase slowly, especially for the poorest a. Democratic Republic of Congo b. Nigeria 100 100 80 80 60 60 Percent Percent 40 40 20 20 0 0 35 40 45 50 55 60 35 40 45 50 55 60 Age (months) Age (months) c. Sierra Leone d. Togo 100 100 80 80 60 60 Percent Percent 40 40 20 20 0 0 35 40 45 50 55 60 35 40 45 50 55 60 Age (months) Age (months) Poorest 40% Richest 20% Source: Based on Multiple Indicator Cluster Survey 4 data. Note: Graphs show the proportion of respondents who can perform two of the following three tasks: identify or name at least 10 letters of the alphabet; read at least four simple, popular words; know the name and recognize the symbols of all numbers from 1 to 10. Multiple factors contribute to the slow acqui- to gains in school readiness and improvements sition of skills in early childhood. Nutrition is in a range of cognitive skills. Similar effects on one factor, but parenting practices and stimula- skills in early childhood have been found in tion also matter greatly. Poor nutrition in early other developing countries for interventions childhood can hinder the accumulation of focusing on improving parenting practices and cognitive skills12 and affect reading or problem- stimulation through home visits, community- solving skills in adulthood (Maluccio et al. based centers, or preschools.13 2009). Parenting practices and psychosocial Early childhood is also a critical window stimulation also shape cognitive and socioemo- for the development of socioemotional and tional skills. Programs that encourage stimula- behavioral skills, including skills involved in tion as well as nutrition have been shown to forming relationships, communication or be more effective than programs that only socialization, and self-regulation of behav- aim to improve nutrition. They increase skills iors and emotions (table 3.1; Lundberg and as well as earnings in adulthood (Grantham- Wuermli 2012; Tubbs and McCoy 2012; Aber McGregor et al. 2007; Gertler et al. 2013). In and Jones 1997). Stressful experiences such as Mozambique, a recent preschool program led conflicts or other shocks can trigger behavioral 84 Youth EmploYmEnt in Sub-Saharan africa Table 3.1 Developmental tasks, by age by the time individuals are 15 years old, Age Period of life Developmental task whereas many behavioral skills are acquired 0–1 Early childhood Establishing (secure) attachment relationships between the ages of 15 and 30 (figure 3.16). 1–3 Early childhood learning to explore and communicate Psychology has a long history of studying domains of human functioning. In practice, 3–5 Early childhood learning to self-regulate thoughts, behaviors, and emotions socioemotional and behavioral skills are often 6–12 middle childhood learning and reasoning, developing interpersonal taken to encompass a broad set of “soft skills” or and social problem-solving skills “life skills,” which have come to include a mix 13+ adolescence and emerging adulthood Establishing autonomy (renegotiating relationships), of skills and less well-defined individual char- forming identity, setting and achieving goals acteristics such as social skills, self-regulation, Source: Wuermli et al. 2012. goal-setting behavior, personal initiative, achievement orientation, proactivity, action problems and affect socioemotional devel- planning, deliberate practice, or self-confidence. opment through depression, anxiety, or Sometimes behavioral skills also include life self-esteem. Parenting practices and psycho- skills linked with avoiding risky behavior or social stimulation contribute to shaping these more general measures of psychological well- socioemotional skills—as shown recently in being, attitudes, or aspirations. They are some- Mozambique, where children participating times linked to personality characteristics such in a preschool program showed improved as the “big five”: extraversion, emotional stabil- emotional maturity, including self-regulation ity, agreeableness, conscientiousness, and open- (Martinez, Naudeau, and Pereira 2012). ness to experience. These traits can continue to change in adulthood (Roberts and Mroczek Socioemotional and Behavioral Skills 2008). Are a Neglected Part of the Skill Set for Socioemotional and behavioral skills such as Productivity self-regulation have been linked to educational Socioemotional and behavioral skills comple- achievement and those such as self-confidence, ment and extend cognitive skills to improve social skills, or emotional stability have been productivity, but they are often neglected. shown to affect employment outcomes (Heck- Socioemotional skills develop through early man, Stixrud, and Urzua 2006). In fact, em- and middle childhood. Social competen- ployer surveys and qualitative research from cies are consolidated through increasing interactions and socialization in school and in communities (Lundberg and Wuermli Figure 3.16 Behavioral skills take longer to develop than cognitive skills 2012). As children reach adolescence, they further hone their social skills, become 70 increasingly autonomous, and forge a sense of identity (see chapter 2). Depending on 60 % scoring at mean adult level the context in which they live, including the 50 social expectations surrounding them, ado- 40 lescents shape aspirations and set goals for themselves. Concurrently, they experience 30 deep socioemotional changes that can chal- 20 lenge their self-confidence or self-regulatory skills and make them vulnerable to depres- 10 sion, anxiety, or risky behaviors (Lundberg and 0 Wuermli 2012). Long after cognitive skills have 10–11 12–13 14–15 16–17 18–21 22–25 26–30 been acquired, socioemotional and behavioral Age (years) skills remain malleable, reflecting the many Cognitive skills Psychosocial maturity changes and transitions characteristic of ado- Source: Steinberg et al. 2009. Reproduced with permission of the lescence. Cognitive skills are largely acquired authors. Skills for productive Employment 85 various African countries suggest that employ- different behavioral skills are required for dif- ers are looking for soft skills in young hires. ferent occupations and sectors. New research In Botswana, employers report that they seek from the United States suggests that not all workers with skills such as “honesty, commit- behavioral skills matter in the same way across ment and hard work, reliability and punctual- different types of work or even across relatively ity, communication, and team working skills” unskilled tasks. For example, sociability (or (World Bank 2012a). In Lesotho, employers “extraversion”) has higher returns in the ser- rate soft skills (the “appropriate personal char- vice sector—specifically in sales jobs—than in acteristics”) as among the hardest to find in unskilled manual work (Almlund et al. 2011; prospective employees. Of the soft skills they Fletcher 2012). seek when recruiting professionals or skilled The relative importance of behavioral and workers, employers rate “punctuality and reli- cognitive skills is also not well established. ability” and “honesty and trustworthiness” as Fundamentally, many behavioral and cognitive the most important.14 In Sierra Leone, focus skills are interconnected. For instance, the abil- group discussions revealed that the young are ity to regulate one’s behavior or emotions also perceived to lack “work attitudes” appropri- depends on cognitive abilities. Research sug- ate for formal jobs, such as punctuality and gests that for less complex tasks, such as those the ability to follow instructions (Peeters et al. required for relatively “unskilled” work, the role 2009). of behavioral skills can be relatively larger. For Behavioral skills also matter for HEs for more complex tasks, the contribution of cogni- many reasons, including the firm’s dependence tive skills can be relatively higher, although soft on the skills of the owner, who is often the skills matter there as well (Almlund et al. 2011). sole worker (chapter 5). In South Africa and For wage jobs, two personality characteristics, Zimbabwe, for example, the owner’s “entre- in particular, seem to have strong predictive preneurial orientation”—measured by per- power for job performance and wages: con- sonal initiative (proactivity and persistence) scientiousness (the tendency to be organized, and achievement orientation (taking respon- responsible, and hardworking) and emotional sibility for one’s own performance, taking on stability. Almost all such evidence comes from challenges, and setting high goals for one- developed countries, although similar findings self)—are positively associated with success are emerging from developing countries. A in operating a small business.15 A similar link recent study in Peru shows that cognitive skills between personal initiative and business suc- and perseverance—a behavioral skill—have cess was found in Uganda (Koop, De Reu, and similar effects on earnings in wage employment Frese 2000, quoted in Krauss et al. 2005). Other (box 3.4). work has emphasized how the “deliberateness” Overall, socioemotional and behavioral of business owners is linked to success. Exam- skills matter for productivity. Such skills can be ples include the role of “elaborate and proac- shaped through experience in early childhood, tive planning” skills in Namibia, South Africa, through education, as well as through a range and Zimbabwe (Frese et al. 2007) and “deliber- ate practice” (self-regulated and effortful activ- of experiences and programs in adolescence ities showing a willingness to learn) in South and adulthood. The relative effectiveness of dif- Africa (Unger et al. 2009). Self-control— ferent approaches to building these skills needs a behavioral skill that builds on cognitive to be better understood, and this is an area that capacity—also appears to be an important warrants policy attention. influence on the savings and investment behavior of HE owners in Kenya (Dupas and Schooling Can Contribute to Building Robinson 2013). Socioemotional and Behavioral Skills Despite the recognized importance of The link between education and behav- behavioral skills, it is not yet well understood ioral skills is complex. Part of the association which ones matter most or the extent to which between behavioral skills and employment 86 Youth EmploYmEnt in Sub-Saharan africa Box 3.4 Returns to behavioral skills in wage employment: Evidence from Peru A survey of micro and small enterprises in Peru found that topics, including a range of other individual and household employers seek workers with a range of skills. About half of characteristics. the firms said that the main problem in hiring suitable work- Analysis of those data reveals that measures of cogni- ers is the lack of qualified or competent workers (in this case, tive skills correlate with higher earnings. An increase of 1 “qualifications” and “competence” reflect perceived cogni- standard deviation in receptive language, numeracy, work- tive and technical skills). At the same time, about 40 percent ing memory, or verbal fluency was associated with 9–18 cited the lack of skills such as a strong work ethic, ability to percent higher hourly earnings. Behavioral skills were also work in teams, persistence, adaptability, initiative, and other associated with higher earnings. Workers scoring 1 standard socioemotional skills. Other data from the national public deviation higher in the perseverance facet of grit earned 13 employment service show that, in addition to cognitive skills, percent more; those scoring 1 standard deviation higher in employers seek employees with traits related to work ethic, three of the big-five scores (extraversion, emotional stability, reliability, and interpersonal relations—regardless of formal and openness to experience) earned 8 percent more. Not all schooling. Moreover, the results also suggest that behavioral behavioral skills necessarily correlate positively with earnings. skills may be more important for the less educated. A score for “agreeableness-cooperation” that was 1 stan- The study also set out to measure the supply of skills dard deviation higher was associated with 10 percent lower directly. In addition to traditional measures such as years of earnings. education, the study applied a range of standardized and spe- Schooling, cognitive skills, and behavioral skills affect cifically designed instruments. The Peabody Picture Vocabu- earnings through multiple pathways, both direct and indi- lary Test was used to measure receptive vocabulary—the rect. When the role of each factor is evaluated while con- ability to comprehend specific words. Additional cognitive trolling for the other factors, years of schooling, years of tests were designed to measure verbal ability, working mem- work experience, and cognitive skills all have significantly ory, and numeracy or problem-solving skills. Socioemotional positive effects. This finding suggests that measured cogni- skills were measured with self-reported tests for personal- tive skills affect earnings over and above years of schooling ity characteristics related to behaviors. They were measured alone, suggesting that what is learned in school matters. The with scales of the big-five personality factors (openness to behavioral skills that significantly affect earnings (even after experience, conscientiousness, extraversion, agreeableness, accounting for schooling, experience, and cognitive skills) emotional stability), factors that are considered in psychol- are emotional stability and perseverance. Perseverance mat- ogy as characterizing differences in broad personality traits ters as much as cognitive skills. A score that is 1 standard (and associated behaviors), and “grit,” a narrower trait cap- deviation higher on either the cognitive skill or the persever- turing a person’s inclination and motivation to achieve long- ance scale led to a similar increase in earnings (of 8 and 9 term goals (through perseverance of effort and consistency percent, respectively). of interest). The study measured these skills in the context of a household survey that collected information on multiple Source: World Bank 2011b. outcomes stems from their effect on higher Education systems have scope for develop- educational achievement—and the cognitive ing skills other than cognitive skills. Research skills that underlie higher educational achieve- in the United States suggests that elementary ment. Behavioral skills such as perseverance or even preschool programs that enrich the or personal initiative are correlated with suc- early learning environment can have lasting cess in school. It could be that people who are effects on later-life outcomes through posi- inherently more conscientious both acquire tive impacts on behavioral and socioemotional more schooling and perform better at work skills (Almlund et al. 2011). Increased attention (Almlund et al. 2011). Recent work on China to imparting behavioral skills through school- found a positive association between behav- ing may take several forms, including modes of ioral skills and educational achievement, even instruction as well as the modeling of appropri- after accounting for cognitive skills (Glewwe, ate behaviors, including teachers’ behavior, in Huang, and Park 2011). the school environment. Skills for productive Employment 87 Schooling imparts behavioral skills in sev- behavior—might also contribute to improving eral ways. First, school success itself increases students’ behavioral skills through their dem- self-esteem and confers a greater sense of self- onstration effect. Teaching approaches that are determination, as shown in research among more interactive and group based may facilitate high school and college graduates in the United the acquisition of both cognitive and behav- States.16 Second, the way that teaching and ioral skills, for instance. Curricular reforms learning are delivered may influence behav- that explicitly seek to improve behavioral skills ioral skills. Teaching approaches that encourage may also be considered. However, such reforms participation, group activities, and exploration may entail a trade-off requiring a decision instill different mind-sets among students than about what to prioritize. Curriculum reforms approaches that emphasize rote learning. Third, introducing behavioral skills should not come the experience of education and the habits at the cost of neglecting urgent improvements learned in school matter. Teacher absenteeism in the acquisition of basic literacy and numer- on the order of 20 percent, with little or no con- acy skills. Reforms might be more relevant for sequence to teachers, has been documented in higher grades where potential trade-offs with several African countries (World Bank 2003).17 other teaching may be more limited. Given the Students’ exposure to such an environment will limited evidence on the effectiveness of such likely instill a sense that punctuality (one of the approaches and the importance of their devel- skills that some employers say they are seeking) opment in local contexts, they should be intro- is not important. duced in an experimental setup and rigorously There are little data on the extent to which tested before being rolled out widely. education builds socioemotional and behav- Beyond education, other more focused ioral skills, including the soft skills demanded approaches are showing promise in reaching by employers. The education system could be youths and building behavioral skills. They leveraged to build further behavioral skills and include targeted interventions in postconflict reach a large number of children and young settings (box 3.5). Programs that integrate people. Policies that aim to improve the qual- behavioral skills into more comprehensive ity of education—for instance, by decreasing youth employment programs can also improve teachers’ absenteeism and improving their behavioral skills (box 3.6). Chapter 5 (for the Box 3.5 Developing socioemotional and behavioral skills in postconflict settings Children exposed to violence such as rape or killing can ful integration into educational or livelihoods initiatives, and develop conditions such as depression or anxiety that might positive peer, family, and community relationships.” lead to behavioral problems. In Sierra Leone, where the The Sierra Leone pilot intervention was developed long civil war has had deep psychosocial consequences, a through an approach building on rigorous use of qualitative Youth Readiness Intervention has focused on rebuilding methods to ensure that the training modules and interven- skills related to anger management, interpersonal relations, tion strategy were culturally appropriate and could be deliv- and goal setting. The intervention is delivered to groups of ered by local providers. A randomized control trial is under young people. The groups meet for weekly sessions over way to test the intervention’s effectiveness. Such pilots have two months. The intervention seeks to “increase adaptive yet to be tested at scale, and the employment outcomes coping, health-promoting behavior, and the development remain to be documented. Still, this type of well-designed of life skills, such as goal setting and positive self-efficacy; model to foster socioemotional and behavioral skills is an decrease maladaptive coping strategies, such as high-risk example of potential interventions that could be integrated sexual behavior and substance use; decrease trauma-related more systematically into youth employment programs. distress, including aggression, depressive symptoms, social isolation, and poor interpersonal skills; [and] increase proso- Source: Draws from FXB Center for Health and Human Rights 2012, cial behavior, including community involvement and success- including quoted material. 88 Youth EmploYmEnt in Sub-Saharan africa Box 3.6 higher ability—with the pattern being starkest in Kenya and Tanzania. This finding suggests that education has the greatest benefit among Promoting socioemotional and behavioral skills people who start with greater ability or better foundational skills. Employment programs in Malawi and Uganda that focused on female Basic cognitive skills also underlie the devel- empowerment and life skills improved psychosocial well-being and opment of business skills, which are often par- reduced risky behaviors (Bandiera et al. 2013; Cho et al. 2013). A pro- gram integrating training in technical and life skills in Uganda resulted ticularly low for young people and women (Xu in a self-reported increase in routine condom use of almost 50 percent and Zia 2012). A study of small HEs in Ghana and a 29 percent reduction in the likelihood of becoming a mother found that owners with more schooling are (Bandiera et al. 2013). Positive impacts on behaviors was seen in pro- more likely to keep written records: Only 23 grams in Tanzania and Uganda aiming to increase personal initiative percent of HE owners who had not completed and develop a more entrepreneurial mind-set (Glaub 2009; Berge et basic education kept accounts, compared to 52 al. 2011). These results are consistent with others outside of Africa. percent of owners who had completed basic Programs in countries as diverse as Chile, Jordan, Nicaragua, and Peru education (Fafchamps et al. 2011). Low busi- have produced graduates who have greater self-esteem, measured ness skills are observed even among more for- empowerment, mental health, or positive attitudes about the future mal firms. In a survey of owners of small and (Carneiro, Galasso, and Ginja 2010; Valdivia 2011; Macours, Premand, medium enterprises in Ghana, only 27 percent and Vakis 2013; Premand et al. 2012; Groh et al. 2012).a of owners were found to keep business records a. These effects are corroborated by an ongoing evaluation of a comprehensive (Mano et al. 2012). Throughout the world, youth training program—Juventud y Empleo—in the Dominican Republic. Prelimi- financial literacy is associated with higher edu- nary evidence indicates significant effects on future expectations, job satisfaction, and job search attitudes. cational attainment (Xu and Zia 2012). Another study from Ghana measured a range of business practices among HE own- ers, such as whether “the owner keeps writ- HE sector) and chapter 6 (for the modern wage ten business records, has a written budget, has sector) discuss the extent to which programs a sales target, visits competitors’ business to aiming to shape behavioral skills are effective observe prices, asks existing customers if they in improving employment outcomes. had interest in other products.” The number of management practices adopted by the owners Skills Build on Each Other of micro and small enterprises was found to A growing body of research documents how cognitive and socioemotional skills build on Figure 3.17 The returns to schooling are higher for each other, starting in early childhood (Helm- those with greater ability: Returns of an additional ers and Patnam 2011). Cognitive skills such as year of schooling by quantile literacy and numeracy form the foundation for 22 acquiring higher-order and technical skills later 20 in life, whether through more formal education, 18 16 training, or on-the-job learning. Basic cognitive 14 skills are necessary for learning more advanced 12 % return concepts, and better foundational skills lower 10 8 the costs of any additional investments. 6 One way to understand how skills build on 4 2 each other is to observe that the productivity 0 returns to investments in further schooling are Q10 Q25 Q50 Q75 Q90 higher when foundational skills are stronger. Quantile Figure 3.17 presents an example from Ghana, Kenya Tanzania Kenya, South Africa, and Tanzania. Returns Ghana South Africa to schooling tend to be lower for individu- Source: Fasih et al. 2012. Reproduced with permission from the als with lower ability compared to those with authors. Skills for productive Employment 89 vary substantially. In addition, these business In Tunisia, a middle-income country, entrepre- practice scores also predicted business perfor- neurship training was introduced in the univer- mance (Fafchamps and Woodruff 2012). The sity curriculum and led to changes in behavioral adoption of better management practices was skills (Premand et al. 2012). Initial evidence on significantly higher among respondents who a pilot financial education program provided scored better on various cognitive tests. through the schools in Brazil suggests some Although business skills are not taught in positive impacts: 59 percent of students who school, they are correlated with education, benefited from the program saved, compared because they are rarely acquired without basic to 55 percent among the control group, and cognitive skills. A more direct measure of this knowledge about savings increased (Bruhn et effect, however, is the interaction between edu- al. 2013). Similar efforts are under way in some cation and the effectiveness of interventions African countries (for example, in secondary to build business skills. A program to provide education in Uganda). An important challenge business training to small-scale entrepreneurs is that soft skills for self-employment and wage in Tanzania found larger impacts among par- employment may differ, requiring careful atten- ticipants with higher initial cognitive (math) tion to curriculum development and program test scores (Bjorvatn and Tungodden 2010). effectiveness. Basic cognitive skills also complement on- the-job learning for wage workers, includ- ing learning through the use of instructions. Building Skills through Although literate workers can be trained in job specifications and procedures partly through Post-School Training detailed and complex written instructions, such training is rare in Africa, where literacy is low. A wide array of institutions throughout Africa Among firms surveyed in Ghana, Kenya, and offer skills training for nonfarm employment. Zimbabwe, 80 percent said that they rarely use This array of institutions and programs can be technical documentation or procedural manu- described as a market, since it involves those als (Biggs, Shah, and Srivastava 1995). A recent who supply training coming together with study in Ghana found that, in a job involving those who demand training. Informal training the handling of money, more educated workers is normal. Most training is offered by private were more likely to pick up math skills while providers, and the offerings vary in price and working (Aslam and Lehrer 2012). Perhaps quality. The rationale for public investment in because of the complementarity between basic training needs to be made in the context of this education and on-the-job learning, firms in existing market, based on careful analysis of its Africa (as in other parts of the world) are more added value and cost-effectiveness. likely to provide formal on-the-job training to more educated workers (Rosholm, Nielsen, and Pathways from Training to Sectors of Dabalen 2007). Employment Beyond the complementarities between Four main kinds of youth training provide basic skills and business skills, there may be skills for employment outside agriculture: potential for incorporating entrepreneurship apprenticeships, public formal technical voca- education or financial literacy education into tional education and training (TVET), private school curricula. Yet it remains unclear whether formal and informal TVET (“formal” mean- the education system can provide such skills ing integrated into the formal education sys- effectively. Some developed countries have tem; “informal” meaning outside of the formal attempted to include entrepreneurship training education system), and stand-alone programs. at the primary and tertiary level (see Rosendahl Training is delivered through a mix of private Huber, Sloof, and van Praag 2012 for efforts in and public institutions, and each type of train- primary school; Oosterbeek, van Praag, and Ijs- ing tends to lead to a different type of work selstein 2008 for efforts in tertiary education). (nonfarm HE or modern wage sector). 90 Youth EmploYmEnt in Sub-Saharan africa Figure 3.18 Many young people, especially in West Africa, have been an apprentice, whereas experience with TVET is less prevalent a. Ever been an apprentice b. Ever attended a TVET institute Cameroon Côte d’Ivoire Côte d’Ivoire Ghana Ghana Mozambique Rwanda Malawi Sierra Leone Rwanda Senegal Tanzania Sierra Leone Uganda Tanzania Uganda Average Average 0 10 20 30 40 0 10 20 30 40 Share (%) Share (%) Source: Based on standardized and harmonized household and labor force surveys, latest data available (see appendix). Two of the most common forms of train- Apprenticeship was by far the most common ing pursued by young people are apprentice- form (55 percent had been an apprentice), fol- ships and TVET. Apprenticeships are the more lowed by on-the-job formal training in a firm prevalent type of training, particularly in West (25 percent), and formal vocational training (16 Africa. Detailed survey data on apprenticeship percent). An earlier study estimated that tradi- are limited, but in five countries with compara- tional or informal apprenticeships supply 80–90 ble data, 20 percent of young adults ages 24–35 percent of all basic skills training in Ghana, have had experience as an apprentice (figure while public training institutions supply 5–10 3.18), although there is variation across coun- percent (Atchoarena and Delluc 2001). Ghana tries, from 6 percent in Uganda to 35 percent may have as many as four informal apprentices in Ghana. for every trainee in either a formal public or a Enrollment in formal TVET, delivered in the private training center (Darvas 2012; Haan and classroom and leading to a formal degree after Serrière 2002; Monk, Sandefur, and Teal 2008). two to three years, is low throughout Africa. Apprenticeships are widespread elsewhere in Overall, around 4 percent of young people West Africa as well, including in Benin and Côte between 25 and 34 have ever attended formal d’Ivoire (AfDB and OECD 2008). TVET,18 and only 1 percent currently attend.19 Traditional apprenticeship can also be the Because most TVET requires some secondary dominant form of training for nonfarm occu- schooling, the majority of young people lack pations in East Africa. In Kenya, enrollment in the general qualifications even to enroll in a traditional apprenticeships delivered by master technical or vocational institute. craftsmen is much higher than enrollment in The prevalence of apprenticeship in Ghana formal TVET.20 A small survey of 350 informal is well documented (for example, see Atcho- enterprises in Dar es Salaam found that more arena and Delluc 2001; Frazer 2006; Monk, than half of the operators had apprentices, on Sandefur, and Teal 2008). A 2006 urban labor average about two per firm (Nell and Shapiro market survey found that one-third of respon- 1999). dents between ages 16 and 65 had some form Beyond the more traditional apprenticeship of training (Monk, Sandefur, and Teal 2008). and formal TVET models, a broad range of Skills for productive Employment 91 private providers offers various types of skills individuals who have completed primary training. Private providers of informal voca- school or less, while formal TVET is geared tional training (in other words, training outside toward individuals with at least some sec- the formal education system) include for-profit ondary schooling (figure 3.19). For example, private institutes and firms, nongovernmental among young adults ages 25–34 in Uganda organizations (NGOs), and community orga- who had done an apprenticeship, 95 percent nizations. While the distinction with master had no more than a primary education. A craftsmen providing apprenticeships can be study in Ghana found that apprenticeships are fuzzy, the vast majority of providers of informal undertaken primarily by persons with a junior vocational training are self-financing and oper- high school or lower level of education (Monk, ate with little government oversight or support. Sandefur, and Teal 2008). Among those who Many informal private providers are not reg- had entered a TVET program, almost all had istered (World Bank 2003). Many operate at a some education beyond primary. very small scale; these micro training provid- The type of training pursued maps closely to ers provide short, intensive training based on the subsequent sector of employment. Appren- a curriculum of their own design and may ticeship is mainly a pathway to work in an HE, offer certificates (Johanson and Gakuba 2011). since apprentices are most likely to become Assessing their reach is challenging. Household self-employed. By contrast, formal TVET is surveys rarely ask respondents about training mostly a pathway to wage employment. other than apprenticeships, on-the-job train- As a result, the training experience of work- ing, or formal TVET. A recent tracer study of ers in different sectors is quite different. For cohorts of secondary and university graduates young adults ages 25–34 working in the HE suggests that the use of private post-school sector, apprenticeships are the most common training may be rising (Al-Samarrai and Ben- form of post-school training: 32 percent have nell 2007). been an apprentice at some point, compared Apprenticeships (as well as other types of to 30 percent of young adults in the modern informal training) are typically geared toward wage sector and 13 percent in the agricultural Figure 3.19 Apprenticeships are geared toward youths with lower levels of education a. Past apprentices b. TVET graduates Côte d’Ivoire Côte d’Ivoire Ghana Ghana Rwanda Rwanda Sierra Leone Sierra Leone Tanzania Tanzania Uganda Uganda Average Average 0 20 40 60 80 100 0 20 40 60 80 100 Percentage Percentage Less than primary Primary Lower secondary Secondary or higher Source: Based on standardized and harmonized household and labor force surveys, latest data available (see appendix). 92 Youth EmploYmEnt in Sub-Saharan africa Figure 3.20 TVET is geared toward wage work, whereas apprenticeships lead to working in either a household enterprise or wage work a. Apprenticeship b. TVET Côte d’Ivoire Côte d’Ivoire Ghana Ghana Rwanda Rwanda Sierra Leone Sierra Leone Tanzania Tanzania Uganda Uganda Average Average 0 10 20 30 40 50 60 0 10 20 30 40 50 60 Percentage Percentage Agriculture Household enterprise Wage Source: Based on standardized and harmonized household and labor force surveys, latest data available (see appendix). sector (figure 3.20).21 Returns to apprentice- provides practical training in the workplace ships may be particularly high in HEs. A study over a period of a few weeks or months to as from Ghana found that former apprentices much as three or four years. Many apprentice- earn about 49 percent more a year being self- ships build technical skills in a narrow range employed than working as a wage worker, of traditional vocations or crafts, such as met- despite having slightly fewer years of school- alworking, carpentry, mechanics, or tailor- ing and being slightly younger.22 Relatively ing. Some offer certification, but most do not. few young workers enter the HE sector after Apprenticeships can be offered in return for a obtaining formal technical or vocational train- fee or reduced earnings while learning. ing. Only 6 percent of individuals between 25 In an apprenticeship program in Malawi, and 34 in the HE sector have attended a formal master craftsmen were primarily in carpentry TVET institution (figure 3.20). The share of and joinery (19 percent), tailoring (18 percent), workers in the wage sector who have attended auto mechanics (11 percent), and fabrication formal TVET is much higher (17 percent). and welding (11 percent). A large share of the observed apprenticeship training in Rwanda Apprenticeships and Informal Private was concentrated in tailoring (Johanson and Training Gakuba 2011). The distinction between apprenticeship and The duration of apprenticeships can vary informal private training (two of the most greatly. Many youths only spend a few months prevalent forms of youth training) is fuzzy. as an apprentice. In Rwanda, 56 percent of Both types of training encompass a range of HE owners with experience as an apprentice offerings delivered by private providers. reported that their apprenticeship lasted less Apprenticeships in Africa overwhelmingly than a month (Johanson and Gakuba 2011). In occur in small informal firms with a mas- Malawi, participants in an apprenticeship pro- ter craftsman. They are private arrangements gram reported that apprenticeships lasted on between an apprentice and a master craftsper- average 3.3 months; training for auto mechan- son or another relatively skilled person, who ics spanned more than three months, whereas Skills for productive Employment 93 training for hairdressers lasted only three weeks ing vouchers in Kenya, for example, enrolled “For many (Cho et al. 2013). By contrast, in countries such mainly in informal training in tailoring (37 long years, I as Côte d’Ivoire or Ghana, where the institution percent), mechanics (18 percent), hairdressing worked as an of apprenticeship is more established, appren- (9 percent), driving (7 percent), and masonry ticeships can last several years and can be hard (6 percent; Hicks et al. 2011). In northern apprentice and to leave. Uganda, youth groups eligible for cash grants had to follow The range of private informal training in chose training heavily concentrated in a few the masons . . . . Africa is quite wide. Many micro training pro- trades: tailoring (38 percent), carpentry (24 It wasn’t until I viders develop their own teaching programs, percent), metalwork (13 percent), hairdress- turned 30 that market their services, and deliver a mix of ing (8 percent), and business or management I got my license.” theoretical and applied training to individu- (5 percent).24 als in small groups. For example, in Rwanda, The duration of informal training can be Madagascar 97 percent of all training providers are private, quite short. In Northern Uganda, where youths and they account for 90 percent of enrollment engaged in informal training for an average of (figure 3.21; Johanson and Gakuba 2011). about 321 hours over two years, they increased Individuals own about half of all private train- their investment in training to 560 hours after ing enterprises, and associations and coopera- receiving a cash grant. When grant recipients tives own the other half. Just over half of all enrolled with private formal TVET providers, providers are micro training providers, such as the training was substantially longer, ranging associations, cooperatives, or training centers, up to two years (Hicks et al. 2011). enrolling 12 or fewer trainees. Micro train- ing providers enroll only 8 percent of private Formal TVET trainees, but total enrollment in micro training Formal TVET runs parallel to general school- still exceeds enrollment in all public, formal ing at the secondary or tertiary level and meets vocational training in Rwanda (figure 3.21). In the need for intermediate or advanced techni- Tanzania, private training institutions, includ- cal skills. Entry requirements often include hav- ing faith-based organizations and NGOs, ing completed primary or secondary school. produce about three-fourths of all vocational Therefore, participants in formal TVET have graduates, even though most of those insti- substantially more schooling than participants tutions do not operate at capacity (Cojocaru in other forms of post-school training. 2011).23 In Nigeria, technical colleges at the sec- Like apprenticeships, informal training ondary level produce craftsmen and master programs are often concentrated in a limited craftsmen, focusing on traditional technical set of areas. Youth eligible for vocational train- vocations (electricians, vehicle mechanics, and Figure 3.21 Africa has a wide range of informal private training providers, 2009 a. Training institutions, Rwanda b. Enrollment in vocational training, Rwanda Micro private Public Public training providers Semi-public Micro private Semi-public training providers Large private Large private training training providers providers Source: Johanson and Gakuba 2011. 94 Youth EmploYmEnt in Sub-Saharan africa masons). At the tertiary level, vocational insti- often those with larger populations and greater tutions (polytechnics) produce technicians, demand for training (Ghana and Zambia are professionals, and engineers. Business-oriented examples). vocational training is prevalent. Nearly half of Besides pre-employment TVET, post- all polytechnic graduates pursue qualifications employment formal training paid for by employ- in accounting, business studies, marketing, ers can also be a source of technical or vocational and banking and finance (World Bank 2011a). skills. About 30 percent of African firms provide In Rwanda, technical secondary schools pre- formal on-the-job training, a rate comparable to pare students for entering the labor market at other developing regions (box 3.7). roughly the same level as an upper-secondary school graduate. Vocational training centers prepare basic education graduates or dropouts Government Interventions and to enter the labor market (World Bank 2011a). the Post-school Training Market Several types of nongovernmental entities also provide technical and vocational training, Governments the world over are active in including for-profit private institutes and firms skills development, but before designing pub- as well as NGOs. However, data on the extent lic policy, it is essential to assess the rationale to which the private sector provides formal for government intervention. Given the large TVET are scarce; household surveys rarely ask array of training already provided by the pri- respondents to identify whether they attended vate sector, the rationale for governments to a public or a private TVET institute.25 Never- invest in specific programs is not obvious. As a theless, studies suggest that the private sector general principle, public interventions need to is a large and increasingly important provider be based on clearly identified “market failures” of pre-employment TVET. In some countries, and weighted against “government failures.” the majority of trainees are enrolled in nongov- Market failures for skills development take ernmental institutions. Examples include Mali specific forms—all of which can lead to under- (where nongovernmental training accounts investment in training (see the discussion in for two-thirds of all TVET), Tanzania (90 per- Almeida, Behrman, and Robalino 2012; World cent), and Zambia (82 percent). Private techni- Bank 2011a). These failures can be grouped cal institutes in Ghana enroll about six times as into four main categories: many trainees as public institutes (Haan 2001 1. Imperfections in labor markets, such as for Tanzania; Atchoarena and Esquieu 2002 for “poaching externalities,” whereby the firm Mali; Kitaev 2003 for Ghana and Zambia). A that trains an employee loses the benefits of recent study in Ethiopia estimates that 30–50 that training if the worker moves to another percent of TVET students are enrolled in pri- firm, or information and signaling failures, vate institutions (Shaorshadze and Krishnan whereby a potential worker cannot effec- 2013). A recent World Bank report, using sta- tively show her level of skills to a potential tistics from 33 countries in the region, found employer that the private sector currently accounts for about 35 percent of formal TVET enrollment 2. Imperfections in credit markets, whereby it is (Mingat, Ledoux, and Rakotomalala 2010). hard to get financing to pay for training Compared with public institutes, private 3. Coordination failures, occurring, for exam- training providers tend to focus on “light” ple, when workers and firms need to make vocational skills such as business, commercial, investments, but those investments pay off and service skills, possibly owing to the high only if both parties invest, which can lead to fixed costs of providing more industry-oriented neither side making the first move sorts of skills. Private providers in Uganda, for 4. Limited information at the individual level, instance, focus on office qualifications and var- which can lead to too little or too much ious business skills that require only a limited investment in particular types of skills investment (Haan 2001). Private providers also development tend to be concentrated in specific regions— Skills for productive Employment 95 Box 3.7 On-the-job training varies by country and type of firm, and it is not for everyone In the United States, one-quarter to half of human capital between 15 and 30 percent in most countries, but in some accumulation occurs through on-the-job training (Heck- (such as Botswana, Malawi, and Rwanda), the incidence is as man, Lochner, and Taber 1998). Even in developing countries high as 50 percent. (including African countries), many firms provide training to Variations in the rates of on-the-job training across their workers. Estimates from the World Bank’s enterprise Africa do not appear to be related to per capita income surveys indicate that, on average, about 30 percent of formal levels. Since firm surveys tend to focus on formal, officially firms in Africa provide training (figure B3.7.1), only slightly “registered” firms and often miss unregistered firms, some below the average for low- and middle-income countries. of the variation might arise from differences in the share In Africa, the percentage of firms training workers varies of firms that are registered. In most countries around the world, the incidence of training in firms is strongly related Figure B3.7.1 On-the-job training in African firms varies to firm characteristics such as size and export orientation. by country Smaller firms are less likely to provide formal training to employees. So are nonexporters. This pattern is borne out Congo, in Africa as well (figure B3.7.2). Thus African countries with Dem. Rep. Liberia larger, more outwardly oriented firms may have more on- Eritrea the-job training. Niger Central African Figure B3.7.2 On-the-job training in African firms varies Republic by firm size and export orientation, 2006 Sierra Leone Malawi 80 % firms offering formal training Madagascar 60 Togo Mali 40 Burkina Faso Rwanda 20 Chad Lesotho 0 Benin All Exporter Non- Large Medium Small Côte d’Ivoire exporter (100+) (20–99) (5–19) Cameroon Firms Cape Verde Source: World Bank enterprise surveys (average for African countries sur- veyed in 2006). Congo, Rep. Angola Mauritius Firms providing on-the-job training do not train all of their workers. As figure B3.7.2 suggests, African firms that Botswana provide on-the-job training rarely provide it to more than Gabon half of the workforce, perhaps because levels of education Sub-Saharan are low. As in the rest of the world, in Africa workers with Africa All more education and skills are much more likely to receive formal training on the job. A 1980 survey of formal train- 0 10 20 30 40 50 60 ing in enterprises in Dar es Salaam, Tanzania, and Nairobi, Percent Kenya, found that unskilled and semi-skilled manual workers Workers offered formal training Firms offering formal training are significantly less likely to have received formal training Source: Based on World Bank enterprise surveys. from their current employer than skilled production, cleri- Note: Countries ordered by GDP per capita. (continued) 96 Youth EmploYmEnt in Sub-Saharan africa Box 3.7 (continued) cal, and supervisory workers (De Beyer 1990). Surveys con- partly specific to their current firm. On-the-job training is also ducted in Kenya and Zambia in 1995 found that workers specific to the technology used in the firm. with no formal education do not receive on-the-job training The typical worker is more likely to get training through (Rosholm, Nielsen, and Dabalen 2007). on-the-job learning than from classroom training paid for by There are scarce data on the content of on-the-job train- employers. Firm surveys in Kenya and Zambia found that, ing in African firms or on how it may differ from the training in the 12 months prior to the survey, the most common that a young job seeker might receive at a training institute. forms of learning among employees were “instructions from It is likely that the content of on-the-job training has an ele- a supervisor or coworker” and “watching others or learn- ment of firm-specificity. In Dar es Salaam workers who had ing on your own” (Rosholm, Nielsen, and Dabalen 2007). previously been in another firm were almost as likely to get Courses paid for by the employer were less common and formal on-the-job training as those who were on their first about as likely as training at a school or technical or voca- job (De Beyer 1990), suggesting that the training was at least tional institute. These various forms of market failure pro- Existing forms of training are not equally vide general rationales for government inter- available to everyone (figure 3.22). Patterns vention, but their prevalence needs to be care- of training across income groups suggest that fully assessed in specific country contexts. They financial constraints reduce access to training also provide guidance as to the range of activi- among the poorest households. Among youth ties that the public sector might want to sup- from households in the top income quintile, port. For example, credit constraints may pro- 11 percent have ever enrolled in TVET and 2.7 vide a rationale for policies to improve access to percent are currently enrolled. By contrast, in training. Limited or inaccurate information at the bottom income quintile, only 1.6 percent the individual level has also led to underinvest- have ever enrolled in TVET and 0.1 percent are ment in training or suboptimal choice of train- currently enrolled. This inequality in access to ing and can provide a case for public interven- training holds not only for formal vocational tion. Governments need to recognize that there training but also for informal apprenticeships: is a market for training and avoid introducing 25 percent of youth in the top quintile have undue distortions in this market. Overall, there taken an apprenticeship and 7.6 percent are are two broad areas for government interven- currently in an apprenticeship, compared with tion: (1) providing information and facilitat- 7.3 and 2.5 percent, respectively, of youth in the ing access to training and (2) intervening to bottom quintile. ensure the availability of better-quality train- Women also have limited access to train- ing options. Those two areas are discussed next. ing opportunities, and when they do receive training, it often focuses on a limited range of Facilitating Access to Training occupations. Women are less likely than men Public policy should facilitate access to existing to be enrolled in formal TVET or apprentice- training opportunities, including those avail- ships (see also Atchoarena and Delluc 2002). able in the private market. One strategy, for Across the region, 18 percent of individuals example, is to provide information or incen- ages 15–34 have ever been an apprentice, but tives to young people who have the least access only 12 percent of women.26 Women who to training, starting with individuals from the manage to pursue informal technical training poorest households, women, and individuals or apprenticeships tend to end up in heavily in rural areas. Better provision of information concentrated sectors with limited demand, about employment and training opportunities such as tailoring or hairdressing. In Uganda, can start in school. 91 percent of training hours taken by females Skills for productive Employment 97 who were not involved in the Youth Oppor- Figure 3.22 There are stark differences between rich and poor in experience with tunities Program were in tailoring (Blattman, apprenticeships and TVET Fiala, and Martinez 2011). In Kenya, the most 30 popular courses for women were tailoring, hairdressing, and computer packages, whereas 25 men preferred training to become mechanics, Percentage of youths 20 drivers, or masons (figure 3.23). (ages 15–34) Youths in rural areas also have less access 15 to training, since more training providers are located in urban areas and distance to training 10 centers is a constraint for access. Across Sub- 5 Saharan Africa, 25 percent of 15- to 34-year- olds living in an urban area have ever been an 0 apprentice, compared with only 11 percent of Currently Currently an Ever enrolled in TVET Ever an apprentice those living in a rural area. In Kenya, women enrolled in TVET apprentice were significantly more likely than men to cite Quintiles 1st (poorest) 2nd 3rd 4th 5th (richest) proximity to training opportunities as a deter- Source: Based on standardized and harmonized household and labor force surveys (see appendix). mining factor (50 and 43 percent, respectively), suggesting that female participants are more geographically constrained than their male Figure 3.23 Men and women take up different types of vocational training in Kenya peers (Hicks et al. 2011). In the presence of financial and other con- 60 straints on access to training, there is scope for 50 policies to facilitate access to training among Percentage of youths youth, particularly women and the poor. Inter- 40 (ages 15–34) ventions that provide targeted financial incen- tives to increase participation in training have 30 been shown to help. In northern Uganda, a 20 program providing large cash grants to self- created groups increased the proportion of 10 youth enrolled in vocational training from 15 to 74 percent, and recipients also engaged in 0 Tailoring Vehicle related Computers Beauty Business skills Other more intensive training. Some training oppor- tunities were already available in the commu- Male Female nity, but the young people participating in the Source: Hicks et al. 2011. project had not purchased much of this train- ing before. Although youths in the program did not have to buy training, most chose to spend a young people vouchers worth approximately large part of their cash grants on training prior US$460 to encourage them to enroll (Hicks to starting a business. This finding suggests that et al. 2011). Voucher winners who were closer programs that help to finance access to training to private schools were more likely to take up might be effective. Among those who did not training compared to winners who were far- get a grant (the control group), some partici- ther away. Among voucher winners, a random pants did acquire training on their own, but it half received a voucher that could be used only was of much shorter duration. Of the 15 per- in a public (government) vocational institu- cent of individuals who took training without tion, while the other half received a voucher the program, only 6 percent paid for it. that could be used in either a private or a pub- Voucher programs can be effective when lic school.27 The broader choice and access to recipients have the option to choose a pri- private providers increased the use of training: vate training provider. In Kenya, the Techni- 69 percent of individuals who were awarded the cal and Vocational Vouchers Program offered restricted vouchers attended vocational train- 98 Youth EmploYmEnt in Sub-Saharan africa ing, compared with 79 percent of individuals There might also be scope for interventions who were awarded unrestricted vouchers. Win- that provide information on employment and ners of unrestricted vouchers were also more training opportunities to help young people to likely to complete training. decide which training to undertake. Research in Voucher programs can also have an effect on Kenya has shown that young people have inac- skill providers and stimulate the supply of train- curate perceptions about the returns to voca- ing available. In Kenya, a large program that tional training (including misconceptions about provided vouchers to workers in the informal which trades provide the highest earnings) and sector (called Jua Kali) not only increased access that their perceptions have a strong gender bias to training but also led to the emergence of new (Hicks et al. 2011). In such a context, there is a training providers relevant for the HE sector, role for interventions to increase participation such as master craftsmen (see box 3.8; Adams in training by providing information to better 2001; Johanson and Adams 2004). Evidence match trainees to training. In Kenya, young peo- from a smaller-scale pilot of training vouchers ple changed their training choices after receiving in Kenya suggests that programs or schools that information on actual labor market returns— received voucher students were significantly including differences in expected earnings for more likely to expand their course offerings trades dominated by men (such as electrician) (Hicks et al. 2011). Although such programs and women (such as seamstress)—and viewing have not been scaled up yet and would likely inspirational videos about successful female car require some prior identification of eligible mechanics. In particular, the provision of infor- providers, they have the potential to be effective, mation caused more women, especially young given the vast diversity of existing providers. and more educated girls, to take up training in male-dominated trades. Such information failures have long been recognized, but public policy has rarely been Box 3.8 able to address the issue successfully at scale. Many employment programs aim to improve information on the labor market, but most Kenya’s Jua Kali voucher program focus almost exclusively on formal training In the mid-1990s, Kenya’s Jua Kali Program (Kiswahili for “work under providers and wage employment. Despite the the hot sun”) offered training vouchers to HEs operating small fabrica- lack of evidence and thorough testing of such tion or repair workshops. Eligible participants had to pay 10 percent approaches, the cost of providing information of the training cost and received vouchers to cover the remainder. is low and the potential for impact is high. Gov- The vouchers produced a positive supply response, predominantly ernments could systematically disseminate and from NGOs and master craftsmen in the informal sector. These suppli- communicate data on labor market earnings ers developed new programs tailored to the needs of voucher recipi- or training options collected through house- ents and offered the programs at times that suited participants’ work hold surveys or surveys of training providers. schedules. Public institutions showed little interest in adapting their Schools could also provide information on traditional programs to respond to this new source of demand (Adams employment and training opportunities. 2001). In its pilot stage, the Jua Kali Program successfully expanded the supply and reduced the cost of training for workers in the informal sector. There is evidence that the training had a positive effect on par- Intervening to Ensure Better Training ticipants’ earnings and strengthened the capacity of the local Jua Kali Options associations responsible for distributing the vouchers. When the pro- Given that governments have often stumbled gram was scaled up, problems in governance led to high administra- in their efforts to promote skills development, tive costs (Adams 2001; Riley and Steel 2000). Ghana offered a similar interventions to address market failures must voucher program targeted at informal sector enterprises in the early be assessed against the risks of government 1990s, but it failed largely because of flaws related to the marketing failures in the provision of training or training and distribution of vouchers (Johanson and Adams 2004). subsidies. Inefficiencies in public interventions Source: Reproduced from Adams, Johanson de Silva, and Razmara 2013. usually involve challenges in the policy-making process, governance, and institutional arrange- ments—especially as they relate to accountabil- Skills for productive Employment 99 ity—and mismatches between the rationale for Figure 3.24 Training costs and the price paid by government intervention and the ways gov- participants vary by type of training in Kenya ernments actually intervene (such as limited attempts to incorporate the role of information in an effective way). Examples of wasteful insti- Apprenticeship tutions abound. Training is often provided in a fragmented way by a large number of ministries Informal skills training with limited accountability, creating substantial inefficiencies and distortions. Many TVET sys- tems are inefficient and create distortions by Faith-based institutions subsidizing the supply of technical training of similar or inferior quality to the types of train- Formal ing widely delivered by the private sector. Given post-secondary this experience, public intervention should not TVET only be based on a clear understanding of mar- Formal ket failures but also emphasize efficiency and post-secondary TVET quality service delivery. Whenever public pro- (National Youth Service) vision or financing is considered, the efficiency Formal of the institutions involved and the quality of post-primary TVET (Youth Polytechnics) service delivered require careful attention. Given the large number of training institu- Youth Empowerment tions and wide range of private providers, the Program on-the- training available to Africa’s young people var- job training ies greatly in price and quality. In the presence 0 500 1,000 1,500 2,000 2,500 of active training markets, public interventions Cost (US$) need to be selective, performance driven, and Cost range Price range evidence based. Source: Franz 2011. Note: Approximate training cost and price in U.S. dollars. Measuring the Cost of Training Large variations in training costs, as well as in the share of costs borne by participants, pre- fully) cover tuition costs for both government vail across types of training—although sys- and private vocational programs” (Hicks et al. tematic data are hard to come by. Data from 2011). Kenya, illustrated in figure 3.24, give some idea The data from Kenya reflect three patterns of patterns in training costs.28 The data include in training costs that appear to be common in formal TVET (postsecondary TVET, postsec- other countries: ondary National Youth Service, and postpri- mary youth polytechnics) as well as informal • Formal TVET is very expensive. In Ghana, training delivered through faith-based insti- the cost of TVET per pupil is about five “Even if training tutions, private TVET providers, or appren- times the cost of primary education and almost three times the cost of senior sec- was available for ticeships. Training costs range from US$113 for the cheapest private TVET to US$204 for ondary education, in line with international free, it would be apprenticeships, US$1,942 for National Youth averages, in which TVET costs range from hardly possible Service, and US$1,704 for the most expensive two to three times the cost of secondary for me to get private TVET. The share paid by participants education (Adams et al. 2009). In Mozam- time to attend, ranges from all (in private TVET or appren- bique, TVET costs four times more than lest I lose the ticeships) to nothing (National Youth Service). secondary school (Fox et al. 2012). little income that The vouchers supplied in the Technical and • The cost of private training varies substan- I get daily in Vocational Vouchers Program were approxi- tially. The median cost of private training in mately US$460, which the program designers northern Uganda, for example, ranges from order to survive.” calculated was “sufficient to fully (or almost US$24 to US$444, depending on the type Tanzania 100 Youth EmploYmEnt in Sub-Saharan africa of training (Blattman, Fiala, and Martinez more likely to have flush toilets and electricity 2011). than their public, rural, and informal counter- • Informal apprenticeship is often among the parts, but they had comparable instructional cheaper training options. Sometimes pay- capital per student. ment is not even required. Among Rwan- Systematic evidence on the quality of dan owners of HEs who have been trained TVET institutions is hard to compile. Key through apprenticeships, only 40 percent indicators of the quality of training are atten- paid for the apprenticeship (Johanson and dance, dropout rates, and graduation rates. Gakuba 2011). Better attendance leads to better learning (Bjorvatn and Tungodden 2010), and high- quality programs have fewer dropouts. Many Measuring the Quality and providers do not track the percentage of par- Effectiveness of Training ticipants who complete training, however; The quality of various training providers is when data are available, it is common to find rarely measured directly. A survey of training that 25–50 percent of participants do not providers in Kenya found few differences in complete training. Yet high dropout rates are observable quality between public, private for- not inevitable (box 3.9). mal, and private informal training providers The share of enrollees who graduate from (Hicks et al. 2011). When young people were a training program (in other words, complete offered vouchers to take vocational training, and pass the final exam) is often reported to they had a slight preference for public train- be very low. For example, the pass rates for the ing centers (56 percent) compared to private Malawi craft and advanced craft examinations providers (44 percent). Instructors in public are less than 70 percent. For the trades, the pass institutions were more likely to have taken the rate is closer to 50 percent, and it is even lower secondary school exam and completed college, (on the order of 10–20 percent) in the trades yet the profiles of instructors in all institutions for which delivery is more complex, such as the were similar in many respects (they had the electrical and mechanical trades (World Bank same practical experience, for example). Dif- 2013). Moreover, the constraints facing young ferences in infrastructure also appear limited. women tend to lead to even higher dropout Urban and formal private institutions were rates for them (Cho et. al. 2013). Box 3.9 Can incentives improve training quality and participation? Beneficiary assessments consistently highlight serious issues trainees participated in a variety of contests and competi- with the quality of training. For example, a qualitative assess- tions. All of the measures contributed to a retention rate sur- ment of the Youth Empowerment Project in Kenya found passing 95 percent and an attendance rate of 90 percent. that the implementation of various training modules varied Incentives to provide good-quality training can also substantially in quality, and those variations played a key role improve attendance. In a business skills program in Tanzania, in participants’ evaluation of the modules and decisions to attendance, subjective evaluations of course quality, and self- drop out (KEPSA 2012). reported knowledge were higher when professional trainers Dropout rates can be limited by motivating training partic- did the teaching (Berge et al. 2011). A qualitative assessment ipants as well as providers. In Liberia, the Economic Empow- of training in life skills under the Liberia Youth Employment erment of Adolescent Girls and Young Women (EPAG) Pro- Project found large variations in quality among community gram used innovative strategies to ensure high participation trainers (World Bank and Republic of Liberia 2012). In the and training quality (Word Bank 2012b). Participants signed EPAG Program, training providers were incentivized through commitment forms, received small stipends contingent on performance bonuses, and frequent and unannounced visits attendance, were offered free child care, and were assigned to monitor the quality of training also helped to make the to small teams with mentors. In addition to those incentives, training program effective (World Bank 2012b) Skills for productive Employment 101 Box 3.10 Impact evaluation to build the evidence base on youth employment programs Impact evaluations provide rigorous evidence of a program’s Impact evaluations are often implemented at the scale effectiveness by estimating the program’s effects on final of pilots, with the result that the scalability and ultimate outcomes, based on an estimate of the counterfactual (the affordability of the evaluated programs remain a matter of outcomes that would have prevailed for program beneficia- debate, even when the results are positive. Moreover, it may ries if they had not participated in the program; see Gertler not always be possible to achieve similar positive results in et al. 2010). The hallmark of sound impact evaluations is other contexts. Despite these caveats, impact evaluations that they ensure that the counterfactual is credible—that provide critical information about whether a program can the only difference between treatment and control groups is work. When scaled-up programs are evaluated rigorously, participation in the program. the results provide evidence on their effectiveness in “real- Randomized assignment is the gold standard for impact world” conditions. evaluation. It generates fully comparable treatment and Few interventions to support young workers in Africa comparison groups by assigning program benefits randomly have been subject to impact evaluations over the years (Bet- (for example, by lottery) among equally eligible individuals, cherman, Olivas, and Dar 2004; Cho and Honorati 2012). households, or communities. Impact evaluations can rely on The impacts of a small number of programs targeting the other methods, although those methods typically require self-employed were evaluated between 2002 and 2012, and additional (often untestable) assumptions. Most impact eval- recent studies add somewhat to this tally, but more high- uations require baseline data as well as a solid compari- quality impact evaluations are greatly needed. Such studies son group, and they are best designed before a program is would specifically evaluate the best ways of designing youth implemented, to ensure that baseline data are collected and employment programs (including the best components to valid comparison groups identified. package together), the cost-effectiveness of those programs at scale, and their general equilibrium effects. Another indicator of quality is the prob- evaluations are required to compare employ- ability of securing a job after graduation. A ment outcomes among graduates to counter- 2008 study from Nigeria (Billetoft 2010), for factual employment outcomes, as well as to example, estimates that less than 30 percent of assess cost-effectiveness by comparing mea- polytechnic graduates secure a paid job within sured impacts to costs. Overall, the evidence their area of competence after completing their from Africa to guide specific programs for education. Polytechnic graduates with techni- improving skills remains thin. Building that cal or scientific training do better, however, evidence base is an urgent priority for action than those with more general training oriented (box 3.10). toward white-collar jobs. A study of public Given the large array of skills training avail- TVET graduates in Tanzania in 1996 found that able in the private sector, public financing or only 14 percent had found work upon gradua- direct provision of training needs to be selec- tion (Fluitman 2001). Studies in the mid-1990s tive, performance driven, and evidence based. in Mali and Madagascar estimated employ- Governments need to have rationales for the ment rates of 44 and 45 percent, respectively, value-added of public intervention in the mar- one to three years after graduation (Johanson ket for skills, and they must be sure not to dis- and Adams 2004). In Uganda, although two- place privately provided training. In addition, thirds of TVET graduates were working, only the objectives of funding or providing training 31 percent had a “permanent” job (Johanson need to be clear; it is particularly important to and Adams 2004). clarify whether such training aims to provide Aside from tracking employment outcomes pathways to productive employment in HEs or among graduates more systematically, impact wage employment. 102 Youth EmploYmEnt in Sub-Saharan africa Conclusion: A Skills Agenda promising and worthy of public investment. for Youth More evidence on what works in improving employment outcomes among youth, includ- This chapter has reviewed aspects of the skills ing careful evaluations, is clearly needed across agenda that cut across the three sectors of the board. employment: agriculture, HEs, and the modern Chapter 4 discusses the scope and priori- wage sector. The most urgent policy action is to ties for skills training in the agricultural sector. ensure that children and youths acquire foun- Traditional agricultural TVET and extension dational skills through quality basic education. have had a mixed record. However, new models The abysmal performance of education systems of service delivery that empower farmers and throughout Africa results in low worker pro- allow them to choose across a range of provid- ductivity, and the lack of foundational skills ers appear promising. These models include prevents youths from acquiring additional participatory farmer field schools as well as skills and reaching their full potential. It is a beneficiary-driven models to deliver extension pressing priority to improve substantially the services, build skills, and facilitate access to quality of schooling so that it results in actual information among young people. learning and skills acquisition for the growing Chapter 5 reviews the evidence on the effec- cohorts of youth who will enter the labor force tiveness of training for the HE sector. Young in the next 10 years. Early childhood develop- people often face multiple constraints in start- ment and nutrition must also be promoted ing a business. Programs attempting to build actively to ensure that children are more ready one skill at a time (such as technical skills, to learn when they enter school. business skills, or behavioral skills alone) have Having a range of skills, beyond cognitive had limited impacts. “Integrated” interventions and technical skills, matters for productive that build a range of complementary skills employment. Behavioral and socioemotional together are more promising. Especially prom- skills are often a neglected skill set. More atten- ising are “bundled interventions” that deliver tion is required to identify and build the behav- integrated skills training along with assistance ioral skills that contribute to productivity— with accessing start-up capital. Finally, infor- including the skills that employers demand. mal training delivered by private providers is There is room to leverage education systems to normal in the HE sector, so there is scope for impart behavioral skills as well as to consider governments to leverage NGOs and private including a behavioral skills component in providers to support youth through demand- youth employment programs. driven, performance-based models. Skills “markets” are active in Africa, operat- Chapter 6 outlines specific policy recom- ing through the private provision of appren- mendations to build skills for the modern wage ticeships in the HE sector, on-the-job training sector. Overall, the experience with TVET has in the modern wage sector, and other channels. been disappointing. Governments in Africa In this context, there is a role for governments should focus on support for public goods in to provide information to youth about training TVET, such as quality assurance and infor- and occupational choices as well as to facilitate mation, and they should facilitate access to greater access to existing training opportunities training for poor and disadvantaged youths. among disadvantaged groups, such as women Post-school vocational training should only be and the poor. provided selectively, based on careful targeting The scope for direct government interven- and demand-driven models that link employ- tion in the skills market is more limited. It ers and training providers. To the extent that needs to be well motivated by clearly identify- governments support specific training options, ing the rationale for public sector engagement, those options should emphasize portable skills and it needs to be evidence based. Chapters 4–6 rather than the firm- or job-specific skills that review the evidence for each sector and iden- employers should already have an incentive to tify interventions and approaches that are more provide. Programs for disadvantaged youths Skills for productive Employment 103 that integrate training with internships show been concentrated at lower levels of education.” promise—but the challenge is to make them As Teal (2010) finds, “Returns to education, cost-effective. measured both by macro production functions Overall, education and skills matter. The and by micro earning functions, are highest starting point in meeting the youth employ- for those with higher levels of education.” He argues, “Growth has been more closely linked to ment challenge is to improve human capital investment in physical capital than in education, by providing education and allowing youths to and this may well reflect the fact that education acquire the skills needed for productive work. is most valuable when it is linked to technol- At the same time, it is only the starting point. ogy which requires higher skills.” See also Moll Governments cannot solely “train their way” (1996); Appleton, Hoddinott, and Krishnan to more and better jobs for youth. Policies (1999); Schultz (2004); and Lassibille and Tan addressing the economic and business envi- (2005). ronment have a critical role to play, and chap- 6. This categorization is only indicative, and many ters 4–6 examine the relative roles of policies skills are interconnected. For instance, self- that promote skills and those that promote the control is a behavioral skill that also builds on business environment for the agricultural sec- cognitive ability. tor, the HE sector, and the modern wage sector. 7. In Botswana form-two and in South Africa grade-nine students were tested, corresponding Notes to nine years of schooling; in Ghana form-two 1. Only in Ghana do self-employed workers with students were tested, corresponding to eight incomplete primary education earn significantly years of schooling. more (20 percent) than self-employed workers 8. Other studies find that differences in a range without education. of cognitive skills explain part of the earnings. 2. Low differentials in earnings premiums can- See, for example, Glewwe (1991) in Ghana; Moll not be explained by the recent surge in primary (1998) for computational skills in South Africa; school enrollment and completion in Africa. Denny, Harmon, and O’Sullivan (2003) for Most of that surge happened in the late 2000s, functional literacy; Boissiere, Knight, and Sabot when the cohorts benefiting from increased (1985) in Kenya and Tanzania; Heckman and enrollment had not reached the labor force. Vytlacil (2000) in the United States; or Azam, 3. In Ghana, the mean wage for workers with Chin, and Prakash (2010) in India. incomplete primary schooling was lower than 9. Ozier (2010) is based on the fact that in Kenya the mean wage of workers without any edu- the probability of admission to a government cation, but the difference was not statistically secondary school rises sharply at a score close significant. to the national mean on a standardized eighth- 4. The general patterns for earnings increments grade examination. The causal effect of school- hold for both men and women, as well as across ing is estimated by comparing those who score urban and rural areas. In Rwanda, Tanzania, and just below the national mean to those who score Uganda, earnings differentials tend on average just above it. to be higher and steeper for women. In Rwanda 10. See www.sdindicators.org. and Uganda, earnings differentials are higher in 11. For Madagascar, see Fernald et al. (2011); for urban areas, while in Ghana and Tanzania, dif- Mozambique, see Naudeau et al. (2010); for ferentials are higher in rural areas. other parts of the world, see Paxson and Schady 5. See Aromolaran (2006); Rankin, Sandefur, and (2007); Case, Lubotsky, and Paxson (2002). Teal (2010). For example, Rankin, Sandefur, and 12. Walker et al. (2007); Engle et al. (2007). See also Teal (2010) find clear evidence of convexity in the Maternal and Child Undernutrition Series the returns to education for the self-employed of The Lancet, http://www.thelancet.com/series/ in urban Ghana and Tanzania, along with low maternal-and-child-undernutrition. average returns. Söderbom, Teal, and Hard- 13. For home visits, see Attanasio et al. forthcoming; ing (2006) find convex earnings functions in Macours et al. forthcoming. For community- both Ghana and Tanzania. They suggest that based centers, see Martinez, Naudeau, and “convexity may be part of the explanation as Pereira (2012). For preschools, see Attanasio to how rapid expansion of education in Africa and Vera-Hernández (2004); Behrman, Cheng, has generated so little growth if expansion has and Todd (2004); Berlinski, Galiani, and Gertler 104 Youth EmploYmEnt in Sub-Saharan africa (2006); Berlinski, Galiani, and Manacorda individuals are willing to fund apprenticeships (2008). because they can reap the returns to the specific 14. Based on World Bank, Lesotho skills and training if they manage to acquire the capital to employment survey, 2011. start their own firm and replicate the technology 15. See Krauss et al. (2005), who study a sample of and business practice of the apprenticeship firm. owners of firms employing seven persons on Apprentices are constrained only by capital from average. All had been active for more than a year, becoming apprenticed entrepreneurs. and half had started with less than US$1,000; 37 23. In northern Uganda, when youth groups were percent of the firms were informal. Krauss and provided cash to pay for training as part of her co-authors measure business success as busi- the Youth Opportunities Program, 33 percent ness growth or number of employees. They find chose to obtain training from local artisans and that other skills (learning orientation, autonomy 32 percent chose to obtain it from informal orientation, competitive aggressiveness, innova- training institutions. In Nigeria, the volume tive orientation, risk-taking orientation) are less of informal training far exceeds that of formal strongly associated with business success than TVET. Despite being short and of varying qual- entrepreneurial and achievement orientation. ity, informal training is in great demand (Bil- 16. Heckman, Stixrud, and Urzua (2006). The Rot- letoft 2010, 185). Kenya also has a remarkably ter Scale measures the degree of control that large, diverse array of private training providers individuals feel they possess over their lives. (Hicks et al. 2011; Franz 2011). 17. www.sdindicators.org. 24. Youths who did not receive cash grants chose a 18. The remainder of this discussion focuses on slightly different mix of training. Short courses young adults ages 25–34, because younger indi- on business and management (27 percent) or viduals may still be in school, and results includ- agribusiness and farming (7 percent) were more ing them would not reflect the apprenticeship or common. TVET experience accurately. 25. In the statistics based on household surveys used 19. In Ghana, only 12.7 percent of students who in this chapter, for example, it is not possible to advance beyond lower-secondary education apportion TVET enrollment between public enroll in a TVET institution (World Bank 2009). and private providers or even to know with cer- Low enrollment partly reflects limited capacity; tainty whether the numbers correspond to both only 5 percent of junior secondary students public and private TVET or to public TVET could have expected to obtain a place in a pub- alone. lic TVET institution. If private TVET capacity 26. A recent review of training in Malawi estimates is included, the percentage rises to 7.2 percent. that, over the past 10 years, female enrollment In Nigeria, less than 1 percent of university in apprenticeships was 21–35 percent of male enrollment is in a technical college and about enrollment (World Bank 2013). 20 percent is in a more advanced polytechnic 27. While the program was mainly for “industrial (World Bank 2011b). Rwanda’s vocational train- courses,” it also allowed students to enroll in ing centers enroll just over 10,000 trainees; of more academic courses (for example, computer these, about 4,700 are enrolled in a public train- training) and to cover fees up to the level of the ing center—a modest number compared to the average two-year industrial course. roughly 260,000 students enrolled at the second- 28. These costs exclude opportunity costs, which ary level (World Bank 2011a). can be large for training of long duration. 20. An estimated 100,000 youths are enrolled in formal TVET, compared with 150,000 in tradi- References tional apprenticeships (Franz 2011). Abdulai, Awudu, and Wallace E. Huffman. 2005. 21. Other studies point to similar patterns. In Nige- “The Diffusion of New Agricultural Technolo- gies: The Case of Crossbreeding Technology in ria, it is estimated that more than half of small Tanzania.” American Journal of Agricultural Eco- business operators acquired their skills in the nomics 87 (3): 645–59. informal sector from master craftsmen or mas- ter trainers (Billetoft 2010). In Rwanda, 25 per- Aber, Lawrence, and Stephanie Jones. 1997. “Indi- cators of Positive Development in Early Child- cent of HE owners report having been trained as hood: Improving Concepts and Measures.” In an apprentice (Johanson and Gakuba 2011). Indicators of Children’s Well-being, edited by 22. Frazer (2006) finds that apprenticeship train- Robert Hauser, Brett Brown, and William ing increases an individual’s productivity in Prosser, 395–427. New York: Russell Sage the current firm but not in any other firm. Still, Foundation. Skills for productive Employment 105 Adams, Arvil V. 2001. “Assessment of the Jua Kali tional Institute for Educational Planning (IIEP) Pilot Voucher Program.” World Bank, Washing- and the United Nations Educational, Scientific, ton, DC. and Cultural Organization (UNESCO), Paris. Adams, Arvil V., Sara Johansson de Silva, and Atchoarena, David, and Paul Esquieu. 2002. “Pri- Setareh Razmara. 2013. Improving Skills Develop- vate Technical and Vocational Education in Sub- ment in the Informal Sector: Strategies for Sub- Saharan Africa: Provision Patterns and Policy Saharan Africa. Direction in Development Series. Issues.” International Institute for Educational Washington, DC: World Bank. Planning (IIEP) and the United Nations Edu- Adams, Arvil V., Harold Coulombe, Quentin cational, Scientific, and Cultural Organization Wodon, and Setarah Razmara. 2009. “Education, (UNESCO), Paris. Employment, and Earnings in Ghana.” In Ghana: Attanasio, Orazio, Emla Fitzsimons, Sally Job Creation and Skills Development, Vol. 2. Grantham-McGregor, Costas Meghir, and Marta Washington, DC: World Bank. Rubio-Codina. Forthcoming. “Stimulation and Aedo, Christian, and Ian Walker. 2012. Skills Childhood Development in Colombia: The for the 21st Century in Latin America and the Impact of a Scalable Intervention.” Institute for Caribbean. Washington, DC: World Bank. doi: Fiscal Studies, University College, London. 10.1596/978-0-8213-8971-3. Attanasio, Orazio, and Marcos Vera-Hernández. AfDB (African Development Bank) and OECD 2004. “Medium- and Long-Run Effects of Nutri- (Organisation for Economic Co-operation and tion and Child Care: Evaluation of a Community Development). 2008. African Economic Out- Nursery Program in Rural Colombia.” Working look 2008. Paris: AfDB and OECD. doi: 10.1787/ Paper EWP04/06, Institute for Fiscal Studies, aeo-2008-en. University College, London. Al-Samarrai, Samer, and Paul Bennell. 2007. Azam, Mehtabul, Aimee Chin, and Nishith Prakash. “Where Has All the Education Gone in Sub- 2010. “The Returns to English-Language Skills in Saharan Africa? Employment and Other Out- India.” IZA Discussion Paper 4802, Institute for comes among Secondary School and University the Study of Labor, Bonn. Leavers.” Journal of Development Studies 43 (7): Bandiera, Oriana, Robin Burgess, Narayam Das, 1270–300. Selim Gulesci, Imran Rasul, and Munshi Sulai- Almeida, Rita, Jere Behrman, and David Robalino, man. 2013. “Can Basic Entrepreneurship Trans- eds. 2012. The Right Skills for the Jobs? Rethinking form the Economic Lives of the Poor?” IZA Training Policies for Workers. Washington, DC: Discussion Paper 7386, Institute for the Study of World Bank. Labor, Bonn. Almlund, Mathilde, Angela Lee Duckworth, James Banerjee, Abhijit V., and Esther Duflo. 2005. J. Heckman, and Tim Kautz. 2011. “Personal- “Growth Theory through the Lens of Develop- ity Psychology and Economics.” IZA Discussion ment Economics.” In Handbook of Economic Paper 5500, Institute for the Study of Labor, Growth, edited by Philippe Aghion and Steven N. Bonn. Durlauf, 473–52. Amsterdam: Elsevier. Appleton, Simon, John Hoddinott, and Pramila Behrman, Jere, Yingmei Cheng, and Petra Todd. Krishnan. 1999. “The Gender Wage Gap in Three 2004. “Evaluating Preschool Programs When African Countries.” Economic Development and Length of Exposure to the Program Varies: A Cultural Change 47 (2): 289–313. Nonparametric Approach.” Review of Economics Aromolaran, Adebayo B. 2006. “Estimates of and Statistics 86 (1): 108–32. Mincerian Returns to Schooling in Nigeria.” Berge, Lars Ivar Oppedal, Kjetil Bjorvatn, and Oxford Development Studies 34 (2): 265–92. Bertil Tungodden. 2011. “Human and Financial Aslam, Monazza, and Kim Lehrer. 2012. “Learn- Capital for Microenterprise Development: Evi- ing by Doing: Skills and Jobs in Urban Ghana.” dence from a Field and Lab Experiment.” NHH CSAE Working Paper 2012-15, Centre for the Department of Economics Discussion Paper Study of African Economies (CSAE), Oxford. 1/2011, Chr. Michelsen Institute (CMI), Bergen. Atchoarena, David, and Andre Delluc. 2001. “Revis- Berlinski, Samuel, Sebastian Galiani, and Paul iting Technical and Vocational Education in Gertler. 2006. “The Effect of Pre-Primary Educa- Sub-Saharan Africa.” Report for the World Bank, tion on Primary School Performance.” Working International Institute for Educational Planning, Paper WP06/04, Institute for Fiscal Studies, Uni- Paris. versity College, London. ———. 2002. “Revisiting Technical and Vocational Berlinski, Samuel, Sebastian Galiani, and Marco Education in Sub-Saharan Africa: An Update on Manacorda. 2008. “Giving Children a Better Start Trends, Innovations, and Challenges.” Interna- in Life: Preschool Attendance and School-Age 106 Youth EmploYmEnt in Sub-Saharan africa Profiles.” Journal of Public Economics 92 (5–6): “Financial Education and Behavior Formation: 1416–40. Large-Scale Experimental Evidence from Brazil.” Besley, Timothy, and Anne Case. 1993. “Modeling World Bank, Washington, DC. Technology Adoption in Developing Countries.” Bruns, Barbara, Deon Filmer, and Harry Anthony American Economic Review 83 (2): 396–402. Patrinos. 2011. Making Schools Work: New Evi- Betcherman, Gordon, Karina Olivas, and Amit Dar. dence on Accountability Reforms. 2004. “Active Labor Market Programs: New Evi- Washington, DC: World Bank. doi: dence from Evaluations with Particular Attention 10.1596/978-0-8213-8679-8. to Developing and Transition Countries.” Social Card, David. 1999. “The Causal Effect of Education Protection Discussion Paper 04012, World Bank, on Earnings.” In Handbook of Labor Economics, Washington, DC. Vol. 3, edited by Orley Ashenfelter and David Biggs, Tyler, Manju Shah, and Pradeep Srivastava. Card, 1802–59. Amsterdam: Elsevier. 1995. “Technological Capabilities and Learn- Carneiro, Pedro, Emanuela Galasso, and Rita Ginja. ing in African Enterprises.” Technical Paper 288, 2010. “The Impact of Providing Psycho-Social World Bank, Washington, DC. Support to Indigent Families and Increasing Bigsten, Arne, Anders Isaksson, Mans Soderbom, Their Access to Social Services: Evaluating Chile Paul Collier, Albert Zeufack, Stefan Dercon, Mar- Solidario.” Evaluation report, University College, cel Fafchamps, and Jan Willem. 2000. “Rates of London. Return on Physical and Human Capital in Afri- ca’s Manufacturing Sector.” Economic Develop- Case, Anne, Darren Lubotsky, and Christina Pax- ment and Cultural Change 48 (4): 801–27. son. 2002. “Economic Status and Health in Childhood: The Origins of the Gradient.” Ameri- Billetoft, Jorgen. 2010. “Labor Market Trends and can Economic Review 92 (5): 1308–34. Skills Development.” In Putting Nigeria to Work: A Strategy for Employment and Growth, edited Cho, Yoonyoung, and Maddalena Honorati. 2012. by Volker Treichel, 167–202. Washington, DC: “Entrepreneurship Programs in Developing World Bank. Countries: A Meta Regression Analysis.” Policy Bjorvatn, Kjetil, and Bertil Tungodden. 2010. Research Working Paper 6202, World Bank, “Teaching Business in Tanzania: Evaluating Par- Washington, DC. ticipation and Performance.” Journal of the Euro- Cho, Yoonyoung, Davie Kalomba, Ahmed Mushfiq, pean Economic Association 8 (2–3): 561–70. and Victor Orozco. 2013. “Gender Differences in Blattman, Christopher, Nathan Fiala, and Sebastian the Effects of Vocational Training: Constraints Martinez. 2011. “Can Employment Programs on Women and Drop-Out Behavior.” Policy Reduce Poverty and Social Instability? Experi- Research Working Paper 6545, World Bank, mental Evidence from a Ugandan Aid Program.” Washington, DC. Columbia University, New York. Clemens, Michael. 2004. “The Long Walk to School: Boesel, David, Nabeel Alsalam, and Thomas M. International Education Goals in Historical Per- Smith. 1998. Research Synthesis: Education and spective.” CGD Working Paper 37, Center for Labor Market Performance of GED Recipients. Global Development, Washington, DC. Washington, DC: U.S. Department of Education, Cloutier, Marie-Hélène, C. Reinstadtler, and Isa- January. bel Beltran. 2011. “Making the Grade: Assessing Boissiere, M., J. B. Knight, and R. H. Sabot. 1985. Literacy and Numeracy in African Countries.” “Earnings, Schooling, Ability, and Cogni- DIME Brief, World Bank, Washington, DC. tive Skills.” American Economic Review 75 (5): http://go.worldbank.org/15Y7VXO7B0. 1016–30. Cobb-Clark, Deborah A., and Michelle Tan. 2010. Boone, Peter, Ila Fazzio, Kameshwari Jandhyala, “Noncognitive skills, occupational attainment, Chitra Jayanty, Gangadhar Jayanty, Simon and relative wages” Labour Economics 18 (1): Johnson, Vimala Ramachandrin, Filipa Silva, 1–23. and Zhaoguo Zhan. 2013. “The Surprisingly Dire Situation of Children’s Education in Rural Cojocaru, Alexandru. 2011. “Tanzania: Skills and the West Africa: Results from the CREO Study in Informal Sector.” World Bank, Washington, DC. Guinea-Bissau (Comprehensive Review of Edu- CREATE (Consortium for Research on Educational cation Outcomes).” NBER Working Paper 18971, Access, Transitions, and Equity). 2010. “Com- National Bureau of Economic Research, Cam- plementary Education and Access to Primary bridge, MA. Schooling in Northern Ghana.” CREATE Ghana Bruhn, Miriam, Luciana de Souza Leão, Arianna Policy Brief 2, September. http://www.create-rpc. Legovini, Rogelio Marchetti, and Bilal Zia. 2013. org/pdf_documents/Ghana_Policy_Brief_2.pdf. Skills for productive Employment 107 Cunha, Flavio, and James Heckman. 2007. “The Experiment in Ghana.” CSAE Working Paper Technology of Skill Formation.” American Eco- 2011-11, Centre for the Study of African Econo- nomic Review 97 (2): 31–47. mies, Oxford. Darvas, Peter. 2012. “Demand and Supply of Tech- Fafchamps, Marcel, and Christopher Woodruff. nical and Vocational Skills in Ghana.” World 2012. “Identifying and Relaxing Constraints to Bank, Washington, DC. Employment Generation in Small-Scale African De Beyer, Joy. 1990. “The Incidence and Impact on Enterprises.” University of Oxford. Earnings of Formal Training Provided by Enter- Fasih, Tazeen, Geeta Kingdon, Harry Anthony prises in Kenya and Tanzania.” Economics of Edu- Patrinos, Chris Sakellariou, and Mans Soderbom. cation Review 9 (4): 321–30. 2012. “Heterogeneous Returns to Education in the Labor Market.” Policy Research Working Delors, Jacques. 1996. Learning: The Treasure Paper 6170, World Bank, Washington, DC. Within. Report to UNESCO of the International Commission on Education for the Twenty-first Fernald, Lia C. H., Ann Weber, Emanuela Galasso, Century. Paris: United Nations Educational, Sci- and Lisy Ratsifandrihamanana. 2011. “Socioeco- entific, and Cultural Organization. nomic Gradients and Child Development in a Very Low Income Population: Evidence from Madagas- Denny, Kevin J., Colm P. Harmon, and Vincent car.” Developmental Science 14 (4): 832–47. O’Sullivan. 2003. “Functional Literacy, Edu- cational Attainment, and Earnings: A Multi- Fletcher, Jason M. 2012. “The Effects of Personality Country Comparison.” School of Economics, Traits on Adult Labor Market Outcomes: Evi- University College, Dublin. dence from Siblings.” IZA Discussion Paper 6391, Institute for the Study of Labor, Bonn. DeStefano, Joseph, Audrey-Marie Schuh Moore, David Balwanz, and Ash Harwell. 2006. “Meet- Fluitman, Fred. 2001. “Working, But Not Well: ing EFA: Reaching the Underserved through Notes on the Nature and Extent of Employment Complementary Models of Effective Schooling.” Problems in Sub-Sahara Africa.” ITC Occasional EQUIP 2 Working Paper, U.S. Agency for Inter- Paper, International Training Centre, Interna- national Development and Academy for Educa- tional Labour Organization, Turin. tion Development, Washington, DC. Foster, Andrew D., and Mark R. Rosenzweig. 1995. De Wolf, Stefan, Yves Rolland Rakotoarisoa, Lau- “Learning by Doing and Learning from Others: rence Vanpaeschen, and Honoré Rabekoto. 2008. Human Capital and Technical Change in Agri- Madagascar: Le Grand Livre des Petits Métiers: culture.” Journal of Political Economy 103 (6): Portraits of Daily Life Professions. Belgium: Snoek 1176–209. Publishers. ———. 2010. “Microeconomics of Technology DFID (Department for International Develop- Adoption.” Yale University, Economic Growth ment, UK). 2012. “School for Life ‘Literacy for Center, New Haven, CT. Change.’” DFID Annual Review (December). Fox, Louise, Lucrecia Santibañez, Vy Nguyen, Duflo, Esther. 2001. “Schooling and Labor Market and Pierre André. 2012. “Education Reform in Consequences of School Construction in Indone- Mozambique: Lessons and Challenges.” World sia: Evidence from an Unusual Policy Experiment.” Bank, Washington, DC. American Economic Review 91 (4): 795–813. Franz, Jutta. 2011. “Realizing the Youth Dividend Dupas, Pascaline, and Jonathan Robinson. 2013. in Kenya through Skills for the Informal Sector: “Savings Constraints and Microenterprise Devel- Institutional Assessment of Skills Development opment: Evidence from a Field Experiment in and Youth Employment Promotion Programmes Kenya.” American Economic Journal: Applied Eco- and Projects.” World Bank, Washington, DC. nomics 5 (1): 163–92. Frazer, Garth. 2006. “Learning the Master’s Trade: Engle, Patrice, Maureen Black, Jere Behrman, Apprenticeship and Human Capital in Ghana.” Meena Cabral de Mello, Paul Gertler, Lydia Journal of Development Economics 81 (2): 259–98. Kapiriri, Reynaldo Martorell, Mary Young, and Frese, Michael, Stephanie I. Krauss, Nina Keith, the International Child Development Steer- Susanne Escher, Rafal Grabarkiewicz, Siv Tonje, ing Group. 2007. “Strategies to Avoid the Loss T. Luneng, Constanze Heers, Jens Unger, and of Developmental Potential in More Than 200 Christian Friedrich. 2007. “Business Owners’ Million Children in the Developing World.” The Action Planning and Its Relationship to Business Lancet 369 (9557): 229–42. Success in Three African Countries.” Journal of Fafchamps, Marcel, David McKenzie, Simon Applied Psychology 92 (6): 1481–98. Quinn, and Christopher Woodruff. 2011. “When FXB Center for Health and Human Rights. 2012. Is Capital Enough to Get Female Microenter- “Building Youth Readiness: An Integrated prises Growing? Evidence from a Randomized Approach to Promoting Healthy Functioning 108 Youth EmploYmEnt in Sub-Saharan africa among War-Affected Youth in Sierra Leone.” Haan, Hans Christian, and Nicholas Serrière. 2002. Harvard University, Cambridge, MA. “Training for Work in the Informal Sector: Fresh Gertler, Paul, James Heckman, Rodrigo Pinto, Evidence from West and Central Africa.” Inter- national Training Centre, International Labor Arianna Zanolini, Christel Vermeersch, Susan Organization, Turin. Walker, Susan M. Chang, and Sally Grantham- McGregor. 2013. “Labor Market Returns to Early Hanushek, Eric A., and Ludger Woessmann. 2012. Childhood Stimulation: A 20-Year Follow-up to “Do Better Schools Lead to More Growth? an Experimental Intervention in Jamaica.” NBER Cognitive Skills, Economic Outcomes, and Working Paper 19185, National Bureau of Eco- Causation.” Journal of Economic Growth 17 (4): nomic Research Cambridge, MA. 267–321. Gertler, Paul, Sebastian Martinez, Patrick Premand, Heckman, J., L. Lochner, and J. Taber. 1998. Laura B. Rawlings, and Christel M. J. Vermeersch. “Explaining Rising Wage Inequality: Explora- 2010. Impact Evaluation in Practice. Washington, tions with a Dynamic General Equilibrium DC: World Bank. Model of Earnings with Heterogeneous Agents.” Review of Economic Dynamics 1 (1): 1–58. Glaub, Matthias E. 2009. Training Personal Initiative to Business Owners in Developing Countries: A Heckman, James A., Jora Stixrud, and Sergio Urzua. Theoretically Derived Intervention and Its Evalu- 2006. “The Effects of Cognitive and Noncogni- ation. Ph.D. thesis, Justus-Liebig University, tive Abilities on Labor Market Outcomes and Social Behavior.” Journal of Labor Economics 24 Giessen. (3): 411–82. Glewwe, Paul. 1991. “Schooling, Skills, and the Heckman, James, and Edward Vytlacil. 2000. “Iden- Returns to Government Investment in Educa- tifying the Role of Cognitive Ability in Explain- tion: An Exploration Using Data from Ghana.” ing the Level of and Change in the Return to Living Standards Measurement LSM76, World Schooling.” Review of Economics and Statistics 83 Bank, Washington, DC. (1): 1–12. ———. 2002. “Schools and Skills in Developing Helmers, Christian, and Manasa Patnam. 2011. Countries: Education Policies and Socioeco- “The Formation and Evolution of Childhood nomic Outcomes.” Journal of Economic Literature Skill Acquisition: Evidence from India.” Journal 40 (2): 436–82. of Development Economics 95 (2): 252–66. Glewwe, Paul, Qiuqiong Huang, and Albert Park. Hicks, Joan Hamory, Michael Kremer, Isaac Mbit, 2011. “Cognitive Skills, Non-Cognitive Skills, and Edward Miguel. 2011. “Vocational Education and the Employment and Wages of Young Adults Voucher Delivery and Labor Market Returns: A in Rural China.” Paper presented at the annual Randomized Evaluation among Kenyan Youth.” meeting of the Agricultural and Applied Eco- Report for Spanish Impact Evaluation Fund nomics Association, July 24–26, Pittsburgh. (SIEF) Phase II, World Bank, Washington, DC. Gove, Amber, and Peter Cvelich. 2010. “Early Read- Huffman, Wallace E. 1977. “Allocative Efficiency: ing: Igniting Education for All: A Report by the The Role of Human Capital.” Quarterly Journal Early Grade Learning Community of Practice.” of Economics 91 (February): 59–79. Research Triangle Institute, Research Triangle Hungi, Njora, Demus Makuwa, Kenneth Ross, Park, NC. Mioko Saito, Stéphanie Dolata, Frank van Cap- Grantham-McGregor, Sally, Yin Bun Cheung, San- pelle, Laura Paviot, and Jocelyne Vellien. 2010. tiago Cueto, Paul Glewwe, Linda Richter, and “SACMEQ III Project Results: Pupil Achieve- Barbara Strupp. 2007. “Developmental Potential ment Levels in Reading and Mathematics.” in the First Five Years for Children in Developing Working Document 1, Southern Africa Con- Countries.” The Lancet 369 (9555): 60–70. sortium for Measuring Educational Quality (SACMEQ), Harare. http://www.sacmeq.org/ Groh, Matthew, Nandini Krishnan, David McKen- downloads/sacmeqIII/WD01_SACMEQ_III_ zie, and Tara Vishwanath. 2012. “Soft Skills or Results_Pupil_Achievement.pdf. Hard Cash? The Impact of Training and Wage Subsidy Programs on Female Youth Employment Johanson, Richard K., and Avril V. Adams. 2004. in Jordan.” Policy Research Working Paper 6141, Skills Development in Sub-Saharan Africa. Wash- World Bank, Washington, DC. ington, DC: World Bank. Haan, Hans Christian. 2001. “Training for Work in Johanson, Richard, and Theogene Kayiranga the Informal Sector: Fresh Evidence from East- Gakuba. 2011. “Rwanda: Training for the Infor- ern and Southern Africa.” International Train- mal Sector.” World Bank, Washington, DC. ing Centre, International Labour Organization, KEPSA (Kenya Private Sector Alliance). 2012. “Ben- Turin. eficiary Assessment of Kenya Youth Empower- Skills for productive Employment 109 ment Project (KYEP) Training and Internship Mano, Yukichi, Alhassan Iddrisu, Yutaka Yoshino, Component.” KEPSA, March. and Tetsushi Sonobe. 2011. “How Can Micro and Kitaev, Igor, with contributions from J. Glover, A. Small Enterprises in Sub-Saharan Africa Become Melomey, Th. Coleman, and B. Kaluba. 2003. More Productive? The Impacts of Experimental Basic Managerial Training.” World Development “Phase II: Synthesis of Main Findings from Two 40(3):458–68. Case Studies Carried out in Ghana and Zambia on Private Technical and Vocational Education Martinez, Sebastian, Sophie Naudeau, and Victor and Training (TVET).” International Institute Pereira. 2012. “The Promise of Preschool in for Educational Planning (IIEP) and UNESCO, Africa: A Randomized Impact Evaluation of Paris. Early Childhood Development in Rural Mozam- bique.” World Bank and Save the Children, Koop, Sabine, Tamara De Reu, and Michael Frese. Washington, DC. 2000. “Sociodemographic Factors, Entrepreneur- ial Orientation, Personal Initiative, and Envi- Mattero, Minna. 2010. “Second Chance Education ronmental Problems in Uganda.” In Success and for Out-of-School Youth: A Conceptual Frame- Failure of Microbusiness Owners in Africa: A Psy- work and Review of Programs.” World Bank, chological Approach, edited by M. Frese, 55–76. Washington, DC. Westport, CT: Quorum. Mingat, Alain, Blandine Ledoux, and Ramahatra Krauss, Stefanie I., Michael Frese, Christian Fried- Rakotomalala. 2010. Developing Post-Primary rich, and Jens M. Unger. 2005. “Entrepreneurial Education in Sub-Saharan Africa: Assessing the Orientation: A Psychological Model of Success Financial Sustainability of Alternative Pathways. among Southern African Small Business Own- Washington, DC: World Bank. ers.” European Journal of Work and Organiza- Moll, Peter G. 1996. “The Collapse of Primary tional Psychology 14 (3): 315–44. Schooling Returns in South Africa, 1960–90.” Kuepie, Mathias, Christophe J. Nordman, and Oxford Bulletin of Economics and Statistics 58 (1): Francois Roubaud. 2009. “Education and Earn- 185–209. ings in Urban West Africa.” Journal of Compara- ———. 1998. “Primary Schooling, Cognitive Skills, tive Economics 37 (3): 491–515. and Wages in South Africa.” Economica 65 (258): 263–84. Lassibille, Gerard, and Jee-Peng Tan. 2005. “The Returns to Education in Rwanda.” Journal of Monk, Courtney, Justin Sandefur, and Francis Teal. African Economies 14 (1): 92–116. 2008. “Does Doing an Apprenticeship Pay Off? Evidence from Ghana.” CSAE Working Paper Lee, Jean N., and David Newhouse. 2012. “Cogni- 2008-08, Centre for the Study of African Econo- tive Skills and Youth Labor Market Outcomes.” mies, Oxford. Background paper for World Development Report 2013: Jobs. World Bank, Washington, DC. Mullis, Ina V.S., Michael O. Martin, Pierre Foy, and Alka Arora. 2012. TIMSS 2011 International Lundberg, Mattias, and Alice Wuermli, eds. 2012. Results in Mathematics. TIMSS & PIRLS Inter- Children and Youth in Crisis: Protecting and Pro- national Study Center, Lynch School of Educa- moting Human Development in Times of Eco- tion, Boston College, Chestnut Hill, MA, and nomic Shocks. Washington, DC: World Bank. doi: International Association for the Evaluation of 10.1596/978-0-8213-9547-9. Educational Achievement (IEA) IEA Secretariat, Macours, Karen, Patrick Premand, Norbert Schady, Amsterdam, the Netherlands. and Renos Vakis. Forthcoming. “Experimental Naudeau, Sophie, Sebastian Martinez, Patrick Pre- Evidence from an Early Childhood Parenting mand, and Deon Filmer. 2010. “Cognitive Devel- Intervention in Nicaragua.” World Bank, Wash- opment among Young Children in Developing ington, DC. Countries.” In No Small Matter: The Impact of Macours, Karen, Patrick Premand, and Renos Vakis. Poverty, Shocks, and Human Capital Investment in 2013. “Demand Versus Returns? Pro-Poor Tar- Early Childhood Development, edited by Harold geting of Business Grants and Vocational Skills Alderman, 9–50. Washington, DC: World Bank. Training.” Policy Research Working Paper 6389, Nayar, Reema, Pablo Gottret, Mitra Pradeep, Gor- World Bank, Washington, DC. don Betcherman, Yue Man Lee, Indhira Santos, Maluccio, John A., John Hoddinott, Jere R. Mahesh Dahal, and Maheshwor Shrestha. 2012. Behrman, Reynaldo Martorell, Agnes R. “More and Better Jobs in South Asia.” South Quisumbing, and Aryeh D. Stein. 2009. “The Asia Development Matters Report 66229, World Impact of Improving Nutrition during Early Bank, Washington, DC. Childhood on Education among Guatemalan Nell, Marian, and Janet Shapiro. 1999. “Traditional Adults.” Economic Journal 119 (537): 734–63. Apprenticeship Practice in Dar es Salaam: A 110 Youth EmploYmEnt in Sub-Saharan africa Study.” Consultancy report prepared for GTZ/ sion Paper 6512, Institute for the Study of Labor, VETA, GTZ/VETA, Dar es Salaam. Bonn. Nicholson, Sue. 2006. “Accelerated Learning Pro- Rosholm, Michael, Helena Skyt Nielsen, and grams in Post Conflict Settings: A Discussion Andrew Dabalen. 2007. “Evaluation of Training Paper.” Save the Children US, Westport, CT. in African Enterprises.” Journal of Development Oosterbeek, Hessel, Mirjam van Praag, and Auke Economics 84 (1): 310–29. Ijsselstein. 2008. “The Impact of Entrepreneur- Schultz, T. Paul. 2004. “Evidence of Returns to ship Education on Entrepreneurship Competen- Schooling in Africa from Household Surveys: cies and Intentions: An Evaluation of the Junior Monitoring and Restructuring the Market for Achievement Student Mini-Company Program.” Education.” Journal of African Economies 13 IZA Discussion Paper 3641, Institute for the (supplement 2): ii95–148. Study of Labor, Bonn. Shaorshadze, Irina, and Pramila Krishnan. 2013. Ozier, Owen. 2010. “The Impact of Secondary “Technical and Vocational Education and Train- Schooling in Kenya: A Regression Discontinuity ing in Ethiopia.” Working Paper, International Analysis.” Development Research Group, World Growth Centre, London School of Economics, Bank, Washington, DC. February. Paxson, Christina, and Norbert Schady. 2007. “Cog- Shonkoff, Jack P., and Deborah Phillips, eds. 2000. nitive Development among Young Children From Neurons to Neighborhoods: The Science of in Ecuador: The Roles of Wealth, Health, and Early Childhood Development. Washington, DC: Parenting.” Journal of Human Resources 42 (1): National Academy Press. 49–84. Söderbom, Måns, Francis Teal, and Alan Harding. Peeters, Pia, Wendy Cunningham, Gayatri Acharya, 2006. “The Determinants of Survival among and Arvil Van Adams. 2009. “Youth Employment African Manufacturing Firms.” Economic Devel- in Sierra Leone: Sustainable Livelihood Oppor- opment and Cultural Change 54 (3): 533–55. tunities in a Post-Conflict Setting.” World Bank, South Sudan, Government of, Ministry of General Washington, DC. Education and Instruction. 2011. Education Sta- Premand, Patrick, Stefanie Brodmann, Rita tistics for the Republic of South Sudan: National Almeida, Rebekka Grun, and Mahdi Barouni. Statistical Booklet, 2011. Juba. 2012. “Entrepreneurship Training and Self- Steinberg, Laurence, Elizabeth Cauffman, Jennifer Employment among University Graduates: Woolard, Sandra Graham, and Marie Banich. Evidence from a Randomized Trial in Tunisia.” 2009. “Are Adolescents Less Mature Than Adults? Policy Research Working Paper 6285, World Minors’ Access to Abortion, the Juvenile Death Bank, Washington, DC. Penalty, and the Alleged Apa ‘Flip-Flop.’” Ameri- Pritchett, Lant. 2013. The Rebirth of Education: can Psychologist 64 (7): 583–94. From 19th-Century Schooling to 21st-Century Teal, Francis. 2010. “Higher Education and Eco- Learning. Washington, DC: Brookings Institu- nomic Development in Africa: A Review of tion Press. Channels and Interactions.” CSAE Working Rankin, Neil, Justin Sandefur, and Francis Teal. Paper 2010-25, Centre for the Study of African 2010. “Learning and Earning in Africa: Where Economies, Oxford. Are the Returns to Education High?” CSAE Tubbs, Carly, and Dana Charles McCoy. 2012. Working Paper 2010-02, Centre for the Study of “Early Childhood Development: An Introduc- African Economies, Oxford. tion.” In Children and Youth in Crisis: Protecting Riley, Thyra A., and William F. Steel. 2000. “Kenya and Promoting Human Development in Times Voucher Programme for Training and Business of Economic Shocks, edited by Mattias Lundberg Development Services.” In Business Development and Alice Wuermli, 106–15. Washington, DC: Services: A Review of International Experience, World Bank. edited by Jacob Levitsky, ch. 12. London: Inter- UNESCO (United Nations Educational, Scientific mediate Technology Publications. and Cultural Organization). 2012. EFA Global Roberts, Brent W., and Daniel Mroczek. 2008. “Per- Monitoring Report: Youth and Skills: Putting sonality Trait Change in Adulthood.” Current Education to Work. Paris: UNESCO. Directions in Psychological Science 17 (1): 31–35. Unger, Jens M., Nina Keith, Christine Hilling, Rosendahl Huber, Laura, Randolph Sloof, and Michael M. Gielnik, and Michael Frese. 2009. Mirjam van Praag. 2012. “The Effect of Early “Deliberate Practice among South African Small Entrepreneurship Education: Evidence from a Business Owners: Relationships with Education, Randomized Field Experiment.” IZA Discus- Cognitive Ability, Knowledge, and Success.” Jour- Skills for productive Employment 111 nal of Occupational and Organizational Psychol- Country Study, Policy Note, World Bank, Wash- ogy 82 (1): 21–44. ington, DC. Uwezo Tanzania. 2011. “Are Our Children Learn- ———. 2012a. “Botswana: Policy Note 2: Labor ing? Annual Learning Assessment Report.” Market Signals on Demand for Skills.” World Uwezo Tanzania, Dar es Salaam. http://www. Bank, Africa Region, Human Development uwezo.net/wp-content/uploads/2012/08/ Department, Washington, DC. TZ_2011_AnnualAssessment-Report.pdf. ———. 2012b. “Preliminary EPAG Midline Report: Valdivia, Martin. 2011. “Training or Technical Assis- Economic Empowerment of Adolescent Girls tance? A Field Experiment to Learn What Works and Young Women (EPAG) Project in Liberia.” to Increase Managerial Capital for Female Micro- World Bank, Washington, DC. entrepreneurs.” Grupo de Análisis para el Desar- ———. 2013. “The Skills Development System in rollo (GRADE), Lima. Malawi.” World Bank, Washington, DC. Walker, Susan P., Theodore D. Wachs, Julie Meeks ———. Various years. World Development Indica- Gardner, Betsy Lozoff, Gail A. Wasserman, tors. Washington, DC: World Bank. http://wdi. Ernesto Pollitt, and Julie A. Carter. 2007. “Child worldbank.org. Development: Risk Factors for Adverse Out- World Bank and Republic of Liberia. 2012. “Com- comes in Developing Countries.” The Lancet 369 munity Works Life Skills Training Assessment: (9556): 145–57. Youth, Employment, Skills Project Mid-Term Welch, F. 1970. “Education in Production.” Journal Review.” World Bank, Washington, DC. of Political Economy 78 (1): 35–59. Wuermli, Alice, Rainer K. Silbereisen, Mattias World Bank. 2003. World Development Report 2004: Lundberg, Michele Lamont, Jere R. Behrman, Making Services Work for Poor People. New York: and Larry Aber. 2012. “A Conceptual Frame- Oxford University Press. work.” In Children and Youth in Crisis: Protecting and Promoting Human Development in Times ———. 2009. Ghana: Job Creation and Skills Devel- of Economic Shocks, edited by Mattias Lundberg opment. Vol. 1: Main Report. Washington, DC: and Alice Wuermli, 29–102. Washington, DC: World Bank. World Bank. ———. 2011a. “Challenges and Options for Tech- Xu, Lisa, and Bilal Zia. 2012. “Financial Literacy nical and Post-Basic Education in South Africa.” around the World: An Overview of the Evidence World Bank, Washington, DC. with Practical Suggestions for the Way Forward.” ———. 2011b. “Raising Productivity and Reduc- Policy Research Working Paper 6107, World ing the Risk of Household Enterprises.” Ghana Bank, Washington, DC. Chapter 4 Agriculture as a Sector of Opportunity for Young Africans Efforts to accelerate agricultural growth and technology, and connections through trade that improve food security have often been sepa- spurred diversification and structural change. rated conceptually from efforts to create jobs for Faster growth in nonagricultural sectors has young people. This damaging compartmental- drawn enough labor out of agriculture to cause ization, if it persists, will limit Africa’s ability the share of employment in agriculture to fall. to reap the benefits of its youth dividend. Agri- Labor shifted out of agriculture because pro- culture, already Africa’s largest employer, is the ductivity gains on the farm saved labor (push most immediate means of catalyzing economic factors), and higher productivity and earnings growth and employment for young people. To opportunities off the farm attracted labor (pull realize this potential, farming must shift rapidly factors).1 from its present status as occupation of last resort Africa is not following that trajectory. In and low productivity to one of technical dyna- many African countries, income generated mism and recognized opportunity. With higher through the extraction of natural resources priority accorded to accelerated implementation and urban construction and services has raised of well-designed programs of public investment gross domestic product (GDP) without draw- in agriculture, continued progress on regulatory ing significant numbers of workers out of agri- and policy reform, and further specific measures culture. Even under optimistic assumptions, to include young people, agriculture will absorb the cohort of young Africans now entering the large numbers of new job seekers and offer the labor force is likely to exceed the number meaningful work with large public and private that can be absorbed into jobs in manufactur- benefits. ing and services (see chapter 1). Over the next few decades, young people will continue to Agriculture—once the predominant sector apply their energies and talents to agriculture, in most of the world’s economies—has histori- on or near the farmstead of their birth (Proc- cally played a lesser role as economies devel- tor and Lucchesi 2012). The critical question oped the accumulated wealth, innovations in is how African youths and their countries can 113 114 Youth EmploYmEnt in Sub-Saharan africa benefit from agricultural employment. The US$313 billion in 2010 to US$1 trillion in 2030 answer depends on whether governments take (World Bank 2013). Food imports surged ahead the policy and investment decisions that will of exports as recently as 2003, and they have lift the constraints on agricultural productivity. continued to climb. The growth in imports The reasons for Africa’s slow growth in agri- has been variously attributed to the failure cultural productivity are known. Cropping sys- of agricultural production to keep up with tems based on wheat and irrigated rice, which population growth (which is incorrect, as per registered spectacular productivity gains in capita production has risen over this period), South and East Asia, are not suited for most climate change, and other supply-side factors. environments south of the Sahara. Africa’s Supply matters, but the fundamental point complex agro-ecologies and highly diverse is that rapid growth in population, incomes, production systems demand a level of original and urbanization is increasing the demand for research comparable to that undertaken else- imported food faster than the supply of domes- where in the world, but Africa is only begin- tically produced substitutes is growing. ning to reverse decades of neglect and under- Rapidly growing demand creates oppor- investment in agricultural research. It will take tunities for suppliers. For example, urbaniza- time for the benefits to be felt in earnest. The tion could be good news for local agriculture. effects of low productivity in African agricul- The denser patterns of settlement seen as rural ture are also well known. Low productivity is communities grow and merge3 can reduce partially responsible for the high food prices marketing costs for agricultural producers in prevalent in much of Africa, where expendi- the hinterlands and raise the returns to invest- tures on primary food products can account ments in processing raw products. Nor is for as much as half of consumers’ expenditures growth in demand limited to Africa’s expand- (OECD and FAO 2012). High food prices also ing domestic markets. Global food prices are curtail competitiveness by increasing the cost at their highest point in several decades, and, of labor. barring significant changes in policies related In the interim and until research specific to biofuels, food prices are expected to remain to Africa’s environments is available, growth high for at least the rest of the decade. in agricultural productivity will need to come The opportunities presented by this grow- from wider use of superior technologies that ing local and international demand for food have worked elsewhere—improved seeds, are likely to be as varied as African agriculture breeds, cropping methods, conservation prac- itself. On average agriculture is a sector of low tices, and equipment. Over the past decade, labor productivity and high employment, but more farmers across Africa have started to in reality it is exceptionally heterogeneous. adopt such technologies,2 although not as rap- Even in developed countries, agriculture is suf- idly as farmers in other regions. Recent invest- ficiently heterogeneous to raise questions about ments and policy reform may catalyze more what constitutes a farm. In Africa every farm rapid adoption, but the levels of investment, lies somewhere along wide continuums of farm pace of implementation, and quality of pro- size, capital intensity, use of mechanical and grams in Africa have not yet sufficed to deliver biological technology, and degree of commer- a large shift in productivity. cialization. The first requirement for creating good opportunities for young people in African agriculture is to gain a detailed understanding Agriculture: Potential of the sector by peeling back the averages to Opportunity, Room to Grow reveal the dispersion of participants’ activi- ties, command of assets, and use of skills. This The opportunity that farming represents in task demands close, continuous attention to Sub-Saharan Africa is clearly evident in the gender issues, given the importance of women region’s trade accounts. The value of Africa’s and girls in Africa’s agricultural labor force food markets is projected to increase from (box 4.1). agriculture as a Sector of opportunity for Young africans 115 New opportunities for African agriculture to Box 4.1 benefit from changes in local and national mar- kets will emerge from segments of the farming structure that have been underdeveloped in the Women and girls: A major force within past but now have room to grow. For example, Africa’s agricultural labor force in many African capitals, rice is twice as expen- sive as it is in Asian exporting countries. The African women and girls work in agriculture as farmers on their own price of maize, the main food staple in Eastern land, as unpaid workers on family farms, and as paid laborers on other farms and in agricultural enterprises. They are involved in both and Southern Africa, is 30–40 percent above crop and livestock production at subsistence and commercial levels. export prices in South Africa, the United States, Across developing countries, women comprise 43 percent of the and the Black Sea region (table 4.1). If local agricultural labor force, on average; this figure ranges from around producers become more competitive, they can 20 percent in Latin America to 50 percent in parts of Africa and Asia. capture thriving domestic and regional mar- Some Sub-Saharan African countries have seen women’s share of the kets. Measures that reduce the costs of produc- agricultural labor force rise significantly in recent decades due to con- tion (such as the dissemination of improved flict, the human immunodeficiency virus/acquired immune deficiency technology) and marketing (such as invest- syndrome (HIV/AIDS), migration, and livelihood diversification, but ments in transportation and infrastructure) regional data conceal wide differences. The share of women in the can increase profitability and reduce food costs. agricultural labor force ranges from 36 percent in Côte d’Ivoire and Even in countries that are relatively well linked Niger to more than 60 percent in Lesotho, Mozambique, and Sierra Leone (FAO 2011). Nevertheless, one regional trend is clear—it is usu- to world markets, increased local production ally the male member of the agricultural household who moves into a can lower food prices, because international nonfarm activity first (Fox and Sohnesen 2012). prices do not translate directly into local mar- ket prices (Minot 2011). Lower food prices not only help consumers but also temper demands for higher wages in the nonfarm sector, which How countries choose to increase agricul- attracts new investment in manufacturing and tural productivity will influence whether this services. New investment creates new jobs, fuel- virtuous cycle is sustained, with benefits to ing a virtuous cycle (box 4.2). young people and the economy as a whole, or Table 4.1 Wholesale prices of unprocessed maize and rice in selected countries (average, January–April 2012) Maize Rice Wholesale price Wholesale price Market (US$ per ton) Market (US$ per ton) Africa Ethiopia 390 benin 1,055 Kenya 393 burkina faso 738 malawi 400 madagascar 593 mozambique 378 mali 690 rwanda 318 mozambique 865 South africa 293 niger 850 tanzania 334 Senegal 810 togo 453 togo 1,097 uganda 334 uganda 1,368 Zimbabwe 300 International benchmark International benchmark india 378 black Sea region 267 thailand 556 united States 276 Vietnam 434 Source: FAO Global Information and Early Warning System. 116 Youth EmploYmEnt in Sub-Saharan africa Box 4.2 When agriculture is more productive, economies can grow Despite differences in the global context and national eco- Figure B4.2.1 Higher total factor productivity helped U.S. nomic circumstances, the effects of agriculture on the U.S. farmers to compensate for declining terms of trade, 1975–2009 economy over the past 45 years may provide useful lessons 1.8 for Africa at this juncture, when so many young people 1.6 are poised to enter the agricultural workforce. From 1960 1.4 to 2005, the United States produced more agricultural 1.2 commodities more cheaply than ever. The real prices of 1.0 Index most agricultural commodities declined 20–50 percent 0.8 (see table B4.2.1). Even though the cost of food was ris- 0.6 ing throughout the world, most food prices were still lower 0.4 in 2010 (in constant U.S. dollars) than they were in 1960. 0.2 0.0 For the economy as a whole, low prices of primary food 1975 1980 1985 1990 1995 2000 2005 2010 products meant that it was much cheaper to process food into an array of consumer products (an activity that gener- Total factor productivity (1992=1) ated new jobs). The low prices enabled consumers to spend Terms of trade (prices of farm outputs/prices of farm inputs) (1990–92=1) more money on items other than food. Higher agricultural productivity contributed to broad-based income growth in Source: Executive Office of the President 2011. the United States. Even while the prices of commodities they produced declined, farmers and other agricultural workers preserved Table B4.2.1 Percentage change in prices of selected food their livelihoods. As real agricultural prices fell and the cost products in the United States, 1960–2010 of inputs and factors of production rose, farmers in the Period Wheat Maize Sugar Beef United States and other developed countries still managed 1960–2005 –43 –52 –19 –23 to profit (and so did agricultural wage earners) by increas- ing their productivity. They used inputs more efficiently and 2006–2010 8 41 50 22 changed their mix of outputs. In the United States between 1960–2010 –24 –18 24 –10 1975 and 2010, total factor productivity rose 2.2 percent a Source: World Bank Pink Sheets. year, which was low by historic standards but sufficient to maintain profitability (see figure B4.2.1). whether it is short-lived. As demand for food African leaders who are framing strategies rises, growth in total factor productivity will be for agriculture now should be aware that the necessary to keep real food prices in check and circumstances they face are quite different maintain the capacity to create jobs.4 Unless from those that shaped traditional experiences serious attention is paid to interventions that of development and structural change. African can truly sustain productivity, such as agri- agriculture is developing in a context of high cultural research, the development of farm- global food prices, potential for growth in ing skills, and the adoption of new and better area and yield, few nontradable manufactured varieties, growth in output will come through goods, and shifts in comparative advantage in increased use of purchased inputs such as fertil- the developed world in favor of technology- izer and agro-chemicals. The resulting growth intensive services and products (Losch, in output may be rapid for a short period if Fréguin-Gresh, and White 2012). In this con- the inputs boost yields, but it will be costly, text, if African farmers change the technology increase the real price of food, and ultimately they use and the mix of commodities they pro- erode potential gains to producers, consumers, duce, agriculture’s share in African GDP could and society at large (box 4.3). remain larger than has historically been the case agriculture as a Sector of opportunity for Young africans 117 or even grow with development. Furthermore, Box 4.3 the cost of withdrawing labor from agriculture seems to have increased over time around the world, and this is another factor arguing that Compromising growth and job creation Africa’s experience may be different (Timmer without improving long-term food security and Akkus 2008). Over the past 50 years, the point at which wages in agriculture converge Since 2008, several African countries have introduced subsidy pro- with those in nonagricultural jobs has been grams that encourage farmers to use fertilizer on staple crops without supporting the use of more productive varieties and better manage- reached at later and later stages in the transfor- ment practices. As a result, farmers have devoted much of their land mation of successful economies, perhaps sug- to staple crops, crowding out higher-value, more labor-intensive crops gesting that globally industry is becoming less and slowing the pace at which agriculture creates jobs and expands and less able to absorb labor. productivity. In some cases, the production of staple crops surged Recent cautiously optimistic trends suggest briefly, only to drop off sharply once the expensive (and often politi- that the sources of agricultural growth in Africa cized) distribution of subsidized fertilizer came to an end. Early evi- may be changing. Between 1960 and 2008, dence suggests that, even while the subsidies were in place, the yield cultivated area grew faster than yields (Fuglie response of varieties in use was considerably less than anticipated. 2011). About 40 percent of the growth in yield Much can be done to improve the production of staple food crops came from greater use of purchased inputs and without compromising growth and job creation in this way. The les- 60 percent from changes in total factor pro- son is that fertilizer use should be one part of a broader program to promote gains in productivity and encourage diversification into ductivity. Between 2006 and 2008, when Afri- higher-valued and more labor-intensive products. can governments began to give greater atten- tion to agriculture,5 yield growth dominated area growth, and total factor productivity rose. These recent developments suggest that if gov- ernments would intensify and sustain their livelihood, and their perceptions varied widely “Farming is a efforts to raise factor productivity, they could across Africa. Within three broad categories of good job because secure lower prices for consumers, higher earn- jobs (nonfarm wage, nonfarm nonwage, and it is where I can ings for farmers, and good opportunities for agriculture), “family farming” was the desirable young people to enter farming. job named most often. Yet with the exception of get food to eat a women’s group in North Darfur, none of the and live well.” focus groups from South Africa, Sudan, or Togo Liberia Recognizing the Opportunity in mentioned any farming activities as a good job Agriculture for Young People (Petesch and Caillava 2012). Within the same broad categories, farming followed only illegal Many young people know little of the oppor- and antisocial jobs as the “worst job.” Focus tunities and dynamism possible in farming group interviews with urban young people sug- today. When young rural Africans in 32 focus gest that agriculture virtually disappears from groups were asked about the best and worst mention as the “best job.” To attract young peo- ways to earn a living in their community, they ple, agriculture will need to be more dynamic rarely mentioned agriculture as the “best job” and appealing than it is now, and young people (although it was not considered the worst). will need to view the sector more positively (IDS Participants described good jobs as those that 2012). The farms that offer attractive opportu- command good pay and respect, two features nities will have to be quite different from those not typically associated with farming under that most young Africans know. the conditions most familiar to young Afri- cans. They described bad jobs as those that Patterns of Land Use, Farm Size, and offer poor, insecure returns and are physically Profitability damaging or demanding or illegal. The farms that many young Africans know Even so, the focus groups gave mixed mes- from childhood are small and worked with sages regarding the desirability of farming as a back-breaking labor and little mechanization. 118 Youth EmploYmEnt in Sub-Saharan africa Figure 4.1 The young do not typically own land sible to farm larger areas, but trypanosomiasis 80 and other animal diseases constrain the use of draft animals in many parts of Africa. 70 Alternatively, land may be abundant but vir- % who own at least one plot of land tually impossible to acquire because of ambi- 60 guities in the transactability of land through 50 purchase, sale, leasing, inheritance, assignment under traditional rules, and mortgage (World 40 Bank 2012b). When constraints on the opera- 30 tion of land markets raise the cost of accessing new land, a young person reaching adulthood 20 may simply farm a portion of the family’s origi- nal holding rather than secure a new allotment. 10 It is not unusual for the continuous fragmenta- 0 tion of small holdings to persist alongside the 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60+ acquisition of large tracts by outside investors, Age (years) whether domestic or foreign. Tanzania Uganda Malawi A third consideration is that small, labor- Source: Based on data from the Living Standards Measurement Study–Integrated Survey on intensive farms can be economically appro- Agriculture. priate, efficient, and profitable under certain conditions. Recent evidence based on a geo- Holdings of 1 to 2 hectares predominate, and graphically wide and heterogeneous set of data the most common implements are the hand confirms an inverse relationship between maize hoe and machete (Nagayets 2005). According to yield and farm size, supporting the premise World Bank data from three countries, owned documented in earlier studies that small farms land increases with age of the farmer (figure are often productive in the African context 4.1)—the average plot size, even for older farm- and that smallholders do not necessarily forgo ers, often remains well under 1 hectare. economies of scale (Larson et al. 2012). His- This pattern of land use is seen whether torically, primary production of staple com- land is scarce or abundant, although for differ- modities has not exhibited increasing returns ent reasons. Where settlement is dense and land to scale, and smallholders who voluntarily form “If you have only is scarce, as in Rwanda and Malawi, holdings producer groups can capture scale economies a small parcel per household and per worker are small and where they do exist—for example, in the mar- for eight people, shrinking with population growth. Under these keting of their produce and access to informa- circumstances, investments in irrigation, appli- tion (see box 4.4; Morris, Binswanger-Mkhize, not everyone can cation of purchased inputs, improved varieties, and Byerlee 2009). eat.” Madagascar high-valued crops, double and triple cropping, The most desirable farm size, however, is an terracing, and other practices can increase the economic issue and not a matter of principle productivity of land and incomes. Investments or ideology. Where relevant costs of produc- that make it easier to reach markets increase the tion are readily divisible, smallholders will do demand for agricultural products and reduce as well as or better than others. Where costs are the cost of transporting them. The returns to not divisible for whatever reason, smallholders intensification rise, and more such investments will be at a disadvantage but will still be very take place. numerous. In that case, programs that facilitate But why are farms so small in areas where adjustments in farm size or address the indivis- land is abundant, as in much of Africa? Farm ible costs will be constructive. size is often limited to the amount of land that a household can farm manually, because machin- Getting Young Africans the Farms ery is expensive, cannot be purchased without They Need financing, and can be challenging to own and Even where small farms are demonstrably effi- use collectively. Animal traction makes it pos- cient, agricultural productivity cannot grow agriculture as a Sector of opportunity for Young africans 119 if more family labor crowds onto them. The Box 4.4 income that 1 or 2 hectares can generate is rarely sufficient to pull all members of a house- hold out of poverty. For agricultural produc- Producer organizations and the transition to tivity and incomes to grow, young people will modern supply chains need to be able to acquire more land, and young workers will need to be able to leave the farm Rural productive alliances can bring producer organizations and of their birth for other forms of employment. commercial buyers together to increase income and employment via participation in modern supply chains. These alliances have been Since mobility out of farming has been shown to bring about higher agricultural incomes and increased rural low in Africa, much of the land is now held employment, especially for agricultural workers and women working by aging farmers despite the large cohort of in postharvest activities (World Bank 2012a). Farmers have also ben- potential new entrants. Constraints to inter- efited from employment opportunities generated by public-private generational transfers of land are particularly partnerships that enhance agricultural productivity. For example, a costly when land is scarce or young people have successful model in Latin America that sought to increase competi- difficulty acquiring holdings to start farming tiveness along the entire value chain for cassava (production, process- on their own. Where old-age pensions do not ing, and utilization) worked with farmer groups and cooperatives exist and rental markets are poorly developed, (among many others) and ultimately expanded training and jobs for elderly farmers often retain control over hold- farmers in cassava-based agro-industries. ings that would be managed more efficiently by Source: World Bank 2012a. younger, more innovative, and energetic farm- ers (box 4.5). More fluid land markets would create bet- ter opportunities for young people to practice more productive and managerially demanding Box 4.5 agriculture. As processors and urban consum- ers demand quality and traceability in agricul- tural produce and as changing weather patterns Options for establishing or leaving a farm undermine the validity of traditional “rules of in Kenya thumb” for the agricultural cycle, agriculture Kenya’s young people have great difficulty establishing themselves requires a more sophisticated level of manage- as farm operators. According to a large national sample drawn from ment. Young people are well suited to acquire participants in the Kenya Agricultural Productivity and Agribusiness and exercise managerial expertise, and they can Project, people whose primary economic occupation is farming are do so in many ways, but the managerial acu- in their late 50s, on average. In most cases, these people are also the men of an individual farmer is as indivisible as principal decision makers on the farm. a tractor. Each creates economic pressures to Men who identify themselves primarily as farmers usually farm as amalgamate very small farms into larger units their first occupation; they have a spouse who works in the household or develop new networks of producers to share and on the farm but does not earn significant outside wages. Women costs. For this reason, increased fluidity of land who identify themselves primarily as farmers may or may not have an adult male in the household contributing wage earnings to the house- markets, in particular through land rentals, is hold income. Women farmers with wage-earning adult males in their essential for a new generation of African farm- household do very well in farming—in most cases, better than men. ers to take advantage of opportunities emerg- In contrast, single women who manage farms are, on average, about ing in agriculture. Producer organizations may 10 years older than other farmers, and their earnings are the lowest. need to innovate in the delivery of managerial These women probably retain control of land because the cost of hold- services, an area in which they have not been ing it is low in the absence of land tax, and they have no other way active in the past. to feed themselves as they age. This information suggests that elderly When factor endowments and the charac- women and land-hungry young farmers could benefit from participat- teristics of technology and markets imply that ing in a program that eases intergenerational land transfers while pro- the optimal farm size is larger than what is viding some kind of social safety net for elderly landholders. observed, constraints on capital and land mar- Source: Torkelsson 2012. kets impose a high burden of inefficiency on rural people. Although smallholders may not 120 Youth EmploYmEnt in Sub-Saharan africa have the skills or appetite for risk to manage gering the change. The conditions in Africa as much as 100 hectares, many could prob- offer ample opportunities for simultaneous ably handle 5–10 hectares if they had access increases in average farm size and in employ- to machinery to work it, particularly if public ment. The fact that average farm size in Africa investments were made in infrastructure that is now declining is a worrisome indicator that would make farming more profitable. The constraints on land markets are already damag- incentives for young people to remain in school ing the prospects for young people and becom- and acquire basic numeracy and literacy skills ing stronger (Djurfeldt and Jirström 2013). would increase if intermediate-size farms were among the possible options and were known to require those basic skills for successful opera- Agricultural Career Paths for tion. Intermediate-size farms can emerge only the Future if land markets are more active. One concern is that an expansion of farm The young people who will look to agriculture size in Africa could displace labor precisely for employment are familiar with traditional when demography requires agriculture to agriculture, but given the changes under way absorb labor. In parts of the world where farms in the sector, they are likely to experience their have expanded from very small (2 hectares and working years in ways different from their par- less) to mid-size holdings (5–100 hectares), ents. They will also have different requirements labor has often been displaced. This displace- for support if they are to succeed. Young farm- ment need not occur in Africa because under- ers will face four generic paths to agricultural used land can be brought into production— employment: continuing on the family plot but Africa can still expand at the extensive margins with a different mix of enterprises; establishing of farming without compromising forest area. their own operations on new land; combining Larger farms need not be less labor intensive farming with part-time other work; or taking than small ones when both area and employ- wage work on large or mid-size commercial ment can expand simultaneously. However, if holdings. Although these four basic paths cover farm size grows through consolidation on land many options, the diversity of African agricul- that is already farmed and is accompanied by a ture ensures that some young people will face capital subsidy that reduces the cost of mecha- other choices. For example, young people in nization, as occurred in Brazil, then bigger pastoral areas confront a different set of chal- farms could be expected to displace labor. If lenges and opportunities. the change is occasioned by shifts in technol- The four basic pathways to employment in ogy and markets that require greater manage- agriculture vary in their requirements for land, rial skill, formerly independent farm operators capital, and skills (table 4.2). The first two— might become hired workers or outgrowers full-time employment on the family farm and on larger, technically more sophisticated hold- full-time farming on a new holding—are the ings. Thus the effects of changing farm size on most prevalent. Among households surveyed employment are specific to the factor endow- in nine African countries in 2008, 51 percent ment in a given market and to the forces trig- reported that inheriting land already under cultivation was the most common means for young people to obtain land, while 16 percent Table 4.2 Pathways for agricultural employment and their requirements would be allocated land not previously culti- Need for Need for Need for vated, 9 percent would rent or borrow land, Type of employment land capital skills and 12 percent would buy land (Proctor and full-time on existing family holding none medium medium Lucchesi 2012). full-time on new holding high high high For each pathway to become a more produc- part-time combined with household enterprise tive source of employment, policy makers will low medium high (processing, trading, sales of services) need to use a range of approaches to improve Wage work off the family farm none none medium or high young people’s acquisition of land, capital, and agriculture as a Sector of opportunity for Young africans 121 skills. Evidence from various African countries household with several young adults should be “Two [of our suggests some approaches that show promise. able to undertake the required work. children] Thus even if young people are absorbed have become Pathway 1: Full-Time Employment on into the farm of their birth as a young adult, farmers as well an Existing Family Holding a change in management of the household enterprise could make this absorption more . . . the others For young people with no other options, the default outcome is to remain on the family rewarding for the individuals and the family. took another holding and simply farm a portion, essentially A combination of pooled off-farm earnings, a direction. It is subdividing an already small parcel. Others shift in farming technology to higher-value and best that we are choose not to leave. Eventually many can expect more commercial products, and aggregation of not in the same to inherit a portion of the land, but if siblings household labor at peak periods could allow small farms to absorb young adults construc- profession; that are in the same position, the holding will be tively. An emphasis on extension programs that way we can help small. These youths need capital and skills to make the most of their small holdings through focus on the household as an enterprise and do out one another.” higher-value agriculture. Young people who not just offer technical and economic advice on Madagascar foresee this as their future, however, may have crops or livestock could help this group. little incentive to invest in skills, since they will The view sketched above of the small house- not have the power to use them as long as the hold farm evolving as young adults become parental generation retains recognized rights of economically active provides an important decision. perspective on the conceptual understanding Families in this situation may find them- of youth employment. A young person in, for selves in increasingly difficult circumstances, example, northern Uganda who is a member with alienated young people resentful of their of such a household and has also benefited elders’ continued control over resources. With from the Youth Opportunities Program of the some guidance and mentoring, however, fami- northern Uganda Social Action Fund might lies could turn this situation to advantage by have acquired vocational skills as a hairdresser managing the household as an enterprise with (see discussion in chapter 3). She might see her a portfolio of activities. Many households primary occupation as “hairdresser,” but she already support small natural resource–based would also be an equity holder in a small farm enterprises (selling eggs and poultry, process- enterprise (if her earnings were applied in part ing cassava or grain, collecting reeds, making to investment on the farm) and an occasional bricks) as adjuncts to their primary farm- laborer at times of peak demand. Her economic ing enterprise. The difference here is that the security would come from the farm earnings household takes a strategic approach enabling as well as from her trade. To create space for the small household farm to evolve and sup- the contributions of young people like this, the port multiple generations and families. In this parental generation would need to view the way, pathway 1 resembles pathway 3, but the farm as an enterprise. emphasis is on full-time agriculture as part of a diversified, multigenerational family business. Pathway 2: Full-Time Employment on In this pathway, the skills and labor of mul- New Farm Starts tiple young adults in the household could allow A second group of young people will succeed for specialization. If there is demand for their in leaving the farm of their childhood and labor, those capable of earning off-farm wages establishing a new and separate holding, ide- could do so, thus easing the household’s capital ally larger than the parcel they left. Those more constraints. Those with sufficient skill to man- likely to succeed in such an undertaking would age higher-value agriculture could sharpen the probably be relatively experienced in farming specific skills required through short courses and hence on the older end of the age range or focused training. Some superior technolo- for “youth.” They would also have the high- gies, such as conservation tillage, require high est potential returns in the form of increased investment of labor at peak periods, and a productivity. These young farmers would 122 Youth EmploYmEnt in Sub-Saharan africa have the greatest need for land, start-up capi- mechanized field operations, and advice can be tal, and advisory services or training to assist met by young men and women with the capital with technical and managerial challenges. Few and skills to start a small business. These young young farmers will be able to assemble the ele- people may not have the capital to acquire a full ments required to establish a new farm without array of farm machinery, but they could offer assistance. services on a paid basis by purchasing or leasing New holdings may be in the localities where a limited selection of equipment. Young people young people already live and on land newly would also need the particular skills to deliver available for cultivation through clarification the services and maintain the machinery. of ownership, conversion of marginal or graz- ing land, or public investment in irrigation and Pathway 4: Wage Work off the improvement. Alternatively, new holdings may Family Farm be farther away, in which case establishing the The seasonal nature of agriculture creates new farmstead will require relocation. demand for part-time wage work at peak peri- Resettlement is often controversial. Experi- ods even on small farms. In a heterogeneous ence globally and in Africa attests to the impor- farm structure with significant numbers of tance of strict adherence to voluntary decision large holdings, wage work on a regular basis is making on the part of participants, careful also observed. Most of this work pays relatively selection, full information for all stakehold- little and requires very little skill, and few young ers, effective support services for new arrivals, people aspire to be low-skilled day laborers. For and adequate investment in infrastructure. An the very poorest, however, paid work, even if assessment of several decades of public sup- undesirable, is a better option than no work. port for resettlement in Indonesia found mixed Therefore, it is anticipated that a fourth group results tending toward the negative. Improve- of young people will take wage work, whether ments in the incomes and access to public ser- formal or informal, on large commercial farms vices of settlers were offset by disappointing or in the processing and service sectors. These outcomes in agricultural production, environ- young people need skills to handle a range of mental degradation, and resentment against tasks and equipment. At a minimum, for the newcomers on the part of indigenous inhabit- most basic low-skilled work, they need good ants (World Bank 2012a). Preliminary results health to withstand often grueling working con- regarding a program of market-assisted land ditions. Such wage work could fit into the liveli- reform in Malawi, in contrast, indicate more hood strategies described above in combination positive outcomes (Chirwaa 2008). If local with other activities, or it could be a temporary “My first job was young people can secure access to land in or option until better opportunities appear. a contract job near their community, this approach is clearly Not all wage work is poorly paid or low that paid 300 simpler. If relocation is required, lessons of past skilled. Some very large enterprises, both in experience should be fully weighed. primary production and in processing, require [Liberian] dollars per day . . . I a range of skills depending on the technical Pathway 3: Part-Time Farming and sophistication of the production process and didn’t enjoy it. I Household Enterprises types of machinery used. Drivers, machine just did it for the A third group of young people may be indepen- operators, mechanics, quality testing techni- money, worked dent part-time farmers, either managing their cians, and others will be required in increas- on someone else’s own holdings or contributing to family opera- ing numbers in the future, and these jobs are farm . . . I just tions described under pathway 1, with enough often better paid than unskilled day labor. For forced myself capital to establish themselves as a seller of ser- example, Red Fox Ethiopia, a floriculture firm vices, a trader, or an occasional wage worker. located outside of Addis Ababa, draws labor because of food Higher-value agriculture will use services more from the surrounding rural areas and towns . . . I’m still doing intensively and create employment for those and offers employer-provided transport to it. It was my first who can provide them (box 4.6). Demand for work, life and health insurance, and a subsi- job.” Liberia transport, plant protection, veterinary services, dized cafeteria (box 4.7). agriculture as a Sector of opportunity for Young africans 123 Box 4.6 High-value agriculture and opportunities for employment off the farm The labor market effects of high-value agriculture are docu- Processing food for local markets is another growth area mented in several countries in Africa. In 1985, just 14 per- for rural employment. As new towns proliferate through in cent of the agricultural and food exports from African coun- situ urbanization, demand for processed foods increases, tries consisted of high-value agricultural products; by 2005, along with investment in processing. Greater attention to the share had risen to 30 percent, with many jobs created food safety in public policy will improve conditions within along the way. In Madagascar, the export of vegetables has processing plants, since the conditions that assure safe relied on about 10,000 smallholders contracted for procure- products also contribute to improved hygiene and safety for ment of primary produce. In other instances, production of workers. The regulatory bar for wages and working condi- primary produce has been vertically integrated with large tions cannot be set so high that it stifles investment, elimi- estate farms, as is the case with tomato and bean exports in nates jobs, and depresses demand for primary production. Senegal; the jobs created are for wage earners in processing The surge in food imports in Africa since 2003 reflects the units and pack houses. Table B4.6.1 gives several examples underdevelopment of local food-processing capacity. Rem- of employment created in export horticulture chains. Hor- edying this underdevelopment will create jobs and raise ticultural work is generally labor intensive and effective at returns to investments in primary agriculture. alleviating poverty, especially among women. Table B4.6.1 Employment in horticultural supply chains in Sub-Saharan Africa Number of employees in the fresh fruit and Percentage of Country Commodity Year of survey vegetable agro-industry female employees cameroon bananas 2003 10,000 — côte d’ivoire bananas and pineapples 2002 35,000 — Kenya flowers 2002 40,000–70,000 75 fruits and vegetables 2002 2,000,000 Senegal cherry tomatoes 2006 3,000 60 french beans 2005 12,000 90 South africa Deciduous fruit 1994 283,000 53 uganda flowers 1998 3,300 75 Zambia flowers 2002–03 2,500 35 Vegetables 2002–03 7,500 65 Source: Maertens, Minten, and Swinnen 2009. Note: — = not available. Lifting Key Constraints on Capital, agricultural research and infrastructure to the Land, and Skills weak rural investment climate—is also integral to raising agricultural productivity and creat- Constraints on the acquisition of capital, land, ing jobs, but those constraints are not specific and skills block young people’s progress along to opportunities for young people and are not the four basic pathways to employment in addressed here. agriculture. To create opportunities commen- surate with the number of young people who Financial Services will need employment, those constraints must Access to capital and credit for smallholders has be removed or relieved, as discussed next. The been a perennial problem and the subject of removal of other constraints—from the lack of analysis for decades.6 Small farmers in Africa, 124 Youth EmploYmEnt in Sub-Saharan africa Box 4.7 Red Fox Ethiopia: More technically sophisticated wage work Red Fox Ethiopia was established in 2003 by a German Red Fox controls the end-to-end supply chain by having entrepreneur with long experience in the flower business. its own importing company, transportation services, and dis- Red Fox Ethiopia produces and exports more than 150 vari- tribution networks in the international market. The company eties of unrooted young plants, mainly to France, Germany, imports its fertilizer and other agro-chemicals and sources Italy, and the United States. In 2009, it exported 127 million packaging materials and plastic bags locally. The presence cuttings valued at US$10 million. of a well-established customer network enables the firm to The firm started on 8 hectares in Koka, 95 kilometers book orders in advance and produce accordingly, resulting in from Addis Ababa. The factory’s area increased incremen- minimal wastage and risks of price fluctuations. tally to 35 hectares in 2009, and the firm is acquiring Red Fox plans to strengthen its market leadership and additional land to bring its area to 65 hectares. The firm consolidate its special expertise in producing unrooted young has 1,300 employees, 450 of whom are hired on a sea- plants. In addition, it plans to diversify into fruit production, sonal basis for three to four months at a time. Expatriate in partnership with another firm that has knowledge and professionals currently run the operation, but the own- experience of the sector. ers plan to replace them slowly and smoothly with local professionals. Source: Sutton and Kellow 2010. like their counterparts elsewhere, work in risky attention nonetheless to identify, replicate, and environments that are expensive for financial scale up the successful approaches. institutions to serve. Most such farmers have little or no usable collateral and little experi- Institutions and Organizations ence with financial services. A history of pub- Various actors offer financial services, includ- lic intervention in credit markets has created ing bank and nonbank financial institutions, expectations that defaults on agricultural loans insurers, and payment service providers. As will carry little penalty to the borrower. All of commercial banks tend to limit their outreach these challenges for financial institutions seek- in rural areas, alternative institutions such as ing to serve smallholders are relevant for young self-help groups, savings and credit associa- farmers and compounded by their lack of tions, and cooperatives have emerged to fill the experience. Not everyone will be able to access gap and to address both credit risk (usually credit, although many farmers can benefit higher in agriculture than in other sectors) and “You may have from a wider array of financial services such as covariant risks specific to agriculture—weather innovative insurance and money transfer. Yet small farm- and other climatic risks, pest and disease epi- ideas, but there ers, particularly if they are young, need capital demics, and so on (see focus note 3). is nothing you to adopt the technologies and secure the land and equipment that will allow them to become Access to Credit can do if you more commercially active. Because finance is Allowing alternative forms of collateral, such can’t afford to so important and the potential client base is as chattel mortgages, warehouse receipts, and get a loan from so large, banks and nongovernmental organi- the future harvest, can ease the credit market. the bank. It is zations (NGOs) continue to experiment with The OHADA7 Uniform Act on Secured Trans- difficult getting innovations to overcome the barriers to provid- actions, in effect in 17 Sub-Saharan African a loan from the ing sustainable financial services to large num- countries, was amended at the end of 2010 to bers of smallholders. A brief review of some of allow borrowers to use a wide range of assets bank, as you these new products and services follows. Many as collateral, including warehouse receipts and have to have of the innovations discussed are still being movable property such as machinery, equip- security. It’s tested; their performance and sustainability on ment, and receivables that remain in the hands hard.” Rwanda a large scale are unknown. They warrant close of the debtor (AgriFin 2012). Even where the agriculture as a Sector of opportunity for Young africans 125 regulatory framework allows collateralization, Many governments and development partners however, assets may not be attractive for vari- use matching grants in a variety of schemes, ous reasons, and banking practices require time including efforts to promote improved tech- for adjustment. nologies, empower farmers to hire service pro- Leasing also offers young farmers some viders, strengthen linkages with private firms relief, as it requires either no or less collateral through productive partnerships, and provide than typically required by loans. Most rural rural infrastructure for common use (AgriFin leases are financial (unlike operating leases); 2012). Grant schemes carry well-known risks the price of the asset is amortized, and the les- of diversion and elite capture, and their success see can purchase the asset at the end of the lease depends crucially on their design, including period for a small price (IFPRI and World Bank transparent rules for participation, checks and 2010, brief 6). A notable example is DFCU balances in monitoring at the local level, and Leasing in Uganda, which provided more than clear expectations regarding accounting and US$4 million in farm equipment leases in 2002 auditing. The expectation and encouragement for items such as rice hullers, dairy process- of savings should be a key feature of grants to ing equipment, and maize milling equipment. individual beneficiaries. Africa has widespread CECAM in Madagascar8 leased more than experience with grant programs, yet few have US$2.8 million in 2002–03 to rural microen- focused specifically on the needs of young terprises, with an average US$945 per lease participants. In Sri Lanka, the Gemi Diriya (Kloeppinger-Todd, Nair, and Mulder 2004). Program allocates a portion of its Livelihood Individuals in pathways 2 and 3, who may need Fund for one-time grants of US$46–US$92 to new equipment to start their venture, would generate income and help clients to start an particularly benefit from leasing. Despite leas- economic activity without incurring the risk of ing’s clear potential to relieve constraints on a loan (World Bank 2007b). Young people are access to mechanical technology, few firms have one of the program’s target groups. Just over chosen to enter this business. 10 percent of participants are destitute young Young farmers’ simultaneous needs for people (World Bank n.d.). finance and information can be addressed by linking agricultural credit to extension services, Contracting Arrangements an approach followed by BASIX Social Enter- Some outgrower arrangements offer prefinanc- prise Group, a livelihood promotion institution ing of inputs and assured marketing chan- based in India. Initially established to provide nels. In Mozambique, Rwanda, Tanzania, and microcredit to the rural poor, BASIX now pro- Zambia, Rabo Development (a subsidiary of vides rural households with financial services Rabobank) provides management services and and advice in managing crop and livestock technical assistance to financial institutions, enterprises. Almost 1,000 service providers work which, in turn, finance supply chains with with more than 25,000 villages in India under a range of agricultural clients. Participants the program. BASIX’s research has shown that include commercial farmers, farmers with farmers prefer cost-saving and risk-reducing little commercial presence, and an intermedi- interventions to yield-enhancing ones that ate group of farmers with ambitions to grow require more investment. The combination of commercially. Rabo takes particular interest in financial services and information or mentor- linking this last group to finance through con- ing allows the financial institution to identify tract farming under financial arrangements the products in greatest demand, such as sav- that limit the risk of default or side selling.10 ings, money transfer, and insurance rather than Kenya’s DrumNet Project is piloting a similar credit (IFPRI and World Bank 2010, brief 13). supply-chain approach to promote agricultural lending among 3,000 farmers in the horticul- Grants tural and oilseed subsectors. Risks of default Matching grants can promote employment are reduced through cashless direct payment to and employability among young people.9 the input supplier via a bank transfer once the 126 Youth EmploYmEnt in Sub-Saharan africa product is delivered to the buyer (IFPRI and traditional insurance by being available through World Bank 2010, brief 14). well-trusted yet innovative channels and by offering low premiums, products with simple E-Transfers and Payments designs, flexible payments for premiums, and Electronic and mobile technology are rapidly prompt settlement of claims. For example, bringing banking services to rural areas (where more than 11,000 Kenyan maize farmers, some the regulatory environment permits). For with as little as 1 acre, have obtained insur- example, Kenya’s M-PESA service has trans- ance policies that cover significant losses when formed rural banking there. This service allows drought or excess rain destroys their harvest. users to transfer money safely via their mobile Similarly, BASIX and a commercial insurer in phone without requiring a bank account. Ini- India provide weather insurance based on a tially intended to enable wage earners to send rainfall index to smallholders to improve their money home to families in rural areas, M-PESA access to credit. Payments are triggered when now allows customers to pay bills (utilities, rainfall at local weather stations exceeds a mini- school fees, and others), repay loans, and pay mum threshold; insurance contracts secure the insurance and microinsurance premiums. A repayment of loans (IFPRI and World Bank new business feature allows companies to pay 2010, brief 9). employees via M-PESA. Equity Bank in Kenya recently offered all M-PESA users the option Loan Guarantees to open a savings account, using M-PESA to Banks reluctant to enter the business of agri- deposit and withdraw funds (IFPRI and World culture can sometimes be induced to do so Bank 2010, brief 8). Young people are especially through partial guarantee schemes that protect quick to adopt innovations based on mobile their losses in cases of default. The Alliance for phones when they have access. a Green Revolution in Africa has established The use of biometrics is being explored in an innovative financing initiative operating in the context of credit markets in countries where Kenya, Mozambique, and Tanzania. The initia- unique identification systems do not exist (mak- tive provides partial guarantees that result in ing it difficult for banks to spot repeat default- lower interest rates on loans to smallholders. ers). Biometric identification allows lenders to Since 2009, it has provided US$160 million in withhold new loans from known defaulters and financing to smallholder agriculture. Rabo- to grant loans to known responsible borrowers. bank’s Rabo Sustainable Agriculture Guaran- An experiment in Malawi linked higher repay- tee Fund issues partial credit guarantees and ment rates with the use of fingerprint scanning provides other financial products to mitigate among paprika farmers (IFPRI and World Bank the risks of financial intermediaries, allowing 2010, brief 9). Biometric tools that reduce the them to offer better prices and terms for com- costs of identifying borrowers and diminish mercial finance to grow and export agricultural default rates can enhance outreach to hard-to- produce.11 serve clients. Such measures are unlikely to be introduced solely to foster the employment of Rural Finance Targeted to Young People young people, but they are yet another example None of these innovations in rural finance of how measures that generally facilitate agri- is relevant exclusively to young people. Nor cultural growth have specific, significant ben- should young people be segregated as a group efits for the young. and offered financial services designed specifi- cally for them. The risks of working with this Insurance client base are high, and separating young peo- Innovations in microinsurance are also under ple from a larger pool for sharing risks would way. The International Labour Organization make them even less attractive to financial (ILO) estimates that microinsurance in Africa institutions. Rather, any and all innovations in almost doubled between 2006 and 2009 from finance that facilitate sustainable outreach to a very small base. Microinsurance differs from small farmers and rural entrepreneurs should agriculture as a Sector of opportunity for Young africans 127 be supported. When necessary, additional fea- of land. Accelerated land registration facili- tures should be added to enable these programs tates land rental markets, which make it easier to serve young people. for the poor to access land on rental terms . . . Land access for the poor can also be improved by redistributing underused and unused agri- cultural land to them. Land Policies That Benefit the Young Policies and programs to improve access to land can include special provisions to assist Of the many aspects of land administration young people. Several are described next.16 that require attention in Africa, the two that matter most to young entrants to the labor Systematic Inventory and Registration force are the need to improve security of ten- of Land ure and the need to relax controls on rental. Systematic land registration is a prerequisite for Land redistribution will also influence young creating employment in agriculture through people’s access to land. In general, policies and any of the four pathways. Notable efforts are measures that help the poor to gain access to yielding results for various categories of ten- land will also help young people. ure, but the pace of activity does not reflect the High food prices and the resulting spike in urgency of the problem. demand for land add urgency to the challenges Only 10 percent of occupied land in Africa is of improving land governance for all citizens formally registered (World Bank 2012b). State and applying appropriate safeguards to protect ownership of land is widespread, but even the land rights of the poor. When arrangements state-owned land is not fully documented, and for governance are weak, the rights of tradi- long-term use and occupancy by individuals or tional users may be overlooked or abused, con- groups blur ownership claims and limit invest- sultation with communities about impending ment. For example, in Ghana in 2000, the state transactions may be limited, and transparency owned about 40 percent of urban and periur- may be constrained (IDS 2012). ban land, most of it undeveloped (Kasanga and Decentralized land administration can Kotey 2001). Periurban land is often in transi- empower local communities, expedite deci- tion from agricultural to nonagricultural use. sions on land management and uses (highly It is well placed to offer high returns in inten- desirable for individuals pursuing pathways sive horticulture, tank aquaculture, and pig or 1 and 2), and help to clarify the legal rights poultry production, but it requires significant of landowners and tenants amid the surge in investment. Holders of periurban land often demand for land.12 Various models for decen- have other income streams and are linked to tralization exist (see Bruce and Knox 2009, the financial system, so they can, in principle, for example). Their success depends on their make the required investments. They will not design, implementation, and prevailing local do so, however, if their ownership or the dura- conditions. tion of their tenure is ambiguous. The Land Governance Assessment Frame- work13 and the Voluntary Guidelines on the Inventory and registration of individual land Responsible Governance of Tenure of Land, Fish- rights. Several countries are making progress eries, and Forests in the Context of National Food in formally documenting individual land- Security14 have been developed to assist deci- ownership. By the end of 2012, Rwanda had sion makers at the country level and guide the demarcated all 10.5 million land parcels in the formulation of land tenure projects and poli- country and registered and prepared leases cies.15 According to the World Bank (2012b), for at least 83 percent of them. Of the almost Sound land policies can safeguard the liveli- 1 million leases collected as of March 2012, hoods of the very vulnerable by giving them 7 percent were claimed by women, 5 percent access to land and income-earning opportuni- by men, 83 percent by married couples, and ties through rental markets or redistribution 1 percent by other legal entities (World Bank 128 Youth EmploYmEnt in Sub-Saharan africa Box 4.8 securing an agreement with an outside inves- tor (that will generate jobs within the com- munity) or allocating a portion of communal Documenting land rights: Encouraging land to young people for new farm starts. As investment and reducing the cost of noted in a recent World Bank review (World land transfers Bank 2012b), registration can be very slow if there are no clearly defined community own- The economic literature has long held that more secure tenure will ers of land and if new formal entities have to increase investments in land. Evidence from Ethiopia and Rwanda be developed. Demarcation of communal appears to confirm this finding and to highlight improvements in land boundaries requires time and financial environmental management as well. Other estimates suggest that resources. Registration needs to be followed certification-induced investment increased output in Ethiopia by up with resources to plan for land use and to about 9 percentage points (Deininger, Ali, and Alemu 2011). Invest- delineate common property resources (such as ment and productivity improvements were also found in Benin, where households participating in rural land use plans planted more peren- grazing land). nial crops than nonparticipants (Selod 2012). The same documentation of rights that strengthens tenure can Inventory of state land. The extent of state reduce the cost of transactions. By 2010, both Ghana and Rwanda landownership in Africa is largely unknown, as had reduced the cost of transferring property to less than 1 percent of most lands have not yet been surveyed and reg- property value (World Bank 2010b). istered. Some governments have started inven- torying state-owned land, including recent efforts in Ghana and Uganda. Underused or poorly used state land can be auctioned to 2012b). Ethiopia used a participatory public the private sector in ways that combine large- process to award certificates for more than scale operators and small and medium farm- 25 million parcels in rural areas throughout ers in innovative relationships—with care to the country, with noted benefits, including avoid disenfranchising indigenous users such “reduced conflicts, empowerment of women, as herders and subsistence communities. Long- increased individual and community invest- term occupants can be formally (legally) rec- ment, and improved security” (World Bank ognized as owners (as in Kenya), and land can 2012b). Madagascar has issued 75,000 certifi- be made available to land-poor farmers (as in cates akin to traditional land titles; Tanzania Malawi), including the young (for Kenya, see issued about 27,000 certificates of customary United Republic of Kenya 2010; World Bank rights of occupancy in two districts. A pilot 2011; for Malawi, see World Bank 2004; Tchale program in Ghana registered nearly 10,000 2012). Individuals in pathway 2 are most likely land parcels in periurban areas, and a similar to benefit from these programs, and underused program in Uganda registered 10,000 parcels state-owned land is a clear source of supply for in three districts. Benin, Burkina Faso, and young people showing promise in farming. Côte d’Ivoire have been piloting various rural land use plans—plans fonciers ruraux—as Reforms in Land Rental Markets another way to establish individual land use For the very poor, the landless, the young, and rights. While methods have differed and suc- migrants, land rental is the gateway to agricul- cess has varied, these efforts have done much tural employment and eventual landownership. to establish smallholders’ land rights (box 4.8). For those pursuing pathway 1 (perhaps hoping to acquire additional land to expand family Inventory and registration of communal holdings) and pathway 2, rental is a workable land. Where legal provisions recognize cus- approach to gain access to land. Worldwide, tomary tenure and communal land, it may be evidence demonstrates that introducing more appropriate to register communal land long-term leases or certifying land rights can than individual holdings. Registration of com- increase land rental activity,17 because people munal land can be an important first step in with secure rights are more likely to offer tem- agriculture as a Sector of opportunity for Young africans 129 porary use of their land to others. In turn, well- Box 4.9 functioning land rental markets can facilitate labor mobility, increase efficiency by transfer- ring land to more productive users, increase Mexico’s program to speed intergenerational equality, and enhance structural transforma- land transfers tion. Rental can be particularly helpful in eas- ing the intergenerational transfer of land while Most land in Mexico was held in common under unclear tenure still providing income to elderly owners (box arrangements until reforms initiated in the 1990s. Heavy restrictions on the transfer of rights to common land from one generation to 4.9). The most common restrictions on rental another limited young farmers’ access to land. In the early 2000s, with markets, such as ceilings on rental rates or pro- support from the World Bank and as part of a wider set of reforms hibitions against absentee landownership, are in land administration, the Government of Mexico initiated its Young often introduced to safeguard smallholders’ Rural Entrepreneurs and Land Fund Program (Programa para Jóvenes interests, but instead they may lock land into Emprendedores Rurales y Fondo de Tierra) to accelerate the intergen- inefficient patterns of use, greatly disadvantag- erational transfer of land. This successful program provided credit for ing prospective young users (Deininger 2003). rural youths without land to acquire underused common land. The Land rental markets have promoted com- young people were trained and received technical assistance in set- mercial farming in Ghana and created new ting up their farming activities. The program also helped older land- opportunities elsewhere in West Africa (for owners who transferred land to young farmers to gain access to social Ghana, see Amanor and Diderutuah 2001; welfare schemes for their retirement. for elsewhere in West Africa, see Estudillo, Source: FAO, IFAD, and MIJARC 2012. Quisumbing, and Otsuka 2001). In Sudan land rental markets facilitated the transfer of land to smaller producers (Kevane 1996). In con- trast, Uganda’s rental markets largely ceased to function in the 2000s due to severe ceilings on be seen in Malawi, South Africa, and Zambia, rent and controls on the eviction of tenants. In and each of which has drawn on lessons from Ethiopia, restrictions on land rental markets programs in Brazil. in all regions except Amhara not only reduced opportunities to use land more productively Malawi’s pilot land reform program. Malawi but also may have inhibited development of recently piloted a land reform program in four the nonfarm sector, as individuals who took districts in which underused land from former nonfarm jobs perceived that the risk of los- tea estates was made available to smallholders ing their land through redistribution was high wishing to relocate from densely settled areas.19 (Deininger et al. 2003). Patterned after Brazil’s market-based approach to land reform, the pilot had three key ele- Redistribution of Agricultural Land ments: (1) communities voluntarily acquired Land redistribution programs can profoundly land from estate owners, the government, or and positively affect the poor, but their suc- private donors; (2) resettlement and on-farm “But even for a cess depends critically on their objectives and development included transportation of set- farmer, without design. If operated at sufficient scale, they can tlers, establishment of shelter, and purchase change the income distribution and increase the of basic inputs and advisory services; and (3) education, forget incentives and opportunities for investment by redistributed land was surveyed and registered, about good poor households. If poorly designed, they can initially under group title, with the expecta- production.” transfer land to persons poorly suited to farm tion that individual titles would be provided to Tanzania it and can discourage investment by heighten- beneficiaries on demand in the future. A cap ing uncertainty about future redistribution. As on the maximum amount of a grant that could individuals in pathway 2 have the greatest need be spent to acquire land improved the bargain- for new land, they will have the most to gain or ing power of beneficiaries in relation to land lose from approaches to redistribution. Exam- sellers,20 and access to advisory services signifi- ples of approaches to land redistribution can cantly lowered the failure rate. 130 Youth EmploYmEnt in Sub-Saharan africa Although the program did not focus explic- similar to collective farms and experienced the itly on attracting young people, most partici- deficiencies in internal management that are pants were young. They preferred to relocate common under such arrangements. within or close to their original home, which preserved sociocultural ties and supported Zambia’s irrigation development and support young people when they were forming families project. In this project (approved in 2011 and and needed links to the older generation. The receiving just over US$200 million from all program distributed an average of more than funding sources), smallholders can exchange 1.5 hectares to each of 15,142 rural house- small parcels for holdings of 3–5 hectares as holds, increased agricultural incomes 40 per- part of a larger scheme that will join small pro- cent a year on average for beneficiaries between ducers, large commercial operators, and mid- 2005–06 and 2008–09, and had positive effects size farmers in a shared area. Management of on surrounding communities. irrigation services for the entire scheme will be contracted to a concession. To ensure that South Africa’s land reform program. Despite small and mid-size farmers earn enough to having had programs in place for a decade and pay irrigation fees, professional farm manage- a half, South Africa has made little progress ment services will be available to assist with in providing growing numbers of underem- production and marketing. The selection of ployed rural young people with land that they smallholders taking on the mid-size parcels is can farm.18 At the end of apartheid in 1994, not complete, so the age distribution of par- South Africa’s new government introduced ticipants is unknown, but this opportunity is tenure reform, restitution, and redistribu- expected to appeal to young people with prior tion of land. The redistribution program was experience in farming. designed to transfer land through market- mediated transactions to historically disad- vantaged South Africans who wanted to enter Enhancing Skills and Building a farming. After disappointing results, a new Better Educational Foundation program launched in 2001 provided graduated areas of land and start-up grants depending on Rural children need to go to school and learn. the amount of the beneficiary’s contribution. Better-educated farmers are more likely to Those who could contribute little (and most of adopt modern farm inputs and technolo- that in kind) received a base allotment of land gies, make better use of purchased inputs and and a grant of R 20,000 (almost $3,000). Those labor, choose technologies more effectively, who could contribute more or leverage a bank and respond rapidly to changes in markets or loan could receive larger holdings and a grant to natural calamities (Schultz 1988). Basic edu- of up to R 100,000 (about $14,000). Although cation can significantly improve the efficacy of the goal was to redistribute 30 percent of the agricultural training. The relationship between land by 2014, as of March 2011, only 6.27 mil- education and agricultural development cuts lion hectares (7.2 percent of land owned by both ways, and the two are mutually reinforc- white African farmers) had been redistributed ing, with demand for schooling rising as rural to black African farmers, and many recipients incomes increase. struggled to manage that land well. The pro- To equip young Africans with the skills gram had limited success because allocations to thrive in all four pathways to agricultural of land and start-up capital were not accom- employment, schools must do a better job of panied by advisory services or technical assis- providing the basic skills for any endeavor. tance. Those who received land through the Foremost among these are reading, writing, program were prohibited from subdividing numeracy, and the ability to use digital tech- it, even though repeal of the prohibition was nology to access and interpret information announced several times. Beneficiaries of small (box 4.10). Beyond these basics, the skills allotments were forced into group structures required for individuals in pathways 1 and 2 agriculture as a Sector of opportunity for Young africans 131 Box 4.10 Information and communication technologies: Altering the flow of agricultural information For African farmers operating in an environment of rapid cli- price and weather information in Colombia and India failed matic and economic change, the age-old questions of what to have a substantial impact on crop choices, revenues, or to plant, when to plant, and how to plant have assumed profits (Camacho and Conover 2011 on Colombia; Cole and immense importance. Answers are proliferating, as rural Hunt 2010 on India). These findings suggest the need to radio, television, the Internet, and mobile services offer an experiment with different content and methods of delivery. expanding array of information about specific technologies More attention should be given to the underinvestigated, and practices, climate change, disaster management, early distinct, and potentially complex issue of using ICT to sup- warning (for drought, floods, and diseases), prices, natural port agricultural production and production technologies, as resource management, production efficiency, and market distinct from marketing information. access (World Bank 2012a). For newer ICTs such as mobile phones and the Internet to Researchers are rigorously testing the effectiveness of dif- convey agricultural information effectively, the content and ferent information and communication technologies (ICTs) at mode of delivery may need to change, but what about the reaching and benefiting farmers, focusing primarily on the users themselves? Much has been made of young people’s transmission of price information.a Radio, a long-standing ready adoption of newer ICTs, as well as the capacity of ICTs method of communicating with farmers, has helped them (newer and older) to break through barriers to the acquisi- to obtain better farm-gate prices by providing information tion of information and skills, such as distance, the inability on commodity prices. Internet stations with agricultural mar- to read and write, or the expense of producing and dissemi- ket information (e-Choupals) have done the same in Mad- nating audiovisual information. Much has also been made hya Pradesh, India. Mobile phones have allowed fishermen of the potential for interactive ICTs to provide agricultural and onshore buyers in Kerala, India, to communicate supply recommendations tailored to an individual farmer’s circum- and price information, resulting in higher profits, lower con- stances. Yet much depends on whether the individual user sumer prices, and less waste. Among grain traders in Niger, of ICTs is able to frame relevant questions based on learning the introduction of mobile phone coverage in two markets acquired in good primary schools, coupled with practice in reduced the price variation between the two, ultimately imagining states of the world other than those already expe- increasing traders’ profits, decreasing prices paid by consum- rienced. As discussed in chapter 3, most African educational ers, and increasing total welfare. Cell phone coverage also systems are not delivering high-quality basic education, even reduced within-year price variation for producers. though enrollment is increasing. Among rural youths who Actors in the public and private sector are interested in lack a basic cognitive foundation on which they can build, using mobile phones to deliver information to farmers in the benefits of ICTs may be slower to emerge. developing countries. A rigorous evaluation of one program found that text messages provided by Reuters Market Light a. See Svensson and Yanagizawa (2009) on radio and farm-gate prices; did not have a robust effect on producer prices or input Goyal (2010) on Internet stations in India; Jensen (2007) on fishers in use among farmers of five crops in Maharashtra, India (Faf- Kerala; Aker (2008) on traders in Niger; and Aker and Fafchamps (2010) champs and Minten 2012).Similar interventions to supply on phone coverage and variation in producer prices. may differ from the skills required in pathways from processing and marketing to agricultural 3 and 4. The majority of farmers, who will have research. little more than a primary school education, This section describes recent initiatives and will need access to effective agricultural exten- changes in “schools of thought” regarding the sion services to sharpen their skills and clearly role and delivery of both basic and agricultural convey their requirements for information education, agricultural extension, and other and technology to service providers. Finally, innovative models of training and research a growing and diversifying agricultural sec- and development. As with some of the finance tor will create jobs that demand increasingly mechanisms discussed earlier, many of these advanced technical and professional skills, programs are experimental. They have not 132 Youth EmploYmEnt in Sub-Saharan africa been rigorously evaluated or tested for their without schooling, assuming that both have effectiveness or sustainability on a larger scale. access to the same assets, and this effect is magnified in environments undergoing rapid Schooling and Learning technical change (Rosenzweig 2010). Schooling To the extent that schooling raises literacy and thus enhances learning, and a dynamic agricul- numeracy skills and enhances the ability to tural sector provides opportunities to apply it. process agricultural information, an educa- Education offers spillover effects when tion effect can exist independent of the design uneducated farmers are able to observe the of school curriculum. Returns to such skills choices and outcomes of their better-educated “I decided to get are particularly magnified in a modernizing neighbors. This type of social learning is usu- married after agricultural sector, where access to advanced ally inferred from observed behavior or out- technology complements an understanding of comes over time. For example, in Ghana, social my stepmother learning played an important role in diffusing how to use it. The decision to adopt new tech- refused to pay nologies is an investment decision if significant knowledge on pineapple cultivation among school fees for costs are incurred in obtaining information and farmers. In this case, the experiences of the me, and the learning about the performance of one or more farmers and their neighbors influenced profit- job I could new technologies, while the returns are distrib- ability and adoption rates (Conley and Udry find there was uted over time. Furthermore, only a small share 2010). of new technologies will be profitable for any The importance of female labor in agri- only farming.” culture brings into sharp focus the urgency given farmer to adopt. Given the degree and Tanzania multiple sources of uncertainty facing farmers, of improving access to schooling for girls and effective schooling may help them to make bet- women. There is widespread recognition of the ter decisions to increase farm profitability. need to improve both basic education and agri- Returns to schooling in rural areas depend, cultural vocational education for women and in part, on the pace of technological innovation to enhance rural women’s access to extension in farming. A large body of literature has shown services. that more educated farmers are the first to adopt new seed, tillage practices, fertilizer, and Postsecondary Education in Agriculture animal breeds (for example, see Welch 1970; Alongside jobs in primary production, a grow- Huffman 1977; Besley and Case 1993; Foster ing and diversifying agricultural sector creates and Rosenzweig 1996; Abdulai and Huffman demand for skilled labor in rural areas inside 2005). Moreover, farmers with primary educa- and outside of the sector. Africa’s existing agri- tion tend to earn higher profits than farmers cultural vocational schools can play a construc- tive role in training skilled personnel for jobs in processing, marketing, machinery opera- Box 4.11 tion and repair, transport, logistics, and quality control—provided that students have suffi- Vocational agricultural education: cient preparation to benefit from that training (box 4.11). A poor substitute for general education The number and quality of trained techni- To remedy gaps in skills, many African governments and their devel- cal and professional personnel in agriculture opment partners in past years focused on vocational agricultural train- are critical factors in agricultural development, ing, often at the expense of primary, secondary, or postsecondary because a sector undergoing structural trans- education. More often than not, attempts to substitute vocational formation has an expanding need for skills. education in agriculture for general education have failed. The returns In addition to gaining technical skills, workers to such training have been mixed at best, largely because the indi- need to master teamwork, communication, dili- viduals undertaking the training and the neighbors who could learn gence, creativity, and entrepreneurship. In many from their example lacked sufficient basic education to make use of cases, these behavioral “soft skills” are learned more advanced, specific skills. through mentoring and through the standards of performance set in the formal workplace. agriculture as a Sector of opportunity for Young africans 133 As formal nonfarm employment increases, a settings around the world. Their needs then “Many who larger cohort of young people can be expected were quite similar to those of young Africans received training to acquire behavioral skills in this way. now entering the labor force in rural and urban commented Institutional infrastructure for agricul- areas. Much of what has been learned about tural higher education and training has been effective extension methods may be used to that their in place in Africa since the 1960s but has not design advisory services and mentoring pro- farming efforts strengthened sufficiently over time to meet the grams for young people generally, in farming are now ‘more enormous demands evident now. Sub-Saharan and in other spheres of activity. efficient and Africa now has more than 200 public universi- Early models of agricultural extension were sustainable.’ ” ties (compared with 20 in 1960), about 100 of centralized, public, and linear. The basic model Ethiopia which teach agriculture and natural resource was one in which a trained extension worker management. Private universities complement traveled over a large area to convey messages to this public capacity (World Bank 2007a). Much farmers, who then applied the advice to improve stronger national and regional institutions are their operations. The deficiencies of this model needed to train future professionals and lead- became clear over the years, especially in Africa. ers with appropriate technical and functional The foremost deficiencies involved cost, qual- skills. ity, and relevance. Traditionally designed agri- Women face particular obstacles in obtain- cultural extension programs are now a rarity, ing the education and training to become more although the term is still used and applied to successful in agriculture, whether as farmers, nontraditional approaches. Newer programs entrepreneurs, providers of extension infor- empower farmers to specify the informa- mation, or leaders in agricultural research and tion they require and to select the provider education. Few women graduate from agricul- (box 4.12; see Davis 2008). The provision of tural education programs, few women become information is still recognized as a public good, agricultural extension workers, and women are and the government assumes a share of the cost, often marginalized during agricultural events, particularly for small farmers and the poor. The activities, and programs. These problems are advice may be delivered by public officials, pri- widely evident, although detailed gender- vate advisers, NGOs, or the media, depending disaggregated data are only available sporadi- on farmers’ needs. The new systems are decen- cally or are not reported at all (World Bank tralized, integrated with the private sector, coor- 2009). Efforts are under way to give stronger dinated with agricultural research, and tailored recognition to the role of women in agricul- to local contexts. Extension is understood to be ture, to increase the number of female students part of a broader innovation system. in agricultural schools and colleges, and to pro- Agricultural extension services can con- vide resources for extension services directed to tribute significantly to young farmers’ success, women farmers. An innovative program was but the design of successful programs is still an launched in 2008 by the Gender and Diversity open empirical question. Several approaches Program of the Consultative Group on Inter- have been tried and reviewed in different con- national Agricultural Research. The African texts, but rigorous assessment is elusive (Davis Women in Agricultural Research and Develop- 2008). For example, participatory and group- ment Program seeks to strengthen the research based approaches are gaining traction. These and leadership skills of African women in agri- methods have the potential to overcome bar- cultural science, empowering them to contrib- riers to participation, foster inclusiveness, and ute more effectively by establishing mentoring lead to more demand-driven services. Many partnerships, building science skills, and devel- African countries have pluralistic extension oping leadership capacity. services (which have a variety of service pro- viders), including Kenya, Mozambique, and Agricultural Extension Uganda. Agricultural extension arose to address farm- Many factors in addition to the mode of ers’ needs for information in a wide array of providing extension services affect agricultural 134 Youth EmploYmEnt in Sub-Saharan africa Box 4.12 Innovations in agricultural extension: Relying on farmers to improve service delivery The question of whether agricultural extension services are nicator. Small performance-based material and social incen- best provided by the public sector (the traditional model), the tives were distributed to a subset of the treatment villages. private sector, or a range of providers continues to generate The results from these large pilots suggest that female debate in light of the mixed results obtained with the tra- farmers can be as productive as male farmers in teaching ditional model.a This uncertainty has complicated efforts to their peers about a new technology and getting them to widen the adoption of improved agricultural technologies adopt it. Adding a woman communicator to a male-cen- and increase agricultural productivity. Newer models of exten- tered model for delivering extension advice can add value sion, which are driven by farmers and reinforce the quality of and change the numbers of male and female beneficiaries. service delivery through incentives and other innovations, are Evidence on the use of performance-based incentives sug- proving more effective than traditional methods, but they, gests that providing service to the community is more costly too, encounter difficulties and constraints to effectiveness. for women leaders, as they are more responsive to incen- The impact of innovative approaches that mobilize farm- tives. Finally, despite performing as well as, and in some ers to improve returns to agricultural extension was recently cases better than, male communicators, female communica- evaluated on a large scale. In Malawi and Mozambique, tors still suffer from discrimination and are rated as worse randomized controlled trials tested multiple modalities for teachers than men. implementing peer and lead farming. In both experiments, Overall, these results suggest that development projects communities nominated lead or peer farmers, who were that place the adoption of new agricultural techniques at the trained to use sustainable land management techniques and center of their theory of change may consider using peer and mandated to communicate those techniques to other farm- lead farming interventions to boost their returns. Given that ers in their village through demonstrations. female leaders appear to be as productive as male leaders in In Malawi, the social status (peer versus lead) and gender getting farmers to learn about and adopt new techniques, of the communicator were subject to random assignment, empowering women to take on leadership roles may not and a small performance-based material incentive was given only increase equity but also add value. Performance-based to a subset of the communicators. The project designated incentives can play an important role in getting women lead- “shadow” communicators in control villages to provide ers to devote additional time and effort to working with their a counterfactual. In Mozambique, lead farmers—mostly community. men—were already designated by the project team in all project villages at baseline. The intervention trained a ran- a. For example, Birkhaeuser, Evenson, and Feder (1991) found no signif- dom subset of these lead farmers in sustainable land man- icant relationship between the provision of traditional extension services and farm productivity in Africa, whereas Evenson (2001) and Dercon et agement. To add a gender variation, a woman lead farmer al. (2009) identified some successes. Anderson and Feder (2003) pro- was added to a random subset of treatment villages, since it pose an organizational inquiry into which model of extension (public or was not possible to demote the previously chosen commu- private) can deliver superior results. performance. Spillover effects are hard to cap- Each pathway to agricultural employment ture or isolate. Selection bias may enter even has particular needs for advice and training in controlled environments, and programs per- that may determine which approach will work forming well at scale can be subverted by cli- best. For example, programs of competence- entelism and patronage (Anderson and Feder based training in Ethiopia and Uganda in high- 2004). While most experts would agree that value export crops (horticulture and floricul- advisory services or extension of some kind ture, respectively) are providing a workforce are vital, particularly in light of the challenges for these demanding subsectors. Individuals in faced by young people entering agriculture in pathways 3 and 4 (engaged in wage work either Africa, the profession does not have a clear view part or full time) might benefit most from such on the best approach to program design. training. For those in pathways 1 and 2, farmer agriculture as a Sector of opportunity for Young africans 135 field schools may be useful. They exist in many help they need, coupled with careful evalua- When asked countries and are “a participatory method of tion and transparent display of user satisfaction how their days learning, technology development, and dis- with the various channels of information. The differ from semination based on adult-learning principles alternative approach of seeking a new, highly such as experiential learning” (World Bank structured, unitary style of extension system, their parents’ 2012a). A recent study in East Africa found widely applicable to all, is not likely to deliver generation, that farmer field schools are especially benefi- good results. In designing a skills agenda for one young man cial for women, people with low literacy levels, Africa’s rural young people, distinctions should offers, “Those and farmers with medium-size landholdings. be made between the needs of the stock of days were days Field school participants had significant dif- young adults in the labor force today—whose for farming and ferences in outcomes with respect to the value elementary education is incomplete—and the of crops produced per acre, the livestock value children who will flow into the labor force in that is it. Now gained per head, and agricultural income per the future. The very young need better schools we do farming, capita (Davis et al. 2010). For those in path- with more learning as a matter of highest we do business, way 1 who adopt a more corporate approach urgency. Those who are already beyond school we even go to to family farming, the shift to entrepreneurial age will need a mix of short-term remedial pro- school in great family farms can be aided by local agribusiness grams for applied literacy and numeracy and numbers.” development services, which are advisory ser- access to mentoring, apprenticeships, and flows vices with a business orientation. Although the of “just-in-time” information, perhaps deliv- Tanzania provision and use of these services are still rela- ered through electronic media. tively new, anticipated impacts for smaller-scale farmers and entrepreneurs include enhanced rural income (both directly and through Current Agricultural Programs employment) and enhanced small-scale entre- Deliver Too Little, Too Slowly, preneurial activity (World Bank 2012a). Such to Meet the Needs of Africa’s services could also assist young people seeking Young People to combine self-employment with part-time farming (pathway 3). As early as 2003, African heads of state met Producer organizations can be a highly in Maputo, Mozambique, and pledged to give effective means of building members’ skills renewed attention and resources to agriculture. and entrepreneurial expertise by improving The pledge was made under the rubric of the access to advice and training. Once again, how- Comprehensive African Agriculture Develop- ever, the ability to self-organize and participate ment Programme (CAADP) of the African effectively in such organizations requires the Union and the New Partnership for Africa’s fundamentals of a solid education. Development.21 The CAADP framework rec- ognizes the breadth of the agricultural agenda Priorities for a Diverse Skills Agenda and the corresponding need for multiple entry The skills agenda to meet the needs of Africa’s points and complementary public investments young people is diverse, and the resources to in several areas (box 4.13). address that agenda are highly constrained. Until food prices spiked in 2008, the com- Priority should be accorded to improving the mitment to increase public spending on agri- quality of basic education and keeping young culture was not implemented widely, but people in school long enough for them to between 2003 and 2008, the technical work to acquire basic skills. Agricultural programs in design a framework for reinvestment in agri- tertiary education must be strengthened to culture under CAADP proceeded. When rising produce a new generation of scientists and food prices caught the attention of global and teachers in all fields. In the intermediate arena African leaders, the conceptual framework was of extension and outreach, emphasis should be available to address the long-standing neglect placed on providing resources to the final users of key public goods and services. The frame- of information, so that they can seek out the work does not specifically recognize the unique 136 Youth EmploYmEnt in Sub-Saharan africa Box 4.13 Overview of the Comprehensive African Agriculture Development Programme (CAADP) CAADP’s four pillars are complementary. The first pillar, land ity and profitability provide a cushion of savings for hard and water, addresses the design of programs and invest- times. ment required to improve land administration, sustainability Finally, the agricultural technology pillar underpins the of land use, and management of water through irrigation other three. Modern agriculture is science based, and pro- and water harvesting and storage. ducers at all levels of sophistication benefit from improved The second pillar identifies investment and reforms in systems to generate and spread improved technologies. policy and regulations needed to improve smallholders’ Some of these entail breeding improved crops and animals access to markets. Many of these interventions focus on rural to address changing demand or agro-ecological conditions infrastructure, including roads, rail transport, and power or to allow producers to select a desired level of risk. Oth- (both on- and off-grid), but important regulatory measures ers emphasize new systems of management and rotation, also require attention, such as regulation of the trucking to reduce costs of inputs, enhance soil health, and capture industry and food safety standards. These measures aim to carbon for additional revenue streams. reduce marketing costs to make farming more profitable, The CAADP framework is applied to help countries and while reducing food prices for net buyers to accelerate job regions to improve the quality of their agricultural planning creation. and policy making and to use them as the basis for scaled- The third pillar addresses measures that will make up investment in the sector. CAADP offers political, tech- agriculture less risky for those with a commercial orienta- nical, and financial support for countries and regions that tion and strengthen the resilience of the very poor when engage in this process, through a partnership of continental shocks hit. Diversification, affordable insurance products, and regional African institutions in collaboration with other and rural safety nets can help people to manage risks, and stakeholders, including civil society, the private sector, and higher income levels associated with growth in productiv- Africa’s development partners. demography of Africa, nor does it make spe- and private sectors, domestically and interna- cific provisions for young farmers, but its key tionally, under strategic initiatives already in features can be enriched to address a youth place. No new or separate strategy is required, agenda. but the current slow pace of implementa- Complementing the largely public ele- tion, if continued, will fail young people and ments of CAADP, local and international pri- compromise Africa’s future. Existing com- vate investors are expressing growing interest mitments must be accorded focused atten- in the opportunities in agriculture. The Afri- tion, with improved quality of public spend- can Union has declared a “Decade on Youth ing, more efficient approaches to increasing Development in Africa 2009–2019.” The United production of food staples, more attention to Nations General Assembly has called for mem- meeting demand for high-quality products by ber states to prepare a “National Review and the growing urban middle classes, continued Action Plan on Youth Employment,” and the progress on policy and regulatory reforms, and African Union, Economic Commission for improved data and tracking of progress. With Africa, African Development Bank, and the better-implemented public programs, private ILO have recently proposed a “Joint Initiative investment will accelerate, and opportunities on Job Creation for Youth in Africa.” The initia- for young people will increase. Some proac- tives demonstrate the level of attention being tive additional attention to meet the specific given to the youth employment issue (Proctor needs of the large group of young people may and Lucchesi 2012). Each of these organiza- be required, but doing so without more effec- tions also emphasizes agriculture. tive programs in general will be counterpro- Thus efforts to reach out to Africa’s young ductive. Successfully mobilizing the talents of farmers can draw on resources from the public young people will increase the likelihood that agriculture as a Sector of opportunity for Young africans 137 CAADP and other ongoing initiatives will meet years old—the “mature young.” Energy, cre- their ambitious goals. ativity, and strength are attributes that Africa’s young people have in abundance. The agricul- ture that attracts them will have to be profit- Harnessing Agriculture’s Youth able, competitive, and dynamic. These same Dividend characteristics are needed for agriculture to deliver growth, to improve food security, and Agriculture—already Africa’s largest to preserve a fragile natural environment. With employer—is changing, and the large num- much higher priority accorded to the imple- bers of young people entering the sector will mentation of well-designed programs of public accelerate the pace of change. Africa’s leaders investment in agriculture, continued progress recognize that agriculture is a source of growth, on regulatory and policy reform, and a mod- an instrument for improved food security, and est overlay of attention to assure the inclusion a means to steward valuable natural resources. of young people in Africa’s agricultural renais- As the potential for the sector to absorb the sance, the sector’s youth dividend can be col- large numbers of new job seekers and to offer lected and widely shared. meaningful work with public and private ben- efits becomes clearer, agriculture will gain even Notes more attention from policy makers. 1. Whether an economy’s agricultural labor force This attention will be necessary, because the rose or fell in absolute numbers as the relative sector’s ability to create jobs will not be real- share of the sector declined depended on birth and death rates in rural areas, migration, and ized without modifications to public programs. the size and labor intensity of sectors that were Present levels of public investment are not suf- growing more rapidly than agriculture. See Tim- ficient. The quality of investment is inadequate mer and Akkus (2008). to yield high returns. Too much has gone into 2. For example, farmers grow modern improved short-term palliatives, such as fertilizer sub- varieties of food crops on an estimated 35 per- sidies, without complementary attention to cent of all planted area, compared to just 23 per- improved technologies and management prac- cent in 1998 (Renkow and Byerlee 2010). tices and long-term investments in research 3. A band of settlements of 10,000 or more inhab- and infrastructure. The investment climate still itants now stretches from Djibouti to Dakar, cannot attract the private firms needed in mar- with few gaps in between. Another rings Lake keting, processing, input supply, and finance. Victoria, and another marks the Kinshasa- Brazzaville corridor. Public policies governing trade, the introduc- 4. Agricultural total factor productivity is grow- tion of new varieties, licensing and intellectual ing at just over and under 3 percent annually in property rights, and taxation offer weak incen- Southeast Asia and South America, respectively. tives to producers and innovators. Since 2000, total factor productivity in Sub- Detailed agendas in each of these areas are Saharan Africa has been higher than the aver- beyond the scope of this chapter, but the future age in the four prior decades but is still short of Africa’s young people is at present hostage to of being transformative. Estimates range from the wide gap between rhetorical commitment just under 1 percent annually to just over 2 per- to the importance of agriculture and actual, cent annually, owing to severe deficiencies in the effective attention accorded to it by Africa’s underlying data; see Fuglie (2011); Nin-Pratt, leaders. Efforts to address constraints to land, Johnson, and Yu (2012). 5. After food prices spiked in 2007–08, govern- capital, and skills will have to be redoubled and ments began to set ambitious growth targets for accelerated, and features to make programs agriculture. At 8–10 percent, those rates exceed friendly to the needs of the young introduced. the 6 percent target set by the African Union Although farming is often done by the through the Comprehensive Africa Agriculture elderly, the profession’s requirements for Development Programme and the rates recently energy, innovation, and physical strength make observed for the entire region (3.8–4.0 percent; it ideally suited for persons who are 25–45 World Bank 2012c). 138 Youth EmploYmEnt in Sub-Saharan africa 6. This section draws heavily on AgriFin (2012) for transportation, water, shelter, and farm and IFPRI and World Bank (2010). development. 7. Organisation pour l’Harmonisation en Afrique 21. See http://www.nepad-caadp.net/. du Droit des Affaires (Organization for the Har- monization of Business Law in Africa). 8. Caisses d’Epargne et de Crédit Agricole References Abdulai, Awudu, and Wallace E. Huffman. 2005. Mutuelles, a cooperative agricultural financial “The Diffusion of New Agricultural Technolo- institution. gies: The Case of Crossbreeding Technology in 9. This section draws from World Bank (2010a). Tanzania.” American Journal of Agricultural Eco- 10. IFPRI and World Bank (2010, brief 4); see also nomics 87 (3): 645–59. http://www.rabobank.com/content/products_ AgriFin. 2012. “Making Finance Work for Africa: services/business_clients/professionalproducts/ Policy Brief on Agricultural Finance in Africa.” raboagrifund/index.jsp. Report sponsored by the African Union, GIZ, 11. See http://www.rabobank.com/content/products_ and BMZ. Washington, DC. services/business_clients/professionalproducts/ Aker, Jenny C. 2008. “Does Digital Divide or Pro- raboagrifund/index.jsp. vide? The Impact of Cell Phones on Grain Mar- 12. For a perspective on equity issues arising from kets in Niger.” CGD Working Paper 154, Center heightened global interest in Africa’s farmland, for Global Development, Washington, DC. see Deininger and Byerlee (2011). Aker, Jenny C., and Marcel Fafchamps. 2010. “How 13. See Deininger, Selod, and Burns (2011) for more Does Mobile Phone Coverage Affect Farm-Gate on this relatively quick and innovative tool. Prices? Evidence from West Africa.” Working 14. Endorsed by the Committee on World Food Paper, Tufts University, Department of Econom- Security in May 2012, the Voluntary Guidelines ics and the Fletcher School; Center for Global promote secure tenure rights and equitable Development; University of Oxford, Center for access to land, fisheries, and forests as a means the Study of African Economies. http://www. of eradicating hunger and poverty, supporting aeaweb.org/aea/2011conference/program/ sustainable development, and enhancing the retrieve.php?pdfid=629. environment. They set out principles and inter- Amanor, Kojo S., and Maxwell K. Diderutuah. nationally recommended standards for respon- 2001. “Share Contracts in the Oil Palm and Cit- sible practices. They are a framework that actors rus Belt of Ghana.” International Institute for can use when developing their own strategies, Environment and Development, London. policies, legislation, and programs, allowing Anderson, Mary, Dayna Brown, and Isabella Jean. government authorities, the private sector, civil 2012. “Time to Listen: Hearing People on the society, and citizens to judge whether their pro- Receiving End of International Aid.” CDA Col- posed actions and the actions of others consti- laborative Learning Project, Cambridge, MA. tute acceptable practices. Anderson, Jock, and Gershon Feder. 2003. “Rural 15. Encouraging collaboration with the Sustain- Extension Services.” Policy Research Working able Commodity Roundtables can also help to Paper 2976, World Bank, Washington, DC. increase the extent to which crop production ———. 2004. “Agricultural Extension: Good Inten- systems meet voluntary environmental and tions and Hard Realities.” World Bank Research social criteria, including those of the Roundtable Observer 19 (1): 41–60. for Responsible Soy, the Roundtable on Sustain- Besley, Timothy, and Anne Case. 1993. “Modeling able Palm Oil, the Better Sugarcane Initiative, Technology Adoption in Developing Countries.” and a variety of forest certification processes. American Economic Review 83 (2): 396–402. 16. The following sections draw heavily from World Birkhaeuser, Dean, Robert Evenson, and Gershon Bank (2012b). Feder. 1991. “The Economic Impact of Agricul- 17. For example, in China, the Dominican Republic, tural Extension: A Review.” Economic Develop- Ethiopia, Nicaragua, and Vietnam. ment and Cultural Change 39 (3): 607–50. 18. This section is adapted from Lahiff and Li Bruce, John W., and Anna Knox. 2009. “Structures (2012); World Bank (2012b). and Stratagems: Decentralization of Authority 19. Adapted from Tchale (2012); World Bank over Land in Africa.” World Development, special (2012b). issue on the limits of state-led land reform 37 20. Each family received a grant of US$1,050, man- (8): 1360–69. aged directly by the beneficiary; up to 30 per- Camacho, Adriana, and Emily Conover. 2011. “The cent was for land acquisition, and the rest was Impact of Receiving Price and Climate Informa- agriculture as a Sector of opportunity for Young africans 139 tion in the Agricultural Sector.” Working Paper De Wolf, Stefan, Yves Rolland Rakotoarisoa, Lau- IDB-WP-220, Inter-American Development rence Vanpaeschen, and Honoré Rabekoto. 2008. Bank, Washington, DC. Madagascar: Le Grand Livre des Petits Métiers: Chirwaa, Ephraim W. 2008. “Land Tenure, Farm Portraits of Daily Life Professions. Heule, Belgium: Investments, and Food Production in Malawi.” Snoek Publishers. Discussion Paper 18, prepared for Institutions Djurfeldt, Agnes Andersson, and Magnus Jirström. and Pro-Poor Growth (IPPG). 2013. “Urbanization and Changes in Farm Size in Cole, Shawn, and Stefan. Hunt. 2010. “Information, Sub-Saharan Africa and Asia from a Geographi- Expectations, and Agricultural Investment: Evi- cal Perspective: A Review of the Literature.” A dence from a Field Experiment in India.” Work- Foresight Study of the Independent Science and ing Paper, Harvard University, Cambridge, MA. Partnership Council, Consultative Group on International Agricultural Research, Washing- Conley, Timothy, and Christopher Udry. 2010. ton, DC. http://www.sciencecouncil.cgiar.org/file “Learning about a New Technology: Pineapple admin/templates/ispc/documents/Strategy_and_ in Ghana.” American Economic Review 100 (1): Trends/2013/Foresight.Andersson.pdf. 35–69. Estudillo, Jonna P., Agnes R. Quisumbing, and Davis, Kristin. 2008. “Extension in Sub-Saharan Keijiro Otsuka. 2001. “Gender Differences in Africa: Overview and Assessment of Past and Land Inheritance and Schooling Investments in Current Models and Future Prospects.” Journal of the Rural Philippines.” Land Economics 77 (1): International Agricultural and Extension Educa- 130–43. tion 15 (3): 15–28. Evenson, Robert. 2001. “Economic Impacts of Agri- Davis, Kristin, Ephraim Nkonya, Edward Kato, cultural Research and Extension.” In Handbook of Daniel A. Mekonnen, Martins Odendo, Richard Agricultural Economics. Vol. 1a: Agricultural Pro- Miiro, and Jackson Nkuba. 2010. “Impact of duction, edited by Bruce L. Gardner and Gordon Farmer Field Schools on Agricultural Productiv- C. Rausser, 573–628. Amsterdam: Elsevier. ity, Poverty, and Farmer Empowerment in East Africa.” IFPRI Discussion Paper, International Executive Office of the President. 2011. “Economic Food Policy Research Institute, Washington, DC. Report of the President 2011.” U.S. Council of Economic Advisers, Washington, DC. Deininger, Klaus. 2003. Land Policies for Growth and Poverty Reduction. Washington, DC: World Fafchamps, Marcel, and Bart Minten. 2012. “Impact Bank. of SMS-Based Agricultural Information on Indian Farmers.” World Bank Economic Review Deininger, Klaus, Daniel A. Ali, and Tekie Alemu. 26 (3): 383–414. 2011. “Impacts of Land Certification on Tenure Security, Investment, and Land Market Participa- FAO (Food and Agriculture Organization). 2011. tion: Evidence from Ethiopia.” Land Economics Women in Agriculture: Closing the Gender Gap 87 (2): 312–34. in Agriculture. State of Food and Agriculture Report. Rome: FAO. Deininger, Klaus, and Derek Byerlee. 2011. “Rising Global Interest in Farm Land: Can It Yield Sus- FAO (Food and Agriculture Organization), IFAD tainable and Equitable Benefits?” World Bank, (International Fund for Agricultural Develop- Washington, DC. ment), and MIJARC (Mouvement International de la Jeunesse Agricole et Rurale Catholique). Deininger, Klaus, Songqing Jin, Berhanu Adenew, 2012. “Facilitating Access of Rural Youth to Agri- Samuel Gebre-Selassie, and Mulat Demke. 2003. cultural Activities.” Farmers’ Forum Youth Ses- “Market and Non-Market Transfers of Land in sion, Rome, February 18. Ethiopia: Implications for Efficiency, Equity, and Nonfarm Development.” Policy Research Paper Foster, Andrew D., and Mark R. Rosenzweig. 1996. 2992, World Bank, Washington, DC. “Technical Change and Human-Capital Returns Deininger, Klaus, Harris Selod, and Anthony Burns. and Investments: Evidence from the Green 2011. “The Land Governance Assessment Frame- Revolution.” American Economic Review 86 (4): work: Identifying and Monitoring Good Practice 931–53. in the Land Sector.” ARD Series, World Bank, Fox, Louise, and Thomas Pave Sohnesen. 2012. Washington, DC. “Household Enterprises in Sub-Saharan Africa: Dercon, Stefan, Daniel O. Gilligan, John Hoddinott, Why They Matter for Growth, Jobs, and Liveli- and Tassew Woldehanna. 2009. “The Impact of hoods.” Policy Research Working Paper 6184, Agricultural Extension and Roads on Poverty World Bank, Washington, DC. and Consumption Growth in Fifteen Ethiopian Fuglie, Keith. 2011. “Agricultural Productivity in Villages.” American Journal of Agricultural Eco- Sub-Saharan Africa.” In The Food and Financial nomics 91 (4): 1007–21. Crises in Sub-Saharan Africa: Origins, Impacts, 140 Youth EmploYmEnt in Sub-Saharan africa and Policy Implications, edited by David R. Lee Africa.” IFPRI Discussion Paper 01059, Interna- and Muna Ndulo. Cambridge, MA: CABI. tional Food Policy Research Institute, Washing- Goyal, Aparajita. 2010. “Information, Direct Access ton, DC. to Farmers, and Rural Market Performance Morris, Michael, Hans Binswanger-Mkhize, and in Central India.” American Economic Journal: Derek Byerlee. 2009. Awakening Africa’s Sleeping Applied Economics 2 (3): 22–45. Giant. Washington, DC: World Bank. Huffman, Wallace E. 1977. “Allocative Efficiency: Nagayets, Oksana. 2005. “Small Farms: Current Sta- The Role of Human Capital.” Quarterly Journal tus and Key Trends.” Information brief prepared of Economics 91 (February): 59–79. for the Future of Small Farms Research Work- IDS (Institute of Development Studies). 2012. shop, Wye College, June 26–29. “Young People and Agriculture in Africa.” IDS Nin-Pratt, Alejandro, Michael Johnson, and Bulletin, special issue 43 (6). Bingxin Yu. 2012. “Improved Performance of IFPRI (International Food Policy Research Insti- Agriculture in Africa South of the Sahara: Taking tute) and World Bank. 2010. “Innovations in Off or Bouncing Back?” IFPRI Discussion Paper Rural and Agriculture Finance: Focus 18.” Wash- 01224, International Food Policy Research Insti- ington, DC. tute, Washington, DC. Jensen, Robert. 2007. “The Digital Provide: Infor- OECD and FAO (Organisation for Economic Co- mation (Technology), Market Performance, and operation and Development and Food and Agri- Welfare in the South Indian Fisheries Sector.” culture Organization). 2012. Agricultural Outlook Quarterly Journal of Economics 122 (3): 879–924. 2012–2012. Annex B, Table B.1. Paris: OECD; Rome: FAO. Kasanga, Kasim, and Nii Kotey. 2001. “Land Man- agement in Ghana: Building on Tradition and Petesch, Patti, and Ines Rodríguez Caillava. 2012. Modernity.” International Institute for Environ- “Voices of Young Villagers in Sub-Saharan ment and Development, London. Africa.” Background paper, World Bank, Wash- ington, DC. Kevane, Michael. 1996. “Agrarian Structure and Agricultural Practice: Typology and Application Proctor, Felicity J., and Valerio Lucchesi. 2012. to Western Sudan.” American Journal of Agricul- “Small-Scale Farming and Youth in an Era of tural Economics 78 (1): 236–45. Rapid Rural Change.” International Institute for Environment and Development, London. Kloeppinger-Todd, Renate, Ajai Nair, and Annabel Mulder. 2004. “Leasing: An Underutilized Tool in Renkow, Mitch, and Derek Byerlee. 2010. “The Rural Finance.” ARD Discussion Paper 7, World Impact of CGIAR Research: A Review of Recent Bank, Washington, DC. Evidence.” Food Policy 35 (5): 391–402. Lahiff, Edward, and Guo Li. 2012. “Land Redistri- Rosenzweig, Mark. 2010. “Microeconomic bution in South Africa: A Critical Review.” Work- Approaches to Development: Schooling, Learn- ing Paper for the “Land Administration and ing, and Growth.” Journal of Economic Perspec- Reform in SSA” study, World Bank, Washington, tives, American Economic Association 24 (3): DC. 81–96. Larson, Donald, Keijiro Otsuka, Tomoya Matsu- Schultz, T. Paul. 1988. “Education Investments and moto, and Talip Kilic. 2012. “Should African Returns.” 1988. In Handbook of Development Eco- Rural Development Strategies Depend on Small- nomics, Vol. 1, edited by Hollis Chenery and T. N. holder Farms? An Exploration of the Inverse Srinivasan, ch. 1, 543–630. Amsterdam: Elsevier. Productivity Hypothesis.” Policy Research Work- Selod, Harris. 2012. “Formalizing Rural Land ing Paper 6190, World Bank, Washington, DC. Rights in West Africa: Results from an Impact Losch, Bruno, Sandrine Fréguin-Gresh, and Eric Evaluation in Benin.” Paper prepared for the Thomas White. 2012. Structural Transformation conference “Land and Poverty 2012,” World and Rural Change Revisited: Challenges for Late Bank, Washington, DC, April 23–26. Developing Countries in a Globalizing World. Sutton, John, and Nebil Kellow. 2010. An Enterprise African Development Forum series. Washington, Map of Ethiopia. London: International Growth DC: World Bank. Centre. Maertens, Miet, Bart Minten, and Jo Swinnen. Svensson, Jakob, and David Yanagizawa. 2009. 2009. “Growth in High-Value Export Markets in “Getting Prices Right: The Impact of the Market Sub-Saharan Africa and Its Development Impli- Information Service in Uganda.” Journal of the cations.” LICOS Discussion Paper 245, Katho- European Economic Association 7 (2–3): 435–45. lieke Universiteit, Leuven. Tchale, Hardwick. 2012. “Pilot Redistributive Land Minot, Nicholas. 2011. “Transmission of World Reform in Malawi: Innovations and Emerging Food Price Changes to Markets in Sub-Saharan Good Practices.” Working Paper for the “Land agriculture as a Sector of opportunity for Young africans 141 Administration and Reform in SSA” study, World Region Human Development Department, World Bank, Washington, DC. Bank, Washington, DC. Timmer, Peter, and Selvin Akkus. 2008. “The ———. 2007b. “South Asia: Livelihoods Learning: Structural Transformation as a Pathway out of Community-Managed Microfinance: A New Poverty: Analytics, Empirics, and Politics.” Work- Model from Sri Lanka.” Series 1, Note 5, World ing Paper 150, Center for Global Development, Bank, Washington, DC. Washington, DC. ———. 2009. Gender in Agriculture Sourcebook. Torkelsson, Asa. 2012. “Sex Disaggregated Data on Washington, DC: World Bank. Agriculture, Water, and Food Security: Lessons ———. 2010a. “Designing and Implementing from Kenya.” World Water Week Seminar on Agricultural Innovation Funds: Lessons from Global Practice in Promoting Gender Equality in Competitive Research and Matching Grant Proj- the Water Sector, Stockholm, August 30. ects.” Report 54857-GLB, World Bank, Washing- United Republic of Kenya. 2010. “The Constitu- ton, DC. tion of Kenya 2010.” National Council for Law ———. 2010b. Doing Business 2011: Making a Reporting, Nairobi. Difference for Entrepreneurs. Washington, DC: Welch, Finis. 1970. “Education in Production.” World Bank. Journal of Political Economy 78 (1): 35–59. ———. 2011. “Kenya Informal Settlements Improvement Project.” Project Appraisal Docu- World Bank. n.d. “Sri Lanka Community Develop- ment, World Bank, Washington, DC. ment and Livelihood Improvement Gemidiriya Project: Analysis of Youth Inclusion and Partici- ———. 2012a. Agricultural Innovation Systems: An pation in Gemidiriya.” World Bank, Washington, Investment Sourcebook. Washington, DC: World DC. Bank. ———. 2004. “Malawi Community-Based Rural ———. 2012b. “Land Administration and Reform Land Development Project.” Project Appraisal in Sub-Saharan Africa: From Piloting to Scaling Document, World Bank, Washington, DC. Up.” World Bank, Washington, DC. ———. 2007. “Cultivating Knowledge and Skills ———. 2012c. World Development Indicators to Grow African Agriculture: A Synthesis of an 2012. Washington, DC: World Bank. Institutional, Regional, and International Review, ———. 2013. “Growing Africa: Unlocking the 2007.” Report No. 40997-AFR. Agriculture and Potential of Agribusiness.” AFTFP/AFTAI Rural Development Department and Africa Report, World Bank, Washington, DC. FOCUS Safety Nets and Pathways to NOTE 2 Productive Employment S ocial safety nets are programs that aim to ment at a critical period when youth may be increase households’ consumption of basic completing their education, transitioning to commodities and facilitate access to basic work, or already working on their own account services (Grosh et al. 2008; Monchuk 2014). in low-productivity occupations. As such, They can also help households to cope with safety nets with these complementary compo- shocks such as an economic crisis, drought, nents have the potential to improve produc- or illness. The coverage of safety net programs tive employment for the current generation of targeted to poor and vulnerable households is youth. growing rapidly in Africa (Monchuk 2014; Sub- barao et al. 2013; McCord and Slater 2009). The most common safety nets include public works Short-Term Benefits of Safety Net and cash transfer programs. Participants in Programs public works programs receive cash or food in exchange for engaging in labor-intensive work African governments rely extensively on pub- to build or maintain public goods such as local lic works programs to provide short-term infrastructure. Some longer-term public works employment. Public works programs are com- programs ensure temporary employment to the mon throughout the region and typically offer poor at critical junctures, such as lean periods temporary employment for periods ranging in the agricultural cycle. Cash transfers provide from 10 days to 6 months.2 For example, the temporary relief to households to reduce pov- Malawi Social Action Fund has offered tem- erty. These traditional safety nets can improve porary employment to more than 800,000 productivity in the long term through a range individuals for 12 days on average since 1995 of channels (Alderman and Yemtsov 2013). (World Bank 2011). The Tanzania Social Action For instance, safety nets have been shown to Fund offers cash transfers through short-term increase human capital by improving nutrition employment in a labor-intensive public works and access to education and health services.1 program, which has offered an average of 75 Such improvements for today’s children can days of employment to more than 200,000 ben- contribute to higher labor productivity for eficiaries since 2000 (World Bank 2010). One tomorrow’s youth. public works program of longer duration is Because safety net programs explicitly tar- Ethiopia’s Productive Safety Net Program. Par- get the poor or vulnerable, they often reach the ticipants, who can remain in the program for core of the population engaged in low-produc- up to five years, work during the lean season tivity employment, particularly in agriculture on labor-intensive projects building commu- and household enterprises. Many governments nity assets, such as soil and water conservation also deliver complementary productive inter- structures. The program reaches more than 7 ventions through safety net programs, with million people and operates with an annual the explicit objective of fostering productive budget of nearly US$500 million (Gilligan, employment in the short to medium term. The Hoddinott, and Taffesse 2009). intention is to allow poor households to build Public works programs have well-docu- pathways to raise productivity in the farming mented short-term benefits, including on par- sector, diversify livelihoods, or enter into non- ticipants’ income. To work the hours required farm household enterprises. If complementary by these labor-intensive programs, beneficiaries productive interventions are targeted to youth, usually forgo other income-earning opportuni- they can pave a pathway to productive employ- ties (Subbarao et al. 2013). For this reason, ben- 142 focus note 2 143 eficiaries’ earnings generally increase less than a range of outcomes such as food security or the actual wages they receive (which are usu- nutrition, and increasing the use of health or ally based on estimated market rates for wage education services (see Fiszbein and Schady employment or a minimum wage). Argentina 2009 for a review). There is also growing evi- initiated a public works program in 2002 after a dence showing that cash transfer beneficiaries severe economic crisis pushed unemployment are able to save and enter into more produc- and poverty to record heights. Net earnings tive activities, even if these programs were not (after subtracting income forgone from other designed with that objective in mind.3 For activities) were estimated to be two-thirds of instance, in Malawi, cash transfers facilitated the benefits offered by the public works pro- investments in agricultural assets and livestock gram during the crisis and only one-third of and reduced adults’ participation in low-skilled those offered after the crisis had dissipated labor (Covarrubias, Davis, and Winters 2012). (Galasso and Ravallion 2004). In Mexico, cash transfers led poor rural house- In low-income African countries, nearly holds to invest in productive assets such as farm all individuals are working, and very few are animals and land for agricultural production, unemployed (see discussion in chapter 1). In increasing agricultural income by 10 percent this context, public works do not necessarily (Gertler, Martinez, and Rubio-Codina 2011). substantially increase overall employment. At Beneficiaries were also more likely to start non- the same time, public works programs tend agricultural household enterprises, particularly to be oversubscribed despite paying low wages the production of handcrafts for sale. and can strongly increase the earnings of ben- A new generation of safety net programs eficiaries. Few evaluations measure net earn- is trying to be more explicit about maximiz- ings gains from participating in public works ing their productive potential in the short to programs in Africa. medium term. Overall, public works programs are first and foremost social protection programs that provide temporary employment. They are Safety Nets Plus Explicit generally not designed to foster sustainable Productive Components productive employment beyond beneficiaries’ participation in the program. Some evidence A growing number of safety net programs suggests that traditional public works pro- attempt to create the conditions for benefi- grams can open pathways to employment in ciaries to access pathways to more productive new activities, even if that is not their primary employment or sustainable livelihoods in the goal. In Liberia, participants reported using short to medium term. Policy makers start with 14.2 percent of the income from public works the proven ability of safety net programs to for farm and nonfarm investments (Andrews stabilize household consumption and add pro- et al. 2011). In Ethiopia, beneficiaries from the ductive interventions that may lead to sustain- public works program were more likely to enter able, productive employment for the current in nonfarm business activities and less likely generation. For example, “public works plus” to work as day laborers (Gilligan, Hoddinott, delivers complementary services such as links and Taffesse 2009). Evidence from an impact to financial services or training in a range of evaluation in Sierra Leone shows that partici- skills. Some “cash transfer plus” programs take pation in a cash-for-works program increased a similar approach. participation in informal saving groups, led to accumulation of assets such as livestock, and Financial Services fostered the creation of new household enter- A growing number of public works programs prises (Rosas and Sabarwal 2013). aim to strengthen financial inclusion or build Cash transfer programs also primarily linkages to other intermediate services. Par- focus on immediate poverty relief, and there ticipants in Rwanda’s Vision Umurenge Pro- is solid evidence documenting their short- gramme are given access to bank accounts and term impacts on reducing poverty, improving encouraged to save some of their earnings. 144 Youth EmploYmEnt in Sub-Saharan africa Beneficiaries of Tanzania’s Productive Safety and My Community, Making a Living, Manag- Net Program can participate in small groups ing Money, My Workplace, My Health, and My that promote community savings to increase Future. their ability to save for future needs and invest- Some programs deliver a package of benefits ments; a similar approach is being considered including cash transfers together with a range for a public works program in Mozambique. A of complementary productive interventions. public works program in Côte d’Ivoire includes For instance, the “targeting the ultra-poor” or a component that fosters saving by deliver- “graduation model” is currently being tested ing payments to a bank account (during its in 10 countries around the world including pilot phase, the program awarded a matching Ethiopia and Ghana (Hashemi and Montes- grant to participants who had saved a certain quiou 2011). The intervention targets the poor amount). For more discussion on financial ser- and offers basic consumption support similar vices, see focus note 3. to a safety net program. It also facilitates sav- ings, transfers assets to allow the poor to enter Capital into higher-productivity farming activities (for Participants in Ethiopia’s public works program example, livestock) or start a household enter- are linked to the Household Asset Building Pro- prise, and provides skills training and regular gram to help them to make their farms more technical assistance to beneficiaries. The stated productive and to increase their long-term food objective of this approach is to help households security. They obtain “at least one of several to graduate from extreme poverty through pro- productivity-enhancing transfers or services, ductive employment. including access to credit, agricultural extension services, technology transfer (such as advice on food crop production, cash cropping, livestock Do Complementary Interventions production, and soil and water conservation), Open Pathways toward and irrigation and water harvesting schemes” (Gilligan, Hoddinott, and Taffesse 2009). Productive Employment for Youth? Skills Training Many public works programs also provide Despite their potential to increase productiv- short-term training in basic literacy, skills for ity and facilitate entry into new employment microenterprise development, technical skills, opportunities, few rigorous evaluations have or life skills. In South Africa’s Expanded Public shown that expanded safety net programs Works Program, beneficiaries receive two days lead to employment or productivity gains.4 In of training each month in literacy and numer- Ethiopia, households with access to the public acy, vocational skills, and business skills. In works program and complementary interven- Côte d’Ivoire, the public works program imple- tions are “more likely to be food secure and mented by the national roads agency includes are more likely to borrow for productive pur- sensitization in basic life skills, training in basic poses, use improved agricultural technologies, business skills to help participants to set up a and operate nonfarm own business activi- household enterprise, and sensitization to wage ties” (Gilligan, Hoddinott, and Taffesse 2009). employment opportunities to help youth to Their incomes grow, and distress sales of assets prepare themselves and search for a wage job. decline. Elsewhere in Africa, evidence on the In Liberia, the community works component productive impacts of “public works plus” of the Youth Employment and Skills Project programs is limited, especially evidence on the provides eight sessions of life skills training effectiveness of short training modules and during the 40-day works program. Each of links to financial services. An evaluation testing these sessions lasts two to three hours and is alternative complementary training targeted to led by 1 trainer for every 25 trainees. The life young beneficiaries of a public works program skills training has six main sections: Myself is under way in Côte d’Ivoire. focus note 2 145 Figure F2.1 Impacts of conditional cash transfers plus business grants or conditional cash transfers plus skills training in Nicaragua a. Consumption and employment b. Profits and wages 14 1,400 12 1,200 10 1,000 Change in local currency (C$) 8 800 % increase 6 600 4 400 2 200 0 0 –2 –200 –4 –400 Consumption (log) Nonagricultural Nonagricultural Profits in nonagricultural Wage in private jobs self-employment wage job self-employment Conditional cash transfers plus training Conditional cash transfers plus grants Source: Macours, Premand, and Vakis 2012. Note: Impacts are measured as differences between beneficiaries and a control group. The impact on consumption and entry in a nonagri- cultural wage job is not significant for beneficiaries of the “conditional cash transfers plus training” package, while the impact on entry in a nonagricultural wage job and on wages in a private sector wage job is not significant for beneficiaries of the “conditional cash transfers plus grant” package. Evidence is also thin on the effectiveness training delivered to cash transfer recipients was of productive components embedded in cash to increase wages in private wage jobs. But over- transfer programs. In Niger, monitoring data all the average impact on income and consump- suggested that beneficiaries save a substantial tion across all beneficiaries was not significant. share of their earnings through savings groups Early results from impact evaluations of the created by the program. A new cash transfer “graduation model” also suggest substantial program in Cameroon encourages beneficia- productive impacts. For instance, a program ries to participate in awareness and training that targets the ultra-poor with a package activities to learn about generating income, including large asset transfers and intensive understand how to access microfinance, and training was effective in inducing entry into acquire business skills. Impact evaluations are self-employment in Bangladesh. The com- under way in both countries. bined intervention was successful in inducing Although the effectiveness of complemen- a change from agricultural wage labor to small tary productive interventions has not been businesses, raising annual income 36 percent ascertained in Africa, promising evidence is on average (Bandiera et al. 2012). These results emerging elsewhere. In rural areas of Nicaragua, are consistent with evidence emerging from an evaluation tests the relative effectiveness of similar programs in Honduras, Pakistan, and complementing a cash transfer program with West Bengal. business grants or vocational training (see fig- ure F2.1; Macours, Premand, and Vakis 2012). Two years after the end of the program, results Safety Nets as Vehicles to Deliver show that the business grant enabled cash Interventions Aimed at Improving transfer recipients to enter nonagricultural self- Youth Employment Outcomes employment and increased profits in household enterprises as well as overall income and con- Traditional safety nets do not seek primarily to sumption. The main effect of the vocational foster productive employment for the current 146 Youth EmploYmEnt in Sub-Saharan africa generation of youth, but they have the potential to those of a program delivering cash to benefi- to do so. These programs target the very house- ciaries amounting to the labor and capital costs holds that are engaged in low-productivity of a public works program. activities. As such, safety net programs have the potential to deliver complementary interven- References tions opening pathways toward more produc- Alderman, Harold, and Rusla Yemstov. 2013. “How Can Safety Nets Contribute to Economic tive employment for the poor. There are not Growth?” World Bank Economic Review. necessarily any equity-efficiency trade-offs in Andrews, Collin, Prospère Backiny-Yetna, Emily targeting productive interventions to the poor: Garin, Emily Weedon, Quentin Wodon, and in fact, poor households often would benefit the Giuseppe Zampaglione. 2011. “Liberia’s Cash most from productive interventions (Macours, for Work Temporary Employment Project: Premand, and Vakis 2013). The effectiveness of Responding to Crisis in Low-Income, Fragile these complementary interventions for opening Countries.” Social Protection and Labor, World Bank, Washington, DC. such pathways is yet to be evaluated rigorously in Africa, but their potential to help individu- Andrews, Colin, Mirey Ovadiya, Christophe Ribes Ros, and Quentin Wodon. 2012. “Cash for Works als to “graduate” to more productive and secure in Sierra Leone: A Case Study on the Design and livelihoods deserves to be considered as part Implementation of a Safety Net in Response to a of inclusive employment strategies. Greater Crisis.” Social Protection Discussion Paper 1216, attention could be paid to targeting productive World Bank, Washington, DC. interventions to youth within poor households Asfaw, Solomon, Silvio Daidone, Benjamin Davis, benefiting from such programs. Josh Dewbre, Alessandro Romeo, Paul Winters, Katia Covarrubias, and Habiba Djebbari. 2012. “Analytical Framework for Evaluating the Pro- Notes ductive Impact of Cash Transfer Programmes on 1. Safety nets can also create positive externalities Household Behaviour? Methodological Guide- on the local economy. lines for the From Protection to Production 2. See Subbarao et al. (2013) for a review. For Project.” Working Paper 101, International Policy instance, Liberia and Sierra Leone rolled out Centre for Inclusive Growth, Brasilia. cash-for-work programs in 2008 to cushion Bandiera, Oriana, Robin Burgess, Selim Gulesci, the impacts of soaring food prices (Wodon and Imran Rasul, and Munshi Sulaiman. 2012. “Can Zaman 2010). In its first phase, the program in Entry-Level Entrepreneurship Transform the Sierra Leone reached 16,000 beneficiaries, who Economic Lives of the Poor?” Mimeo, London worked for approximately 50 days, for six to School of Economics. eight hours a day, on road rehabilitation, refor- Blattman, Christopher, Nathan Fiala, and Sebastian estation, soil conservation, and cultivation of Martinez. 2011. “Employment Generation in rice and alternative crops (Andrews et al. 2012). Rural Africa: Mid-Term Results from an Experi- The program in Liberia offered, on average, mental Evaluation of the Youth Opportunities 40 days of temporary employment to 17,000 Program in Northern Uganda.” Social Protection households, which mostly rehabilitated public Discussion Paper 66523, World Bank, Washing- agricultural land and cleaned and cleared roads, ton, DC. drains, and public spaces (Andrews et al. 2011). Covarrubias, Katia, Benjamin Davis, and Paul The program was later scaled up to cover 45,000 Winters. 2012. “From Protection to Production: beneficiaries. Productive Impacts of the Malawi Social Cash 3. Cash transfers can contribute to raising growth Transfer Scheme.” Journal of Development Effec- through channels other than their direct impact tiveness 4 (1): 50–77. on beneficiaries (Alderman and Yemstov 2013). Fiszbein, Ariel, and Norbert Schady. 2009. Con- For instance, simulations also suggest that they ditional Cash Transfers: Reducing Present and have productive impacts on the local economy Future Poverty. Washington, DC: World Bank. (Asfaw et al. 2012). Galasso, Emanuela, and Martin Ravallion. 2004. 4. Pilots delivering cash grants to beneficiaries have “Social Protection in a Crisis: Argentina’s Plan had large impacts on employment and earn- Jefes y Jefas.” World Bank Economic Review 18 ings (for example, Blattman, Fiala, and Marti- (3): 367–99. nez 2011). It remains unclear how the impacts Gertler, Paul J., Sebastian W. Martinez, and Marta from a public works program would compare Rubio-Codina. 2011. “Investing Cash Transfers focus note 2 147 to Raise Long-Term Living Standards.” Ameri- Monchuk, Victoria. 2014. Reducing Poverty and can Economic Journal: Applied Economics 4 (1): Investing in People: The New Role of Safety Nets in 164–92. Africa. Washington, DC: World Bank. Gilligan, Daniel O., John Hoddinott, and Ale- Rosas, Nina, and Shwetlena Sabarwal, 2013. “Pub- mayehu Seyoum Taffesse. 2009. “The Impact of lic Works as a Productive Safety Net in a Post- Ethiopia’s Productive Safety Net Programme and Conflict Setting? Evidence from a Randomized Its Linkages.” Journal of Development Studies 45 Evaluation of Sierra Leone’s Cash for Work (10): 1684–706. Program.” World Bank, Washington, DC. Grosh, Margaret, Carlo del Ninno, Emil Tesliuc, Subbarao, Kalanidhi, Carlo del Ninno, Colin and Azedine Ouerghi. 2008. For Protection and Andrews, and Claudia Rodríguez-Alas. 2013. Promotion: The Design and Implementation of “Public Works as a Safety Net: Design, Evidence, Effective Safety Nets. Washington, DC: World and Implementation.” World Bank, Washington, Bank. DC. Hashemi, Syed M., and Aude Montesquiou. 2011. Wodon, Quentin, and Hassan Zaman. 2010. “Reaching the Poorest: Lessons from the Gradu- “Higher Food Prices in Sub-Saharan Africa: ation Model.” CGAP Focus Note 69, Consultative Poverty Impact and Policy Responses.” World Group to Assist the Poor, Washington, DC. Bank Research Observer 25 (1): 157–76. Macours, Karen, Patrick Premand, and Renos Vakis. World Bank. 2010. “Project Paper on Second Addi- 2012. “Transfers, Diversification, and Household tional Financing Credit in the Amount of SDR Risk Strategies: Experimental Evidence with 23.1 Million (US$35 Million Equivalent) in Pilot Implications for Climate Change Adaptation.” Crisis Response Window Resources to the United Policy Research Working Paper 6053, World Republic of Tanzania for the Second Social Bank, Washington, DC. Action Fund Project.” World Bank, Africa Region ———. 2013. “Demand Versus Returns? Pro-Poor Social Protection Unit, Human Development, Targeting of Business Grants and Vocational Washington, DC. Skills Training.” Policy Research Working Paper ———. 2011. “Implementation Status and Results 6389, World Bank, Washington, DC. of Malawi Third Social Action Fund (MASAF3) McCord, Anna, and Rachel Slater. 2009. “Overview APL II (P110446).” Report ISR7009, World Bank, of Public Works Programmes in Sub-Saharan Washington, DC. Africa.” Overseas Development Institute, London. Chapter 5 Creating Productive Employment for Youth in the Household Enterprise Sector Despite the large share of employment in house- has not left the household economy. Employ- hold enterprises (HEs) and the potential for sub- ment in HEs is growing especially rapidly for stantial growth, few governments recognize that those who have not completed a secondary “the informal will be normal” or develop effec- education—the majority of new entrants into tive policies and programs to help youth to create Africa’s overwhelmingly young, burgeoning productive enterprises. Young people who lack the workforce. These young people generally lack education for a wage job recognize the potential the education for a wage job but can pursue the in HEs, but they often struggle to amass the sav- economic opportunities that HEs offer—as a ings, skills, and market know-how to enter and full-time activity or together with agriculture consistently succeed in the HE sector. At a time in a mixed-livelihood strategy. when the formal wage sector is just beginning to Development processes in other regions gain ground, supportive national strategies are have included HEs in parallel with large-scale needed to facilitate entry into the HE sector, pro- manufacturing, so the growth of HEs in Africa vide an environment to make it more productive, is not surprising (figure 5.1). Such enterprises “Because I am and realize the corresponding benefits for youth are still an important source of employment not educated, employment and economic growth. in low-income countries where nonfarm wage I could not jobs have grown rapidly, such as in Bangladesh be employed. In Africa today, most employment in low- and Cambodia. In Asia and Latin America, Therefore, I had income households comes from household- HEs provided an important pathway for sur- based activities. These activities include family plus labor to move out of agriculture as farm no other option farming and very small nonfarm enterprises, productivity improved. In Africa’s low- and than employing which we call household enterprises (HEs; see middle-income countries, where the mod- myself.” box 5.1). As countries have become richer, the ern wage sector has not grown fast enough to Tanzania labor force has shifted out of agriculture, but it absorb everyone who is prepared to leave agri- 149 150 Youth EmploYmEnt in Sub-Saharan africa Box 5.1 What is a household enterprise? Household enterprises (HEs) are unincorporated, nonfarm on hours worked or output, although they may share in businesses owned by households. From an employment per- the profits. Some call these enterprises microenterprises, spective, they include self-employed people running unin- but the term is not used uniformly (for example, in Grimm, corporated businesses and family members working in those Knorringa, and Lay 2012). Some authors categorize both businesses. When employment in the HE sector is tabulated, self-employment and larger, more substantial businesses as employees from outside the family are not included in this microenterprises (and may include partnerships and firms category; they constitute wage workers. The owners them- that employ workers regularly). In this report, only HEs that selves account for most of the employment in the sector. hire workers (less than 4 percent of all owners) are classified Family helpers make up only 11 percent of employment in as microenterprises. the sector (not counting wage workers). From an enterprise Some refer to HEs as informal enterprises. The Interna- perspective, they follow the same pattern. Most HEs in Africa tional Labour Organization (ILO) considers everyone working today are pure self-employment, and only 10 percent have in an unincorporated business with less than five employees hired someone outside of the family (see figures B5.1.1 and as having informal sector employment (ILO 1993; see Fox and B5.1.2). Therefore, employment in the sector grows primarily Pimhidzai 2013 for a discussion). The label “informal sector” through the creation of new enterprises. ignores the household nature of these businesses, however, HEs may also be referred to as own-account work- and the ILO definition does not distinguish between the fam- ers (another term for self-employed) or as contributing or ily (which gets the profits from the enterprise and absorbs unpaid family workers, indicating that the family members the losses) and outside wage workers (who are paid for a work in the enterprise but receive no cash payment based task but not expected to share in profits or losses). Figure B5.1.1 Most people working in the HE sector are owners Figure B.5.1.2 Most HEs are family operations 100 90 80 Family helper 70 11% 60 Percent 50 40 30 20 Owner 89% 10 0 Self-employed Self-employed With 1–4 5+ employees with family employees helpers Source: Fox and Sohnesen 2012. Source: Fox and Sohnesen 2012. culture, the HE sector is larger than in compar- between higher rural HE income and lower ator countries. The HE sector has generated the poverty have been observed in several Asian majority of new nonfarm jobs in most low- and countries (Haggblade, Hazell, and Reardon lower-middle-income countries of Africa, even 2010; Lanjouw, Quizon, and Sparrow 2001). during times of high economic growth (figure The shift in primary economic activity out of 5.2). This trend is expected to continue. agriculture—including day labor—and into HEs are an important way for households running a nonfarm business contributed sub- to move out of poverty. Strong correlations stantially to poverty reduction in Bangladesh creating productive Employment for Youth in the household Enterprise Sector 151 Figure 5.1 Household enterprises are an important share of nonfarm employment in low- and middle-income countries 100 90 Household enterprise share of nonfarm 80 70 employment (%) 60 50 40 30 20 10 0 Sub- Lao PDR Bangladesh Cambodia Sub- Bolivia Vietnam Nicaragua Philippines Mongolia Saharan Saharan Africa Africa Low-income countries Lower-middle-income countries Source: Based on standardized and harmonized household and labor force surveys for Sub-Saharan Africa (most recent data available) and on household and labor force surveys for countries in other regions (see appendix). Figure 5.2 In Africa, as labor moved out of agriculture, more people moved into the HE sector than into private wage employment 25 Percentage point change in employment share 20 15 10 5 0 –5 –10 –15 –20 –25 Nigeria Rwanda Ghana Uganda Tanzania Senegal Côte d’Ivoire Agriculture Wage Household enterprise Source: Based on standardized and harmonized household and labor force surveys (see appendix). and Uganda over the last decade (see Inchauste for youth to take advantage of the opportuni- and Olivieri 2012 on Bangladesh; Fox and Pim- ties these businesses present. Youth often lack hidzai 2012 on Uganda). even the small amount of savings needed to In Africa, most government strategies over- start a trading business, and banks and micro- look the HE sector, partly because policy mak- finance entities rarely lend to start-ups. A ers may know little about the sector or its role small-scale processing or service venture could in creating youth employment and support- require more capital than a trading venture, as ing economic growth. Urban policies often well as some technical skills that school leavers actively discourage HEs, making it challenging may not have. Young people may also need to 152 Youth EmploYmEnt in Sub-Saharan africa develop business and behavioral skills, if they agriculture and into the services sector, as the have not acquired them in school or through majority of HEs engage in trading (figure 5.3). experience. In short, it is difficult for youth to At the same time, HEs have contributed to find pathways into the HE sector. employment in the industrial sector, where they Facilitating entry into the HE sector and engage in manufacturing (primarily transform- providing an enabling environment to make it ing agricultural goods or natural resources into more productive are fundamental policy chal- products such as charcoal, flour, roof thatching, lenges. Research across Africa and other regions or bricks) and artisanal activities (producing shows that these challenges can be met; the key custom furniture and ironwork, dressmaking is to recognize the potential. By changing the and tailoring, or construction). In the services fragmented approach to the HE sector into a sector, HEs engage in food services (making “One who does coherent, coordinated approach—with a strong and selling snacks or meals), transport, and not have a focus on sustainable employment for youth— personal services (barbering and hairdressing). nonagricultural government, donors, nongovernmental organi- Partly because of local opportunities to pro- enterprise zations (NGOs), and other stakeholders could cess agricultural products, manufacturing is a stays idle for lift the constraints on productivity and open common activity for rural HEs. Although street most of the dry opportunities for sustainable employment for vendors and local markets are the most visible everyone in this sector. signs of HE activity, many businesses operate season when This chapter outlines the opportunities and out of the owner’s home. agricultural challenges facing HE owners in Africa today. It HEs survive and grow because they pro- activities are explores why young people often struggle to vide low-cost goods and services demanded few. Therefore, enter the sector and describes policies and pro- in a growing but less developed economy that I started my grams that could change this reality and make lacks a modern services sector. In urban areas, enterprise of the sector more productive. mobile retail traders and HEs in market stalls provide the services found in convenience stores selling soft and and malls in richer countries. The lower-quality alcoholic drinks.” The Household Enterprise Sector goods manufactured by HEs typically will not Tanzania Today be demanded as incomes rise and mass-pro- duced or higher-quality goods enter the market. HEs have been an important part of the recent For this reason, services tend to dominate the shift of value added and employment out of sector and to persist longer than manufacturing. Figure 5.3 Most HEs are in the trading sector a. Female b. Male Mining/natural resources/construction/energy Mining/natural resources/construction/energy 4% 5% Other services 15% Other services 24% Manufacturing Manufacturing 27% 26% Wholesale/retail Wholesale/retail 46% 54% Source: Fox and Sohnesen 2012. creating productive Employment for Youth in the household Enterprise Sector 153 Although HEs account for a larger share Table 5.1 Rural versus urban household enterprises of employment in urban areas, 60 percent of Indicator Urban Rural HEs in Sub-Saharan Africa are located where Share of households owning an hE 53 36 the population is—in rural areas. HEs are dif- Share of hEs as primary activity for owner 82 42 ferent in rural and urban areas (table 5.1).1 It Share of hE owners working more than 30 hours a week 79 45 is quite common for rural households with an Share of hE owners operating their own business at least 10 months a year 69 61 HE to operate a farm, either for subsistence and food security or as a commercial venture. The Share of hE owners operating their business at home 30 43 majority of these households consider the HE Share of hE owners who are female 55 47 to be a secondary activity. Most HEs in rural Share of hE owners with hired labor 8 11 areas operate during at least 10 months of the Source: Fox and Sohnesen 2012, based on data from Cameroon (2001), the Republic of Congo (2009), Ghana (2005), Mozambique (2008), Rwanda (2005), and Tanzania (2005). year, but for less than 30 hours a week. Often rural HE owners can do business only on week- ends, when markets are open and foot traffic is in the labor force, so has the level of education “All of us have heavier. Owners in rural areas are more likely among HE owners. Figure 5.4 shows that the the maximum to operate their business at home; in rural areas youngest owners are much less likely than their of primary without electrification, HEs usually cannot parents’ generation to have no education. Still, operate after sunset. only 29 percent of young adults (ages 25–34) education, Urban HEs are almost always the primary employed in the HE sector have completed and this cadre activity of the owner, but some owners have more than primary school, much less than the represents the multiple businesses or lines of work. When HEs 59 percent in the modern wage sector. majority who are a secondary activity, the primary activity is are roaming in usually a wage job. As a result, urban owners The Business: Constraints and the streets. It is work long hours; 47 percent work more than 50 hours a week. Almost all of those who hire Opportunities impossible for paid labor are based in urban areas. Women are such people to as active in the urban HE sector as men, but As businesses, Africa’s HEs are quite hetero- get employment they are slightly underrepresented in rural HEs. geneous. Even in the trading sector, some are because there are Women and men clearly gravitate toward dif- very small-scale operations, using little human better educated ferent types of enterprises. For example, females or physical capital, whereas others sell higher- individuals.” are more likely to do tailoring, and men are value products, have a substantial inventory, more likely to work in construction (Fox and and provide a much higher income. For exam- Tanzania Sohnesen 2012). The result of this gender seg- ple, a study of female HE owners in Accra, regation is that women are less likely to work in Ghana, found “high-profit” females, who had the sectors where earnings are higher (box 5.2). much higher average profits and capital stock, Basic education is a key pathway for youth to working in roughly the same sectors as low- transition into the HE sector. Individuals who profit females (Fafchamps et al. 2011). enter the labor force with a completed primary Profits vary by sector as well. An analysis education are the most likely to become an HE of data from seven capital cities in West Africa owner, and primary education is the main for- found monthly profits ranging from US$70 mal opportunity for developing skills that most for petty traders to US$107 for construction young people now entering the HE sector will workers (Grimm, Krüger, and Lay 2011). That ever have. Whereas 35 percent of young adults same analysis found very high returns to capital (ages 25–34) who work in agriculture have for many self-employed and family businesses, never been to school, 79 percent of those in the especially at low levels of investment (Faf- nonfarm HE sector have at least some school- champs et al. 2011 found similarly high rates ing. Although the majority of HE owners in in Accra). Monthly marginal returns on capital Africa do not have primary education (reflect- are estimated to be about 70 percent for those ing the low levels of education among the older with capital stock valued at less than US$150. generations), as the level of education has risen Although most businesses report capital con- 154 Youth EmploYmEnt in Sub-Saharan africa Box 5.2 Why do females earn less? Occupational segregation in the household enterprise sector Gender segregation is common among HE owners. Both Evidence also suggests that men and women have dif- women and men are likely to engage in trading, but women ferent information or expectations about particular enter- are much more likely to do tailoring or catering (lower-paying prises. In Kenya, a recent experiment with vocational activities) and men to engage in carpentry, metalworking, and training vouchers found that prior to enrollment males repair (higher-paying activities). For example, owners of HEs overwhelmingly preferred traditionally “male-dominated” surveyed in the Kassida workshop area of Kampala, Uganda, courses such as motor vehicle mechanics, whereas women indicated that even in metalworking, males almost completely almost exclusively chose traditionally “female-dominated” dominate the higher-earning fabrication activities, while more courses such as hairdressing (Hicks et al. 2011). The pro- females engage in scrap metal processing, where earnings are gram administrators randomly provided half of the partici- lower. One of the sectors with the lowest earnings—selling pants with information on the actual returns to training, drinks—is almost completely dominated by females. highlighting the higher returns in male-dominated trades The reasons for this gender segregation are complex. They and using “soft persuasive” methods, such as a video of include social norms, the lack of female role models who have female auto mechanics, to encourage females to pursue entered occupations traditionally performed by men, and more traditionally male-dominated trades. Females exposed constraints on time and money, which can be more binding to the information were almost 9 percentage points more for women than for men. For example, studies in South Asia likely to express a preference for a male-dominated course, and Kenya found that when women choose to work outside and they were 5 percentage points more likely to enroll in the home, they spend more time traveling on slower modes one. Younger and more educated females were especially of transport, which limits them to employment options that likely to prefer male-dominated fields. This experiment sug- are closer to home (Uteng 2011; Gulyani, Talukdar, and Jack gests that efforts to provide information and encourage- 2010). Regardless of the cause, women tend to earn less than ment have potential to reduce occupational segregation and men from HEs (Fox and Sohnesen 2012). increase women’s earnings. Figure B5.2.1 Earnings are higher in male-dominated subsectors than in female-dominated subsectors 400 $371 350 Average monthly earnings (US$) $296 300 250 200 $148 150 $128 100 $86 50 0 Saloons Scrap metal works Catering Metal fabrication Electrical 90% 24% 98% 4% 8% Share of women Source: Campos et al. 2013. straints, returns do not appear to increase with Banks and microfinance institutions in Africa the level of capital stock. rarely lend to start-ups. Almost all owners People need savings to start an HE, because report that they started their business with their it is virtually impossible for them to get credit. own savings and a loan or grant from their fam- creating productive Employment for Youth in the household Enterprise Sector 155 Figure 5.4 Younger HE owners tend to have more acy can add as much as 40 percent to their earn- education than older ones ings, controlling for sector, location, and other characteristics (see the discussion in chapter 35 3; Fox and Sohnesen 2012). While some of 30 this earnings premium probably reflects other Percentage of age cohort 25 characteristics of primary school completers, it suggests that the basic skills acquired in 20 primary school—such as literacy and numer- 15 acy—matter for productivity in HEs. Develop- ment policies and programs often confuse HEs 10 with small and medium enterprises (SMEs) or 5 micro, small, and medium enterprises. This is 0 a mistake (box 5.3 explains why). Even though No education Primary Primary Secondary + there is substantial heterogeneity among HEs, incomplete complete a fundamental difference between HEs and Ages 15–24 Ages 25+ SMEs is that while HEs often persist for a long Source: Fox and Sohnesen 2012. time (more than five years), they are not ori- ented toward employment growth (Fox and Sohnesen 2012). Even if productivity improves, ily or friends (table 5.2). Very few applied for a these businesses rarely grow beyond HE status. loan after starting their enterprise. To the extent Data from West Africa show that capital grows that these enterprises have access to financial quickly to a steady state, and even after 10 years services, those services tend to be informal. of operation, the capital stock of the business Suppliers’ credit (informal loans at high interest rates) is common for trading enterprises. Other Table 5.2 Sources of capital for household enterprises: Start-up and credit for sources include informal local systems such as operations rotating savings and credit associations and vil- Source Male Female All lage savings and loan associations, which pool Start-up capital a savings within a village and lend small amounts personal savings 79.9 71.6 75.1 for a short term, usually no more than one year (see focus note 3). Microfinance has not official or formal 1.3 0.8 1.0 yet reached this sector, even for working capi- relative or friends 6.6 13.9 10.8 tal. Despite the high returns to capital, current microfinance, nGo, cooperative 1.3 1.3 1.3 owners, those who did not start a business, and informal 3.0 3.2 3.1 those who closed a business all reported that other 7.9 9.2 8.7 limited access to capital is the most important total 100.0 100.0 100.0 business constraint they face. Completing primary education is a key Credit for business operation b driver of profits for HE owners. Compared to official or formal 17.0 15.8 16.3 having no schooling at all, having only a few relative or friends 39.9 43.0 41.7 years of education does not add to earnings in microfinance, nGo, cooperative 28.7 21.4 24.7 the HE sector. This outcome reflects the fact informal 10.3 13.1 11.9 that in many African countries a few years of other 4.0 6.6 5.5 basic education fail to provide basic skills such as literacy and numeracy. Because 45 percent total 100.0 100.0 100.0 Source: Fox and Sohnesen 2012, based on World Bank harmonized household surveys. of young adults in the HE sector have not com- Note: “Official or formal” includes commercial banks and public programs. “Other” includes loans pleted basic education, the lack of basic skills from employers and other unspecified loans. “Informal” includes credit from customers or suppliers, informal moneylenders, rotating savings and credit associations, village savings and loan associations, contributes to their low earnings. By contrast, and savings and credit cooperatives. youth who have completed primary school and a. Includes Cameroon (2001), the Republic of Congo (2009), Ghana (2005), Rwanda (2005), and Uganda (2005). reached functional levels of literacy and numer- b. Includes Ghana (2005), Rwanda (2005), Tanzania (2005), and Uganda (2005). 156 Youth EmploYmEnt in Sub-Saharan africa Box 5.3 Small and medium enterprises are not the same as household enterprises Policies and programs to encourage business start-ups often Relationships between the enterprise and the state with place HEs and SMEs under one umbrella, often called micro, respect to registration and payment of taxes depend in part small, and medium enterprises. But HEs and SMEs have on the rules and practices in each country. important differences. Differences in registration and taxation requirements: Differences in where they locate: • In many countries, it is legal to run HEs without any busi- • HE are usually not located in a building (establishment) in ness registration or license. However, local governments a business or commercial area. They may locate in a regu- may require HEs to get a trading permit or license or to lar market stall, on a usual corner, or in the owner’s house. register for a place in a public market stall. Many have no fixed location. • Most countries require SMEs to have a business license, • SMEs operate from a fixed location. They may be a sole and their hiring is subject to national labor laws. At the proprietorship or have multiple owners, but they usually same time, SMEs are known for hiring workers infor- do not operate from the household, and they separate the mally—that is, without paying payroll taxes or enrolling business from the household accounts. them in a mandatory social insurance system. In part this practice reflects the lower labor productivity in the SME Differences in number of employees: sector, which calls for lower real wages. • Most HEs do not hire labor. Differences in access to financial services: • SMEs by definition hire labor. Definitions of size vary, but in Sub-Saharan Africa a business with 5–20 employees is • HEs usually combine household and business finances. considered small, and a business with 21–50 is considered The HE is part of the household’s livelihood portfolio, in medium. However, some global studies classify businesses which funds move back and forth between the HE and with less than 250 employees as small (see, for example, other activities (such as purchasing farm inputs). Ayyagari, Demirgüç-Kunt, and Maksimovic 2011). Differ- • SMEs are likely to have a relationship with a bank or micro- ences in classification often lead to confusion in policy finance institution (as a saver or borrower), but they depend discussions. on their own and their family’s savings for capital as well. remains the same (Grimm, Knorringa, and use informal retail networks, including HEs, as Lay 2012). Most enterprises never hire another vendors to reach consumers (for example, sell- worker (Fox and Sohnesen 2012). ing bottled beverages or mobile phone cards). This steady state is not simply the result of Depending on location and sector, the mar- capital constraints. It arises for multiple reasons. ket for their product may be limited, which First, consistent with their growth-oriented restricts growth. approach, SMEs tend to start not as HEs but Third, HEs tend to operate in sectors that as SMEs, with more assets and employees at are relatively easier to enter, so they face a lot the beginning. From the start, they are a differ- of competition. Expansion would require more ent type of business. The owners have demon- capital and constitute a big risk. HE owners strated that they have an important skill—the report having to work hard to keep customers ability to hire and manage labor outside the and stay ahead of the competition. In a West family. Managerial skill is often considered a Africa study, 60 percent of HEs reported that key indicator of entrepreneurial potential (De competition and lack of customers are major Mel, McKenzie, and Woodruff 2012b; Gelb et threats to their existence (Grimm, Knorringa, al. 2009). and Lay 2012). Second, almost all HEs rely on households, Fourth, even if the market for their prod- not other businesses, as their customers. HEs uct has growth potential, expansion for a self- are rarely connected to larger value chains. employed HE owner means taking on employ- The exception occurs when large wholesalers ees, with associated management costs that creating productive Employment for Youth in the household Enterprise Sector 157 owners usually do not want to handle. HEs are Sub-Saharan Africa is still mostly rain-fed, “Despite the likely to take on apprentices, but not full-time resulting in substantial seasonal underemploy- difficulties … in employees. ment. Another reason is that farmers still use a day’s work, I Fifth, HEs operate in a risky environment, very basic production techniques, which make make as much with limited opportunities to hedge their it hard for them to get ahead. as a civil servant risk. Given the retail nature of their sales, HEs In urban areas, some HE owners earn more depend on growth in income both from agri- than they could expect to make in a wage job, does in a month.” culture and from wages and salaries to survive. either because their lower education and skills Madagascar This dependence means that their incomes are limit their access to a wage job or because wage pro-cyclical; a negative shock to the local agri- rates are still low in Africa. It is usually the cultural sector can hurt them as well. Expan- roughly 10 percent of HE owners who have sion may only increase risk. the managerial skills to hire workers whose Seventh, in addition to business risk, HEs earnings exceed those from wage employment face household risk. In surveys, business risk (Fox and Sohnesen 2012). Other benefits cited tends to be listed as more important, although by HE owners include the opportunity to be females tend to cite household events such as their own boss and flexible hours (Falco et al. an illness in the family as a reason for exiting 2012). the HE sector. In addition, HEs rarely sepa- rate the business accounts from the household accounts. Constraints for Young People to Finally, HE owners have other responsibili- Enter the HE Sector ties (in farming or the household, for example), so they do not have additional time to devote Owners of HEs tend to range between the ages to their business. In their risky environment, of 25 and 40 (figure 5.5). Despite having more giving up subsistence farming to devote more education than older workers, young people time to the business may not make sense for struggle to enter the HE sector. They are held the household. back by multiple constraints, including insuf- For creating employment, the implication of ficient capital or savings, inadequate informa- these differences is that employment grows in tion on markets and input supply, and the lack the HE sector because an individual or house- of a range of skills, such as literacy, numeracy, hold seizes a business opportunity and creates and business, behavioral, and technical skills. a new enterprise, not because HEs hire a young Often the most important constraint is lack of person looking for work. capital. Young people may leave school with Even though HEs will not transform them- the aspiration to start a business, but with- selves into SMEs, they are often a good business out savings, they find it very difficult to start opportunity in Africa. In African countries they a business. tend to be found in richer areas, and house- Many young people, including those still in holds with these enterprises are less likely to be school, work in the HE sector, but not as busi- poor—they tend to be clustered in the middle ness owners. They contribute work to a fam- quintiles (Fox and Sohnesen 2012). Stron- ily business without receiving a regular wage. ger evidence from recent panel data in a few This experience can be a pathway to sustain- Eastern and Southern African countries indi- able employment in the sector (see chapter 2). cates that adding an HE does make household Working in a family business can provide essen- income and consumption grow faster regard- tial on-the-job training—in business or tech- less of wealth level, suggesting that HEs help to nical skills—and build the informal networks reduce poverty. necessary for success. These young people may For many rural households, the hourly also be earning the trust of their family, who income from an HE is higher than the hourly might eventually supply start-up capital for an income from the agricultural sector. One rea- enterprise. Studies in West African capital cit- son for this discrepancy is that agriculture in ies show a strong correlation between having a 158 Youth EmploYmEnt in Sub-Saharan africa Figure 5.5 Household enterprise owners tend to be over 25 years of age HE. Some occupations, including trading (the most common business), do not require spe- 35 cific technical skills, but some do (for example, 30 construction, manufacturing, repair, and per- 25 sonal services such as hairdressing). Young people acquire these skills through training 20 Percent or apprenticeships, mostly supplied by private 15 providers rather than public institutions. Gov- ernments need to recognize that, just as infor- 10 mal employment is normal, informal training is 5 normal for youth in the HE sector. Most such 0 training consists of private courses and appren- 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–65 66+ ticeships, as discussed extensively in chapter 3 Age (years) and summarized in box 5.4. Distribution of household Share of age group that are Two issues related to apprenticeship have a enterprise owners household enterprise owners bearing on young people’s ability to enter and Source: Fox and Sohnesen 2012. remain in the HE sector. First, the skills offered through apprenticeships are often quite narrow, limited to a specific production technology, self-employed parent and being self-employed and difficult to transfer to another occupation. (Pasquier-Doumer 2013). Second, the time spent in an apprenticeship Another common way for youth to enter the can make it challenging for young people to sector is to gain skills and experience. Techni- accumulate the savings they will need to apply cal skills are not a prerequisite for starting an their new skills in an enterprise of their own. Box 5.4 Informal training is normal for youths seeking to operate a household enterprise Young people working in HEs are substantially more likely to narrow set of technical skills—such as tailoring, carpentry, have gone through informal skills training or an apprentice- vehicle repair, or hairdressing—that are used primarily in the ship than formal vocational training. Among 25- to 34-year- HE sector. In addition to being specific to the sector of activity, olds employed in the HE sector, 32 percent have been an such skills may be specific to the technology used by the mas- apprentice at some point, but only 6 percent have gone ter (Frazer 2006). Ghana’s highly developed apprenticeship through formal vocational training. system is a good example (Monk, Sandefur, and Teal 2008). Private, informal training is the primary source of tech- Apprenticeships take many forms and vary in duration. nical skills for HEs. The spectrum of providers includes for- If a family lacks cash to compensate the master, the master profit private institutes and firms, NGOs, religious and com- may take on the apprentice for a lower fee but require more munity organizations, and individuals. The vast majority of time (labor) from the apprentice in return. Many apprentice- these providers develop their own teaching programs, are ships are informal, although some are based on a contract. self-financing, and operate on a very small scale with little To enforce this contract, the first year of the apprentice- government oversight or support. Training is generally short ship often involves little training, just unskilled labor. The and intensive, and it may offer a certificate. In some coun- time spent in an apprenticeship without earning money tries, informal private training providers enroll more trainees may account for the lower payoff to apprenticeship that is than public institutes providing formal vocational training. observed in countries such as Ghana, where this practice is Apprenticeship, in which an experienced enterprise owner common. Unless young people can save money during their (master) teaches skills on the job, is by far the most common apprenticeship, they still face barriers to entering the sector. and important institution providing training. Apprentices gen- After completing an apprenticeship, youth often report that erally have at most a junior high school education. The scope a lack of capital prevents them from applying their newly and content of apprenticeships are heavily focused on a basic, acquired skills. creating productive Employment for Youth in the household Enterprise Sector 159 In addition to capital and skills, a third barrier little guidance on which interventions could to entry is young people’s general lack of infor- work on a large scale—that is, beyond small- mation about opportunities in the HE sec- scale pilots. What is clear is that, although tor, especially relative to wage and salary jobs. young people encounter specific constraints in Young people participating in a small survey in the HE sector, reducing the multiple obstacles urban Tanzania in 2005 reported that they had that all HE owners encounter in entering and spent an average of five years idle or doing odd earning a living in the sector can generate pro- jobs while seeking a wage or salary job in an ductive employment for many. The next sec- enterprise. Three-fourths of the respondents tion focuses on how governments and policy ended up self-employed (Bridges et al. 2013). makers can support productive employment Could this search time be reduced by provid- for the large numbers of people who will spend ing better information about opportunities for their working lives outside of the wage econ- self-employment or programs to help young omy in the HE sector. people to enter the HE sector? The short answer is that the large number of public and nongovernmental projects under- Creating and Sustaining taken in Africa to help young people to enter Productive Employment in and remain in the HE sector have provided lit- Household Enterprises tle evidence on the best ways to facilitate entry and raise the earnings of HEs (box 5.5 presents The HE sector has developed with little public examples from Rwanda). In particular, there is support. Public policy has neglected this sec- Box 5.5 Rwanda: Many programs to support household enterprises, but little information on results In 2010 the World Bank and Government of Rwanda inven- HEs. While this fragmentation is common in low-income toried the major public and nongovernmental programs to countries, it means that the whole is less than the sum of support HEs and then conducted focus group discussions its parts. around the country to elicit HE owners’ perceptions of the Several organizations contacted for the inventory com- programs. Of 19 NGO and 7 government programs identi- plained that their products were not being taken up. Pro- fied, most government programs targeted SMEs, not HEs; grams directed at women noted that social attitudes limiting most NGO programs targeted HEs and emphasized training. the scope of women’s activities hindered participation. Some Some NGO programs combined training with small grants to NGOs may not be providing a product appropriate for the start a business. NGO programs tended to support specific majority of the clients or government policies may not be groups, such as women, youths, people with human immu- supportive of their efforts. Training cannot be effective if it nodeficiency virus/acquired immune deficiency syndrome cannot be used productively. (HIV/AIDS), and so on, which meant that many participants Since systematic evaluation of many of these projects is felt excluded. Few participants had received any support, but rare, it is difficult to measure their effectiveness. Some pro- those who had were generally appreciative. Most programs grams could not even provide assessments of beneficiaries focused on providing support to set up a business; partici- or basic monitoring data. A few had evidence on outcomes, pants felt that assistance is also needed after a business is however. The literacy programs had data on the number of started. women who passed the test. Better monitoring and evalu- Given the huge number of HEs operating in Rwanda, the ation would help the government to identify which pro- demand for assistance is high, but support is relatively mod- grams might be scaled up effectively and to help the NGO or est because funding is limited, as reported by many nongov- agency involved obtain financing to do so. ernmental and civil society organizations. These limitations partly reflect the tremendously fragmented effort to support Source: World Bank and IPAR 2012. 160 Youth EmploYmEnt in Sub-Saharan africa tor—both by failing to support youth to enter In many cases, public support is needed only it and by failing to provide the business climate to facilitate entry and to encourage the private to sustain incomes and productivity. Govern- initiatives that already serve the sector. In other ment strategies tend to identify the SMEs as a cases, information gaps or market failures indi- source of employment, but not the HE sector. cate a need for public intervention in the form Often the main obstacle to supporting the of regulation or targeted financial support. HE sector is an implicit bias against HEs, which are not necessarily attractive in the eyes of National Strategies Employment policy, including youth employ- public authorities, who sometimes even chase ment, is embedded in national growth strat- them out of business areas in capital cities. The egies. In Africa, most national strategies perception of HEs as unworthy of public sup- explicitly recognize that private enterprise port renders systematic efforts to support them is the key agent in economic growth and job politically challenging. In some development creation, but they focus on farms, SMEs, and circles, HEs have been criticized for not offer- larger firms, despite the large share of employ- ing the income and benefits of wage and salary ment in HEs and the potential for substantial employment, so national governments hesitate growth. Limited national support for the HE to include them in their strategies, despite the sector as an agent of economic development evidence that they contribute to growth, reduce trickles down to subnational governments, poverty, and provide better opportunities than which often exclude HEs from local develop- other occupations such as agriculture. Recog- ment (box 5.6). Mechanisms to organize the nizing that “the informal will be normal” is the sector and give its constituents a voice are first step in developing effective policies and not developed, and national institutions lack programs to help youth to create sustainable channels to provide support to HEs. Because enterprises. national and local policies, programs, and Analysis of the sector and its multiple con- projects that could support productive HEs straints yields five key areas where public policy are not designed with the sector in mind, gov- can be effective in supporting the creation and ernments miss opportunities to improve the growth of HEs:2 incomes and prospects of HEs and encour- age entry into the sector for many who could • National strategies that recognize HEs, benefit from it. The result is that HEs and the encourage the sector, and give HEs a voice employment they provide are invisible to pol- in developing national and local strategies icy makers. • Urban policies that provide adequate loca- To the extent that HEs are even on the stra- tions where HEs can work and sell their tegic radar screen, they are seen as entities to products, along with essential support ser- be transformed into SMEs and “formalized.” vices such as lighting, water, sanitation, and In Tanzania, a law from 1972 outlawing busi- security nesses that operate without a fixed premise is still on the books. This law effectively makes • Financial sector policies and programs that more than 80 percent of HEs illegal (a fact con- encourage private providers and NGOs to improve household access to financial ser- veniently ignored when the government col- lects taxes or fees from HE owners). The objec- vices, including savings and credit tive of national economic policy (2008)—“to • Support for programs that tackle multiple empower the informal sector to become formal constraints, either by building a range of in order for them to access finance, training, or skills (technical, business, and behavioral any other business development service”—runs skills) or by combining skills with capital directly counter to HEs’ own aspiration not to • Support for programs that improve access be transformed into an employer, but rather to to markets by integrating HEs into value survive and bring sufficient cash income into chains the household. It is not clear how a sector that creating productive Employment for Youth in the household Enterprise Sector 161 is virtually regulated out of business by law Box 5.6 could become “formal” in the sense of comply- ing with the law. Rwanda’s development strategies (Vision Need for a comprehensive approach 2020 and the Economic Development and In 2007 Victor Tokman, one of the first researchers on informal enter- Poverty Reduction Strategy), while support- prises, published a seminal article called “Modernizing the Informal ive of the informal sector in general, do not Sector” (Tokman 2007). He observed, recognize HEs as key economic actors. HEs either are not differentiated from SMEs or There is general agreement over the need to pay attention to are excluded entirely from government sup- the informal sector because of its importance to employment port (for example, the SME policy is limited to and poverty issues. There are also an increasing number of pro- grammes aimed at supporting similar informal activities in highly enterprises with hired labor). The disconnect diverse national contexts . . . Nevertheless, to the extent that it between Rwanda’s national strategies and the fails to embrace a shared strategic vision, this is a limited con- characteristics of the HE sector creates a poor sensus that hinders the effectiveness of policies implemented in business environment in which HEs are largely this area. While often adequate on an individual basis, they are overlooked and their specific needs are not insufficient and produce limited effects by failing to respond to a addressed. Even so, the fact that HEs are legal more comprehensive approach. in Rwanda gives the government large scope to broaden national policies specifically to include them. Ghana, where HEs have been included in tative mechanisms between the government the national strategy and institutions for many and the private sector. Aside from providing years, provides a positive example of what can mechanisms for dialogue at the strategic level, be achieved (box 5.7). Its rich history as a trad- the development of HE associations helps to ing economy, dating back before colonization, improve information flows and access to tech- is one of the reasons why it has some of the nologies and markets, allowing national and most developed public and private institu- international distributors to integrate informal tions to support informal enterprises. For agents into their value chains. example, one objective detailed in the national Poverty Reduction Strategy Paper of 2006 is Urban Policies and the Household to “enhance productivity and income/wage, Enterprise Sector with equal opportunities for men and women HEs are much more common in urban areas, in all sectors of the economy, including the and urban residents are more likely to identify informal economy” (Republic of Ghana 2006). their business as their primary and only activity The National Board for Small Scale Industries (although rural HEs are an important pathway (NBSSI) explicitly includes the self-employed out of agriculture, as detailed in chapters 2 and within its mandate and has used technical assis- 4). Youth unemployment and idleness are also tance and donor funding to pilot, refine, and more frequent in urban areas. Yet the business scale up locally developed programs to sup- climate for HEs in urban areas is rarely sup- port this sector. Rather than shunning HEs, portive and often hostile. The most frequent the trade union movement in Ghana opened complaint relates to the lack of functional space its doors to them, taking in nascent organiza- in which to do business. Other complaints refer tions such as the Ghana Traders Association to authorities’ petty corruption in the context (which at the time primarily covered Accra of enforcing regulations or collecting taxes and City) to give a collective voice to this impor- the lack of services such as security. tant economic group. In 1996, the Ghana Trade Urban authorities have five responsibilities Union Congress adopted a policy to encourage that affect the entry and productivity of HEs: the organization of informal enterprises and informal economy workers and to support • Controlling the use of public space (side- their integration into the organized consul- walks, streets, parks, and the like) 162 Youth EmploYmEnt in Sub-Saharan africa Box 5.7 Ghana’s integrated approach to HE development Ghana’s development strategy explicitly acknowledges the One lesson learned as Ghana has sought to target youth contribution of HEs to employment absorption, income self-employment more forcefully is that too many initia- growth, and local economic development. The strategy tives, with too little coordination, are under way in various places a growing emphasis on enabling youths and women ministries and nongovernmental programs. In response, the to obtain the skills and capital they need to succeed in busi- Ministry of Employment and Social Welfare, with support ness. The approach is sustainable because it is decentralized. from the World Bank, has developed a National Strategy Local officials support HEs’ growth because they recognize and Action Plan for Informal Enterprises. The plan estab- that their political survival and the generation of internal lishes a national policy framework, identifies strategic areas revenue depend on a vibrant local economy that includes for action by every ministry and agency from within their HEs. At the national level, the government collaborates with own resources and programs, and provides for coordination donors to provide supportive programs and policies that cre- through a National Committee on the Informal Economy. ate a consistent framework, disseminate lessons learned, Although the process of submitting the action plan to the and provide targeted funding. The approach has been cabinet was interrupted by the recent election, consulta- developed in phases since the late 1990s, largely through tive validation with key stakeholders has disseminated some the Rural Enterprise Programme of the Ministry of Trade and of its key messages and principles. The following areas are Industry, with funding from the International Fund for Agri- identified for concerted, coordinated action: cultural Development and the African Development Bank. • Continue to work with local governments to improve The local business advisory centers are the cornerstone of the policy and business environment, especially the legal, the system. Overseen on a technical basis by the NBSSI, they regulatory, and fiscal environment and mechanisms for receive financial support from local governments (the munic- dialogue ipal or district assemblies). The NBSSI funds up to two staff • Reduce vulnerability, particularly through infrastructure members, whose responsibilities include monitoring and and secure sites for business liaising with the NBSSI. The business advisory centers match • Improve access to finance and business development donor and government projects and programs to the needs services of local clients and help to strengthen local associations of • Raise productivity and widen access to markets, especially household, small, and medium enterprises. The associations by upgrading the apprenticeship system and standardiz- have several important roles: expressing members’ needs, ing the skills and qualifications system organizing training, liaising with authorities, disseminating • Use social protection to reduce household vulnerability information, and in some cases assisting authorities in man- aging markets and collecting market fees. “The authorities • Developing and enforcing rules on the use owners complain that their interests are not are harassing us of private space (zoning regulations) being adequately addressed. • Providing urban services to support local eco- For HEs, the lack of secure premises (a and confiscating nomic development, including local roads, problem frequently exacerbated by outright our assets. I have legal or extralegal harassment from local street lighting, public transport, and security decided to open authorities) delays start-ups, which especially my business • Vetting businesses in order to protect con- handicaps youth. In a survey of tailors and sumers (for example, ensuring that taxi driv- at night. I do ers know how to drive, professionals have dressmakers in capital cities in West Africa, business when required training, and restaurant kitchens are 43 percent of enterprises in operation for less than one year reported the lack of an adequate the government hygienic) locality for their business as a major problem is asleep. I earn • Setting and collecting fees and revenue to (Grimm, Knorringa, and Lay 2012). This prob- a living for my support local activities lem also prevents businesses from expanding. family by doing Authorities have to balance many interests The most common and violent conflicts occur this.” Tanzania in executing these responsibilities, and HE when police exert control over the space where creating productive Employment for Youth in the household Enterprise Sector 163 hawkers and other traders attempt to reach sense of insecurity and vulnerability among “I did well potential customers. These conflicts usually HEs (as borne out by Lyons and Msoka 2007; before the local erupt because local authorities fail to recognize Liviga and Mekacha 1998; Sisya 2005). authority moved the importance of these trading businesses to Overlapping and insecure land rights com- the market. the local economy and do not provide adequate plicate the problem of finding space. In Dar es Now I have space for them to operate. Local authorities do Salaam, the local government is not allowed to not always consider how rapidly HEs can be develop its own rules on the use of land next difficulty feeding expected to grow as urbanization increases. to national roads because it does not own the my family and In some cases, authorities have attempted to land—the Ministry of Transportation owns it. cannot afford to develop markets or other work space for HEs, The local government is therefore required to send my children but without consulting them, so the space enforce rules set by the Ministry of Transpor- to school.” developed has been unsuitable. Different busi- tation and not allowed to develop and enforce Rwanda nesses have different needs. Traders and per- rules that might support local enterprises in sonal service providers need premises with foot using the land. Land tenure systems based on traffic, not outside of town but in central busi- a combination of customary and common ness districts (on sidewalks, especially through law restrict the development of efficient land the creation of pedestrian-only streets), at bus markets and can prevent HEs from obtaining stops and terminals, near major road intersec- land. In Nairobi, efforts to establish secure land tions, and other places that are convenient for tenancy for slum dwellers who run HEs are people to shop and transact business. Repair complicated by multiple land rights and land shops and manufacturers such as metalworking disputes (World Bank 2013). operations need to cluster to realize agglom- Another problem limiting HEs’ productiv- eration efficiencies and share technology, and ity is the poor quality of urban services they urban areas need to provide suitable sites for receive. The construction and maintenance them to do so. Industrial estates designed for of markets, where traders and service provid- large firms can include HEs in planning and ers can congregate to sell and customers can space allocation, for the convenience of work- gather to purchase, is one of the most impor- ers and the firms. tant urban services needed by HEs, which pay Cities that fail to anticipate the growth of a fee to locate in the market. In most urban HEs and proactively identify locations for their areas, governments are simply not creating activity enter destructive cycles of “deconges- markets fast enough to keep up with popula- tion.” The precipitating event is usually a forth- tion growth. In addition, although fees are paid, coming political or sporting event that will security and sanitation may not be provided as swell crowds and heighten security concerns. expected. As demand for valuable urban land Authorities “decongest” the city by mobilizing goes up, developers pressure local governments police or other security forces to evict traders to close markets rather than to develop higher- from the city center and other lucrative areas, density, multiple-use solutions. In the end, HEs sometimes confiscating their inventory and lose customers. Other key services that can other assets, or demolishing their temporary affect their productivity are transport (taken by business structures. Eviction is rarely perma- HE owners to reach markets and by customers nent. HEs retreat, but many gradually amass to reach HEs) and water supply. the capital to return to their trading location A common perception is that HEs do not until the next eviction. pay their fees and taxes, so they do not deserve Forceful eviction from the more lucrative these services. Nothing could be further from areas where traders make spot sales presents the truth. Analysis suggests that although HEs a compelling case for the failure of local gov- may be exempt from national corporation taxes ernments to support HEs (box 5.8 presents an or value added tax registration, the majority of example from Tanzania). Even though local HEs pay local business taxes at a higher rate authorities are simply enforcing the law, their than large businesses (Fox and Sohnesen 2012). operations are counterproductive: they increase Taxes, fees, and local rules regarding land use rather than reduce poverty and heighten the are reported to be sources of petty corruption, 164 Youth EmploYmEnt in Sub-Saharan africa Box 5.8 Weakening the local economy by perpetuating the machinga cycle in Dar es Salaam Dar es Salaam has limited business premises in its high-rent, even though most such regulations apply only to formal lucrative commercial areas. Owners of HEs take to the streets enterprises. HEs have little knowledge of the tax code or at the borders of these areas as mobile operators (machin- registration requirements and no place to complain. Many gas), crowding sidewalks and roads, even if zoning laws pro- report paying large sums to unscrupulous officials to avoid hibit them from doing so. The proliferation of cars, foot traf- having their merchandise confiscated. fic, and traders, especially during rush hours, is more than city In response to the lack of market space, the city of Dar authorities can manage. To curb the congestion that mobile es Salaam recently took out a loan and built a seven-story vendors and service people bring to towns and cities, the building for use by itinerant traders. Dubbed the “mach- authorities, especially in Dar es Salaam, Arusha, and Mwanza, inga complex,” the project was estimated to cost T Sh 13 regularly engage in “clean-up” operations, in which mobile billion (approximately US$13 million), but it does not meet traders are especially vulnerable targets. Faced with eviction, the needs of the machingas, who were not consulted in its some HE operators abandon their business. Local authorities design. Among other issues, the building has no elevators deploy disproportionate force to drive out those who remain. to move customers or merchandise. No parking is avail- HE owners cite harassment by local government and law able. The building was designed to provide 10,000 work enforcement officers as their biggest problem. More than 60 spaces (compartments or stalls), but in the end, only 6,500 percent of 622 operators interviewed for one study said that rooms were built. Fewer rooms increased the average cost, forceful eviction is their most vivid experience of government requiring city authorities to raise rents to repay their loan. intervention. The same study catalogues machingas’ losses When the complex opened, only the bottom floor was filled from local eviction and relocation policies: physical capital by vendors, who then complained to authorities that the (kiosks), operating capital (fines, confiscation of inventory), machingas selling on the sidewalk were undercutting them. customers and supply lines (through increased distance), and Eventually, the upper floors were rented out to other ten- trading time (through jail sentences or time spent rebuilding ants who prize the downtown location but do not need capital to return to business). Repeated cycles of loss and to locate goods on-site for sale. Meanwhile, the machinga hardship resulting from these policies increase the poverty problem has not been solved. and deprivation of household members. Local officials also subject HEs to inspections to check compliance with licensing, taxation, and other regulations, Source: Kweka and Fox 2011. especially by police. Failure to pay the requested government is not elected by the citizens, but bribe can result in disproportionately negative it is stronger outside the capital, where local consequences (such as confiscation of inven- authorities are elected and responsibility for tory). Often owners of HEs do not even know local economic development has been decen- the current fee schedule or how much they tralized. Outside Accra, local authorities focus should pay. Corrupt local officials can then take more on creating space for HEs to operate advantage of them. and less on controlling space to exclude them Two overarching factors contribute to the (box 5.9). Improvements in the organization of poor business environment in urban areas for HEs at the local level, facilitated by the locally HEs: national policies and strategies exclude recruited and managed business advisory offi- HEs, and local governments are not accountable cers, have reportedly enhanced communication to their citizens. In Rwanda, national policies between HEs and elected local officials. that exclude HEs and strictly limit their loca- tions in urban areas create indifference to HEs’ needs among local authorities. The authorities Easing Access to Credit to Start and see their job as enforcing national regulations Sustain a Household Enterprise to control HEs, not to support their develop- HEs need capital to start a business and work- ment. In Ghana, the cooperation between HEs ing capital to maintain it. Most HEs struggle and local authorities is poor in Accra, where the to obtain capital. Over and over in surveys creating productive Employment for Youth in the household Enterprise Sector 165 Box 5.9 Clustering household enterprises for the benefit of all In the city of Bechem, Ghana, national commitment and For the businesses: public-private-donor collaboration supported the establish- • They gain a secure location. ment of a mini industrial site for metalworking, car repair, • Through clustering, trades find it easier to obtain com- and carpentry workshops. Prior to the project, these work- mon services and participate in value chains (particularly if shops were scattered around the city, some occupying valu- project-funded support services such as training or finance able downtown locations. The city endeavored to obtain a are located on-site). suitable parcel of land outside the most congested areas and • Access to business development services improves, because to develop it for the HEs, allowing the workshops to form training providers easily reach clustered clients. productivity-enhancing clusters and the city to assign use of • Access to electricity, water, sewerage, and garbage removal the valuable city-center land to higher-density, higher-value improves. projects. • Business associations become stronger. The project, which required multiple consultations with HE associations, took nearly four years to develop, but it For local authorities: finally opened in 2012 with 37 masters and 68 appren- tices. The rest of the space is being filled as new HEs move • Activities that local governments prefer to locate outside in. The site provides work space, electricity, and water. the town center, such as metalworking, carpentry, and Once the site was open, seven spare parts dealers and auto repair, are moved and clustered. some food vendors (also HEs) moved in behind the main • Businesses are easier to tax and less prone to avoid taxes, tenants to supply support services. Organizations offering because they see value for money in the public services training in areas such as basic engineering skills and elec- they receive. tronic engine servicing have also reached out to the HE For clients: owners, as they can easily reach a critical mass for their services. • Suppliers concentrate in one location, giving easier access When done with adequate prior consultation and prepa- to regional as well as local markets. ration of facilities, such sites achieve mutually satisfactory objectives. Source: William Steel, personal communication. asking about the major constraints to starting needs and supply is widest for young people or expanding a business, HE owners list capi- looking to start an HE. tal constraints at the top (see Fox and Sohne- At the root of this problem is the lack of sen 2012; Grimm, Krüger, and Lay 2011). In financial inclusion for households in Africa (see Mozambique, in a survey of HE owners who focus note 3). Most households in Africa do not had to close their business, 56 percent reported have access to a bank or microfinance institu- lack of capital or liquidity as the primary rea- tion (MFI), either for saving or for credit. Many son (Fox and Sohnesen 2013). households rely on informal savings groups to Unable to obtain credit through formal help them to save for business or household sources, they use their own savings and infor- emergencies and to obtain short-term credit mal sources of credit to meet their needs when savings are not enough. Young people (box 5.10). A complicating factor for HEs is have started to join these groups, and it is one that business and household finances are often of the reasons why more than 25 percent of linked, so efforts to smooth household con- youth in the low-income and lower-middle- sumption (to pay for lumpy expenditures such income countries report having saved some as school fees or home repairs) may compete funds in the last 12 months (FINDEX data). with efforts to maintain business liquidity and Young people with access to a bank still have potentially undermine the sustainability of the trouble getting a loan for start-up capital. Even business. Access to credit would help to balance MFIs in Africa, which target a lower-income these competing needs. The gap between credit population than banks, prefer to lend to indi- 166 Youth EmploYmEnt in Sub-Saharan africa Box 5.10 Sources of credit used by households to start a business in Tanzania • I got my capital, amounting to T Sh 250, from my mother • My initial capital of T Sh 5,000 was given by my husband, in 1986 and started by selling local brew known as and I started selling vegetables. wanzuki.a • I started selling cooked food and local brew with locally • I started with a capital of T Sh 50,000, which I obtained borrowed money (T Sh 10,000). From this business I from my father in April 2007. obtained a capital of T Sh 30,000. Then I decided to • I started with a capital of T Sh 20,000 in 1994. I got the change business by opening a genge (fresh food stand). capital from selling maize which I had produced. • I am operating a grocery where I am selling beer and • I had accumulated a capital of T Sh 50,000 from my soft drinks and had started with an initial capital of wages as I was formerly employed. T Sh 200,000, half of which was given by my mother and • I started with supplier’s credit (mali kauli) to sell goods: 1 the other half was obtained from agricultural production. kilogram of sugar, 2 kilograms of rice, 1 liter of cooking • I borrowed materials for starting cooked food vending: 3 oil, tea leaves, salt, and firewood that I collected from the kilograms of rice, 2 kilograms of meat, cooking oil (a half bush. liter), and salt. I obtained my initial capital amounting to • I started by hiring a sewing machine for T Sh 5,000 per T Sh 8,000 from my uncle, and I decided to do a business month in 2007. I worked with that machine until I got of keeping and selling guinea fowls. enough money (T Sh 120,000), which I used to buy my own sewing machine. • I obtained the initial capital amounting to T Sh 5 million Source: Kweka and Fox 2011. from selling cashew nuts, and the profit from our shop. a. Wanzuki is a local brew made of honey and yeast. Sometimes sugar, We then bought a milling machine. tea leaves, and yeast are used to make it when honey is not available. viduals with a salary so that they have a better to existing businesses in Sri Lanka helped to chance of repayment. Some MFIs will lend to improve sustainability, providing evidence of households based on collateral (such as house- capital constraints both at start-up and during hold assets) or the guarantee of a relative or operation (De Mel, McKenzie, and Woodruff friend who has a salary. In this case, an initial 2012b). savings deposit is often required. The main problem with the pilot grant pro- If young people are not creditworthy grams for youths to start a business relates to because they do not have enough savings or the availability of funds: funds are not available a track record with a formal savings institu- to provide the 5 million or so youths who are tion, are grants an option to help them to expected to start a business every year in Africa start a business? Most of the NGO-led pilots over the next 10 years with a grant of US$100 in Africa that help people to start a business (about the average size of grant used in the use grants instead of loans, and grants have pilot programs). And even if youth grants on been tried in South Asia and Latin America that scale were affordable, what about adults as well. In almost all cases, they facilitated who want to start a business? Excluding them entry into the sector. But in no cases did the could be politically difficult. In Tanzania, gov- experiment use only grants for start-up capi- ernment involvement in grant programs to tal—either business training and support support business creation led to poor target- services or vocational training was supplied ing, suggesting that governments should tread as well. Other experiments used matching carefully in this area (Kweka and Fox 2011). grants plus financial education to encour- A more broad-based strategy would be to age saving so that youths would have start-up expand financial inclusion (especially savings) capital. These approaches also produced posi- by reducing the costs of financial services and tive results, but the target groups were quite improving the range of products available to well off by African standards. Grants provided HE start-ups. creating productive Employment for Youth in the household Enterprise Sector 167 What Should Governments Do to incomes (car repair, metalworking). But Help Equip Youth with Skills for the effect of the training and the effect of other, unmeasured personal characteristics the HE Sector? associated with choosing the sector are so comingled that the returns to the training Training programs are the most common are hard to identify. government and donor intervention to sup- port HEs—both for facilitating entry and • Many people who take training do not prac- for improving incomes—whether targeted tice their skill for various reasons, including at youth or not.3 Programs provide technical the inability to finance an enterprise. training in a specific sector (such as tailoring, As a result, more systematic and care- metalworking, operating a bakery), business or ful evaluations are needed, including impact financial literacy skills (such as basic account- evaluations that measure outcomes among ing or money management), behavioral and program participants and a relevant com- life skills, or a mixture of skills. Programs tar- parison group (chapter 3, box 3.9). Only a few geted specifically to youth focus primarily on programs designed to foster self-employment the skills needed to enter the sector and may were subjected to an impact evaluation in the include all four types of skills listed above. Pro- past decade. More high-quality impact evalu- grams targeted at existing HEs tend to focus ations are needed, including studies that spe- on the business skills needed to strengthen cifically evaluate the best ways to design youth “With some or expand an enterprise, moving toward the training the employment programs (especially the best goal of improving earnings and productivity. future will be components to package together and the most The good news for youth employment is that programs designed to facilitate entry appear to effective agencies for delivering programs), the better. Instead of cost-effectiveness of those programs at scale, simply resoling have had some success (more than those tar- geted at existing entrepreneurs), so there are and their general equilibrium effects. shoes, I would Despite the limited evidence, information some positive models. like to start Despite the large number of training pro- from recent impact evaluations and other stud- ies offers a starting point to guide policy and manufacturing grams, evidence of their effectiveness among them.” Uganda HEs in Africa remains thin. Most training pro- identify promising approaches to help youth grams operate on a small scale, do not collect to enter and stay in the HE sector. This section monitoring data on dropout or graduation reviews evidence on the effectiveness of three rates, and are unable to track outcomes, let major types of training—technical skills, busi- alone outcomes for a comparison group. Even ness skills, and behavioral skills. It concludes the larger programs have not systematically that programs combining multiple interven- documented success (or failure). Analysis of tions—different types of training or training household survey data on HE earnings is not plus capital—are more successful than pro- able to isolate specific returns to training, for grams offering one intervention. None of the several reasons: combined interventions has been scaled up, however, and cost is an issue. Finally, we con- • Participation in training programs is poorly sider what the limited evidence suggests about captured in household survey data, par- the potential role for government and public ticularly for the more prevalent types of policy in building skills for more productive apprenticeships or private training. and sustainable HEs. • The product itself is very heterogeneous, even in the same sector or skill. Building Technical Skills • The majority of existing HE owners did not As discussed here and in chapter 3, technical participate in any training, and training is training, through apprenticeships and other not needed to start a business in, for exam- types of private training outside of the formal ple, trading. Training is needed to enter spe- education system, is the most popular form of cific sectors, which often do provide higher training for HEs. Informal training is very het- 168 Youth EmploYmEnt in Sub-Saharan africa erogeneous. Training can be just a few months vehicle repair, tailoring, and hairdressing. The of on-the-job learning with a skilled HE owner, apprenticeships lasted 3.3 months on aver- or it can last for years; it can combine classroom age, including a week of life skills and business instruction with on-the-job learning. Many training; 63 percent completed their appren- informal training programs or apprentice- ticeships, and a small number received start-up “Our landlord ships build narrow sets of technical skills (for capital. The estimated cost per participant was had a wood example, in tailoring, carpentry, mechanics, US$800. Participants were encouraged to start workshop, and I or hairdressing). Little is known about which their own business upon completion of the passed in front modalities are the most effective and about the apprenticeship. Results showed no improve- best way of promoting them to ensure that they ments in labor market outcomes or returns of it every day improve skills and productivity. in the short term, although the training did when going Household survey data show an association develop skills and improve business practices to school. I between private apprenticeships and increased (Cho et al. 2012).4 started stopping employment and earnings—for example, in there to play, Nigeria, Rwanda, and Tanzania (Van Adams, Building Financial Literacy and and the owner Johansson de Silva, and Razmara 2013). Business Skills Apprenticeships in Ghana offer a return in Evidence suggests that financial literacy and introduced me, specific activities, such as construction, espe- general business skills are weak in the HE little by little, to cially in rural areas, but not in other activities, sector. According to a 2008 survey of micro, “the profession.” such as tailoring (Fox and Sohnesen 2012). small, and medium businesses in Zambia, just Madagascar Apprenticeships also offer returns to work- 27 percent kept up-to-date financial accounts ers who apprentice in a firm and then leave to (FinMark Trust 2011). In a sample of tailors found their own business. The former appren- and dressmakers in Ghana, only 17 percent tice basically replicates both the technology reported keeping any written financial records, and business practice of the firm, but as a self- only 7 percent said that they had spent money employed individual (Frazer 2006). These for- on marketing their services during the previ- mer apprentices earn about 49 percent more ous year, and only 30 percent rated their shop a year than those who stay on as employees. as very organized (Karlan, Knight, and Udry Even though primary school completers are 2012). In a sample of microfinance clients in the group most likely to get an apprenticeship, Tanzania (most of whom were SMEs, not HEs), returns to apprenticeships appear to be higher two-thirds kept records, but only half engaged for individuals with lower levels of academic in marketing to attract customers (Berge, Bjor- achievement, suggesting that apprenticeships vatn, and Tungodden 2011). can provide technical skills even to those The fact that many operate in the HE sector without a basic education. Among currently without strong financial literacy and business employed people who have no formal edu- skills might suggest that young people do not cation but who have had an apprenticeship, need those skills to enter. Yet the lack of those earnings are 50 percent higher (Monk, Sand- skills could limit the sustainability and produc- efur, and Teal 2008). tivity of the enterprise, including those run by The only impact evaluation of an appren- young people. Can business or financial literacy ticeship program in Africa comes from a small- training boost the creation and productivity of scale pilot that was part of a national appren- HEs? The evidence on programs providing only ticeship program in Malawi. The pilot program financial literacy or general business skills is mixed. targeted 1,900 low-income school dropouts. Very few studies have been done in Africa (see The great majority (84 percent) had completed McKenzie and Woodruff 2012 for a review).5 primary school or less. Masters (with their Basic financial literacy is the most funda- own HE) selected through the Technical Edu- mental business skill for HEs (see the review cation and Vocational Education and Training in Xu and Zia 2012). Financial literacy can Authority (TEVETA) provided training to par- encompass many concepts, but among HEs ticipants in occupations such as bricklaying, in low-income countries it entails the most creating productive Employment for Youth in the household Enterprise Sector 169 basic levels of financial awareness, the ability literacy had no significant effects on business to understand the financial aspects of the busi- start-up in Bosnia and Herzegovina, in a sam- ness, and the ability to access financial services. ple where 85 percent of the participants had Substantial inequalities in financial literacy completed secondary school (Bruhn and Zia prevail. Women and less educated people con- 2011). In a lower-income context, a “Start and sistently measure lower (on average) on tests of Improve Your Business” Program in Sri Lanka financial literacy. Poor financial literacy could for urban women ages 25–45 with a second- contribute to low access to the formal financial ary education increased their likelihood of system—banking in particular. launching a business (De Mel, McKenzie, There is very little evidence that interven- and Woodruff 2012a). An eight-day business tions to build financial literacy in low-income training course in Pakistan, targeted to new countries affect the earnings of HEs or the microfinance clients with much less educa- capacity of youth to enter this sector. Evalu- tion, yielded no impacts on business creation ations are under way for generic programs among households that did not operate an (including some in Ghana, South Africa, and enterprise at baseline (Giné and Mansuri Uganda), but results are not available. One 2011). Efforts to evaluate the integration study from Indonesia found that financial liter- of entrepreneurship training in secondary acy had no effect on promoting saving of train- school are under way in Uganda. ees overall, although impacts were detected on The record on general business training for trainees with the very lowest levels of initial existing HEs is even weaker. While evaluations financial literacy (Cole, Sampson, and Zia 2011; show that business training has changed busi- Xu and Zia 2012, 27). ness practices, it has rarely improved produc- General business skills training ranges from tivity or survivorship of the typical small busi- very basic “rule-of-thumb” programs to more nesses in the HE sector. Studies show that some advanced classroom programs designed to help programs have changed business practices and small businesses to grow, which are often called knowledge, such as recording sales and record- business development services programs. One ing money taken for household needs (HEs problem with such programs is that they are in Pakistan); keeping records of withdrawals usually aimed at SMEs, not at the types of busi- from the business, reinvesting profits into the nesses found in the HE sector in Africa. Often business, and innovating in business (HEs and the programs target microfinance clients, mean- SMEs in Peru); improving business practices ing that other constraints, such as start-up capi- among the “Start and Improve Your Business” tal and access to work space, have already been Program participants (HEs in Sri Lanka); and surmounted. For these reasons, it is important separating personal and business expenses, to sift through the growing evidence on this keeping accounting records, and formally cal- training to determine if it is applicable to youth culating revenues (both HEs and SMEs in the employment in African HEs. Dominican Republic).7 Several programs piloted in other regions More “intensive” programs have somewhat have facilitated entry into the HE sector for larger impacts. In Peru, intensive personalized youth, but these programs were designed for business training for microfinance clients (most more educated individuals, not for young of whom had businesses larger than an HE) people with only a primary education, which improved business practices and outcomes, is the target group in Africa. Hands-on busi- whereas simple classroom-based instruction ness training and coaching provided to uni- yielded no impacts (Valdivia 2011). The evi- versity students in Tunisia increased their dence that “more is better” does not always entry into self-employment, although the hold, however. In the Dominican Republic, a impact was small in absolute terms (Premand rule-of-thumb approach that taught simple et al. 2012).6 The training also improved busi- rules for financial decision making was com- ness and behavioral skills. In contrast, com- pared to a more formal approach that focused prehensive business training and financial on the fundamentals of financial accounting. 170 Youth EmploYmEnt in Sub-Saharan africa All of the program’s impacts came from the the binding constraint. Given the expense of rule-of-thumb training. In both the Domini- these programs for both the providers and the can Republic and Pakistan, follow-up visits beneficiaries, it may not be justified or feasible yielded no additional impacts (Drexler, Fischer, to expand this kind of program to HEs. At least, and Schoar 2010 for the Dominican Republic; more careful cost-effectiveness is required. Giné and Mansuri 2011 for Pakistan). But changes in business practices do not Building Life Skills for Work always translate into improvements in produc- Limited socioemotional, or limited behavioral tivity or business survivorship. In none of the skills, or even low aspirations can prevent studies just mentioned (the Dominican Repub- potential entrants from finding and seizing lic, Pakistan, Peru, Sri Lanka) did changes in opportunities in the HE sector. Among other business practices measurably increase sales, hurdles, the lack of these skills or aspirations revenues, profits, or employment. In Pakistan, may make it hard for youth to attempt to enter some impacts were measured among men, and the sector or take up interventions designed in Peru some impacts were measured in par- to overcome barriers to entry, such as skills ticularly bad months, but overall the evidence training or capital. Because behavioral skills is weak that training improved business perfor- are still malleable among young people, vari- mance. In Ghana, a program delivering busi- ous programs have been shown to affect them. ness skills training and support to existing tai- Changing the mind-set and attitudes of young loring businesses did not lead to higher profits, people (especially young women) seems to despite short-term improvements in business help them to transition into the HE sector. practices (Karlan, Knight, and Udry 2012). Changes in behavioral skills may be part of the One exception to this pattern comes from benefits of apprenticeship or other on-the-job Sub-Saharan Africa, but not from HEs. An training. evaluation of a high-quality intensive busi- Young females face specific challenges in ness training program in Tanzania targeted to entering the HE sector (see chapter 2), and there SME microfinance clients (who already had is some evidence that behavioral skills training, employees) yielded profits that were 29 per- combined with social support, can help them cent higher among males but had no effect to overcome those challenges by teaching skills, on profits for females. The effect came mainly influencing attitudes toward the future, or through business expansion but also through unlocking aspirations. These particular com- better management of employees. In this case, ponents, included in a broader program, may simply graduating clients to a larger loan did reduce risky behavior or early pregnancy in not improve outcomes (Berge, Bjorvatn, and females and allow them to develop a plan to Tungodden 2011).8 Still, it is not clear that this enter the HE sector while they are young. type of intensive training would pay off even Partly because behavioral skills training is for male HE owners (or that they would have often one part of a comprehensive program, the time to attend such a program). there is little evidence that programs focused Clearly the heterogeneity in results across only on building behavioral skills can improve settings suggests that much remains to be labor market outcomes. However, programs learned about building business capacity— that aim to shape behavioral skills together with both how to do so and for whom. Positive other skills have shown promise. impacts suggest that the lack of business skills may be a constraint in some settings and that Building a Range of Complementary interventions may improve practices—and Skills sometimes business outcomes. At the same Youth often face multiple skill constraints in time, the results confirm that growth-oriented entering the HE sector. Programs attempting businesses are rare in the HE sector, so perhaps to build one skill at a time may have limited sustainability, not growth, is the better out- impact, but evidence from pilots that build a come to target. The studies also suggest that range of complementary skills together is more skills may not be the only constraint or even promising. These programs include “bundled creating productive Employment for Youth in the household Enterprise Sector 171 interventions” that deliver behavioral, business, cash grants to youth groups for investment and technical skills training as part of a com- capital and vocational training. The program prehensive package of support. had large, sustainable impacts on employment For example, in Uganda, BRAC delivered and earnings, especially for male partici- training in behavioral skills and technical skills pants.11 Monthly real earnings increased by targeted to adolescent girls ages 14–20.9 The U Sh 17,785 (about US$9.88) after two years program increased employment 32 percent, and U Sh 19,878 (US$11) after four years, cor- mainly through entry into self-employment, responding to increases in income of 49 and and it shaped life skills and reduced risky 41 percent, respectively, relative to the control behavior (Bandiera et al. 2012). The direct cost group. The average grant was US$374 per participant was US$18 per eligible girl, or (U Sh 673,026) per group member, so the esti- approximately US$85 per participating adoles- mated return to the average grant was 35–39 cent girl. percent. In Liberia, a program for adolescent girls and Another pilot program in northern Uganda young women combined either technical and targeted to very poor women provided four behavioral skills or business and behavioral skills. days of business skills training, an individual The program was highly effective in increasing start-up grant of US$150, and regular follow-up employment and income among adolescent by trained community workers. The program girls, most of whom had some secondary edu- led to a large increase in income of 98 per- cation (World Bank 2012). The training pack- cent, or US$6.50 a month. The program was age combining business and behavioral skills particularly effective for the poorest women. was more effective at facilitating entry into self- Additional technical assistance through employment than the training package com- follow-up had little additional impact (Blatt- bining technical and behavioral skills.10 Direct man et al. 2013). The program cost US$688 cost per participant was US$1,221 for the busi- per beneficiary. ness and behavioral skills training, leading to These results have two main implications an increase in monthly income of US$75 on for designing programs to facilitate produc- average (a 115 percent increase). tive employment for youth in the HE sector. The rural enterprise projects of the Inter- First, to facilitate entry, integrated strategies national Fund for Agricultural Development tackling multiple constraints—such as build- (IFAD), operating in several African countries, ing a range of skills or combining skills with finance demand-driven, privately provided increased access to capital and urban space— technical training to improve the productiv- are more promising.12 Both behavioral skills ity and sustainability of existing enterprises and business skills are best included in these and to help new entrants to start an enterprise integrated packages. Second, it is not clear (box 5.11). While no impact evaluation is avail- that current HE owners need more training to able, tracer studies on beneficiaries in Ghana raise their productivity or sustainability. Most suggest that both objectives can be achieved at operators do not list skills—business or tech- very low cost. Although the project specifically nical—as their main problem. They are much targeted employment growth through business more likely to mention the need for finance or expansion, most of the employment created customers (access to markets, market infor- was through start-up self-employment. mation, or work space). But when technical or business training is offered at little or no Combining Skills and Capital cost, participation is substantial. For privately Consistent with the evidence that multiple provided, demand-driven technical training, constraints hinder entry into the HE sector, the tracer studies on the Jua Kali Program in bundled interventions combining skills train- Kenya (chapter 3, box 3.7) and the IFAD pro- ing with capital are also promising. gram in Ghana (box 5.11) suggest that benefits To stimulate economic development and can accrue to trainees. For business training, employment in northern Uganda following the outcomes on productivity are elusive, espe- the cessation of hostilities, a program delivered cially for owners of HEs. 172 Youth EmploYmEnt in Sub-Saharan africa Box 5.11 Rural enterprise projects bolster skills and business development in Ghana and Senegal The International Fund for Agricultural Development (IFAD) repaid. But the participation of banks and MFIs in the credit supports rural enterprise projects to provide the skills and component was low owing to creditworthiness issues with other resources that help rural people, especially women and the start-ups. (Typical reasons for such low participation are young people, to create and develop local businesses that discussed in focus note 3.) Microenterprise subcommittees provide income and employment off the farm. Projects may established within local governments help to sustain project include the following components: gains (see box 5.7). Although the second phase specifically targeted employment growth through business expansion, • Business advisory centers provide a range of business devel- most of the employment created was through self-employ- opment services, including business orientation seminars, ment in a start-up (sometimes following a paid apprentice- community-based skills training, small business manage- ship). This result reinforces the point that targeting start-ups ment training, literacy and numeracy training, and infor- is the most promising avenue for creating employment. mation and referral services. The Promotion of Rural Entrepreneurship (PROMER) Proj- • Rural technology facilities support master craftspersons, ect in Senegal, implemented by IFAD in 2006, sought to traditional apprenticeships, and the promotion, dissemi- reduce rural poverty by fostering and consolidating profit- nation, production, and repair of technology for rural HEs able rural HEs and microenterprises capable of offering and microenterprises. stable jobs. PROMER focused on strengthening and profes- • Rural financial services, offered in conjunction with finan- sionalizing rural entrepreneurship and improving the overall cial institutions, include credit for on-lending to small rural political, legal, and institutional environment for such enter- businesses and training for financial institutions to provide prises. For its target population—rural youths and women financial services to vulnerable groups. ages 18–35 who were poor, unemployed, and out of school • Support for rural household and microenterprise orga- and who either operated or wanted to start an enter- nizations includes support for local trade associations to prise—PROMER provided a combination of technical and build partnerships with stakeholders and support for for- management training and funding. Technical skills training mulating and strengthening polices through a working primarily involved agribusiness, including food processing, group on enterprise development. and provided skills to 700 entrepreneurs in metalworking, In Africa, IFAD has implemented rural enterprise projects equipment manufacturing, textile and clothing production, in Ghana, Madagascar, Rwanda, and Senegal. Impacts and and hygiene and quality monitoring. Management training challenges of projects in Ghana (the longest-running project) was provided to about 500 entrepreneurs. The project cost and Senegal are highlighted here. CFAF 10 billion from 2006 to 2013. The Rural Enterprise Project in Ghana was implemented PROMER helped to create 240 enterprises, consolidate in two phases between 1993 and 2012. During the second 665 enterprises, create 3,750 jobs, and teach 458 people to phase (2003–12), the project collaborated with Ghana’s read. It usually takes enterprises three to five years to reach NBSSI and the Ghana Regional Appropriate Technology their full potential, and finding a niche in the economy can Industrial Service to develop 53 business advisory centers and be critical for success. Through the project, for example, a 13 rural technology facilities in 24 districts at a cost of US$30 baker started to make traditional bread, which was in high million, financed by IFAD, the African Development Bank, the demand in periurban areas but not supplied by modern bak- Government of Ghana, and contributions from project bene- eries. Apprentices trained informally by the baker opened 20 ficiaries (62 percent of whom were women). Almost half (47 enterprises of their own and created about 84 jobs. percent) of participants reported that they had increased their Other rural enterprises were not as successful. Some profits as a result of better record and bookkeeping prac- developed products that succeeded in rural markets but not tices, and 37 percent reported that they had improved their in urban markets because of high transport costs, poor mar- management and marketing skills. Following technical train- keting, or poor quality. Most entrepreneurs reported chal- ing in skilled occupations such as processing palm oil, pro- lenges in maintaining quality and continuous production. ducing tie-dyed and batik cloth, and making leather goods Rural enterprises had trouble finding an urban location in and soap, 22,000 new enterprises were established, 63 per- which to present their products, a result that highlights the cent headed by women. About 15 percent of the start-ups problem of multiple constraints. Exhibitions have played a later became inactive because they lacked a market, working major role in bringing some rural products to a wider mar- capital, or raw materials; their processing equipment failed; ket, especially furniture, and have attracted better contracts or they had personal reasons for closing the business, such as that have led to modest job creation. moving to another area. About 4,300 loans were disbursed through participating rural banks and MFIs; 87 percent were Sources: IFAD 2011; Senegal Ministry of Agriculture 2011. creating productive Employment for Youth in the household Enterprise Sector 173 Government’s Role in Building Skills mentioned in box 5.2, young people in Kenya The role of public policy in building young had misperceptions about the returns to voca- people’s skills for productive employment in tional training. the HE sector needs to be considered carefully. There is limited justification for govern- The accumulating evidence that building skills ments to provide directly the type of training in can help to facilitate entry and raise earnings in basic technical skills that is delivered routinely HEs, along with the fact that a range of skills by private providers active in countries today. matters for productivity, does not constitute a Governments should not create additional dis- case for public financing or provision of skills tortions in training markets. Attempts to limit training. Not every type of training is cost- or overly regulate the many small training effective, and the private sector routinely pro- providers have proven ineffective and should vides many types of training. For all of these be avoided. However, private providers rarely reasons, governments must carefully assess and provide other relevant skills such as behavioral justify whether there is scope for directly financ- or business skills. There may be scope for pub- ing or providing specific training opportunities. lic policy to encourage the provision of these The rationale for public intervention needs to complementary skills, which are particularly be informed by well-identified market failures relevant for the HE sector (curriculum devel- and assessed against risks of “government fail- opment is one option for doing so). ures” in providing or financing training. Overall, the appropriate role for govern- If young people cannot afford to finance ment is likely to be greater for financing than their training, governments may need to help. for providing training. Governments can Youths from the poorest households have the leverage the private sector, including NGOs, to least access to training. For example, young deliver the most promising models. The range people from the richest quintile of the wel- of training delivered by private providers is fare distribution are three times more likely to very diverse, and so are the content, duration, apprentice than those from the poorest quintile and quality of that training. Given the lack of (chapter 3). Gender patterns in participation data on the full range of privately provided in training and the types of training selected training, the degree to which training prices are also strong. Females are less likely to pur- accurately signal the quality of training oppor- sue training; when they do, they often choose tunities remains unclear. When governments training that builds a narrow set of skills, such finance training, they should strongly consider as tailoring or weaving. In this context, public performance-based contracting of private pro- policies to facilitate inclusion can be put in viders based on outcomes or at least key indica- place. Government interventions have shown tors of quality. For instance, Liberia’s Economic that financial barriers to training opportunities Empowerment of Adolescent Girls and Young can be overcome through interventions such as Women Project (EPAG) Program for adoles- vouchers (see box 3.7 on the Jua Kali Program cent girls achieved a dropout rate of less than in Kenya) or cash grants targeted to youth 5 percent through innovative design elements groups (the Youth Opportunities Program in that ensured high participation and training northern Uganda; Blattman et al. 2013). quality (see World Bank 2012; and box 3.7 in Information failures lead not only to under- chapter 3). Training providers were incentiv- investment in training but also to investment ized through performance bonuses, and the in suboptimal types of training. Poor informa- program included frequent and unannounced tion can also contribute to a misalignment of monitoring visits to check on the quality of aspirations, expectations, and attitudes toward provision. self-employment among young people. In this context, government involvement in providing National Skills Strategies better information on existing training options Many governments develop national skills and employment opportunities is warranted. strategies, but too often they focus exclusively For instance, information on training oppor- on the formal labor and training markets. Often tunities or on earnings in different occupations these strategies fail to recognize the importance can influence women’s choice of training. As of the nonfarm HE sector and fail to acknowl- 174 Youth EmploYmEnt in Sub-Saharan africa edge the prevalence of private providers and bine skills training with increased access to the existence of training markets. capital. The strategies could identify areas where • Experiment with demand-driven financ- failures exist in the HE training market, such as ing for training. Such experiments would failures in information provision, and consider include using techniques that have shown the role that government programs and poli- some results in other programs, such as cies could play in addressing them, mindful of vouchers. the cautions noted earlier. Governments could also use these strategies to provide information • Prioritize financial assistance and programs targeting the poor and women. Targeting about market needs and successful approaches the poor and women should be a priority to the multitude of donors and NGOs active in for reasons of efficiency and equity. Recent this area. In developing these strategies, govern- impact evaluations found that programs ments could consider the following options: for the poor and for women may have high • Focus programs for youth on the transi- returns (Macours, Premand, and Vakis tion into the nonfarm HE sector. Programs 2013). Programs that are simply offered appear most effective at facilitating entry “on demand,” however, may not reach into the nonfarm HE sector, including pro- the poor or women. Indeed, those groups grams that enable young people to transi- might not aspire to participate in such tion out of agriculture. Evidence is weaker programs, and they are more likely to lack on how to raise the earnings of those who the social networks that facilitate access to already operate HEs. Policies aimed at facil- them. To design inclusive programs that itating entry are more likely to yield the reach the poor and women, explicit efforts highest earnings and productivity gains. are required, as well as close attention to the barriers that females face in entering the • Adopt an evidence-based and learning-based HE sector and earning an adequate return. approach. At the very minimum, encourage all programs to track and report outcomes. Pilot programs such as BRAC’s program This monitoring should be done through in Uganda (targeting females) should be provider organizations, not through expen- encouraged as a means of identifying the sive and bureaucratic registration and certi- most effective and scalable approaches. fication processes. Overall, the government’s • Finance the development of a curriculum and role in building skills in the HE sector needs learning materials to teach very basic business to be selective, performance driven, and skills in local languages and incorporate train- evidence based. As discussed, packages or ing in those skills into education and training integrated interventions show the most programs. As part of this effort, the use of promise, although many design questions existing modules—tested and evaluated in remain, especially with respect to scal- the local context—should be encouraged.13 ability, cost, and cost-effectiveness. There- • Consider providing information to primary fore, governments should use a learning- school students and their parents on economic based approach to design, pilot, and evalu- opportunities in the HE sector and on the types ate models to find cost-effective and scal- of training that have had the best outcomes. able interventions. The evidence discussed in this chapter can guide the initial choice of models. Market Access and Voice • Encourage the delivery of “bundled” inter- ventions that tackle multiple constraints. Current and prospective HE owners report Specifically, these interventions include that poor access to input and product markets programs that deliver behavioral, business, is a serious constraint. Although traditional or technical skills training as part of a com- private sector development approaches assume prehensive package or programs that com- that HEs’ exclusion from defined markets and creating productive Employment for Youth in the household Enterprise Sector 175 large value chains cannot be remedied, new gained a dominant market share of Kenya’s “My main research is challenging that assumption. HEs money transfer and e-money market. challenge is stiff can participate in international value chains The second and third are buyer-led models, competition, if producer associations can be created and because the buyer helps to organize the value markets structured to include them. Recent chain so that producers can access a bigger made worse analysis by the Monitor Group highlights how or more secure market. These models require by customers imaginative new “bottom of the pyramid” HE producers to organize in associations or who take long business models in low-income Sub-Saharan cooperatives, with which the buyer works to to pay or don’t Africa can include HEs (Kubzansky, Cooper, obtain a product for a market defined by the pay, including and Barbary 2011). Three distinct business buyer. For example, the buyer provides output the subcounty models are identified: specifications and the necessary training to the association or cooperative and manages and government, • Distribution and sales through informal shops. monitors the quality of the output. The con- whose failure to In this model, enterprises develop a route tract production model has been successful in pay compelled to market that leverages (and sometimes upgrades) informal distribution and sales the handicraft sector (box 5.12). It relies on a me to lay off dedicated entrepreneur at the top of the value workers.” channels to sell products through multiple chain to organize both production and export fragmented or unorganized shops. Uganda into existing distribution channels. The deep • Contract production. This system sources procurement model relies on existing large produce directly from a large number of contractors to reach out to HEs as subcontrac- small-scale producers in (often rural) sup- tors to supply their value chains. ply chains. The contractor organizes the Young people are well placed to participate supply chain from the top and provides in these initiatives. They more easily adopt new critical inputs, specifications, training, and technology and methods brought to the HE sec- credit to suppliers. The suppliers provide tor by the large companies; this was especially assured quantities of specialty produce at true for the mobile phone and mobile money fair and guaranteed prices. value chains, where many youths got their first • Direct procurement. Direct procurement job as sellers of air time or pay-per-use phones. setups bypass traditional intermediar- Governments have usually not played a direct ies and purchase directly from large net- role in facilitating these efforts, as private sector works of low-income producers, often pro- expertise is required. However, by encouraging viding training to meet quality and other partnerships and donor financing of technical specifications. assistance and initial risk capital as well as pro- viding the enabling environment, governments The first model, the producer-led model, is can support these efforts. For example, the probably the best known. It has been used by Rwanda government was a key partner in scal- Coca-Cola, Bayer, mobile phone companies, ing up the Rwanda basket initiative. Challenge and the M-PESA mobile money system, for grant funds from the Department for Interna- example. These companies have been able to tional Development in the United Kingdom forge links with HE traders, using simple meth- helped to develop M-PESA in Kenya, with the ods of distribution to benefit from access to support of Kenya’s central bank. the wide marketing channels that petty trading provides for their products or services. Kottoh Associations (2008) found that the proliferation of trading HE associations are critical to successful devel- in mobile phone credits by hawkers for MTN, opment of the sector. As millions of individu- Vodafone, and other telecommunication giants als operating millions of very small, scattered in Ghana is providing above-average income businesses, HE owners lack a unified voice to the HEs while benefiting the companies. By and struggle to be heard alongside the mul- training small-scale traders to be its agents in tiple players in the development process. Indi- rural and periurban areas, M-PESA quickly vidually, HEs are an easy target for predatory 176 Youth EmploYmEnt in Sub-Saharan africa Box 5.12 Taking a household business to the international market: Gahaya Links and Rwanda’s peace baskets A basket-weaving tradition made its way from the villages ness model. The sisters worked with the weavers to develop of Rwanda to American households, changing the lives of uniform designs and use stronger materials. To expand thousands of Rwandan women basket weavers. For decades, rapidly, the company recruited women through churches, women in Rwanda have produced distinctive, cone-shaped villages, and word of mouth to serve as contract weavers. baskets that are traditionally used to carry wedding gifts. The They established a rigorous training program starting with baskets are handcrafted from enzyme-washed papyrus and master weavers, who could then train others in the area to banana leaf. The traditional zigzag design tells an ancient meet high standards of quality. The Government of Rwanda story of friends walking together, visiting neighboring vil- helped by organizing the self-employed women weavers into lages along the way. The women who weave the baskets cooperatives and building local training centers. Rwanda also at home used to hawk them on the streets in tourist loca- joined the African Growth Opportunity Act, which allowed tions. In 1995, Janet Nkubana, who was managing a hotel Rwandan crafts to enter the United States market duty free. in Kigali, decided to set up a shop on the hotel premises to Participation in a New York trade show in 2005, sponsored market the baskets to hotel guests. She also took some bas- by the U.S. Agency for International Development, facilitated kets to sell in the United States when she visited her sister. connections with international retail stores. Their popularity there encouraged the two sisters to set up Gahaya Links reported annual sales of US$300,000 a factory with the women weavers and market the baskets in 2007; that same year, it contracted with approximately worldwide. The baskets were renamed “peace baskets,” 3,200 women in household basket-weaving enterprises because the weaving groups include Hutu and Tutsi women, across Rwanda to make their product. Gahaya Links issues who find that working together is a healing process. purchase orders based on standard designs, and the weav- Gahaya Links initially started with 27 weavers as employ- ers receive one-third of the proceeds from basket sales. ees. It gained international exposure after a Marie Claire fea- The company thrives because it focuses on quality, training ture on reconciliation through peace baskets brought orders weavers to supply a product consistently valued by interna- for 1,000 baskets. This order was a huge challenge for tional consumers. Gahaya Links, but it provided an opportunity to identify and remedy flaws in their products and to develop a new busi- Source: World Bank 2008. officials or economic agents (box 5.13). Their Ghana to distribute information about new contribution to the local economy may not products being brought to the market. be recognized. Because policies and regula- Effective participatory organizations are tions are developed without their input, these rarely created by governments, but govern- businesses are often over- and underregulated ments can easily undermine them. National (Roever 2006). HE associations make it pos- strategies should recognize the need for effec- sible to manage the consultation required to tive associations. Local government policies and develop and implement successful projects to regulations covering the use and disposition of supply public goods that benefit HEs and the public space should recognize the need to con- local economy (such as development of the sult with HEs and encourage the formation workshop cluster in Ghana, described in box of local associations to facilitate this process. 5.9). Economic agents seeking to work with Donors and other international actors can be HEs to strengthen value chains also need effec- helpful in encouraging and supporting associa- tive producer organizations as partners. HE tions. Donor financing for associations of HEs associations can provide the necessary chan- through the Trade Union Congress in Ghana nels to transmit information about support was helpful in building the capacity of these programs and can be used by private sector groups, and support for creating local HE asso- operators seeking to integrate HEs into their ciations was included in the terms of reference value chains. For example, Procter and Gamble for the business advisory centers supported by works through the hairdressers’ association in IFAD’s Rural Enterprise Project in Ghana. creating productive Employment for Youth in the household Enterprise Sector 177 Box 5.13 Reducing isolation and exploitation through self-help associations Gaining a voice in the national policy dialogue is especially collectors. SEWA representatives participated in crafting challenging for Africa’s youth, whose livelihoods depend national legislation on sexual harassment in the workplace overwhelmingly on activities outside traditional wage in 2010, a law that explicitly covers domestic workers, and employment (and will do so for the foreseeable future). in developing a national policy on vendors and hawkers in These activities are inherently vulnerable to high volatility in 2004. earnings, lack of legal recourse in case of expropriation or Given that nonwage work is likely to persist in Africa and theft, and limited access to safety net programs. These vul- given its strong gender and youth dimensions, finding ways nerabilities are compounded by an inability to influence the to support active engagement offers a better way forward policies that affect their working conditions. For women and than ignoring them, “formalizing” them, or suppressing youths, who are overrepresented in most forms of “vulner- their activities. With rare exceptions (Ghana’s Trade Union able employment”a in Africa, the need for formalized chan- Congress is one), including trade unions in the policy dia- nels that allow access to policy makers and afford a measure logue is unlikely to suffice, because traditional trade unions of visibility in the public sphere is especially pressing. usually exclude nonwage-earning participants in the labor These conditions are not inevitable. Over the past few force. Policies on issues as diverse as natural resource man- decades, new models for collective action have emerged, agement, urban zoning, sexual harassment and discrimi- enabling workers to advocate for their interests directly with nation, the minimum wage, and social safety nets affect buyers, sellers, and the state (often at multiple levels—local, all workers, not just the minority with a formal contract or national, and international). These “worker associations” registered business. Without associations that legitimately occupy a unique space somewhere between an NGO, which represent their interests, the voices of informal workers are provides services to its members, and a labor union, which unlikely to be heard or taken seriously. represents workers in negotiations with employers. Associa- Although an organization of SEWA’s breadth and experi- tions in Africa and other regions are growing in number and ence has yet to develop in Africa, several organizations are establishing this form of organizing as a viable means of applying similar models. The results of these efforts can be improving conditions in nonwage work. as straightforward as improving the visibility of HE owners The world’s largest and probably best-known worker and casual workers in national statistics and as far-reaching association is in India. The Self-Employed Women’s Associa- as including HE owners and casual workers under public tion (SEWA), established in 1972, broke traditional barriers health insurance programs. Associations of street vendors to form organizations representing cigarette rollers, head have been among the most successful to date. For example, porters, and others who did not have “employers” to bar- they have enabled street vendors in Ghana, Kenya, Liberia, gain with and hence were excluded from the trade union and South Africa to receive literacy and skills training, orga- movement. After decades of developing innovative organiz- nize child care and informal schooling, rehabilitate markets, ing modalities to negotiate with municipalities, buyers (such finance their operations, and arrange to source, transport, as cigarette and garment factories), and other actors (such and store merchandise. as forest management councils), SEWA has established itself as a leading force in India’s trade union movement. SEWA’s a. The ILO defines “vulnerable employment” as self-employed and con- activities include obtaining official identification cards for tributing family workers. Those pursuing such activities “are less likely workers to allow them to access state social protection pro- to have formal work arrangements and are therefore more likely to lack grams, programs to develop and upgrade members’ skills, elements associated with decent employment such as adequate social security and recourse to effective mechanisms of social dialogue. Vul- bargaining over piece rates paid to contract workers, provi- nerable employment is often characterized by inadequate earnings, low sion of space and licenses for vendors in marketplaces, and productivity, and difficult conditions of work that undermine workers’ lobbying for better access to forests and markets for gum fundamental rights” (ILO 2010). Conclusion generate enough new nonfarm wage employ- ment to absorb both the new entrants and those Governments need to give serious consider- who seek to leave agriculture. HEs are growing ation to the HE sector’s potential to create jobs as a share of the labor force not because of fail- for youth. Even exceptionally high economic ures in regulation and economic growth, but growth in nonfarm sectors has not and will not because households that manage to stay in the 178 Youth EmploYmEnt in Sub-Saharan africa sector make money—more than those working should be encouraged at the local level so only in agriculture or as casual laborers. that HEs can have a voice. Young people see these opportunities and • Better financial access for households. Regu- are trying to enter the sector, but they are latory reform is needed to reduce the costs frustrated by the lack of capital, information of retail banking, encourage savings, and on markets and opportunities, productivity- develop youth-focused products. NGOs enhancing skills, and locations to work. Because should be encouraged to expand low-cost the HE sector grows when new businesses arise, strategies to help youth to save start-up the strategy for increasing employment in this capital, such as village savings groups. sector is to encourage business creation, sup- • More effective skills strategies. Governments port higher earnings, and increase sustainabil- should focus their efforts on market- ity. Initiatives that address the most important enhancing programs, such as programs to constraints on creating a business—lack of disseminate information about opportuni- finance and space to work—will be the most ties and facilitate access to existing training effective. for disadvantaged youth and not attempt to Most countries have no such strategy for deliver training directly. They should also the HE sector. Instead, they have a fragmented invest selectively in good-quality training youth employment strategy focused primar- programs that tackle multiple constraints, ily on the wage sector and a strategy for SMEs. either by building a range of skills such as This set of strategies is not working. Initiatives technical, business, and behavioral skills or for SMEs rarely reach HEs. Boutique NGO by combining skills with capital. projects cannot reach enough prospective HE • Development of value chains. This activ- owners to make a difference. ity resides primarily with the private sec- A comprehensive national HE strategy tor. Governments should encourage social should focus on approaches that are demand entrepreneurs to include HEs in value chain driven and address the main constraints to development. start-up and sustainability. Elements of a • Associations. Associations are especially comprehensive national strategy for the HE important at the local level. Regulations sector, based on a shared vision of the sector’s on consultation prior to land use planning potential to generate productive employment, or implementation should encourage local include the following: governments to work through independent HE associations. Donors and NGOs should • National strategies. Growth and employ- extend support to these associations, help- ment strategies need to recognize the sec- tor’s potential and propose a supportive ing to develop their capacity. policy framework. The strategy should endorse the creation of independent HE Notes associations to reduce the costs of reaching 1. In many countries, rural households include individual enterprises and to give this sec- households in market towns or periurban areas. tor a voice in government decision making. They are often designated rural because agricul- The specific outcomes to be attained in the ture is still the predominant economic activity. sector should be to expand the number of 2. This section draws from World Bank studies of businesses and increase their productivity. HEs in the Republic of Congo, Ghana, Kenya, Rwanda, Tanzania, and Uganda as well as • Urban policies. HEs also need to be recog- broader reviews of the literature and evidence. nized as an important element in local eco- 3. At the time of writing, the youth employment nomic development. Local governments inventory included 86 youth employment proj- must be accountable to HEs and include ects in Africa, of which 79 provide skills train- them in decisions related to planning, zon- ing. In a review of interventions to support ing, and land use, land markets and land young workers in Sub-Saharan Africa, Rother tenure, and infrastructure. HE associations (2006) found, “In most cases, these programs creating productive Employment for Youth in the household Enterprise Sector 179 include elements targeted at helping young participated in the training; of those who par- people to start their own businesses, com- ticipated, 85 percent took part in the life skills bined with elements of skills development and training and 53 percent took part in the voca- training.” tional skills training (primarily in dressmaking 4. Outside Africa, few studies offer evidence that or tailoring). technical training by itself can facilitate entry 10. The latter package was designed primarily to into self-employment. For instance, skills train- help females to attain a wage job. The main ing had no significant impact among a sample reason for the low impact is that there was little of workfare participants in a welfare-dependent demand in the wage employment sector for the region of Argentina (Galasso, Ravallion, and Sal- graduates. via 2004). 11. Cash grants targeted to youth groups can affect 5. While the number of programs providing busi- occupational choice and improve income. ness skills or financial literacy training has been Blattman, Fiala, and Martinez (2013) evaluate increasing in Africa (see Messy and Monticone the impact of a Youth Opportunities Program 2012; Xu and Zia 2012), only a handful of recent implemented by the Ugandan government in the impact evaluations analyze the effect of such context of the northern Uganda Social Action programs on entry into self-employment. For Fund. The program targeted youth groups of instance, most evaluations quoted by Messy 10–30 persons, who received grants averaging and Monticone (2012) analyze the effect on US$374 per group member. (Participants were intermediary outcomes (such as knowledge of poor by most measures, but the average appli- financial literacy or savings behavior), but not cant was slightly above the average wealth and on employment outcomes (such as entry into education level in the region. For instance, 93 self-employment) or earnings. percent had completed some primary school, 6. In developed countries, Oosterbeek, van Praag, 45 percent had completed some secondary, and and Ijsselstein (2008) studied the impact of only 7 percent had no education.) Of those who entrepreneurship education on students’ com- received the cash transfer, 80 percent enrolled in petencies and intentions. They found that the vocational training, and most invested heavily program did not have the intended effects: the in business assets. The program was effective in effect on students’ self-assessed entrepreneur- changing the type of work that youth perform ial skills was insignificant, and the effect on the in an environment where most employment is intention to become an entrepreneur was sig- outside the modern wage sector: 68 percent of nificantly negative. grant recipients were working in a skilled trade, 7. For Pakistan, see Giné and Mansuri (2011). For compared with 34 percent of the control group. Peru, see Karlan and Validivia (2011). For Sri Income was also higher for grant beneficiaries. Lanka, see De Mel, McKenzie, and Woodruff On average, grant recipients had net income 50 (2012a). For the Dominican Republic, see Drex- percent higher than the control group, a return ler, Fischer, and Schoar (2010). on the initial transfer amounting to 35 percent 8. The business training course consisted of 21 a year. sessions, each lasting 45 minutes, and covered 12. Another example of integrated intervention topics such as Entrepreneurship and Entrepre- can be found in rural Nicaragua, where ben- neurial Character, Improving Customer Service, eficiaries from a cash transfer program also Managing People in Your Business, and Market- received either technical training or a business ing Strategies. Overall, these results suggest that grant. The business grant packages led to entry insufficient human capital among poor micro into self-employment as well as higher aver- and small business owners may be the more age income and consumption. The vocational important constraint to increasing productiv- training package led to entry into self-employ- ity. The results also highlight the need for more ment for the poor and higher wage income for comprehensive measures to promote develop- the nonpoor, but had no impacts on income or ment among female entrepreneurs. consumption on average (Macours, Premand, 9. The program was implemented in urban and and Vakis 2013). Both the training package and rural areas through local “adolescent develop- the business grant package also helped house- ment clubs” in which a female mentor led the holds to protect themselves against shocks by activities. The program targeted girls ages 14–20 allowing them to smooth income through who were still in school or had dropped out. Of diversification (Macours, Premand, and Vakis the girls in the target communities, 21 percent 2012). 180 Youth EmploYmEnt in Sub-Saharan africa 13. Examples of modules in wide use include ILO’s Campos, Francisco, Markus Goldstein, Laura Start and Improve Your Business Program (see McGorman, Ana Maria Munoz Boudet, and http://www.ilo.org/empent/areas/start-and- Obert Pimhidzai. 2013. “Breaking the Metal improve-your-business/lang--en/index.htm) Ceiling: Female Entrepreneurs Who Succeed in Male-Dominated Sectors in Uganda.” United and various materials produced by the Peace Nations Foundation; ExxonMobile Foundation. Corps (http://www.peacecorps.gov/library/ pubindex/); for example, see the Microenter- Cho, Yoonyoung, Davie Kalomba, Mushfiq Mobarak, and Victor Orozco. 2012. “The Effects prise Training Guide (http://files.peacecorps. of Apprenticeship Training for Vulnerable Youth gov/multimedia/pdf/library/M0068_microent. in Malawi.” Working Paper, World Bank, Wash- pdf). ington, DC. Cole, Shawn, Thomas Sampson, and Bilal Zia. 2011. References “Prices or Knowledge: What Drives Demand for Ayyagari, Meghana, Asli Demirgüç-Kunt, and Financial Services in Emerging Markets?” Journal Vojislav Maksimovic. 2011. “Small vs. Young of Finance 66 (6): 1933–67. Firms across the World: Contribution to De Mel, Suresh, David McKenzie, and Christopher Employment, Job Creation, and Growth.” Policy Woodruff. 2012a. “Business Training and Female Research Working Paper 5631, World Bank, Enterprise Start-up, Growth, and Dynamics: Washington, DC. Experimental Evidence from Sri Lanka.” Policy Bandiera, Oriana, Robin Burgess, Selim Gulesci, Research Working Paper 6145, World Bank, Washington, DC. Imran Rasul, and Munshi Sulaiman. 2012. “Can Entry-Level Entrepreneurship Transform the ———. 2012b. “One-Time Transfers of Cash or Economic Lives of the Poor?” London School of Capital Have Long-Lasting Effects on Micro- Economics. enterprises in Sri Lanka.” Science 335 (6071): 962–66. doi: 10.1126/science.1212973. Berge, Lars, Kjetil Bjorvatn, and Bertil Tungodden. 2011. “Human and Financial Capital for Micro- De Wolf, Stefan, Yves Rolland Rakotoarisoa, Lau- enterprise Development: Evidence from a Field rence Vanpaeschen, and Honoré Rabekoto. 2008. and Lab Experiment.” NHH Department of Eco- Madagascar: Le Grand Livre des Petits Métiers: nomics Discussion Paper 1/2011, Chr. Michelsen Portraits of Daily Life Professions. Heule, Belgium: Institute, Bergen. Snoek Publishers. Blattman, Christopher, Nathan Fiala, and Sebastian Drexler, Alejandro, Greg Fischer, and Antoinette S. Schoar. 2010. “Keeping It Simple: Financial Martinez. 2013. “Credit Constraints, Occupa- Literacy and Rules of Thumb.” CEPR Discus- tional Choice, and the Process of Development: sion Paper, Centre for Economic Policy Research, Long-Run Evidence from Cash Transfers in London. http://dev3.cepr.org/meets/wkcn/7/784/ Uganda.” Columbia University, New York. papers/FischerFinal.pdf. Blattman, Chris, Eric Green, Jeannie Annan, and Fafchamps, Marcel, David McKenzie, Simon Julian Jamison, with F. Aryemo, N. Carlson, Quinn, and Christopher Woodruff. 2011. “When M. Emeriau, and A. Segura. 2013. “Building Is Capital Enough to Get Female Microenter- Women’s Economic and Social Empowerment prises Growing? Evidence from a Randomized through Enterprise: An Experimental Assess- Experiment in Ghana.” CSAE Working Paper ment of the Women’s Income Generating Sup- 2011-11, Centre for the Study of African Econo- port (WINGS) Program in Uganda.” Innovations mies, Oxford. for Poverty Action, New Haven, CT. http://www. Falco, Paolo, William F. Maloney, Bob Rijkers, poverty-action.org/sites/default/files/wings_full_ and Mauricio Sarrias. 2012. “Heterogeneity in policy_report_0.pdf. Subjective Well-Being: An Application to Occu- Bridges, Sarah, Louise Fox, Alessio Gaggero, and pational Allocation in Africa.” Policy Research Trudy Owens. 2013. “Labour Market Entry and Working Paper 6244, World Bank, Washington, Earnings: Evidence from Tanzanian Retrospec- DC. tive Data.” Paper prepared for the CSAE confer- FinMark Trust. 2011. “Financial Education in Zam- ence “Economic Development in Africa,” Oxford bia: What Does FinScope Tell Us?” Financial University, Oxford, March. Access Matters 6, Marshalltown, South Africa. Bruhn, Miriam, and Bilal Zia. 2011. “Stimulating http://www.finmark.org.za/publication/ Managerial Capital in Emerging Markets: The financial-education-in-zambia-what-does- Impact of Business and Financial Literacy for finscope-tell-us/. Young Entrepreneurs.” Policy Research Paper Fox, Louise, and Obert Pimhidzai. 2012. “Is Infor- 5642, World Bank, Washington, DC. mality Welfare-Enhancing Structural Transfor- creating productive Employment for Youth in the household Enterprise Sector 181 mation? Evidence from Uganda.” Policy Research Haggblade, Steven, Peter Hazell, and Thomas Rear- Working Paper 5866, World Bank, Washington, don. 2010. “The Rural Non-Farm Economy: DC. Prospects and Poverty Reduction.” World Devel- ———. 2013. “Different Dreams, Same Bed: Col- opment 38 (10) 1429–41. lecting, Using, and Interpreting Employment Hicks, Joan Hamory, Michael Kremer, Isaac Mbit, Statistics in Sub-Saharan Africa: The Case of and Edward Miguel. 2011. “Vocational Education Uganda.” Policy Research Working Paper 6436, Voucher Delivery and Labor Market Returns: A World Bank, Washington, DC. Randomized Evaluation among Kenyan Youth.” Fox, Louise, and Thomas Pave Sohnesen. 2012. Report for the Spanish Impact Evaluation Fund “Household Enterprises in Sub-Saharan Africa: (SIEF) Phase II, World Bank, Washington, DC. Why They Matter for Growth, Jobs, and Liveli- IFAD (International Fund for Agricultural Devel- hoods.” Policy Research Working Paper 6184, opment). 2011. “Promoting Rural Enterprise World Bank, Washington, DC. Growth and Development: Lessons from Four ———. 2013. “Household Enterprises in Mozam- Projects in Sub-Saharan Africa.” Field Experi- bique: Key to Poverty Reduction but Not on the ences 2, IFAD, Rome. Development Agenda?” Policy Research Working ILO (International Labour Organization). 1993. Paper 6570, World Bank, Washington, DC. “Resolution Concerning the International Clas- Frazer, Garth. 2006. “Learning the Master’s Trade: sification of Status in Employment (ICSE), Apprenticeship and Human Capital in Ghana.” Adopted by the Fifteenth International Con- Journal of Development Economics 81 (2): ference of Labour Statisticians.” ILO, Geneva. 259–98. http://www.ilo.org/public/english/bureau/stat/ download/res/icse.pdf. Galasso, Emanuela, Martin Ravallion, and Augus- tin Salvia. 2004. “Assisting the Transition from ———. 2010. “Global Trends in Employment: Workfare to Work: Argentina’s Proempleo Pro- January 2010.” ILO, Geneva. gram.” Industrial and Labor Relations Review 57 Inchauste, Gabriela, and Sergio Olivieri. 2012. (5): 128–42. “Understanding Poverty Reduction in Bangla- Gelb, Alan, Taye Mengistae, Vijaya Ramachandran, desh: A Microdecomposition Approach.” Back- and Manju Kedia Shah. 2009. “To Formalize or ground Paper prepared for the 2012 Poverty Not to Formalize? Comparisons of Microen- Assessment, World Bank, Washington, DC. terprises from Southern and East Africa.” CGD Karlan, Dean, Ryan Knight, and Christopher Udry. Working Paper 175, Center for Global Develop- 2012. “Hoping to Win, Expected to Lose: Theory ment, London. and Lessons on Micro Enterprise Development.” Ghana, Republic of. 2006. “Poverty Reduction NBER Working Paper 18325, National Bureau of Strategy Paper.” National Development Planning Economic Research, Cambridge, MA. Commission. IMF Country Report 06/225, IMF, Washington, DC, June. Karlan, Dean, and Martin Valdivia. 2011. “Teach- ing Entrepreneurship: Impact of Business Train- Giné, Xavier, and Ghazala Mansuri. 2011. “Money ing on Microfinance Clients and Institutions.” or Ideas? A Field Experiment on Constraints Review of Economics and Statistics 93 (2): 510–27. to Entrepreneurship in Rural Pakistan.” World Bank, Washington DC. http://siteresources. Kottoh, M. 2008. “Linking Corporations to House- worldbank.org/DEC/Resources/Money_or_ hold Enterprises: Ghana Telecom (Vodafone Ideas.pdf. Ghana) Value Chain.” New Legon Observer, Institute of Statistical, Social, and Economic Grimm, Michael, Peter Knorringa, and Jann Lay. Research, University of Ghana. 2012. “Constrained Gazelles: High Potentials in West Africa’s Informal Economy.” World Devel- Kubzansky, Michael, Ansulie Cooper, and Victoria opment 40 (7): 1352–68. Barbary. 2011. “Promise and Progress: Market- Based Solutions to Poverty in Africa.” Monitor Grimm, Michael, Jens Krüger, and Jann Lay. 2011. Group. http://web.mit.edu/idi/idi/Africa-%20 “Barriers to Entry and Returns to Capital in PromiseAndProgress-MIM.pdf. Informal Activities: Evidence from Sub-Saharan Africa.” Review of Income and Wealth 57 (5): s27– Kweka, Josaphat, and Louise Fox. 2011. “The 53. doi: 10.1111/j.1475-4991.2011.00453.x. Household Enterprise Sector in Tanzania: Why It Matters and Who Cares.” Policy Working Paper Gulyani, Sumila, Debabrata Talukdar, and Darby 5882, World Bank, Washington, DC. Jack. 2010. “Poverty, Living Conditions, and Infrastructure Access: A Comparison of Slums Lanjouw, Peter, Jaime Quizon, and Robert Sparrow. in Dakar, Johannesburg, and Nairobi.” Policy 2001. “Non-Agricultural Earnings in Periurban Research Working Paper 5388, World Bank, Areas of Tanzania: Evidence from Household Washington, DC. Survey Data.” Food Policy 26 (4): 385–403. 182 Youth EmploYmEnt in Sub-Saharan africa Liviga, Athumani, and Rugatiri Mekacha. 1998. Policy Research Working Paper 6285, World “Youth Migration and Poverty Alleviation: A Bank, Washington, DC. Case Study of Petty Traders (Wamachinga) in Roever, Sally. 2006.” Enforcement and Compli- Dar es Salaam.” Research Report 98.5, Research ance in Lima’s Street Markets: The Origins and on Poverty Alleviation, Dar es Salaam. Consequences of Policy Incoherence toward Lyons, Michal, with Colman Msoka. 2007. “Micro- Informal Traders.” In Linking the Formal and trading in Urban Mainland Tanzania: The Way Informal Economy: Concepts and Policies, edited Forward.” Report for HTSPE Development Con- by Basudep Guha-Khasnobis, Ravi Kanbure, and sulting Services for the Development Partner Elinor Ostrom, 246–62. Oxford: Oxford Univer- Group, Tanzania. sity Press. Macours, Karen, Patrick Premand, and Renos Vakis. Rother, Friederike. 2006. “Interventions to Support 2012. “Transfers, Diversification, and Household Young Workers in Sub Saharan Africa: Regional Risk Strategies: Experimental Evidence with Report for the Youth Employment Inventory.” Implications for Climate Change Adaptation.” World Bank, Washington, DC. Policy Research Working Paper 5137, World Sisya, Marietta R. 2005. “The Impact of Reloca- Bank, Washington, DC. tion of Petty Traders on Income Generation for ———. 2013. “Demand versus Returns? Pro-Poor Local Authorities to Finance Social Services: The Targeting of Business Grants and Vocational Case of Ilala Municipal Council.” MPA thesis, Skills Training.” Policy Research Working Paper Mzumbe University, Morogoro. 6389, World Bank, Washington, DC. Tokman, Victor E. 2007. “Modernizing the Informal McKenzie, David, and Christopher Woodruff. 2012. Sector.” DESA Working Paper 42, United Nations “What Are We Learning from Business Training Department of Economic and Social Affairs, and Entrepreneurship Evaluations around the New York. Developing World?” Policy Research Working Uteng, Tanu Priya. 2011. “Gender and Mobility Paper 6202, World Bank, Washington, DC. in the Developing World.” Background paper Messy, Flore-Anne, and Chiara Monticone. 2012. for the World Development Report 2012: Gender “The Status of Financial Education in Africa.” Equality and Development, World Bank, Wash- OECD Working Paper on Finance, Insurance, ington, DC. and Private Pensions 25, OECD Publishing, Valdivia, Martin. 2011. “Training or Technical Paris. doi: 10.1787/5k94cqqx90wl-en. Assistance? A Field Experiment to Learn What Ministry of Agriculture, Senegal. 2011. “La promo- Works to Increase Managerial Capital for Female tion des micro et petites entreprises rurales: Un Microentrepreneurs.” Grupo de Análisis para el moyen efficace pour lutter contre la pauvreté en Desarrollo (GRADE), Lima. milieu rural; livret de capitalisation (PROME- Van Adams, Arvil, Sara Johansson de Silva, and RII).” Dakar. Setareh Razmara. 2013. Improving Skills Develop- Monk, Courtney, Justin Sandefur, and Francis Teal. ment in the Informal Sector: Strategies for Sub- 2008. “Does Doing an Apprenticeship Pay Off? Saharan Africa. Washington, DC: World Bank. Evidence from Ghana.” CSAE Working Paper World Bank. 2008. “Weaving Peace in Rwanda.” In 2008-08, Centre for the Study of African Econo- Doing Business: Women in Africa, 17–21. Wash- mies, Oxford. ington, DC: World Bank. Oosterbeek, Hessel, Mirjam van Praag, and Auke ———. 2012. “Preliminary EPAG Midline Report: Ijsselstein. 2008. “The Impact of Entrepreneur- Economic Empowerment of Adolescent Girls ship Education on Entrepreneurship Competen- and Young Women (EPAG) Project in Liberia.” cies and Intentions: An Evaluation of the Junior World Bank, Washington, DC. Achievement Student Mini-Company Program.” IZA Discussion Paper 3641, Institute for the ———. 2013. “Time to Shift Gears: Accelerat- Study of Labor, Bonn. ing Growth and Poverty Reduction in the New Kenya.” Kenya Economic Update, World Bank, Pasquier-Doumer, Laure. 2013. “Intergenerational Washington, DC. Transmission of Self-Employed Status in the Informal Sector: A Constrained Choice or Better World Bank and IPAR (Institute of Policy Anal- Income Prospects? Evidence from Seven West- ysis and Research Rwanda). 2012. “Raising Pro- African Countries.” Journal of African Economies ductivity and Reducing Risks of Household Enter- 22 (1): 73–111. doi: 10.1093/jae/ejs017. prises in Rwanda.” World Bank, Washington, DC. Premand, Patrick, Stefanie Brodmann, Rita Xu, Lisa, and Bilal Zia. 2012. “Financial Literacy Almeida, Rebekka Grun, and Mahdi Barouni. around the World: An Overview of the Evidence 2012. “Entrepreneurship Training and Self- with Practical Suggestions for the Way Forward.” Employment among University Graduates: Policy Research Working Paper 6107, World Evidence from a Randomized Trial in Tunisia.” Bank, Washington, DC. Financial Inclusion and the Transition to FOCUS Sustainable Livelihoods for Young People NOTE 3 f “ inancial inclusion” means that individuals households take out a loan to make potentially can obtain a range of financial products— high-yielding but risky investments. savings, credit, insurance, and payment sys- Agricultural households need to save and tems, including systems to send remittances— borrow for a range of reasons—to obtain and at a reasonable cost and on a sustained basis sustain land to farm, to buy or rent equipment, (Gardeva and Rhyne 2011). Financial inclusion and to acquire other inputs that make their remains elusive for many African households, efforts more productive. Households that oper- which generally develop their own strate- ate a nonfarm enterprise also need to save and gies for saving and rely on loans from family, borrow to start and maintain their business. friends, and informal lenders to meet their A complicating factor for farm households need for credit. The gap in access to savings and households operating nonfarm enter- vehicles is at least as important as the gap in prises is that the finances of the farm, house- access to credit. Closing these gaps is challeng- hold enterprise (HE), and household itself are ing for most households, especially for young often linked, so household income smoothing people, whose access to formal financial ser- and credit needs (for example, funds to pay vices is often limited by banking regulations, for lumpy expenditures such as school fees or the lack of products designed to meet their home repairs) may compete with farm or busi- particular needs, and the difficulty of estab- ness liquidity needs, potentially reducing the lishing their creditworthiness. Across Africa, sustainability of the business. For this reason, technology is reducing the cost and increasing the gap in access to credit occurs at both the the accessibility of formal financial services; household and the enterprise level. at the same time, informal savings and credit Young or aspiring farmers and HE owners “It is not easy groups are expanding. These informal groups can have the most difficulty mobilizing capital, for us to access may offer a promising venue for integrating the because their capacity to save and borrow is credit like the delivery of youth programs in communities, often so constrained. In many countries, young adults.” Uganda including programs to foster savings behavior, people cannot access the formal banking system build business and soft skills, empower females, even if they want to, because laws prohibit open- and expand financial inclusion, with the goal ing a bank account before age 18.1 Even where of enabling young people to pursue sustainable no legal obstacles stand in their way, young livelihoods. people can be frustrated by the lack of products designed to meet their particular needs as savers and borrowers of small amounts. Gaps in Access to Savings Despite these limitations, low-income and Credit youths in various situations (at home with par- ents, living away from home, or living on the The ability to save is intricately linked with the street) persist in saving for school, emergen- ability to obtain credit. Without the habit of cies, and less often, to start a business (USAID regularly setting money aside, individuals have 2009). Saving not only is important for meet- difficulty repaying loans. In addition, lenders ing those needs, but having financial assets by will always require borrowers to use a portion itself has positive behavioral effects for young of their savings to pay for part of a proposed people, “increasing future orientation, long- investment (the down payment). Savings can term thinking, planning, and self-efficacy” also serve as a form of self-insurance when (Kilara and Latortue 2012). Research shows 183 184 Youth EmploYmEnt in Sub-Saharan africa that behaviors such as saving and exercising inson 2009). Women sometimes go to extraor- financial discipline are best developed at a dinary lengths to set money aside. For example, young age. If young people do not develop the in the absence of other mechanisms to save, habit of saving, they will have trouble starting market women in India with more established and sustaining farm and nonfarm enterprises businesses report taking loans from informal later on. moneylenders at very high interest rates just so they will be required to put funds aside every day. They resort to this strategy even though Formal Savings Services: they could accumulate much more capital and Issues and Options profits if they were to save a part of their profits every day to self-finance the next replacement A secure place for savings allows people to of inventory (Karlan and Appel 2011). improve their planning, financial decisions, and risk management by keeping their money African Households Save, but Rarely out of temptation’s way (Mullainathan and at the Bank Shafir 2011; Collins et al. 2009). Although low- African households are no different from income households and individuals may have households elsewhere—they know the value very little cash on a given day, a savings habit of savings. Controlling for national income, can transform that cash over time into a larger African adults are actually more likely to save sum, which can then be used for lumpy expen- than non-African adults, especially in low- ditures such as school fees or investment in income countries (figure F3.1). African youths inventory or tools for an HE. But when people in lower-income countries also report sav- lack a secure place to put their money, they are ing more than their counterparts in countries much less likely to save, as the difficulties of with similar income levels. In some countries, daily life for the household and the extended more than 20 percent of individuals ages 15–25 family demand a short-term financial focus, reported that they had been able to save some and multiple requests for cash reduce self- funds in the last year. control. Access to a secure place for savings is Although youths and older adults say that particularly important for women’s ability to they save, they do not use a bank account build up capital for a business, because women for that purpose. Some use informal savings are especially subject to pressure for funds from groups in the village or neighborhood, but other members of the family (Dupas and Rob- many simply keep their cash at home, the least Figure F3.1 Percentage of population ages 15 and older who reported saving any amount in the last 12 months, by savings mechanism a. Ages 15–24 b. Ages 25 and older 50 50 40 40 Percent 30 30 Percent 20 20 10 10 0 0 Low- Lower- Upper- Low- Lower- Upper- Low- Lower- Upper- Low- Lower- Upper- income middle- middle- income middle- middle- income middle- middle- income middle- middle- countries income income countries income income countries income income countries income income countries countries countries countries countries countries countries countries Sub-Saharan Africa Rest of the world Sub-Saharan Africa Rest of the world Formal Informal/club Other Source: Based on FINDEX data. focus note 3 185 safe place. In surveys, young people and adults Box F3.1 express a desire for an affordable, convenient, safe place to keep their money—a service usu- ally provided by a bank. Rural banks in Ghana: Reaching clients who Access to banking services rises with income are underserved by other banks and urbanization. Banking is an industry with economies of scale, so banks go where the Ghana’s rural and community banks (RCBs) were established by the money and the customers are (unless spe- government in the late 1970s to facilitate Cocoa Board payments to farmers. The banks were privatized and became microfinance institu- cial provisions are made; see box F3.1 for an tions in the 1990s, and today they provide various affordable sav- example of rural banking from Ghana). Given ings products (savings accounts, current accounts, daily deposits col- the low incomes and low population density lected by agents who go door to door, and fixed or time deposits), prevalent in Africa, it is not surprising that few credit products (microfinance loans, personal loans, salary loans, and households use formal banking services: only overdraft facilities, for example), and money transfer and payment 24 percent of households in Africa have access services. Ghana’s RCBs reach 2.8 million depositors and 680,000 bor- to a bank account.2 rowers, consisting mainly of farmers, government employees, and But the lack of a bank account is far from the small and microentrepreneurs—clients often underserved by formal whole story in Africa. As in the rest of the world, financial institutions. They are now the largest providers of formal in Africa’s banking industry, technological financial services in rural Ghana. With some additional outreach, they innovation and regulation interact in complex could benefit young people—another underserved population. ways to influence how much it costs to provide Source: IFPRI and World Bank 2010, brief 5. financial services and who can obtain them. Options for Lowering Costs and Expanding Outreach At the firm level, the high fixed costs of operat- population density makes brick-and-mortar “Even if you have ing a bank office (building, machinery, banking branch banking unprofitable, but the con- 100 shillings, you staff, security), the need for reliable infrastruc- struction of cell phone towers and networks can save your ture and services (electricity and communica- has opened opportunities to bank through a tions), and customers’ unwillingness to use a mobile phone, ATM, or point-of-sale device. little money.” bank that is far from their work or home mean In branchless banking, the costs of savings Uganda that high population density is a prerequisite account and bill payment services are as much for traditional banking services to spread.3 as 50 percent lower, and the costs of money Given the fixed costs associated with maintain- transfers are about 20 percent lower, than they ing customer accounts (recording transactions, are in traditional banking by microfinance processing fees and interest), small account bal- institutions (MFIs) and other banks oriented ances can be a losing proposition for the bank toward the poor (McKay and Pickens 2010).4 and for the customer, to whom those costs are The rapid spread of M-PESA in Kenya, which passed in the form of account and transaction increased household access to formal finance service charges. Regulations can compound by 10 percentage points in three years, shows the cost problem when they require proof of the potential (King 2012). The use of mobile identification (birth certificates and similar phones and other new technology to expand documents), which many low- and moderate- access to formal banking is particularly attrac- income individuals lack and which are difficult tive for young people, who are always early to acquire. In Kenya and India, to cite just two adopters.5 examples, technology is helping consumers to At the same time, banks need to use these bypass these problems (box F3.2). new channels for connecting with young peo- Recent technological applications such ple—to offer products attractive to youths and as mobile banking have sharply reduced the bring them in as customers early—even if the costs of retail banking, offering a major new benefits are more likely to materialize over the opportunity to increase household access to long term. This effort may require support as secure savings and payment mechanisms. Low part of financial sector development. 186 Youth EmploYmEnt in Sub-Saharan africa Box F3.2 Use of technology: Bringing secure financial services to new markets in Kenya and India Just as the cellular telephone revolution has lowered the issuers and e-money transactions. Rather than simply block- cost of communication and allowed poor countries in Africa ing Safaricom from operating the system, the central bank and elsewhere to leapfrog over the problem of building a negotiated an agreement under which Safaricom would landline phone network, branchless banking using mobile operate the system. The agreement stipulated that the phones and “micro automated teller machines (ATMs)” that e-money on deposit with Safaricom through M-PESA would read unique bioidentifiers could be used to extend financial be separated from the company’s accounts and cash flow access to the vast “unbanked” population. Already, Africa (in the same way that other nonbank financial institutions, is the second-largest mobile money market, in part because such as stock brokerages, are required to isolate custom- remittances are so important in this region. But as with the ers’ assets from the assets and accounts of the company). rollout of mobile phones, regulatory policy will be as impor- The central bank kept a close watch on the system’s opera- tant as innovation in determining the outcome. tions, including sporadic checks to see that at the close of the day the amount of funds that e-money customers had stored with Safaricom in their phone accounts was equal to Kenya’s Venture into Mobile Money the amount of funds that Safaricom had placed in its trust Since its launch in Kenya in 2007, M-PESA, a mobile phone– account at its partner bank. After three years of this “test- based banking system run by Safaricom, has dramatically and-learn” supervision, the central bank began work on a changed the financial services landscape. M-PESA allows set of e-money and e-payment regulations. registered users to store money in their account by buying phone credits, to transfer money to or receive money from India’s Solution to the High Cost of Verifying other registered users through their mobile phone, as well as Bank Clients to transform their credit balance into cash at any authorized dealer. By building on its existing cell phone record-keeping For low-income populations, a major obstacle to interacting infrastructure and agent network, Safaricom has been able with the formal banking system is the lack of verifiable iden- to offer basic banking services at much cheaper prices than tification. Banking systems need to be able to identify their a branch bank (about 50 percent cheaper, depending on the customers to ensure that financial transactions are secure; service and size of transaction). It has doubled the share of in addition, they have to comply with supervisory reporting the population with access to a secure place to store cash, requirements on transactions, some of which are set inter- and, by operating as a debit card (or electronic wallet), it has nationally as part of global anticorruption and antiterrorism reduced users’ vulnerability to robbery. regulations. Low-income countries are less likely to have reli- M-PESA differs from a bank in two important ways. It able national identification systems, and these databases are does not offer loans (although M-PESA customers can use unlikely to cover lower-income populations and people in the system to link with banks, receive loaned funds, and rural areas. Faced with the high cost of verifying the identity make the required loan payments), and it does not pay inter- of such clients, banks often choose not to serve them at all. est on its accounts. Given that the interest earned on small The Government of India is finding a way around this savings accounts in commercial banks is usually eaten up by problem through the Unique National Identification project, account fees, small savers are better off using M-PESA, even in which identification is based on biometrics rather than without the interest. In 2012, an analysis concluded that standard approaches such as birth certificates and postal M-PESA’s branchless banking system had brought banking addresses. At a very small cost per participant, more than to previously unserved low-income populations and had 300 million people registered in the first two years. Banks “flattened the geographical constraints to access” across have responded enthusiastically to this system, investing in Kenya (King 2012, 28). Despite being a low-income coun- basic biometric readers, which are linked to the database try, Kenya has one of the highest levels of financial access and can instantly identify and verify account holders. Agents in Africa, with mobile money transactions equivalent to 60 in rural areas have been equipped with “micro ATMs”— percent of Kenya’s gross domestic product (GDP) in 2012 point-of-sale devices—that read the client’s biometrics (usu- (Blycroft Ltd. 2012). ally fingerprints) and record the transactions in the bank At least as important as Safaricom’s product innovation database, reducing the need for customers to travel to and successful implementation was the stance of the bank bank branches. Government agencies are using this system regulator, the central bank of Kenya. When M-PESA was to disburse cash transfers to households through low-cost launched, Kenya had no law regulating nonbank e-money bank accounts, and citizens can also pay school fees and (continued) focus note 3 187 Box F3.2 (continued) utility bills through this system. Initially envisaged as a way ing the way for dramatically increased access by providing to simplify administrative procedures and reduce the costs financial services at a much lower cost. of service delivery (including leakages), the system is set to transform the financial services landscape in India, pav- Sources: King 2012; Lauer and Tarazi 2012. Regulation to Increase Financial finance expansion and mobile banking as well Inclusion and Encourage the Use of as programs to support the growth of afford- New Technology able credit to nonfarm businesses in rural areas. As figure F3.2 shows, even at the same income Mongolia has used a dynamic mobile network level, national policies for the financial sec- system and a G2P payment system to broaden tor can produce very different results. Coun- access. Given the rapid evolution of bank- tries such as Kenya, Mauritius, and Rwanda in ing technology and its potential to change the Africa and China, India, Mongolia, and Thai- financial landscape, countries need to develop a land in Asia have achieved much higher finan- regulatory structure that encourages the spread cial access than other countries at similar levels of the technology, while protecting consumers of income. And given income levels, Africa is through interoperability and other regulations not doing badly. Among poor countries (with (see Dias and McKee 2010; Lauer and Tarazi GDP per capita under US$1,000), the major- 2012; World Bank 2013). ity of African countries for which we have data In the CFA system in Western and Central are actually above the trend line of the rest of Africa, regulation has restrained innovation in the world. The exception is countries in the the financial sector. Regional central banks are CFA (Communauté Financière d’Afrique) involved in financial regulation, and changes to zone, which tend to lag the rest of the world in the system require the agreement of all partici- extending bank accounts to households. pants. This arrangement has kept competition Countries with high household access to in the formal banking sector low, and house- finance tend to have one or more of the follow- hold financial inclusion well below the trend ing features: (1) a competitive banking sector, for the rest of Africa and the world. It has ben- which encourages banks to look for customers; efited insiders at the expense of new entrants (2) a vibrant microfinance sector, which tends (World Bank 2013). The penetration of MFIs to specialize in products tailored to the needs is low, and regulations to facilitate their growth of lower-income households; (3) proportion- were not adopted by all countries of the West ate supervision, including a “test-and-learn” African Economic and Monetary Union until approach, which encourages innovation; and 2012 (Riquet and Mbenge 2013). Mobile car- (4) a national strategy. The national strategy riers such as Safaricom (Kenya) cannot start encompasses the first three characteristics. It mobile banking initiatives—only banks have can also include policies to encourage banks to the right to do so, and they have no incentives to offer products that are appropriately designed serve low-income customers. Ideally, a regional and priced for low-income people, along with regulator could spur the development of a large policies that require all government payments market for mobile banking by developing har- to individuals (G2P payments) to be elec- monized cross-border payment systems, but tronic, which encourages financial inclusion. that has yet to happen. There are signs that the For example, Kenya has both a vibrant micro- regulatory environment is adapting, however. finance sector and a policy environment that It is hoped that the new MFI regulations will encourages innovation (box F3.2). Rwanda’s help to create a trustworthy system interested national strategy to encourage financial inclu- in serving the currently unbanked population. sion is reflected in policies to encourage micro- A new financial inclusion strategy, which will 188 Youth EmploYmEnt in Sub-Saharan africa Figure F3.2 Percentage of individuals with an account at a formal financial institution 100 a. Ages 15–24 90 80 MNG MUS 70 CHN 60 THA Percent 50 KEN 40 RSA 30 IND GHA AGO ZWE SWZ 20 UGA NGA BWA RWA 10 BFA COM GAB SEN EGY COG 0 NER MLI TKM 350 3,500 100 b. Ages 25 and over 90 80 MUS MNG 70 THA CHN 60 ZWE RSA Percent 50 AGO 40 RWA KEN IND EGY NGA BWA 30 UGA SWZ GHA COM BFA GAB 20 10 COG SEN NER MLI TKM 0 350 3,500 GDP per capita 2011 (current US$) CFA countries Non-CFA Sub-Saharan African countries Rest of the world Trend of CFA countries Trend of non-CFA Sub-Saharan African countries Trend of rest of the world Source: Based on FINDEX data. Note: The x axis of this figure is on a log scale. require all financial service providers to offer a will remain high. There are fixed staff costs in basic bank account with a package of free ser- loan appraisal. Because most loans are for such vices, is scheduled to be adopted in 2013. Thus small amounts, even the best MFIs report that the prospects for improved household financial they must earn an average of 19 percent a year access are high in this region. on their loan portfolios just to cover operating costs of US$114 per loan (see MIX 2008). Large nonprofit MFIs such as BRAC and Grameen Bank in South Asia cover part of these costs Formal Credit Services: through donations, but the commercially ori- Challenges in Expanding Options ented MFIs that are more common in Africa, for Smallholders, Household such as Ecobank and Equity Bank, have to Enterprises, and Young People charge high interest rates to cover these costs. For banks, lending to smallholders is risky. Even if retail banking costs fall and savings Agricultural cycles are unpredictable. The cli- accounts become cheaper, the costs of lending ents are in remote locations, are often unfa- to smallholders, HE owners, and young people miliar with financial services, and have few focus note 3 189 assets to use as collateral. Young smallholders Lending to HEs has similar problems. HEs “We form groups can offer few assurances to formal lenders, but face high risks of doing business (ranging from to borrow money they are likely to need capital to acquire the fluctuating demand to seizure of goods by to engage in agricultural technologies, land, and equipment authorities), and they are vulnerable to shocks that will allow them to pursue more produc- and demands for cash within the household petty trade, tive and often more commercial agricultural (sickness, for example) and community (funer- pay school fees, livelihoods. Chapter 4 discusses constraints als). HEs may be unable to generate the rates and to build or on financial services in agriculture and details of return required to pay back a loan. Lend- improve on our several ways in which banks and nongovern- ers have to assess these risks and price loans houses.” mental organizations (NGOs) are innovating accordingly. HEs that can generate the returns Uganda to provide them. Financial services alone may to pay the loan back may not maintain a busi- not be a strong springboard for smallholders ness account, however, and thus may not be to improve agricultural productivity and com- able to present themselves as creditworthy, even mercial activity. Some approaches combine the to financial institutions oriented toward serv- provision of financial services with the provi- ing low-income households. sion of agricultural information (planting and Microfinance programs use group liabil- marketing advice, for example; see the discus- ity lending to ameliorate some of the risk and sion of BASIX in chapter 4). Other approaches high cost of lending to people with very limited recognize that integrating smallholders with assets. Because the group guarantees the credit agribusiness—domestic supermarkets, agro- of the individual members, it acts as a support processors, or (further along the supply chain) mechanism for the farmer or enterprise owner exporters—may reduce some of the risk in and as a risk reduction mechanism for the providing financial services to smallholders. lender. The lending methodologies applied by Kenya’s DrumNet Project is one example of a leading MFIs have succeeded in holding their supply-chain approach to agricultural lending portfolio at risk (more than 30 days) below an (box F3.3). average of 3 percent (MIX 2008). Box F3.3 Linking smallholders to supply chains to improve their access to financial services DrumNet (a project of PRIDE AFRICA, a microfinance non- the input credit and served as security for the loans. After profit) sought to improve Kenyan smallholders’ access to signing a fixed-price contract to supply sunflower seed to financial services through a pilot program that integrated Bidco, farmers received credit for inputs from Equity Bank. them into the supply chain for oilseeds. The pilot, which Farmers obtained no cash from Equity Bank; instead, the lasted from 2007 to 2009, involved Equity Bank and Bidco, bank paid input suppliers directly for inputs that farmers pur- a large manufacturer of vegetable oils, fats, margarines, chased with the credit. When farmers delivered their produce and protein concentrates in East Africa. Bidco required an to Bidco, Bidco paid them through DrumNet, which deducted assured supply of sunflower seed. Farmers were recruited the cost of the loan and transferred it to Equity Bank. The to grow sunflowers instead of their typical crop. The other remainder was deposited in the farmer’s account with Equity partners were input suppliers and AgriTrade (which recruited Bank. DrumNet facilitated the financial transactions and com- farmers and managed sunflower production, harvest, and munication via mobile phones, text messages, and e-mail. collection). DrumNet brought the parties together, negoti- More than 2,000 smallholders participated in the sun- ated the contracts on which their collaboration was based, flower pilot. The arrangement between farmers, buyers, and managed the flow of information and financial transac- banks, and retailers, although complex, brought smallhold- tions among them, earning revenue for those services. ers into the formal financial system, integrated them into a Farmer groups (usually about 20–100 farmers in the same supply chain for a commercial crop, and improved efficiency area) opened an account with Equity Bank, through which all throughout the oilseed supply chain. payments were made. Each farmer contributed to a Transac- tion Insurance Fund, which was 25 percent of the value of Source: World Bank 2011. 190 Youth EmploYmEnt in Sub-Saharan africa Box F3.4 Shifting from group to individual liability lending as a positive microfinance innovation: Evidence from the Philippines A central decision in designing credit programs for poor bers. The finding that monitoring decreased but repayment people is whether to rely on group or individual liability. was stable suggests that the screening process changed Individual liability places more responsibility on the lending without worsening repayment. One possibility is that the institution, so borrowers prefer it (Karlan and Appel 2011). removal of group liability may have induced credit officers to Although individual liability lending is becoming more com- increase screening and monitoring, although their workload mon in microfinance, little empirical research has compared did not increase. the two approaches. Between 2004 and 2007, two random- The second trial offered new borrowers group liability in ized controlled trials with Rural Green Bank of Caraga, an some areas and individual liability in others. No statistically or MFI in the Philippines, examined how loan monitoring and economically significant difference was found in repayment repayment differed under group liability and individual liabil- rates. The move toward individual liability lending appears to ity lending. All of the loans (roughly US$18–US$90) went to have distributed the burden for screening, monitoring, and female owners of HEs to expand their businesses. enforcement more equitably between clients and credit offi- The first trial converted group liability lending programs cers. The level of institutional enforcement seen in the two to individual liability, while retaining the group structure. No trials was sufficient for the lender to recover loans without increase in defaults or loss of profits was found for centers group liability and, in some cases, to attract and retain more that switched. Individual liability centers gained more clients clients. Although the experiment was limited to one MFI in (new clients were more likely to remain in the program), one setting, the results imply that other microfinance innova- but because clients’ loans were smaller, the individual liabil- tors that are pursuing strategies for individual (and more flex- ity centers disbursed about the same amount as the group ible) lending to the poor may be moving in the right direction. liability centers. Groups that switched to individual liability spent less time monitoring repayment by other group mem- Source: Giné and Karlan 2013. Yet microfinance groups can require sub- (especially for farmers) or business equipment stantial time and financial guarantees from as collateral, also have downsides. Smallhold- participants. The time required to participate ers and HE owners could lose critical assets just in the group may deter smallholders or owner- when they need them most. MFIs using these managers of an informal enterprise from using strategies in Africa are more selective; they pre- MFI services. If the group is not formed prop- fer to lend to households where a wage earner erly and someone defaults, the downside shock can act as a cosigner for the loan. These require- to smallholders and HE owners, whose margins ments help to explain why HE owners and are already low and who have their own loans farmers, who constantly face liquidity prob- to pay off, is high (box F3.4). The combination lems because of low and uncertain cash flow, of group liability with low flexibility on the part rarely borrow from banks or MFIs and remain of the lender excludes borrowers who want within their informal financial networks. to finance risky projects or investments that will take some time to pay off, creating a bias Informal Services to Fill the Savings toward small, safe projects. In short, microfi- and Credit Gaps nance group lending methodologies that help Although their scope is almost always small, to solve the problems of high cost and risk to informal financial services have sprung up in the financial institutions effectively impose part Africa to fill the gap in savings and credit ser- of these costs on the clients; therefore, they are vices for households. These various savings and suited only for particular types of customers. credit services have a long history and have Alternative strategies, such as lending to developed rules for participation that ensure individuals on more flexible terms but requir- success (box F3.5). They meet a need that can- ing them to pledge key assets such as land not be met through more formal financing focus note 3 191 arrangements, owing to the small amounts and among adults over 25; almost half of adults in numerous small transactions involved, as well lower-middle-income countries who save use as the lack of information on the participants’ these groups. These groups also help young creditworthiness. Informal services are a valu- people to save, and they may explain why the able bridge to financial literacy, a widespread youth savings rate in Africa is the highest in the savings culture, and, eventually, formal finan- world. cial inclusion and more productive enterprises, A key tenet of savings groups that also pro- as African countries continue to develop their vide credit is that members must save before formal financial sectors. they can obtain other services. This practice As seen in figure F3.1, savers in Africa use screens out borrowers who are less likely to formal financial services more than savers any- repay loans. Members in effect pledge their where else in the world, especially in Africa’s savings deposit as partial collateral against lower-middle-income countries, where the the loan. To ensure that members save, some overwhelming majority of households no lon- groups employ a member or other person to ger suffer from extreme poverty. The use of collect funds daily or weekly from members informal savings mechanisms is more common at their homes (a susu collector). Informal Box F3.5 ROSCAs, VSLAs, SHGs, and SACCOs: Examples of informal savings and credit systems Rotating savings and credit associations (ROSCAs) form a cash box with multiple locks. The key holders are not part when individuals agree to save together. They have evolved of the management committee. The group determines the throughout the world—witness West Africa’s tontines, services offered (savings, lending) and corresponding terms Nepal’s dhikuti, and Indonesia’s arisan. Members meet regu- and conditions. At regular meetings, members purchase one larly, and all deposit the same amount of money into a com- to five savings shares; the share price is determined by the mon pot at every meeting. At each meeting, on a rotating group and remains unchanged throughout the year. These basis, one member gets the whole pot. ROSCAs have advan- savings capitalize a loan fund for members, who can borrow tages. They require no bookkeeping (an asset where liter- amounts not exceeding three times their savings. Loans are acy is limited), everyone witnesses the transactions at every given to individuals or groups, for a term not exceeding the meeting, and the sums involved are small. Their accessibility groups’ end point, at a monthly interest rate of 5–10 percent and simplicity make ROSCAs an important source of finan- (though rates as low as 1 percent or as high as 20 percent cial services even where specialized MFIs operate. ROSCAs are reported). Record keeping takes three forms: memoriza- empower their members (the majority of whom are women) tion, passbooks that record only the ending fund balance, and help to build social capital in communities. The struc- or central ledgers that track financial activity. When a cycle ture can be useful in operating informal social protection ends, group members share the savings and corresponding schemes as well, such as burial societies. Yet these associa- interest. Returns can range from 35 to 50 percent; after only tions also have drawbacks. Members cannot always access a few years, a group might manage US$2,000–US$10,000. savings when they want them, and the timing of required Through small periodic contributions, groups may also main- contributions and payouts may not match their cash flow tain a social fund available to the entire community. Mem- needs. The approach favors people with steady incomes who bers may leave the group at any time, under terms decided can contribute consistently. by the group. As a result, unlike ROSCAs, VSLAs intermedi- Village savings and loan associations (VSLAs), also called ate funds between savers and investors within a group over accumulating savings and credit associations (ASCAs), were a short period of time. pioneered in the early 1990s by CARE International, which Aside from CARE, numerous organizations have pro- developed a standard model for VSLAs based on Niger’s ton- moted VSLAs, including Aga Khan Foundation, Catholic tines. Generally 15–25 individuals agree to join forces for sav- Relief Services, Oxfam/Freedom from Hunger, Pact-WORTH, ing and borrowing during a fixed period (usually one year). PLAN, the United States Peace Corps, World Relief, and The group elects a management committee and money World Vision. VSLAs have reached at least 61 countries counters. No one else handles the money, which is stored in in Africa, Asia, and Latin America and have more than (continued) 192 Youth EmploYmEnt in Sub-Saharan africa savings and credit groups limit banking costs limiting losses (little time passes before it is through their own outreach, governance, and clear whether a member can honor an obliga- accounting, and they can recruit staff locally tion). The loans are useful for covering lumpy at relatively low salaries. For this reason, they household expenditures such as school fees or can charge lower interest rates than MFIs. the expenses related to a festival, but they are Larger groups with access to a local bank may not suited to an investment in a farm or in HE opt to protect funds there at an affordable cost equipment, which may require a longer repay- (account charges are shared across members, ment period. Everyone in the group pays the along with any interest earned on the deposit). same interest on loans—even risky loans— These informal services have some disad- which appropriately biases risk downward. vantages. Their local nature and structure limit Even the larger and more sophisticated savings intermediation. They depend on members for and credit cooperatives (SACCOs) may strug- funds and can lend only the money that is avail- gle with profitability, as their membership base able in the group. The lending terms are short, may never be large enough for them to spread rarely exceeding one year and usually lasting out the basic overhead costs of facilities, man- only a few months. This brief time horizon has agement, and security and achieve lower unit the benefit of allowing close supervision and costs than other financial institutions, espe- Box F3.5 (continued) 6 million active participants. The organizations promoting they may belong to the same workplace, community, or the savings groups train the members in group operation religious group). A volunteer board of directors is elected and governance, but because these groups are essentially from among the membership; each member has one vote. self-managed, the risk that funds may be lost through fraud, SACCOs finance their loan portfolios by pooling members’ theft, or borrower default remains. Larger groups in or near voluntary savings rather than seeking outside capital. Ide- urban areas may safeguard their funds in a bank, and CARE ally, members earn higher returns on savings, pay lower is testing the use of mobile banking technology to store and interest on loans, and generally pay fewer fees. SACCOs’ withdraw group funds in VSLAs in East Africa. local nature can prevent them from expanding and reducing Self-help groups (SHGs) are small village groups of 10–20 their unit costs, especially in rural areas. Their democratic women who pool their savings over a few months until they nature requires members to balance borrowers’ preference have sufficient capital to lend to group members or to oth- for low interest rates against the high returns sought by ers in their village. SHGs link with banks and form federa- shareholders, who mainly save. Finally, because SACCOs tions with other villages, allowing them to accumulate more operate far more like banks than cooperatives, their gover- capital for lending. The interest is not distributed back to nance and regulation urgently demand attention. Although members; it is left to grow. Used widely in India, SHGs have their assets form a very small share of the banking system’s potential in Africa, but efforts to establish them have had assets, SACCOs serve a very large (and relatively poor) pop- mixed success. In India they rely on strong social dynam- ulation. A system to guarantee SACCOs’ probity, stability, ics among women within villages and social connections and accountability may not necessarily be costly. It may suf- between villages to catalyze federations. The social struc- fice to institute sound governance and internal controls to tures in African villages are not as conducive to developing protect members’ deposits and limit SACCOs’ exposure to strong women’s groups, and women in Africa are less able risk, but much more research is needed to determine what than women in India to devote the time required to attend works best. meetings, in part because the lower density of settlement requires them to travel longer distances. Sources: Allen and Panetta 2010; Bakiene et al. 2012; Collins et al. 2009; IFPRI and World Bank 2010, brief 3; VSL Associates, “About Us: Savings and credit cooperatives (SACCOs), or credit VSL Model,” http://www.vsla.net/aboutus/vslmodel; “Rotating Savings unions, are member owned, not-for-profit financial coop- and Credit Association,” http://en.wikipedia.org/wiki/Rotating_Savings_ eratives providing savings, credit, remittances, and other and_Credit_Association; “What Is a Credit Union?” http://www.woccu. services to members connected in some way (for example, org/about/creditunion. focus note 3 193 cially in sparsely populated rural areas (5,000 district capitals, presumably in the hands of “Group clients is often considered a minimum efficient the officials there. Only the NGO-organized management of size for MFIs). The agreed rates on loans are VSLAs effectively provided capital to HE own- loans is very bad. typically 3.0–3.5 percent a month for SAC- ers (Kweka and Fox 2011). COs in Uganda, which may still exceed what An especially damaging practice is for gov- One lady in our many smallholders and HE owners can afford ernments or NGOs to use an informal savings group borrowed while still being able to profit from a loan for group to inject capital into the community. T Sh 500,000 an enterprise (Bakiene et al. 2012; Collins et al. Studies show that this practice reduces the and then ran 2009). groups’ incentives to build up their own savings away. I had pool and places the sustainability of the initia- borrowed only The Role of Government: Increasing tive at risk. For this reason, NGOs seem better suited than governments to the task of devel- T Sh 100,000. Access and Protecting Consumers Normally, informal finance exists outside the oping and supporting informal savings groups. How come the scope of government regulation and support, one who had but government, NGO, and donor programs The Role of Informal Financial borrowed have supported the spread of SACCOs and Institutions: Expanding Financial T Sh 100,000 VSLAs through outreach, training, and the Inclusion had to pay for formation of national SACCO member asso- VSLAs and SHGs both hold potential for someone who ciations. Program staff train group members including young people and addressing their on procedures to keep the funds safe and limit capital constraints, particularly if the group had borrowed the loss of funds through theft, fraud, or bad offers mentoring and access to information as T Sh 500,000?” loans, but according to all accounts, reliability well as finance. In Uganda, an innovative strat- Tanzania remains a problem.6 egy is being developed to work through NGOs (US$1 = To protect consumers, governments are to establish VSLAs and promote the inclusion T Sh 2,500) looking for ways to supervise the larger groups, of young people. A donor (the International such as SACCOs, but supervising a large num- Fund for Agricultural Development) will pro- ber of small groups is rarely feasible.7 Gov- vide funds for the Ministry of Finance and Eco- ernment involvement in SACCOs and VSLAs nomic Development to contract with NGOs to can also backfire. Informal groups operate develop and nurture VSLAs. The proposed tar- on trust, supported by a set of procedures to get is to establish 15,000 new VSLAs over five ensure transparency. They keep costs down for years. At least 15 percent of the new members their members because they operate efficiently are expected to be young people, either through and do not have to spend time complying with the formation of youth savings groups or the regulators’ requests for information. For larger incorporation of young people into VSLAs. SACCOs, in which members find it more chal- Project funds will support the staff and materi- lenging to exert proper supervision themselves als to train local groups; no funds will be pro- and the sums involved are substantial enough vided as paid-in capital to the groups.8 to undermine the SACCO program, some sort In addition to providing informal financial of public supervision may be justified. services, savings groups support their members In general, however, governments should in the same way as lending circles in MFIs— keep their involvement in informal savings they provide encouragement and a forum for groups to a minimum. The public tends to members to share experiences. Numerous interpret government involvement in such descriptions of savings groups attest to the programs as a signal that participants can exert importance of this feature (see box F3.6 for an less financial discipline, which undermines example from Mali), which could be a valuable the approach. In Tanzania, a qualitative study means for young people to learn from others documented numerous cases in poor districts and build social capital (Banerjee and Duflo where public funds intended to provide grants 2011; Nimusiima et al. 2012). As savings groups or bring cheap credit to owners of HEs never spread to young people, group members are materialized. Instead, the funds stayed in the approaching their NGO partners for additional 194 Youth EmploYmEnt in Sub-Saharan africa Box F3.6 Village women in Mali: Achieving food and financial security through savings and credit groups Saving for Change (SfC), a village savings and loan program located within the RCT zone in Ségou; four were outside the initiated in rural Mali in 2005, places women firmly in charge RCT zone and had participated in SfC since 2005. In each of their own financial inclusion. The women served by the village in 2009 and 2012, researchers spent two to three program are largely illiterate, poor (mostly living on less days conducting focus group discussions, community inter- than US$1 a day), and living in very isolated areas. By regu- views, and key informant interviews. In the villages familiar larly saving small amounts and lending group funds short with SfC since 2005, these activities provided details on how term to members with interest, each group of about 15–25 community members had adapted the SfC model to meet women accumulates and mobilizes capital without relying local conditions, needs, and livelihood strategies over time. on matching or external funds. By July 2008, SfC had 95,000 The RCT found an overall increase in savings (by 31 per- members; later that year, the program expanded into four cent), a greater flow of credit to women in SfC villages (12 of Mali’s five nondesert regions. By April 2013, SfC served percent more women reported borrowing from the savings 18,804 groups (423,654 members). A rigorous evaluation groups), the accumulation of assets (livestock holdings grew of SfC’s impact was conducted between 2009 and 2012 to in value by 13 percent in SfC villages), and a reduction in document the program’s contribution to household liveli- hunger during periods of food insecurity. Most women cited hood strategies and savings and credit provision. a less tangible impact—the social capital generated through A randomized controlled trial (RCT)—the gold standard participation in a savings group—as one of the program’s for impact assessment—used detailed socioeconomic sur- most highly valued benefits. The groups included younger, veys of 6,000 households in 500 villages in 2009 and 2012 less socially integrated women, although they joined slightly to gather information from SfC participants and nonpar- later than the initial participants. Finally, the savings groups ticipants in treatment and control villages. In the interim, had no measurable impact on business development or researchers repeatedly surveyed a subset of 600 households expansion, perhaps because many groups are located in from treatment and control villages on their financial trans- such remote areas. actions, assets, income‐generating activities, consumption, SfC seems to have improved food and financial security and health. The resulting information provided insights into in these isolated, food-insecure, and impoverished areas. how households were evolving over the course of the study. The fact that some women surveyed in the control villages To interpret and contextualize the RCT data, qualitative formed savings and credit groups on their own, without research elicited detailed information from a small sample external advice from SfC, indicates that these groups offer of 19 villages chosen to reflect some of the variation—in perceived advantages and are well adapted to local needs. geographic location, ethnicity, livelihood strategies, and other variables—across SfC sites. Fifteen of the villages were Source: BARA and IPA 2013. services, including business education and life communities. Through such groups, national skills training. Plan International, an NGO strategies for youth could combine programs active in supporting youth savings groups in to foster savings behavior, build business and West Africa, has been meeting these requests by soft skills, support female empowerment, and contracting with local trainers to provide short expand financial inclusion, with the goal of courses on demand. Based on this experience, enabling young people to pursue sustainable they are developing a holistic training syllabus livelihoods. for their youth savings groups. The content, which is adapted to the needs and learning Notes preferences of young people, is designed to 1. In these countries, a bank account is considered support disadvantaged youths in the transition a contract; the minimum age to sign a contract to adulthood and a sustainable livelihood.9 is 18, and no provision is made for an adult Savings groups may be an effective venue cosigner. to integrate the delivery of youth programs in focus note 3 195 2. See the FINDEX website, http://go.worldbank. Dias, Denis, and Katherine McKee. 2010. “Pro- org/1F2V9ZK8C0. tecting Branchless Banking Consumers: Policy 3. Postal savings account systems were developed Objectives and Regulatory Options.” CGAP to get around this problem by making use of Focus Note 64, Consulting Group to Assist the Poor, Washington, DC. underused rural postal service staff, offices, and security systems to bring savings accounts to Dupas, Pascaline, and Jonathan Robinson. 2009. populations not served by traditional banking “Savings Constraints and Microenterprise Devel- opment: Evidence from a Field Experiment in systems. Kenya.” American Economic Journal: Applied Eco- 4. Gains are even higher when compared with nomics 5 (1): 163–92. account charges at up-market commercial Gardeva, Anita, and Elisabeth Rhyne. 2011. banks. “Opportunities and Obstacles to Financial Inclu- 5. CGAP blog, January 2012 (http://www.cgap.org/ sion: Survey Report.” Publication 12, Center for blog/looking-back-trends-branchless-banking- Financial Inclusion, Accion International, Wash- 2012). ington, DC. 6. See Collins et al. (2009), Karlan and Appel Glisovic, Jasmina, and Mayada El-Zoghbi, with (2011), and Banerjee and Duflo (2011) for dis- Sarah Forster. 2011. “Advancing Savings Services: cussions on how and why this occurs. Resource Guide for Funders.” Technical Guide, 7. This section draws on Glisovic and El-Zoghbi Consultative Group to Assist the Poor, Washing- (2011). ton, DC. 8. Based on draft project document, November Giné, Xavier, and Dean Karlan. 2013. “Group Ver- 2012 (see IFAD 2012). sus Individual Liability: Short- and Long-Term 9. See http://plan-international.org/what-we-do/ Evidence from Philippine Microcredit Lending economic-security. Groups.” World Bank, Washington, DC, June. IFAD (International Fund for Agricultural Devel- opment). 2012. “Republic of Uganda: Project References for Financial Inclusion in Rural Areas Detailed Allen, Hugh, and David Panetta. 2010. “Savings Design Report.” Report 2772-UG, IFAD, Rome. Groups: What Are They?” SEEP Network, Wash- IFPRI (International Food Policy Research Insti- ington, DC. tute) and World Bank. 2010. “Innovations in Bakiene, Amor, Louise Fox, Obert Pimhidzai, Rural and Agriculture Finance: Focus 18.” IFPRI and Elizabeth Mehta. 2012. “How Non-Farm and World Bank, Washington, DC. Enterprises Create Jobs for the Middle Class in Karlan, Dean, and Jacob Appel. 2011. More Than Uganda and How Policies Can Raise Productiv- Good Intentions: How a New Economics Is Help- ity and Reduce Risk.” Policy Research Working ing to Solve Global Poverty. New York: Dutton. Paper, World Bank, Washington, DC. Kilara, Tanaya, and Alexia Latortue. 2012. “Emerg- Banerjee, Abhijit V., and Esther Duflo. 2011. Poor ing Perspectives on Youth Savings.” CGAP Focus Economics: A Radical Rethinking of the Way to Note 82, Consultative Group to Assist the Poor, Fight Global Poverty. New York: Public Affairs. Washington, DC. BARA (Bureau of Applied Research in Anthro- King, Michael. 2012. “Is Mobile Banking Breaking pology) and IPA (Innovations for Poverty the Tyranny of Distance to Bank Infrastructure? Action). 2013. “Final Impact Evaluation of the Evidence from Kenya.” IIIS Discussion Paper 412, Saving for Change Program in Mali, 2009– Institute of International Integration Studies, 2012.” Oxfam America. http://www.oxfam Trinity College, Dublin. america.org/issues/community-finance/files/ Kweka Josephat, and Louise Fox. 2011. “The House- final-impact-evaluation-saving-for-change. hold Enterprise Sector in Tanzania: Why It Mat- ters and Who Cares.” Policy Research Working Blycroft Ltd. 2012. “African Mobile Factbook.” Paper 5882, World Bank, Washington, DC. Africa and Middle East Telecom-Week. http:// blog.bearing-consulting.com/wp-content/ Lauer, Kate, and Michael Tarazi. 2012. “Supervising uploads/2012/10/Africa.Mobile.Fact_.Book_ Nonbank E-Money Issuers.” CGAP Brief (July), .2012.pdf. Consultative Group to Assist the Poor, Washing- ton, DC. Collins, Daryl, Jonathan Morduch, Stuart Ruther- ford, and Orlanda Ruthven. 2009. Portfolios of McKay, Claudia, and Mark Pickens. 2010. “Branch- the Poor: How the World’s Poor Live on $2 a Day. less Banking 2010: Who’s Served? At What Price? Princeton, NJ: Princeton University Press. What’s Next?” CGAP Focus Note 66, Consulta- tive Group to Assist the Poor, Washington, DC. 196 Youth EmploYmEnt in Sub-Saharan africa MIX (Microfinance Information Exchange). 2008. Riquet, Corinne, and Djibril Maguette Mbenge. MFI Benchmark Data | Microbanking Bulletin 17 2013. “Deepening Financial Inclusion in West (August). http://www.themix.org/publications/ Africa.” CGAP blog post, Consultative Group to microbanking-bulletin/2008/08/mfi-benchmark- Assist the Poor, January 25, http://www.cgap.org/ data-microbanking-bulletin-august-2008-issu. blog/deepening-financial-inclusion-west-africa. Mullainathan, Sendhil, and Eldar Shafir. 2011. “Sav- USAID (United States Agency for International ings Policy and Decision-Making in Low-Income Development). 2009. “Youth Savings Account: Households.” In Insufficient Funds: Savings, A Financial Service Perspective; a Literature and Assets, Credit, and Banking among Low-Income Program Review.” MicroREPORT 163, USAID, Households, edited by Michael Barr and Rebecca Washington, DC. Blank, 121–46. New York: Russell Sage Founda- World Bank. 2011. “Module 7: Broadening Small- tion Press. holders’ Access to Financial Services through Nimusiima, Catherine, Fiona Nshemerirwe, ICTs.” In ICT in Agriculture: Connecting Small- Helen Nyamweu, and Skye Dobson, eds. 2012. holders to Knowledge, Networks, and Institutions. 10 Years of Owegatta: A History of the National Washington, DC: World Bank. Slum Dwellers Federation of Uganda (NSDFU) ———. 2013. Global Financial Development Report Narrated by Members. Kampala: Act Together 2014: Financial Inclusion. Washington, DC: Uganda. World Bank. Chapter 6 Raising Productivity in Africa’s Modern Wage Enterprises to Foster Job Growth for Youth Although small (around 16 percent of the labor Governments in Africa can do much to rem- force), the wage employment sector represents edy this situation. The most important step is to Africa’s engine for employment and growth in the improve the business climate through key reforms medium to long term, especially given the sector’s to improve access to finance and infrastructure potential to exploit economies of scale and pro- services, improve trade logistics, and ease regu- duce for export. Yet wage employment is growing latory constraints to entrepreneurship. Many unevenly across Africa. Modern manufacturing reforms are not so expensive in monetary terms firms, in particular, account for only 3 percent of and can deliver huge impacts in the short run by employment and export very little. This limited reducing distortions and increasing efficiency. competitiveness in export markets is mainly the Selective and spatially targeted support to emerg- result of low productivity in the modern wage ing clusters can promote agglomeration econo- sector. It signals the presence of government and mies. For young people to be truly productive in market failures, which vary across the subcon- modern firms, governments should foster a strong tinent but have similar effects. Complementary foundation in basic skills by improving the qual- inputs to labor (electricity, overland transport, ity of general education. In the training sector, and so on) are costly, regulations strangle business governments should focus on “public goods” such processes and the movement of goods, the high as quality assurance and information to foster a costs of financial intermediation starve investors sector that is efficient and relevant to the market of capital, and the small domestic markets and for skills. Programs for disadvantaged youth that trade barriers suppress competition and reduce integrate training with internships show promise, pressure to innovate and improve productivity. as do programs offering managerial training. 197 198 Youth EmploYmEnt in Sub-Saharan africa Alongside employment in household farms Africa’s Modern Enterprise Sector: and firms, discussed in the previous two An Overview chapters, employment in modern enterprises is becoming increasingly important to Afri- Modern nonfarm wage employment is growing can youth. If history is any guide, this trend across Africa, but inconsistently. With the excep- will only continue. As European and North tion of some countries—notably Mauritius and American countries developed, jobs in modern South Africa—nonfarm enterprises account for enterprises gradually replaced other kinds of under 20 percent of the wage employment in jobs, even in agriculture. In developed coun- Africa (figure 6.1, panel a). Despite reductions tries today, the modern wage sector domi- in public sector employment, the private mod- nates employment. In East Asia and China, the ern wage sector still provides less than 10 per- growth of jobs in modern firms is following a cent of employment in most African countries similar pattern. (figure 6.1, panel b). Even by the standards of The modern wage sector, as defined in this Asian and Latin American countries with com- report, includes small, medium, and large firms parable levels of per capita income, employ- that continuously employ five or more work- ment in Africa’s modern nonfarm enterprises ers. It also includes the public sector. Histori- is low, and it is extremely low in comparison cally wage employment grew as jobs in public with larger emerging economies such as Brazil, services and public manufacturing enterprises China, and Indonesia. expanded, but no more. As governments have The expansion of employment in modern sought to curb inefficiency and improve com- wage enterprises has been quite inconsistent petitiveness, they have rolled back employment across Africa. Labor force surveys reveal that in the public sector—in some cases drastically, while the modern wage sector has absorbed a as in Ethiopia and Ghana in the 1990s—which growing share of the labor force in Madagas- has reduced the share of public sector employ- car, Mali, Tanzania, and Uganda, its share in ees among wage and salary earners. For young total employment has changed little in Ethio- people seeking wage employment, a public sec- pia, Kenya, and Zambia and has declined in tor job may be more elusive than ever. Instead, other countries, such as Malawi and Senegal the primary challenge is to sustain high rates (figure 6.2). of investment and job creation in the private Africa’s young people seem to have no spe- sector, where productivity and competitiveness cial advantage when it comes to modern wage really matter. employment (box 6.1). Although the current This chapter closely examines employment generation has more education than its prede- in modern enterprises in manufacturing, ser- cessors, employers seem to value the experience vices, and other nonfarm sectors such as con- brought by older workers. struction because they have the greatest poten- Service jobs, both with and without a for- tial to drive productivity and employment in mal contract, dominate wage employment in the medium to long term, as Africa’s young Africa.1 The single largest share of formal wage workforce grows. Unlike household farms and employment in services consists of jobs in edu- firms, modern enterprises in these sectors are cation, health, and other social services (largely not limited by family size, so they can expand in the public sector). The next largest share to exploit economies of scale, and they are far consists of jobs in commerce (retail and whole- more likely to adopt the new technologies that sale) and transportation (figure 6.3, panel a). permit them to do so. Because modern firms Outside the service sector, manufacturing can impel higher productivity and create jobs, and construction each account for about 10 policy makers are understandably concerned percent of wage jobs. Almost half of all man- with their development. But as shown in the ufacturing employment in modern firms is discussion that follows, the circumstances must found within industries broadly classified as be right for entrepreneurship and productive food and textiles (including leather and leather firms to grow. goods) (figure 6.3, panel b). Some countries, raising productivity in africa’s modern Wage Enterprises to foster Job Growth for Youth 199 such as Cameroon and Rwanda, have sig- Figure 6.1 Africa has less nonfarm wage employment than other regions nificant employment in the wood-processing a. Private and public sector wage employment industry. Employment in high-tech industries, 70 such as machinery and electronics, is quite lim- ited, although employment in the chemicals, 60 LSO CPV plastic, glass, and paper industries is important EGY MNG Share wage/employment (%) 50 PHL LKA GTM in some countries. IDN NIC BOL 40 Modern Manufacturing Delivers 30 COM MRT VNM IND Little Employment, Few Exports MDG KEN SEND GHA COG KHM BGD 20 NGA LBR ZMB Despite their considerable untapped potential TCD RWA 10 BEN LAO for employment growth, modern manufactur- ETH BFA ing enterprises currently account for less than 0 3 percent of total employment in Africa, which 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 is a good deal lower than in other regions of the GDP per capita 2010 (current US$) world (figure 6.4). The ratio of manufacturing wage employment to total employment varies b. Private sector wage employment significantly from one country in Africa to the 70 next (and it is declining in some countries), but there is no indication that Africa as a whole is 60 Share private wage/employment (%) catching up with its comparators. MNG 50 The concern is not just that the manufac- IDN turing sector is so small, but that manufactur- 40 LSO PHL GTM ing firms export so little of what they produce. NIC BOL LKA CPV The percentage of African manufacturing firms 30 EGY exporting is among the lowest in the world; the BGD MRT IND share of African manufacturing output sold 20 KHM VNM COM RWA domestically is among the highest (figure 6.5). LBR TCD LAO COG 10 NGA The poor performance of Africa’s manufac- BEN ZMB turing sector is especially worrying because 0 ETH BFA trade-oriented manufacturing may offer the 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 best chance of quickly creating modern wage GDP per capita 2010 (current US$) jobs. African economies are small, so develop- Sub-Saharan countries Non-Sub-Saharan countries ing an export-oriented sector is more impor- tant to them than to developing countries in Source: Based on household and labor force surveys. other regions like South Asia, where domestic markets are generally much larger (Dinh et al. 2012). Small markets curb productivity and Given the potential significance of trade to growth in employment by limiting the poten- Africa and the fact that manufactured goods tial scale economies that industries can achieve. are inherently more tradable than most ser- Because a small domestic market can support vices, the analysis in this chapter focuses more only a limited number of firms, the result- on manufacturing than on services. The reason ing lack of competitive pressure puts a brake for this approach is not that manufacturing is on investment and growth.2 In the absence of the only alternative for creating productive trade, small economies are also less likely to modern wage jobs in Africa. It is not. Services benefit from agglomeration economies, simply represent an increasingly important share of because manufacturing clusters are much less global trade, and Africa can certainly pursue likely to emerge (Collier and Venables 2008). this means of expanding wage employment. 200 Youth EmploYmEnt in Sub-Saharan africa Figure 6.2 Wage employment is growing A final consideration is that manufactur- inconsistently across Africa ing has been the subject of more research than other sources of wage employment in Africa. 1994 Mali More and better data are available for the 2003 analysis, and the resulting policy lessons can be 1994 applied to firms in other sectors. Burkina Faso 2003 2000 Tanzania How Competitive Is Modern 2006 Manufacturing in Africa? 1999 Ethiopia 2005 Traditional thinking on international trade 1996 contends that the relative abundance of pro- Mozambique duction factors like labor, capital, and natural 2008 resources in a given country strongly deter- 1993 Madagascar mines which of its economic activities are 2001 internationally competitive. Given that Africa 2002 has rich natural resources but relatively poor Uganda 2005 human capital, Africa is thought to have a 1998 comparative advantage in producing pri- Zambia mary commodities for export rather than 2003 manufactured goods for export (Wood and 1997 Malawi Berge 1997; Wood and Mayer 2001). More 2005 recent thinking suggests that factor endow- 1997 ments alone do not determine trade and that Kenya 2005 Africa could potentially become competitive 1995 in manufacturing. Senegal First, aside from labor, capital, and natu- 2005 ral resources, good infrastructure and pub- 0 5 10 15 20 25 30 35 40 45 lic services also contribute to competitive- Share wage/total employment, ages 15–64 (%) ness. Utilities like electricity and water affect how much a firm can produce from a given Source: Based on household and labor force surveys. amount of labor, capital, and raw materials. Transport costs influence the competitiveness Productivity growth in nontraded sectors of a country’s exports by affecting the cost of can also create wage jobs, as demonstrated by sending the output to export markets and recent growth in Africa’s construction sector. the cost of importing inputs. Manufactur- But as East Asia’s rapidly growing economies ing, with its relatively complex supply chains, suggest, trade-oriented manufacturing may relies more heavily on infrastructure services offer the greatest potential to create wage jobs. than agriculture and extractive industries. Equally important is the positive feedback Conceivably, investment in infrastructure from manufacturing to other domestic sectors. could give rise to an internationally competi- A growing manufacturing sector that is driven tive manufacturing sector regardless of factor by trade—and hence not limited by the size of endowments. the domestic economy—will expand domestic Second, newer trade theories suggest that demand for other industries (including services comparative advantage can be acquired over and construction) by raising incomes. Manu- time, though not necessarily in a predictable facturing firms will also demand a range of way. For instance, an industrial cluster can intermediate goods and services as inputs from sometimes become internationally competi- domestic suppliers. tive over time because of agglomeration econ- raising productivity in africa’s modern Wage Enterprises to foster Job Growth for Youth 201 Box 6.1 Where do young people figure in Africa’s wage employment picture? At present, in a cross-section of African countries, the ratio Figure B6.1.1 The share of youth in wage employment of wage employment to total employment among youth is tracks the share in the general population remarkably similar to the ratio in the general population (see figure B6.1.1). The implication is that firms have no special 100 proclivity to hire youth. Youth employment in modern enter- Share of wage employment for population 90 prises simply grows in proportion to the growth of modern 80 enterprises—no more, no less. At the same time, this figure may not tell the whole story. 70 ages 15–34 (%) It may be possible to enhance the impact of growth in wage 60 enterprises on youth employment. Evidence presented in 50 chapter 1 indicates that many individuals move into wage employment only when they are 30 years and older, which 40 could reflect a preference for experienced workers. Policies 30 that facilitate young people’s transition to wage employ- ment, such as programs to support on-the-job training 20 through internships or informational interventions to con- 10 nect youths to firms needing their skills, might enable more 0 young people to fulfill their aspirations for wage employ- 0 10 20 30 40 50 60 70 80 90 100 ment sooner rather than later. The scope for pursuing such Share of wage employment for population policies and evidence of their effectiveness are discussed ages 15–64 (%) toward the end of this chapter. 45 degree line Source: Based on household and labor force surveys. Figure 6.3 Services form the largest share of nonfarm wage employment; within manufacturing, the food and textile industries dominate a. By subsector b. Within manufacturing Electricity and utilities Construction Machinery, electronics Other Manufacturing Commerce Transportation, Metals storage, and Other services communications Food Financial, Chemicals, plastic, insurance, and glass, and paper real estate Alcohol and tobacco Defense, law, Education, health, and Clothing, shoes, and order social services Carpentry and wood and textiles Source: Based on standardized and harmonized household and labor force surveys (see appendix). omies. An agglomeration economy can arise suppliers reduces input costs for the exporting when many input suppliers locate in industrial industry. Knowledge spillovers across firms clusters, and the resulting competition between located near each other constitute another 202 Youth EmploYmEnt in Sub-Saharan africa Figure 6.4 Africa’s modern wage manufacturing sector continues to have a small share in employment a. Africa b. Other regions 1994 1999 Mali India 2003 2009 1994 Burkina Faso 2000 2003 Bolivia 2008 2000 Tanzania 2000 2006 Peru 2010 1999 Ethiopia 2005 2005 Philippines 1996 2010 Mozambique 2008 2000 Guatemala 2002 2006 Uganda 2005 2001 Nicaragua 1998 2005 Zambia 2003 2002 Colombia 1997 Kenya 2010 2005 2001 1993 Brazil Madagascar 2009 2001 2000 1997 Sri Lanka Malawi 2008 2005 1995 2005 Senegal Bangladesh 2005 2010 0 3 6 9 12 15 0 3 6 9 12 15 Share of employment (%) Share of employment (%) Source: Based on household and labor force surveys. potential agglomeration economy (Fujita, Thus comparative advantage in extractive Krugman, and Venables 1999). The develop- industries alone is not an inevitable outcome ment of India’s high-tech information technol- of Africa’s resource endowments. The question ogy clusters is a case in point. then becomes whether other sectors, particu- Third, productivity at the firm level is now larly manufacturing, can become internation- seen as a key determinant of international ally competitive. trade patterns. Firms in the same industry but located in different countries can both export, provided they are productive by international Unit Labor Costs as a Measure of standards and can profitably charge competi- Manufacturing Competitiveness tive prices.3 Firms within the same industry One measure of Africa’s manufacturing com- can also vary significantly in their level of pro- petitiveness is how much an industry has to pay ductivity. The implication is that, aside from workers for producing each unit of output. The resource endowments and factors that broadly specific measure used here is unit labor costs— influence comparative advantage across sec- wages divided by labor productivity (output tors, there are other factors, perhaps equally per worker)—which are a more comprehen- important, that determine competitiveness at sive measure of competitiveness than labor the firm level. productivity or wages alone. Lower unit labor raising productivity in africa’s modern Wage Enterprises to foster Job Growth for Youth 203 Figure 6.5 African firms export relatively little costs indicate a higher degree of competitive- ness (although unit labor costs have potential Middle East and limitations as a single indicator of competitive- North Africa ness; see box 6.2). Sub-Saharan China’s experience illustrates how growth Africa in manufacturing depends on maintaining South Asia international competitiveness. Between 1980 and 2007, China’s manufacturing workforce East Asia and Pacific grew threefold, from 24 million to 72 mil- lion (figure 6.6, panel b). For a long time, World the exceptional performance of China’s man- ufacturing industries benefited from China’s Latin America and the Caribbean low wages relative to the rest of the world— Europe and including Africa. As employment started Central Asia to expand, so did wage rates, which qua- High-income drupled between the early 1980s and 2007. OECD countries But even as China was losing its labor cost 0 20 40 60 80 100 advantage, its labor productivity grew at a Percent rate that compensated for this cost increase, Firms exporting directly or indirectly leaving unit labor costs virtually unchanged. (at least 1% of sales) As shown in figure 6.6 (panel b), China’s Proportion of total sales that are unit labor costs remained below those of domestic sales other emerging economies throughout that Source: Based on World Bank enterprise surveys. period. Box 6.2 Measurement issues and other limitations of unit labor costs Unit labor costs—labor costs per worker divided by labor atively weak in Africa, then its competitiveness will be over- productivity or real (physical) output per worker—have some estimated by revenue-based measures of labor productivity. limitations as a measure of competitiveness. For instance, the To account for price effects, revenue or value added unit labor costs in a shirt-making firm are its wages divided per worker is deflated by an aggregate industry-level price by the number of shirts produced per worker. However, the index. Although this practice adjusts for changes in prices analysis in this chapter approximates labor productivity by over time, it does not adjust for differences in price mark-ups value added per worker, where value added is revenue minus across firms in the same industry. Nor does it adjust for dif- the cost of raw materials. This is standard practice when ferences in mark-ups across countries. measuring labor productivity, and it is largely unavoidable, Another limitation of unit labor costs is that a firm’s com- given that physical output is difficult to measure and to com- petitiveness depends on total costs per unit of output, not pare or aggregate across firms producing different goods. just on labor costs. If the costs of other “indirect” inputs Revenue and value added depend on both physical out- (such as water and electricity) are similar across the coun- put and prices. Using revenue and value added instead of tries being compared, then labor costs per unit of output are physical output can artificially inflate the measurement of what really matter. Evidence that African firms pay relatively productivity when firms charge higher prices than they would more for indirect inputs than firms in other developing coun- be able to charge in perfectly competitive markets (that is, a tries suggests, however, that unit labor costs overstate the price mark-up). To the extent that high value added reflects competitiveness of African firms compared to firms in other a firm’s ability to charge high prices because of low market developing countries. competition, it will overstate that firm’s competitiveness in international markets. If domestic market competition is rel- Sources: Clarke 2011; Eifert, Gelb, and Ramachandran 2008. 204 Youth EmploYmEnt in Sub-Saharan africa Figure 6.6 China’s unit labor costs have remained below those of other emerging economies a. Unit labor costs (3-year averages) b. Employment 1984–86 China China 2005–07 1995–97 Brazil Brazil 2005–07 1991–93 India India 2004–06 1987–89 Korea, Rep. Korea, Rep. 2006–08 0 0.1 0.2 0.3 0.4 0.5 0.6 0 1 2 3 4 5 6 7 8 Unit labor cost Manufacturing (workers, millions) 1990 2007 Source: UNIDO statistics. Manufacturing Is Not Competitive in from West and East Africa; and countries with Most African Countries large and small populations. Data for this sam- Comparing trends in manufacturing employ- ple are compared to data for emerging econo- ment and unit labor costs across Africa and mies, particularly China. emerging economies can help to determine Growth in manufacturing employment whether African countries are competitive or since 1990 has been inconsistent in these Afri- on their way to becoming competitive. The can countries (figure 6.7). Ethiopia and Kenya comparison with China is especially impor- saw significant growth. Although small relative tant, because China maintains the lowest unit to total employment, manufacturing employ- labor costs among large emerging econo- ment in Ethiopia increased from 82,000 in mies and has a large export presence in most 1990 to 135,000 in 2007; over the same period, industries. Potential large competitors such as manufacturing employment in Kenya grew India seem to be reaching for similarly low from 188,000 to 256,000. Cameroon, Ghana, unit labor costs. Because China’s market pen- and Malawi also experienced growth, though etration is so widespread, even African firms to a lesser extent. In contrast, Malawi, Senegal, considering exporting to nearby markets and Tanzania saw relatively little change in the must maintain competitiveness with China. size of the manufacturing workforce. Seen through this lens, the manufacturing Figure 6.8 depicts unit labor costs across the sector in most African countries is not com- sample of African countries and three impor- petitive, although some countries may have tant emerging economies (Brazil, China, and reached competitive levels of unit labor costs. India). As with manufacturing employment, As labor costs rise elsewhere in the world— trends in unit labor costs are mixed in Africa, especially in China—they may create an suggesting that some countries could become opening for African manufacturing. competitive. The countries selected for review are reason- Based on the limited data available, Ghana ably representative of Sub-Saharan Africa. They appears to have unit labor costs that are com- include low-income (Ethiopia, Kenya, Malawi, parable to those in Brazil, China, and India. and Tanzania) and middle-income (Cameroon, Ghana’s level of unit labor costs is consis- Ghana, and Senegal) countries; resource-rich tent with its growing level of manufacturing (Cameroon and Ghana) countries; countries employment. Ethiopia too has low unit labor raising productivity in africa’s modern Wage Enterprises to foster Job Growth for Youth 205 Figure 6.7 Growth in manufacturing wage Figure 6.8 Based on unit labor costs, some African employment has been inconsistent across African countries could become competitive with other countries emerging economies 1990 1991–93 Cameroon Cameroon 2007 2007–09 1990 1991–93 Ethiopia Ethiopia 2007 2005–07 1990 1980–82 Ghana Ghana 2007 2003 1990 1991–93 Kenya Kenya 2007 2000–02 1990 1991–93 Malawi Malawi 2007 2007–09 1990 1991–93 Senegal Senegal 2007 2000–02 1990 1991–93 Tanzania Tanzania 2007 2005–07 0 5 10 15 20 25 30 0 0.1 0.2 0.3 0.4 0.5 0.6 Manufacturing (workers, thousands) Unit labor cost Source: UNIDO statistics. China India Brazil Source: UNIDO statistics. Note: The vertical lines show current unit labor costs in China, India, and Brazil. costs, and the years in which manufacturing rapidly created jobs in Ethiopia coincided with a sharp decline in unit labor costs. In recent Other approaches to measuring competi- years, Ethiopia and Ghana have maintained tiveness consistently suggest that Africa is less unit labor costs almost on par with historical competitive than other developing regions. levels in China and well below levels in other For example, in the World Economic Forum’s major emerging-market economies, including Global Competitiveness Index, which defines Brazil and India. competitiveness as “the set of institutions, Consistent with their disappointing growth policies, and factors that determine the level in manufacturing employment, Senegal and of productivity of a country,” African coun- Tanzania have higher unit labor costs compared tries clearly rank the lowest among developing to the large emerging economies. Tanzanian countries (World Economic Forum 2012). In unit labor costs have declined in recent years, the World Bank’s Doing Business index, which but they are not yet lower than those in Brazil, focuses largely on the regulatory environment, China, or India. Senegal’s unit labor costs have Sub-Saharan Africa ranks the lowest as well remained high for a long time. (World Bank 2012b). The association between unit labor costs and employment trends has some exceptions. Is manufacturing creating jobs in some Unlike Ethiopia and Ghana, Kenya has seen sig- countries? nificant growth in manufacturing employment Growth in total manufacturing employment despite having higher unit labor costs than the has been disappointing in Malawi, Senegal, large emerging economies. Similarly, manufac- and Tanzania, but their wood-processing turing employment has grown in Cameroon industries (excluding furniture) have created and Malawi, even though their unit labor costs jobs. In Malawi, the furniture industry and have been relatively high in recent years. the rubber and plastics industry have also 206 Youth EmploYmEnt in Sub-Saharan africa created jobs. This information suggests that Figure 6.9 Africa does not have a uniform wage industry-specific factors, and not just national advantage over other developing regions determinants, also matter for manufacturing growth. Latin America and Some industries have done well in several the Caribbean countries, while others have fared poorly. The number of jobs in the food, beverage, and Europe and Central Asia wood-processing industries (and to a lesser extent the rubber and plastics industry) has grown substantially in Cameroon, Ethiopia, East Asia and Pacific Ghana, and Kenya. The rubber and plastics industry has generated jobs in Cameroon, Sub-Saharan Ethiopia, Ghana, Kenya, and Malawi. In con- Africa trast, the textile and leather industries have been shedding modern wage jobs in almost Middle East and every country analyzed for this report, except North Africa for Ethiopia (leather and leather goods) and Kenya (textiles). South Asia These findings do not imply that certain industries are destined to do well regardless of 0 2,000 4,000 6,000 8,000 10,000 the country context. For instance, although the Median labor cost (US$) furniture industry has recently created thou- Small firms Medium firms Large firms sands of new jobs in Ethiopia, Ghana, Kenya, and Malawi, it has contracted steadily over the Source: Based on World Bank enterprise surveys. Note: The figure shows median labor costs (per full-time past two decades in Cameroon, Senegal, and employee per year in constant U.S. dollars) across firms. Costs Tanzania. are based on surveys conducted in selected countries during 2008–10. Instead, these findings indicate that indus- tries do well when they are competitive. With some exceptions, employment trends in a given costs in small and medium-size firms are industry tend to mirror trends in its unit labor slightly lower than in East Asia and the Pacific, costs. For instance, the textile industry in Africa they are similar to those in South Asia. Median has higher unit labor costs than its counterparts labor costs in large firms in Africa are higher in China and India, and it is losing jobs. Simi- even than those in East Asia and the Pacific, larly, the leather industry has high unit labor which is significant because large firms tend costs in most African countries examined here, to be the leading exporters in export-oriented except for Ethiopia. In contrast, unit labor costs industries. in the food-processing industry, which has cre- Some African countries have a wage advan- ated thousands of jobs in several African coun- tage over China and India, but it is more than tries in recent years, compare favorably with offset by low labor productivity. Even though those in emerging economies like China and wages have been rising faster in China and India. India than anywhere in Africa over the past two decades, Africa is not closing the unit labor Low Wages Are Not Enough: cost gap with them, because labor productivity Productivity Is the Linchpin of has not been rising as fast in Africa as in China Competitiveness for Africa and India. Can African countries become competitive If Africa cannot compete with large emerg- on the basis of low labor costs alone? It seems ing economies like China or India, could it unlikely. African countries do not have a uni- compete with smaller developing countries form wage advantage over other developing on the strength of low wages? Again, it seems regions (figure 6.9). Although median labor unlikely. Wages are not lower in African coun- raising productivity in africa’s modern Wage Enterprises to foster Job Growth for Youth 207 Box 6.3 Does Africa really have a labor cost advantage? As labor costs in manufacturing continue to rise in China, a age indicators of labor market regulation do not differ sig- common observation is that other developing countries with nificantly between African countries and the comparators. lower labor costs will be poised to compete with China (Lin Another explanation is that prices are generally higher in and Monga 2011). Wages are expected to be lower in coun- African countries. This explanation is supported by a com- tries with significantly lower per capita incomes than China, parison of price levels, so it is possible that firms in Africa including African countries. Recent research using firm-level have to pay relatively high wages to compensate workers for data seems to contradict that view, however, suggesting that the relatively high cost of living. industrial labor costs in Africa are far higher than might be In turn, the high cost of living could result from the domi- expected solely on the basis of gross domestic product (GDP) nance of resource-based industries in Africa. High prices are per capita (Gelb, Meyer, and Ramachandran 2013). Labor a typical feature of resource-abundant countries, because costs per worker are nearly 80 percent higher in firms in Afri- the high income from resources raises the price of nontraded can countries than in firms in other countries at the same goods, including labor. These interactions are often cited level of GDP per capita. to explain why resource-rich countries tend to perform so These findings are partly explained by an “enclave effect”: poorly when they rely on export-led growth alone (Sachs and compared with countries at the same income levels in other Warner 2001). regions, African countries have an enclave of manufacturing Larger firms, in particular, pay higher wages in African firms that have high labor productivity and pay high wages. countries. Research based on matched employer-employee But even after adjusting for this effect, African firms face data from 10 African countries suggests that larger firms pay nearly 50 percent higher labor costs. higher wages partly because labor management is especially Higher labor costs could be explained by labor market problematic in Africa, which has a much higher supervisor to factors such as regulation and unionization, but on aver- worker ratio than elsewhere (Fafchamps and Soderbom 2006). tries than in other developing countries with World Bank’s enterprise surveys.4 To avoid similar levels of per capita income. The exact complications arising from technological dif- reasons for this difference are unknown, ferences across industries, the analysis focuses although some plausible hypotheses have been on the textile industry, which is generally ori- put forward (box 6.3). Given the difficulty of ented toward exports and therefore more com- lowering wages in dollar terms, the inescap- parable across countries than domestically able conclusion is that labor productivity is oriented industries. the linchpin of competitiveness in Africa. This The firm-level data show that labor produc- conclusion is particularly relevant for resource- tivity is particularly low in Africa, especially in rich countries, where wages are driven up by the region’s low-income countries.5 Figure 6.10 the high cost of living. illustrates this point by showing sales per worker, a measure of labor productivity, in the textile industry. For example, the annual turn- Sources of Productivity Gaps in over per employee in Tanzania’s textile indus- African Manufacturing try is just US$9,000, compared to US$25,000 in Malaysia’s garment industry. To understand how African manufacturing can Labor productivity is low in Africa’s low- become internationally competitive, it is neces- income countries partly because workers are sary to pry labor productivity open and exam- not as well equipped with fixed capital (that ine its component parts. This section examines is, plant and machinery) as their counter- components of labor productivity in a group parts elsewhere in the world, such as China, of developing countries, including many from Malaysia, Mexico, and Thailand (figure 6.11). Africa, using firm-level data collected by the The typical textile manufacturer in China, for 208 Youth EmploYmEnt in Sub-Saharan africa Figure 6.10 Labor productivity is particularly low in Figure 6.11 African workers are not as well equipped low-income African countries with fixed capital as their counterparts elsewhere in the world Sales per worker in the textile industry Fixed assets per worker in the textile industry South Africa Namibia China Mexico Malaysia Kenya China Brazil South Africa Mexico Malaysia Thailand Botswana Colombia Morocco Thailand Swaziland Swaziland Mauritania Zambia Zambia Tanzania Botswana Tanzania Morocco Angola Mauritania Namibia Colombia Guinea Nigeria Uganda Nigeria Burundi Congo, Dem. Rep. Gambia, The Rwanda Congo, Dem. Rep. Guinea-Bissau Gambia, The Rwanda Burundi Guinea 0 5 10 15 20 25 30 35 40 0 2 4 6 8 10 12 14 16 2000 US$ (thousands) 2000 US$ (thousands) Source: World Bank enterprise surveys, 2002–08. Source: World Bank enterprise surveys, 2002–08. example, equips every worker with twice as where. The difference in fixed capital between much plant and equipment as a manufacturer firms from Africa and those from other parts in Tanzania. of the world is not as great as the difference Why does African manufacturing use so lit- in output (figure 6.11). For instance, the dif- tle capital per worker? A major reason for low ference in fixed assets between Tanzanian and capital intensity in Africa is that African firms Chinese textile firms is significantly lower than do not have access to the same supply of capi- the difference in output. Namibian textile firms tal as their counterparts in China.6 Although produce significantly lower output per worker access to international capital is particularly than their Chinese counterparts but have simi- important to African manufacturing (given lar levels of fixed assets. low per capita incomes), studies of interna- Even with the same level of plant and tional capital flows suggest that Africa and machinery, it is possible that workers in one other developing regions have limited access to country cannot produce as much as those in international capital markets (see, for example, another. A range of factors other than plant and Kalemli-Ozcan, Alfaro, and Volosovych 2008).7 machinery affect labor productivity. Some of Imperfections in Africa’s domestic credit mar- these factors are internal to a given firm, such as kets, discussed later, are also likely to be wors- the quality of management. Others are external, ening their access to capital. such as the quality of infrastructure. In general, A low level of fixed capital can only partly these other factors are less concrete than plant explain why output per worker is so much and machinery and difficult to quantify. To lower in most African countries than else- distinguish them from machinery and other raising productivity in africa’s modern Wage Enterprises to foster Job Growth for Youth 209 Figure 6.12 African countries have lower productivity than other regions a. Textiles b. All manufacturing and service industries Brazil Brazil Malaysia Thailand Colombia Malaysia China Chile Mexico South Africa Thailand Morocco Morocco China South Africa Mexico Tanzania Colombia Zambia Angola Tanzania Kenya Guinea Botswana Kenya Zambia Uganda Congo, Dem. Rep. Nigeria Swaziland Namibia Nigeria Angola Guinea Gambia, The Uganda Burundi Rwanda Swaziland Gambia, The Rwanda Burundi Botswana Namibia Congo, Dem. Rep. Guinea-Bissau Guinea-Bissau 0 1 2 3 4 0 1 2 3 4 2000 US$ (thousands) Sub-Saharan Source: Based on World Bank enterprise surveys, Africa 2002–08. Non-Sub-Saharan Africa concrete inputs, these factors are often lumped What Determines Productivity at the together under the term “productivity” (as Firm Level? opposed to labor productivity).8 Although productivity is often equated with Since the individual components of pro- production technology, it should be inter- ductivity cannot be measured directly, their preted more broadly. The quality of a firm’s net effect is simply inferred as that part of total management and its organizational efficiency output which cannot be explained by the level (which includes elements such as the quality of measured inputs such as labor, plant, and of factory floor and workplace organization, machinery. Estimated as this residual, produc- workforce incentive and supervision structures, tivity is known to account for a large part of the and supply chain management) also contrib- difference in output per worker across coun- ute to productivity. Productivity is influenced tries (Hall and Jones 1999). by elements of a firm’s human and knowledge When accounting for the difference in labor capital, which cannot be captured in data on productivity between African and Chinese the formal qualifications of its workforce. For firms, productivity is found to be responsible example, firm-specific skills generated by on- for a larger share of the gap than fixed assets. As the-job learning and training for the workforce figure 6.12 shows, the shortfall in productivity can make employees more productive. Knowl- in Tanzania’s textile industry relative to China’s edge and tacit technology, such as a more effi- is slightly larger than its shortfall in fixed assets. cient production process developed through This productivity gap between African and internal research and development (R&D), are Chinese firms persists throughout a range of other firm-level components of productivity, manufacturing industries and services (figure as are computerized processes and databases 6.12, panel b). tailored to a firm’s needs. Firms often become 210 Youth EmploYmEnt in Sub-Saharan africa more productive by adapting a technology to parts of the world (Hao and Hulten 2011; Dutz their context—whether through trial and error et al. 2012). or a more formal R&D process—and such A firm’s productivity is also affected by fac- adaptation also raises productivity. tors external to the firm, such as the quality and Understanding how much these specific reliability of transport and logistics systems factors matter to productivity is very relevant and the supply of utilities, including power, for policy.9 Some studies attempt to unpack telecommunications, and water services.10 Poor productivity by examining how much firms transportation infrastructure can reduce the spend on acquiring specific types of human, efficiency of production by making the supply knowledge, or organizational capital (Corrado, of raw materials less reliable. Similarly, an inad- Hulten, and Sichel 2005; OECD 2010). For equate and unreliable utility supply can inter- instance, a new production process could be rupt production and force workers to remain the result of R&D expenditures incurred over idle. In effect, it increases the amount of labor many years. While the quality of the new pro- and capital needed to produce a given level of duction process itself is difficult to quantify, it output. is likely to be reflected in the R&D expenditure Poor infrastructure also reduces productiv- incurred to develop the process. In developed ity by forcing firms to adopt inefficient and countries, such spending on human, knowl- costly coping mechanisms. Firms may have edge, or organizational capital is at least as high to compensate for the poor quality of public as spending on plant and machinery. It is not services on their own; they might purchase as high in the two emerging economies where generators as back-ups for the public electric- attempts have been made to measure it (Bra- ity supply, for example. A recent study shows zil and China), but it is still quite sizable and that costs related to infrastructure services rising. These differences suggest that variation account for a relatively high share of firms’ in firms’ human, knowledge, and organiza- costs in poor African countries, thus impos- tional capital could account for a large part of ing an extra burden on the competitiveness of the productivity gap between Africa and other African firms (Eifert, Gelb, and Ramachandran 2008). Figure 6.13 shows firm-level indicators Figure 6.13 Africa fares poorly in firm-reported indicators of transport, water, and of the quality of transport, water, and electric- electricity supply ity supply in different parts of the world. Along with South Asia, the Middle East, and North Proportion of products lost to Africa, Sub-Saharan Africa is generally near the breakage or spoilage during shipping to domestic markets bottom of the rankings. Number of water insufficiencies How Unproductive Firms Survive and in a typical month Hurt Industry-Level Competitiveness One way to raise an industry’s competitive- ness is to make all of its firms more produc- Losses due to electrical outages (annual sales) tive. Another way is to let competition between firms run its course and ensure that only the most productive firms thrive. The latter is sig- Number of electrical outages nifgicant because differences in productivity in a typical month across firms in the same industry can be sur- prisingly large, especially in developing coun- 0 5 10 15 20 tries (Syverson 2011; Banerjee and Duflo 2005). Percent According to one study on China and India, if Sub-Saharan Africa South Asia Middle East and North Africa Latin America and the Caribbean manufacturing plants are ranked according to High-income OECD countries Europe and Central Asia their productivity, those near the top of the East Asia and Pacific rankings produce about five times more with Source: Based on World Bank enterprise surveys. the same amount of labor and capital as those raising productivity in africa’s modern Wage Enterprises to foster Job Growth for Youth 211 Figure 6.14 Decomposing textile industry productivity a. Average firm total factor productivity b. Efficiency of firms Brazil China Thailand Malaysia Colombia Brazil Malaysia Colombia Mexico Mexico Morocco Morocco South Africa Burundi China Tanzania Angola Botswana Kenya South Africa Guinea Thailand Zambia Nigeria Tanzania Congo, Dem. Rep. Congo, Dem. Rep. Swaziland Rwanda Uganda Swaziland Zambia Nigeria Kenya Botswana Angola Gambia, The Namibia Uganda Guinea-Bissau Namibia Guinea Guinea-Bissau Gambia, The Burundi Rwanda 0 1 2 3 –0.5 0 0.5 1 1.5 2 2000 US$ (thousands) log Index of allocative efficiency Sub-Saharan Source: Based on World Bank enterprise surveys, Africa 2002–08. Non-Sub-Saharan Africa near the bottom (see, for example, Hsieh and ference between countries can be decomposed Klenow 2009). A reallocation of workers from into these two components.12 the least to the most productive firm in an How much of Africa’s lower productivity industry would thus raise the productivity of is due to the lower productivity of the typical the industry fivefold. firm, and how much is due to the fact that less Indeed, an increase in competition in an productive firms are more likely to survive and industry is often seen to result in such realloca- maintain outsized market shares? Figure 6.14 tion, or shifting, of labor and other inputs from illustrates these components of productivity less to more productive firms. Conversely, mar- for the textile industry in a sample of countries, ket distortions that reduce competition tend including some in Africa. This case is typical of to protect the market share of less productive most manufacturing and service sectors, and it firms, which in turn reduces the (aggregate) shows that the greater survival and abnormally productivity of the industry. large share of unproductive firms (“allocative The productivity of an industry is the sum inefficiency”) in Africa explain a large part of total of the productivity of firms in it. Thus an the productivity gap between African econo- industry could be more productive in Country mies and emerging economies such as Brazil A than in Country B for two reasons. The first and China. is that, relative to Country A, the typical firm in For example, about one-third of the produc- Country B may be less productive.11 The sec- tivity gap between the textile industries in Tan- ond is that, compared to Country A, less pro- zania and China is explained by the fact that the ductive firms may control a larger share of the average Chinese textile firm is more productive market in Country B. The net productivity dif- than its Tanzanian counterpart. The other two- 212 Youth EmploYmEnt in Sub-Saharan africa thirds of the gap is explained by the fact that in Figure 6.15 Trading across borders is expensive and Tanzania, unlike China, unproductive firms have slow in Africa captured too much market share. A comparison of South Africa and China illustrates the sig- Ethiopia nificance of this allocative efficiency even more strongly. The average productivity of South Afri- Tanzania can textile firms is actually higher than that of their Chinese counterparts, but this advantage Zambia is completely overturned by the dominance of unproductive firms within South Africa. Sub-Saharan Africa Potentially, several market distortions could China explain why unproductive firms are more likely to survive and maintain inefficiently large mar- High-income ket shares in Africa. Credit market inefficiency OECD countries is a case in point. In the short run, a work- 0 5 10 15 20 25 30 35 40 45 50 ing capital constraint can prevent a relatively Time to export Cost to export (per productive firm from producing as much as it (days) container, in 100 US$) profitably could.13 Credit constraints can also Source: World Bank Doing Business indicators, 2012. prevent a productive young firm from invest- ing in fixed capital and growing to its optimal size. Larger, older incumbents may have easier It costs 60 percent more to ship to the United access to credit, even if they are less productive. States from Djibouti than from China and about Similarly, political favoritism—such as prefer- the same to ship to Europe, despite the much ential access to land or other inputs granted to greater distance from China. The costs of inland politically connected firms—can enable unpro- transportation from the factory to ports are also ductive firms to capture a large market share. high. A World Bank study estimates that higher Regulations can also affect allocative effi- shipping and inland transportation costs add a ciency. For example, labor market regulations 2.5 percent production cost penalty for textiles can prevent the allocation of labor to its most in Ethiopia and Zambia (Dinh et al. 2012). productive uses by hampering the movement Export competitiveness is also affected by of labor across firms.14 port and terminal handling fees, customs clear- Entry barriers (such as high costs or onerous ance and technical control fees, costs of docu- procedures for establishing a new enterprise) ment preparation and letters of credit, and the can also enable unproductive firms to sustain cost of foreign exchange. These additional trade large market shares by preventing competition costs are exceptionally high in Africa, adding a from entrants who could be more productive. 5.5 percent production cost penalty for textiles A reduction in entry barriers can reallocate in Ethiopia and Zambia, for example. resources away from unproductive incumbents and toward more productive entrants (Chari 2011). Making African Firms Competitive: Priorities for Improving the Business From the Factory to the Market: Climate and Workforce Skills How Poor Trade Logistics Hurt Africa’s Competitiveness Making modern enterprises more competitive When it is costly to transport goods from the by increasing their productivity is critical for factory to international markets, an otherwise creating more modern wage jobs that are acces- productive industry can become uncompeti- sible to Africa’s youth. Increasing productivity tive in those markets. Africa does poorly on involves a complex set of reforms and interven- this dimension of competitiveness (figure 6.15). tions. The most important measures include raising productivity in africa’s modern Wage Enterprises to foster Job Growth for Youth 213 broad interventions that foster improvement in of survey respondents regarded macroeco- the business environment by ensuring macro- nomic instability as a major obstacle to business economic and political stability, strengthening growth, compared to 80 percent of respondents infrastructure services, reducing trade barri- surveyed in 2003 (World Bank 2009). These ers, and improving access to finance and more encouraging conditions deteriorated in the localized interventions that promote competi- wake of the 2008 global recession, demonstrat- tive industrial clusters. Such measures must be ing that the structure of the Zambian economy complemented by steps to improve the human (like the economies of other resource-rich capital of young people, making them more countries in the region) is especially vulnera- employable and productive. ble to the forces of macroeconomic instability. Because the range of potential reforms and Maintaining price and exchange stability and interventions is so large, this section identifies bringing government borrowing under control the priority areas. Actions within each area are over the long term are critical to the smooth classified according to whether they are easier functioning of finance and business investment to implement but have a lower impact (typi- in such countries. cally those that seek to correct a market failure) or are much harder to implement but have a Strengthening Infrastructure Services higher impact (typically those that correct a The poor quality and inadequacy of physi- government failure). cal infrastructure in Africa is the most visible aspect of a deep, pervasive infrastructure prob- Providing a Stable Macroeconomic lem that inhibits the competitiveness of African Framework firms. The more reassuring aspect of this prob- Macroeconomic uncertainty and volatility, lem is that governments can begin to address it such as fluctuations in the cost of borrowing without building new infrastructure. Evidence abroad and changes in inflation and exchange points to government failure as the main source rates, disrupt lending and investment. The of Africa’s infrastructure problem, reflected, volatility of key macroeconomic indicators is for example, in the underpricing of electricity quite pronounced in Africa, because primary and water or the monopoly power of trucking commodities constitute such a large share companies. There is also evidence that infra- of exports and because government finances structure policies and regulations block firms’ depend so greatly on trade in those commodi- access to infrastructure services and undermine ties. Macroeconomic shocks appear to disrupt the incentives for further investment (Briceño- credit flows more frequently, with greater force, Garmendia and Foster 2010). in African countries than in countries where production is more diversified. Dealing with electricity and other public utility When the World Bank assessed Zambia’s shortages. In countries with chronic shortages business environment in 2003, it found that in infrastructure services, established businesses firms faced an extremely high cost of borrow- report that frequent outages cause significant ing, which appeared to be strongly related to losses of revenue. Such shortages particularly high rates of inflation and currency volatility at affect smaller, younger firms and manufactur- that time (World Bank 2004). When the Bank ing firms. Start-ups can wait months to be con- revisited the country with a second assessment nected to the public utility grid, a delay that is in 2007, the situation was radically different. A likely to reduce firm formation and business large debt relief program and a copper price entry rates. boom had dramatically reduced government As electricity shortages have become borrowing and helped to stabilize the currency. increasingly common, governments have Real interest rates had dropped sharply, and as sought to promote long-term solutions in the inflation rate fell to single digits, business the form of large investments in maintenance lending expanded. In 2007, less than 15 percent and in additional generating and transmission 214 Youth EmploYmEnt in Sub-Saharan africa capacity. Such solutions ignore opportunities way networks. But immediate measures to pro- for quicker wins through steps to address the mote more efficient use of the infrastructure causes of underinvestment and inefficiency that already exists can also help. Nor is govern- in the power sector. These include the delib- ment investment the only option for build- erate underpricing of electricity to subsidize ing new networks. Private participation can household consumption, the failure of poorly be expanded if policy reforms create the right managed state-owned operators to collect pay- conditions. Mozambique, where high transport ments, and the absence of a workable legal and costs originate with inefficient railway and port regulatory framework for private investment. services, is both rehabilitating and expanding The specific measures needed to resolve the network as a long-term solution. At the the problem depend on which of these causes same time, it is enacting a series of structural is most prevalent in a given country. Regula- reforms to improve operations and encourage tory agencies could revise power tariffs. State- private participation in the management of owned power companies could be privatized state-owned transport operators. Improving or reorganized on a more commercial footing rail services is the key to cutting transport costs to improve payment collection and minimize in Lesotho, where the chronically dysfunc- transmission and distribution losses. Countries tional rail sector has caused exporters to rely are often advised to reorganize the industry by exclusively on road transport, which costs three separating power generation, transmission, times as much as rail. and distribution into independent opera- tions conducted by independent enterprises. Reducing Barriers to Trade: Trade Cross-border and regional initiatives to supply Liberalization, Costs, and Logistics power are often encouraged because they entail When the domestic market for manufacturing economies of scale. Initiatives could range from and service industries is fragmented or isolated cross-border pooling of power based on exist- from regional and global markets, it is hard ing power grid connectivity among neighbor- for productivity to grow in those industries.15 ing countries to establishing a regional market Such conditions insulate incumbent firms from in electricity as a component of an integrated competition (domestic or foreign), reduce the regional energy market, as planned in Southern entry of potentially more productive firms, and Africa. limit incumbents’ incentives to innovate. A central argument for trade liberaliza- Improving transport infrastructure. Inade- tion is that it increases the competitive pres- quate and costly transport infrastructure is by sure on domestic firms and encourages them far the most important factor in Africa’s excep- to become more productive. The liberalization tionally high trade costs, the fragmentation of of import tariffs by most African countries in the domestic market for manufacturing and the 1990s and early 2000s may have consti- service industries, and the region’s isolation tuted one of the most significant policy devel- from other regional and global markets. Inad- opments in Africa’s recent economic history. equate transport infrastructure is a powerful Although there are no systematic assessments drag on productivity and economic growth, of their impact in the region, anecdotes and manifested in many ways (the port of Dar es studies of the effects of similar reforms in other Salaam is one manifestation; see box 6.4). regions suggest that in Africa they could have Problems with transport infrastructure not led to substantial and widespread productiv- only limit scale economies but also bestow ity gains by exposing local producers to greater market power on firms already in the trade by competitive pressure and reshaping the struc- removing the threat of foreign competition ture of domestic production. and preventing potentially more productive Recent studies of other developing regions firms from entering local markets. and developed economies report that trade lib- Ultimately, the solution to high transport eralization reforms of the kind that took place costs is large-scale investment in road and rail- in Africa lowered domestic prices and mark-up raising productivity in africa’s modern Wage Enterprises to foster Job Growth for Youth 215 rates, increasing competitive pressures on the gain market share from unproductive firms, by large players in the domestic economy. Those encouraging innovation and the adoption of studies also provide evidence that increased better techniques of production, and by facilitat- openness to trade generates productivity gains ing economies of scale (see, for instance, Melitz in three distinct but complementary ways: by and Trefler 2012; Krishna and Mitra 1998). allowing more productive firms to take fuller Another, newer argument for import lib- advantage of their superior productivity and eralization originates in the increasing frag- Box 6.4 The high price of inefficiency at the port of Dar es Salaam The Dar es Salaam Port—East Africa’s second-largest port US$0.39 to US$5.00 (the global price is US$0.60–US$0.80), after Mombasa—is the conduit for about 90 percent of and the ratio between the highest and lowest reported cus- Tanzanian trade, and it is also a gateway for Tanzania’s land- toms values is 152 for rice and 33 for palm oil. locked neighbors, including Burundi, the Democratic Republic Storage tariffs are structured in ways that discour- of Congo, Rwanda, Uganda, and Zambia. For Tanzania espe- age rapid clearance of merchandise from inland container cially, manufacturing, trade, and economic expansion depend depots. After the free storage period of seven days expires, on the efficient movement of goods through the port. each additional day represents a direct additional profit for The main agencies involved in the port’s operations are TPA, TICTS, and the container depots. TPA’s revenues are the Tanzania Port Authority (TPA), the landlord and service higher when TICTS is less efficient. When berths managed provider; Tanzania International Container Services (TICTS), by TICTS are full, some of the container traffic is redirected a private container stevedoring contractor; and the Surface to TPA berths, creating a situation in which the landlord of and Marine Transport Authority (SUMATRA), the multisec- the port (TPA) competes against its own service provider toral regulatory agency. (TICTS). TPA obtained an estimated US$36.5 million in this The port is not efficient by international or East African manner in 2011. standards. In mid-2012, container vessels waited an aver- Such arrangements reduce incentives among port opera- age of ten days for a berth in Dar es Salaam, compared to tors to invest in increased capacity. While a small, well- less than one day in Mombasa. Merchandise took ten days connected coterie benefits from the status quo, Tanzanian on average to clear and exit the port, compared to three to workers, firms, consumers, and the government bear the four days in Mombasa (and 48 hours in many East Asian costs. An uncompetitive manufacturing sector creates fewer ports). Official and unofficial fees are numerous, high, and jobs for workers and produces more expensive goods for inconsistently applied. The additional costs incurred by ship- consumers. Tanzanian agriculture suffers because port ineffi- pers and shipping companies in Dar es Salaam compared to ciency adds an estimated 5.2 percent to the cost of imported Mombasa are equivalent to an additional tariff of 22 percent fertilizer. Policy makers may well ask whether making a single on container imports and 5 percent on bulk imports. Com- large investment to improve efficiency at the port would be pared to the port of Mombasa, inefficiencies at the port of better than continuing to subsidize fertilizer year after year. Dar es Salaam cost Tanzania and its neighbors an estimated The cost of inaction is mounting. The port of Dar es US$2.5 billion a year. Reforms have been initiated but have Salaam will decline in importance as ports and railways in progressed very slowly. neighboring countries become operational and prove more Merchandise fails to clear the port rapidly because pro- efficient. The authorities have been moving to improve port cesses (especially customs clearance) are slow and storage operations, yet more pointed reform is needed, beyond periods are quite long. The rules are not transparent and actions to improve infrastructure (such as building new oversight is poor, creating opportunities for corruption. berths). Efficiency-enhancing reforms would ensure that In Dar es Salaam, where the port charges fees in propor- end users are aware of the costs related to the port’s inef- tion to the value of the merchandise, official port fees are ficiency and participate in decisions related to port reform, 74 percent higher than those in Mombasa, where the port strengthen competition among port operators, and reduce charges flat fees. Customs valuations of goods imported via corruption through streamlined and transparent procedures Dar es Salaam also vary more widely than can be explained and improved oversight. by normal variations in quality and price. For example, the stated customs values for 1 kilogram of fertilizer ranges from Source: World Bank 2013b. 216 Youth EmploYmEnt in Sub-Saharan africa mentation of export value chains. The differ- significantly reduce trade costs. Many coun- ent stages involved in producing a particular tries could considerably reduce trade costs by final good are now often performed in many simplifying customs clearance and import pro- different countries, and a country can special- cedures and increasing the use of inland clear- ize in specific “tasks” in this value chain. For an ance facilities to shorten processing times. African firm to succeed in exporting a manu- Land border crossings remain a major facturing task, it must be able to import all the obstacle to regional integration in Africa. These complementary upstream tasks as easily as any obstacles are common along the gateway corri- international competitor that specialize in the dors serving the landlocked countries, and they same task (Collier and Venables 2007). Import also hinder regional trade and international liberalization and other measures to improve transit. Aside from improvements in transport access to imported inputs can help African infrastructure, improved management of bor- firms to insert themselves in international value der crossing through institutional reform and chains. increasing border coordination could also have Has Africa experienced benefits similar to a major impact on border-crossing times. East those attributed to tariff reductions in other Africa’s one-stop border post (OSBP) initiative regions? Lower tariff barriers should have is a step in this direction. A pilot OSBP between made African firms more productive by open- Kenya and Uganda seems to have produced ing them up to competition and enabling their dramatic results even in the absence of infra- integration within global value chains. Africa’s structure refurbishment (see box 6.5; Fitzmau- experience with the Multifiber Agreement sug- rice and Hartmann 2013). gests that such outcomes did not occur. During the final years of the Multifiber Agreement, the Improving Access to Finance United States imposed strict import quotas on The lack of finance is another serious con- Chinese apparel and awarded duty- and quota- straint on the growth of manufacturing and free access to African apparel. If African firms tradable services in Africa (World Bank 2008, were truly competitive, they should have taken 2012c; IFC 2013; see also Dinh, Mavridis, and advantage of such preferential access to the U.S. Nguyen 2012; Harrison, Lin, and Xu 2013; Li, market. African apparel exports did rise, but Mengistae, and Xu 2011). Firm surveys suggest the rise was temporary and has been attributed that the cause is poor access to formal bank to Chinese firms that shifted their final assem- financing (figure 6.16). A survey comparing bly steps to Africa to avoid the quota (Rotunno, firms in China to firms in Ethiopia, Tanzania, Vézina, and Wang 2012). and Zambia found that the most visible advan- The productivity benefits of tariff liberal- tage enjoyed by Chinese firms was access to ization may have failed to materialize in Africa bank finance at favorable conditions, such as because the costs of trade remain extremely low interest rates and low collateral require- high for reasons unrelated to tariffs, such as ments (Fafchamps and Quinn 2012). Expen- the region’s onerous transport costs. Reducing sive or limited financing suppresses growth such costs could enable African firms to real- in productivity by forcing firms to operate at ize the gains of trade liberalization (Djankov, a suboptimal size or to use substandard tech- Freund, and Pham 2010). nologies. Research suggests that many firms Freight costs are often the largest compo- cannot take advantage of profitable opportu- nent of trade costs in most countries, imply- nities for expansion because they cannot obtain ing that investment in transport infrastructure financing.16 and ports could significantly ease constraints to Access to finance varies by type of economic regional and global trade. Such investment is activity and is generally more constrained for expensive and will take time. In the short run, smaller and younger firms. In Namibia, for tackling nontariff barriers such as inefficient example, banks require small firms (employ- customs administration and high regulatory ing fewer than 30 workers) to provide twice costs of cross-border transactions would also the amount of collateral on average as mid-size raising productivity in africa’s modern Wage Enterprises to foster Job Growth for Youth 217 Box 6.5 Improving land transportation through increased international cooperation and comprehensive procedural reforms: The Malaba border-crossing pilot The long-standing response to the chronically slow move- was included in a border-crossing survey commissioned by ment of goods within Africa has been to build better roads the northern corridor authority. and border facilities, but better physical infrastructure alone The survey found that the border crossing at Malaba has not solved the problem. Twenty-five years ago, only 20 improved dramatically after the reforms. The average time percent of Africa’s main road network was considered to be to cross the border dropped from 24 to 4 hours. The SSATP in “good” condition. Today, almost half of it is considered study estimated that the reforms may have saved up to good. Border facilities have been remodeled. But a lot of US$70 million a year. Notably, these results occurred in the time is still lost at the border, largely due to procedural delays absence of infrastructure refurbishment, which is expected at border crossings. to take place at a later stage. What was so special about the A more recent approach to resolving border delays is Malaba border-crossing reforms? the development of one-stop border posts. OSBPs include The primary insight from this case is that simply chang- more than refurbished facilities; they involve improved coor- ing the procedures is probably not sufficient to improve bor- dination between border agencies from neighboring coun- der crossings. Before launching the reforms, the authorities tries and better coordination among the domestic agencies undertook significant preparatory work to build a culture of managing aspects of transport and border crossings in each cooperation across border agencies (within and between country. Interagency coordination is challenging to achieve, the two countries), develop a legal framework enabling that however, and many OSBPs have not met expectations. cooperation to take place, and install the information tech- The Transit and Transport Coordination Authority of the nology infrastructure making it possible to start the docu- Northern Corridor has been trying to improve OSBPs in East mentation process even before trucks arrive at the crossing. Africa by gathering better data on their performance. The Unlike many other OSBPs, at the Malaba OSBP the authority is supported by the Sub-Saharan Africa Transport reforms targeted all of the key parties involved in border Policy Program (SSATP), an international partnership for pol- crossings: icy reform and capacity building in Africa’s transport sectors. • Border management agencies, through advance prepara- Their efforts include border-crossing surveys to gain insight tion with prearrival lodgment of the customs declaration into the reasons for slow border processing and to docu- and better coordination between agencies at the crossing ment the impact of reforms. • Clearing agents, through mandatory prearrival lodgment During 2011–12, customs authorities in Kenya and of declarations (which used to be optional and rarely used) Uganda modified selected business procedures related to • Truck drivers, through traffic and parking rules to ease border crossings. They had a unique opportunity to observe congestion in the customs zone the impact of those reforms, because the Malaba border post between Kenya and Uganda (one of the pilot OSBPs) Source: Fitzmaurice and Hartmann 2013. firms (World Bank 2010; Barker and Men- banks, the high fixed costs per transaction and gistae 2013). This differential access implies difficulty in attaining scale economies make it that smaller and younger firms are less likely costlier to lend to smaller firms. As bank man- to exploit potentially profitable opportunities agers observed in a survey in Rwanda, the poor than larger and older firms. To the extent that quality of financial statements and business younger firms are a source of innovation and plans, lack of business skills, inability to man- productivity growth, such restrictions pose a age risks, and informality of smaller enterprises particular threat to competitiveness. constitute a big challenge for lenders (World To some degree, however, the problems Bank 2012c). These observations suggest that that smaller and less established firms encoun- the problem could be remedied partly by ter in obtaining credit may reflect the higher improving the management and transparency costs and risks involved in lending to them. For of firms. 218 Youth EmploYmEnt in Sub-Saharan africa Figure 6.16 African firms use relatively little bank The banking sector should be opened up financing with care, however, as the system can become more vulnerable when more banks enter and World begin to compete. For instance, banks may take on too much risk in more competitive East Asia and Pacific environments. It is also hard to open up the financial system to competition if the supervi- Europe and Central Asia sory structure is poor. Reforms to improve the High-income institutions that support the financial sector, OECD countries described next, are more feasible. Latin America and the Caribbean Strengthening Credit Information Middle East and Systems North Africa Well-developed credit information systems South Asia have improved access to finance in advanced economies. One reason why collateral require- Sub-Saharan Africa ments or interest charges are so high in some developing economies is that banks have little 0 5 10 15 20 25 30 35 40 45 information on prospective clients, so they Percentage set high risk premiums. Credit information Firms using banks to Investments systems that permit banks to share customer finance investments financed by banks repayment information can help lenders to Source: Based on World Bank enterprise surveys. evaluate prospective customers and bankable projects (IFC 2013). Most countries in Africa have no credit information system at all. Where Even if lending to smaller or younger firms is credit information systems exist, they tend to inherently more difficult, basic policy reforms be rudimentary, with low coverage of potential to address credit market issues will help them. borrowers (figure 6.17). Indeed, reforms will be particularly helpful to Countries seeking to establish strong credit smaller and less established firms, because they information systems require new regulations are the most disadvantaged by credit market for the licensing, operation, and supervision of imperfections. Critical reforms include steps to credit bureaus. Kenya’s credit information sys- make the banking industry more competitive, tem, for instance, significantly improved after the development of credit information systems, 2008, when new regulations were approved and better creditor rights. on the licensing, operation, and supervision of credit bureaus by Kenya’s central bank.17 By Making the Banking Industry More Competitive The banking sector in much of Africa is charac- Figure 6.17 Coverage of credit bureaus is generally terized by a lack of competition (it is dominated low in Africa by a few large banks) and heavy government involvement (IFC 2013; Bertrand, Schoar, and Kenya Thesmar 2007). Measures that would facili- Nigeria tate new domestic and foreign entry into the industry could increase competition, make the Rwanda lending industry more efficient, and increase South Africa the likelihood that banks will find better ways Tanzania of reaching out to firms. Where government ownership of banking assets is extensive, as in 0 10 20 30 40 50 60 Tanzania, fostering competition may require a % of population reduction in the government’s financial stake. Source: World Bank Doing Business indicators. raising productivity in africa’s modern Wage Enterprises to foster Job Growth for Youth 219 2010 and 2011, respectively, the first two credit passed in 2009 (World Bank 2012c). Angola, bureaus had started to operate. The new leg- Botswana, and Swaziland have among the low- islation allowed banks to share negative credit est global scores on the World Bank’s Doing information by obligating banks to report all of Business indicators of contract enforcement, the nonperforming loans on their books. especially the length of time required to enforce To be truly effective, credit information a standard contract (World Bank 2011b). systems must cover not only all potential bor- Banks’ reliance on real estate and land as rowers but also all prospective lenders, and prime collateral can further complicate the they must include both positive and nega- efforts of smaller enterprises to obtain financ- tive information on borrowers. The scope of ing. Banks typically do not use other assets, credit information systems is commonly much especially movable assets, as collateral for sev- more limited in Africa, however. In Zambia, eral reasons. Contract enforcement is weak, for instance, the credit reference bureau that the registration of assets is poor, and the legal opened in 2007 has yet to expand its sources to framework for creditor rights does not ade- retailers, trade creditors, and utility companies quately support secured transactions in assets. in order to capture a larger share of the popula- Potential reforms include the introduction of tion of potential borrowers (World Bank 2009). electronic property registration and land titling Similarly, in a survey following the regulatory systems, along with the adoption of a more reform in Kenya, the two most prominent con- modern framework of security interests over cerns highlighted by the banks were the need to movable assets. Such a framework would allow share positive as well as negative information banks to register security interests over a wide and the need to incorporate the information range of movable assets—including personal collected by other providers, such as microfi- property, vehicles, machinery, inventory, raw nance institutions and utility companies. The materials, receivables, and intellectual prop- sharing of positive data enables the develop- erty—and to enforce claims on those assets as ment of a credit-scoring system, allowing good collateral. borrowers to establish a track record and access loans at more favorable interest rates. Reducing Constraints to the Entry and Growth of Productive Firms: Strengthening Contract Enforcement Governance Is Key and Creditor Rights Institutions A good business environment should facilitate Legal rights of lenders and borrowers can the entry of new firms and the growth of more facilitate the use of collateral and the ability to productive firms. Unproductive firms are pres- enforce claims in the event of default. Strong sured either to improve their productivity or to creditor rights can expand the supply of loans exit, and overall productivity improves. by providing legal protection for lenders in Credit constraints, poor start-up financing, cases of nonpayment. The legal system should and high costs of access to public utilities are also allow borrowers to use a broader range of major barriers to firm entry and expansion assets as collateral, which should allow them to in Africa. Another critical influence on firm obtain loans on better terms. These legal rights entry and growth is the quality of governance. must also be enforced well in practice. Some Excessive or poorly designed business regula- of the high risk premiums that banks attach tion, corruption, and weak contract enforce- to borrowers in Africa reflect their low confi- ment impose high entry and expansion costs, dence in creditor rights and their enforcement. which reduce productivity through their nega- Contract enforcement institutions are particu- tive effects on entry and competition (Klapper, larly weak in most countries in the region, even Laeven, and Rajan 2006; Djankov 2009; see also upper-middle-income countries. For example, Xu 2011). banks in Rwanda contend that insolvency and Formal entry requirements are probably the bankruptcy procedures do not function well most ubiquitous means by which governments in practice, despite the new legal framework directly regulate entry. Almost everywhere in 220 Youth EmploYmEnt in Sub-Saharan africa Figure 6.18 Africa has the highest formal costs of setting up a business The formal cost of setting up the standard busi- ness (as a percentage of per capita income) is Paid-in 105 percent in Angola and 284 percent in the minimum Democratic Republic of Congo, for example. capital These costs are high by international standards, even compared to South Africa, where the same Cost of starting a cost is a mere 0.3 percent of per capita income. business Corruption in the granting of entry licenses and other permits also hinders the setup and Cost of expansion of firms, and it must be tackled to getting a strengthen competitive forces in Africa. Sur- construction permit vey evidence from Uganda on who has to pay 0 100 200 300 400 500 600 700 800 bribes and how much they have to pay suggests that more profitable firms have to pay larger % of income per capita bribes when seeking access to permits, licenses, Sub-Saharan Africa South Asia High-income OECD countries Middle East and North Africa Latin America and the Caribbean East Asia and Pacific and public services. By raising costs dispropor- Europe and Central Asia tionately for more productive firms, corruption Source: World Bank Doing Business indicators, 2012. acts as a tax on efficiency. This same research Note: Paid-in minimum capital is the amount that the entrepreneur needs to deposit in a bank or with suggests that if firms have better informa- a notary before registration and for up to three months following incorporation. tion and undertake collective action, they can improve their bargaining position when deal- ing with bribe takers and reduce corruption Africa, anyone setting up a new business needs (box 6.6). to obtain an operating license from a local and Better governance can also help by strength- central government authority and to acquire ening contract enforcement. Poor contract legal status to engage in certain transactions enforcement can constrain the growth of pro- from the government registrar. Often, various ductive firms by making trust more important construction permits and site use licenses are than efficiency in determining how production required to establish or expand a firm. Usually is organized. Owners and top management shy the time and costs associated with all of these away from delegating responsibility to other permits and licenses are a substantial share of employees (see, for instance, Cingano and the overall cost of setting up a business, and they can be prohibitive (World Bank 2011b). On average, the formal costs of setting up a Figure 6.19 Business setup costs are higher in resource-rich African countries business are higher in African countries than elsewhere in the world (figure 6.18). Cost of starting a business Reforms to reduce the start-up and expan- sion costs linked to regulation should be a prior- Angola ity for improving the business environment in Congo, Dem. Rep. Africa. Many resource-poor countries in Africa have steadily and significantly reduced business Lesotho start-up costs over the past five years through a Madagascar series of administrative and legislative reforms. For example, in countries throughout South- Malawi ern Africa (including Lesotho, Madagascar, Malawi, and Mozambique), start-up costs have Mozambique converged with or even fallen below those of South Africa South Africa, which have always been low by emerging-market standards. In contrast, start- 0 50 100 150 200 250 300 up costs remain very high in some resource- % of income per capita rich countries (figure 6.19; World Bank 2012b). Source: World Bank Doing Business indicators, 2012. raising productivity in africa’s modern Wage Enterprises to foster Job Growth for Youth 221 Box 6.6 Who must pay bribes? How much? And does it matter? Who must pay bribes and how much? A unique set of survey Subsequently, researchers used the same data to study data on corruption, containing quantitative information on the relationship between bribery payments, taxes, and firm the payment of bribes by Ugandan firms, helps to answer growth. Using industry-location averages to circumvent these questions. The data have two striking features: not all potential problems of endogeneity and measurement errors, firms report that they need to pay bribes, and there is consid- they found that the rate of both taxation and bribery is neg- erable variation in reported graft across firms facing similar atively correlated with firm growth. A 1 percentage point institutions and policies. These patterns can be explained by increase in the bribery rate is associated with a 3 percentage differences in control rights and bargaining strength across point reduction in firm growth, an effect that is about three firms. Firms typically have to pay bribes when dealing with times greater than that of taxation. These results are con- public officials whose actions directly affect their business sistent with the findings from the original research, which operations. Such dealings cannot be easily avoided when, suggest that firms’ investment and technology decisions are for example, a firm must export or import goods or make driven partly by the desire to minimize bribe payments, even use of public infrastructure services. though the decisions adversely affect gross profits. How much must graft-paying firms pay? When the The results have clear policy implications. If the size of quantitative data on corruption are combined with detailed the bribe that a firm needs to pay is an outcome of a bar- financial information from the surveyed firms, the size of the gaining process, collective action on the part of the busi- bribe that a firm needs to pay depends positively on cur- ness community that strengthens the bargaining position of rent and expected future profits and negatively on expected individual firms may be a successful strategy to reduce the alternative returns to capital. In other words, firms’ “ability cost of doing business. Potentially effective measures include to pay” and “refusal power” can explain a large part of the collecting and disseminating information about corrupt variation in bribes across graft-reporting firms. These results practices; informing the private sector and the public about suggest that public officials act as price (bribe) discrimina- service standards, guidelines, and norms of major service tors and the prices of public services are determined partly to providers; increasing the ability of individual firms to commit extract bribes. Moreover, expecting high demand for bribes, to not paying bribes; and recognizing those who are doing a a firm might find it profitable to choose a technology with good job by resisting corruption. higher operating costs per period and thus lower profits, but one that indirectly reduces the size of the bribe that the firm needs to pay. Sources: Svensson 2003; Fisman and Svensson 2007. Pinotti 2012; Bloom, Sadun, and Van Reenen incremental reforms in the investment climate, 2012). Firms also have an incentive to integrate before they are rolled out more broadly. vertically or to enter into contracts only with Firms engaging in similar or linked activi- members of their extended family and social ties often cluster, because clustering has many network. potential advantages. Firms in the cluster can learn production and business management techniques from one another.18 Transport and Localized Interventions to Improve the other types of infrastructure that serve the clus- Wider Business Environment ter improve as clusters grow. Together, the firms To increase competitiveness, the ultimate pol- can attract more consumers, more and better icy objective is to improve the overall business input suppliers, and more skilled workers than environment, but in some cases localized inter- they could attract if they were dispersed. Clus- ventions can lay the groundwork for a com- ters can be especially conducive to the growth petitive, modern enterprise sector to emerge. of specialized input suppliers. They also help Concentrating public resources in selected clus- to reduce the cost and raise the quality of spe- ters of activity is one kind of localized interven- cialized inputs. Investments to improve input tion. Such clusters can also be used to introduce quality or reduce input costs are less risky for 222 Youth EmploYmEnt in Sub-Saharan africa a supplier when several prospective purchasers be concerned mainly with enabling nascent clus- are present; when there is only one purchaser, ters to transition toward self-sustaining growth. the purchaser has the power to bargain those Special economic zones (SEZs) are an gains away. increasingly popular vehicle for supporting In this way, clusters can exhibit “agglomera- localized growth, especially in the form of tion economies”—that is, the firms in a cluster export-oriented clusters. The idea is to support can become increasingly productive as the clus- clustering by concentrating public investment ter grows and matures. Many of the potential (such as infrastructure) and policy reforms productivity benefits from growing clusters in specially marked zones. Some SEZs target occur through spillovers between firms—for locations that are considered to have an inher- example, firms learn from one another or ent advantage for insertion into global value downstream firms benefit from having several chains, such as areas near ports, but the choice input suppliers located nearby. The gains from of a high-potential location may not be criti- being in a cluster thus depend on how many cal to success: a particular location may have other firms locate there, and every firm that no inherent advantage in a sector or task, but it locates in a cluster could be conferring ben- may acquire comparative advantage as a cluster efits on the other firms in that cluster. A clus- starts to develop and agglomeration economies ter that is not out of the ordinary to start with arise (Collier and Venables 2007). As the expe- may become internationally competitive once rience of successful clusters in Africa suggests, it is big enough. But when choosing where to SEZs should focus on supporting incipient locate, firms fail to take this collective benefit clusters to develop beyond a critical threshold, into account. This is a form of market failure, after which they can sustain themselves. SEZs so although clusters can emerge spontaneously, in Africa have rarely been located around pre- they may need some kind of public support to existing clusters, which is thought to be one reach a truly competitive size. reason that their performance has been less Public support may be especially necessary satisfactory than that of SEZs in other develop- to form clusters in manufacturing and tradable ing regions (Farole 2011). services, given their potential export orienta- Creating a superior business environment tion. Because they are not limited by the size within SEZs reduces the availability of public of the domestic market, such clusters can truly resources for improving the business climate exploit the economies of scale from clustering outside of those areas. If there are significant production. As discussed, African firms can increasing returns to clustering, then allocating integrate with global supply chains by special- public resources to SEZs is likely to be supe- izing in specific tasks or steps in a value chain. rior to spreading public resources everywhere To do so, they must be able to import inputs as to provide services at “a uniformly low level” easily and cheaply as their international com- (Collier and Venables 2007). SEZs do not always petitors, and they must have access to the same fulfill their potential, however, as indicated by high-quality physical infrastructure and sup- the varied experience with SEZs worldwide. portive regulations. Thus one option for public Because it is difficult to measure the extent of support is to develop clusters specializing in agglomeration economies rigorously, the causal specific tasks within global value chains.19 impact of promoting clusters through SEZs Box 6.7 looks at the common attributes of is difficult to evaluate. For that reason, every 11 successful clusters in Africa. These clusters potential SEZ should be considered carefully. highlight the role of value chains (linking input SEZs designed to cater to the needs of par- suppliers with downstream firms), knowledge ticular firms or industries to the exclusion of flows, ability to attract a skilled workforce, and others are especially risky. Such SEZs con- coordination. All of the clusters had public sup- centrate resources not just spatially, but also port, although the form of support varied. A in certain lines of activity or firms that are common pattern, however, was the spontaneous deemed to be “winners.” This strategy assumes rather than publicly engineered inception of the that the state has the ability to identify which clusters, suggesting that public support should firms stand to gain the most from agglomerat- raising productivity in africa’s modern Wage Enterprises to foster Job Growth for Youth 223 Box 6.7 Knowledge, technology, and the emergence of successful enterprise clusters in Africa How have successful enterprise clusters evolved in Africa, Africa’s wine industry benefits from the Elsenburg Agricultural and what lessons do they offer for encouraging new clusters Training Institute and the Wine Industry Network for Expertise to develop? A study of 11 clusters that emerged more or less and Technology. spontaneously in various countries and industries provides All successful clusters have a labor force that is more edu- insight into those questions.a cated than the norm for Africa, sometimes far beyond the Except for the Mauritian textile cluster (originating in an norm (in Otigba, 55–60 percent of entrepreneurs are uni- export-processing zone established by the government in the versity graduates). In Suame—Africa’s largest artisanal engi- 1970s), the clusters arose spontaneously because they had neering cluster—75 percent of entrepreneurs have a primary access to major local markets and infrastructure (all clusters); education or less; only 2 percent have a higher education. natural resources (the cut flower, fish, and wine industries); Yet such clusters create employment for increasing numbers and local entrepreneurs with tacit knowledge and basic skills of university graduates, who can be an important force for in trading, design, or manufacturing (the metalwork, com- continued growth and sustainability. puter, and Nnewi auto parts clusters, which started with Government support to clusters takes several forms: trading or repairing and evolved into assembling or manu- defining and implementing sectoral policies, regulations, and facturing). Strong local demand existed for their products, standards; creating a special agency to promote or develop a except for cut flowers and fish, which had strong interna- cluster; establishing institutions to provide technical support tional demand. Governments sometimes promoted clusters and develop capacity; offering incentives in the form of sub- indirectly. For example, the handicraft cluster in Tanzania and sidized land or infrastructure; and forging strategic alliances the Suame automotive cluster in Ghana developed because with industries overseas. the government relocated scattered, unorganized businesses Industrial and professional associations facilitate collec- to clear the city or facilitate spatial planning. tive action and cooperation. For example, Tanzanian clusters Efficiencies from a cluster-based value chain were fun- voice concerns and obtain technical advice through associa- damental. A value chain develops when a cluster achieves tions. Suame associations focus on social welfare. Associa- a certain scale, gains visibility, and expands further to ben- tions for flower producers engage in lobbying, environmental efit from efficiencies derived from a larger clientele, syner- conservation, maintaining standards, and facilitating market gies among firms, the developing knowledge network, and access, among other activities. The Wines of South Africa shared facilities and services, including some provided by Association markets South African wines internationally. governments and donors (warehouses, for example). Amid stiff global economic competition, the survival of To succeed, all clusters had to acquire, adapt, and dis- these clusters depends on overcoming local challenges related seminate knowledge. Horizontal and vertical links among to the critical mass of skills at their disposal, weak local tech- firms—subcontracting relationships, connections with cus- nology institutions (disconnected from the business sector), tomers and suppliers, information and formal collaboration weak government and institutional support, natural resource through joint ventures and franchises, membership in asso- depletion, and the failure to meet international standards for ciations, and the movement of skilled personnel—created product quality and safety. The government has a multifac- knowledge networks. Foreign knowledge and technology eted role to play in enabling clusters to meet these enormous sustain several clusters. Foreign direct investment from Hong challenges. Aside from overall coordination, the public sector Kong SAR, China, and Taiwan, China, brought technology can facilitate the acquisition of knowledge and technology and know-how to the Mauritian textile industry. Nigeria’s (through links with foreign enterprises and improved links Otigba computer village imports computer hardware from between local institutes and businesses), establish a conducive China, Dubai, and Malaysia, with which it maintains techni- institutional regime (clear regulations, standards, and quality cal and production channels. Kenya’s Kamukunji metalwork- assurance) in collaboration with trade and professional associ- ers obtain technology from South and East Asia. Formal ations, strengthen training and skills, provide basic infrastruc- and informal learning and training assume numerous forms. ture, and enact policies that increase consumers’ purchasing Technology-intensive manufacturing clusters are trained by power and demand for high-quality goods. experts (locally and overseas) and offer training on the job or factory site. South African wine producers participate in vini- Source: Zeng 2006. a. The clusters were cut flowers and metalwork (Kenya), fish production culture forums and root stock associations. In most instances, and processing (Uganda), handicrafts and furniture (Tanzania), automo- universities and technology institutes contribute minimally to tive parts and computer hardware (Nigeria), manufacturing and vehicle knowledge and technology flows for clusters. Even so, South repair (Ghana), textiles (Mauritius), and wine and textiles (South Africa). 224 Youth EmploYmEnt in Sub-Saharan africa Box 6.8 Why have special economic zones failed to thrive in Africa, and what are the lessons for the future? Government attempts to launch SEZs in Africa have met entity within the cluster may own more than 20 percent of with little success. Numerous zones and industrial parks the equity, and each firm occupies land in proportion to the have been announced but failed to materialize; others have equity it contributes. The PMC receives fees only if the proj- been “built,” but never occupied. The few SEZs that have ect is accepted. emerged are smaller than those in other regions. For exam- The combination of incentives offered under the scheme ple, SEZs in six African countries supported perhaps 35 firms fosters interdependence among the stakeholders. Institu- on average, versus 350 in Honduras, nearly 300 in Bangla- tional mechanisms reinforce those relationships, including desh, and 3,500 in Vietnam (Farole 2011, 71). Evidence is regular monitoring reports from the PMC to the ministry, emerging that Africa’s SEZs have started to stagnate at low quarterly central government reviews of the scheme, and the levels of growth (Farole 2011, 4). What could help the next appointment of directors on the SPV board by the ministry, generation of SEZs to do better? PMC, and the state. In India, the Scheme for Integrated Textile Parks, inau- The small size of the parks and the grants seems to gurated in 2005 by the Ministry of Textiles, is in its early reduce the incentives for interference at the state and central stages but has successfully built parks and attracted firms level. Because a central ministry rarely has local knowledge to locate and invest in them. The parks appear to be gener- or human resources to monitor at a micro level, devolving ating the benefits associated with clustering. The scheme’s this task to a specialized agency (the PMC) allows the minis- most prominent innovation is to give the park users a much try to focus on its strategic mission and obtain better micro greater role, much earlier, in development of the park. In data. The reporting requirements for PMCs are clear and fact, the first order of business is not to build parks but to standard. Strict rules and methods for grant disbursements organize entrepreneurs. help to limit the abuse of funds. Entrepreneurs form a group (called a special-purpose vehi- Responsibility for engagement and coordination with cle, SPV) to develop a detailed proposal for a park. To assist state agencies lies with local entrepreneurs, who are best with this task, the SPV selects a specialized project manage- positioned to navigate the political economy of their state. ment consultant (PMC) that is stringently vetted by the minis- States issue the clearances that are vital for park devel- try. Guided by the SPV, the PMC designs, costs, and conducts opment, especially for converting land to industrial use. a feasibility study for the project. The SPV makes the final Although limits on state participation in SPV funding curb decisions on location, project design, and costing, because the state’s direct role, the ability to appoint a director to the entrepreneurs constitute the “demand” for the park and the SPV board is an incentive to participate. SPVs also have have the most incentives to select the best alternative. recourse to the center if issues arise with the state. If the central government approves the proposal, it pro- The scheme appears highly adaptable to India’s diverse vides a grant for infrastructure to be shared by park users (up conditions and its textile industry. Although this model may to 40 percent of the cost, up to a maximum of about US$8 not suit circumstances in other industries or countries, it indi- million). The SPV finances the rest. It may solicit state grants, cates that success is more likely when incentives are carefully obtain other state and local contributions (such as subsidized calibrated to the political, institutional, and entrepreneurial land), and even obtain investment from PMCs, but the SPV context. must contribute at least 51 percent of the equity to ensure that members retain managerial control. No single private Sources: Saleman and Jordan 2013; Zeng 2006. ing, which is arguable. The driving principle of Building Skills for the Modern localizing interventions around incipient clus- Wage Sector ters should be one not of picking winners but rather of “letting the winners pick each other As discussed in chapter 3, productivity in mod- by locating near each other.” In other words, it ern wage enterprises depends on a range of skills: is important not to over-engineer who gains cognitive skills, soft skills, and managerial and entry to an SEZ (see box 6.8 for an example of business skills. Policy discussions often revolve a more nuanced approach from India).20 around technical skills, but foundational skills raising productivity in africa’s modern Wage Enterprises to foster Job Growth for Youth 225 acquired through schooling are also important. Several policy recommendations emerge Recent research from Kenya, for instance, sug- from this evidence on the role of skills. They gests that by raising literacy and reasoning skills, emphasize the need to focus on foundational secondary school completion reduces the prob- skills and public goods such as quality assur- ability of low-skilled self-employment by 50 ance and information, the need for govern- percent, while raising the probability of formal ment to focus on building portable rather than employment by 30 percent (Ozier 2013). These job-specific skills, the need for special support skills seem to matter because they facilitate on- to enable poor and disadvantaged groups to the-job learning in firms. Literate workers can acquire skills, the need to develop interna- be trained through written instructions in job tional linkages, and the need for systems that specifications and procedures, for example, and link employers with trainers. The following they require less hands-on technical supervision discussion of these recommendations reviews by others. Such training is rare in Africa, how- interventions that have been tested in Africa ever, because of low literacy levels (Biggs, Shah, and elsewhere. and Srivastava 1995). Employer surveys from some African coun- Focus on Foundational Skills and on tries—Botswana, Lesotho, and Sierra Leone— Public Goods Such as Quality Assurance suggest that employers are also looking for soft and Information skills in young hires. Employers in Botswana, The discussion on public support to build skills for instance, rate reliability and punctuality, for youth employment often takes a short-term commitment and hard work, honesty, and view of the problem and centers on technical teamwork skills as the top skills required when vocational education and training (TVET), but hiring skilled workers (World Bank 2011a). it is equally, if not more, important to build a While there is increasing international evi- solid foundation for individuals to develop dence on the link between skills and produc- their skills as they progress in life. Basic school- tivity, firms in Africa generally do not rate ing is a key determinant of wage workers’ cog- skills as a top constraint. Self-reported short- nitive and soft skills. It is the foundation for ages of skills tend to be higher in relatively the acquisition of additional skills, whether capital-intensive industries. Skills also tend through more formal education, training, or to become a constraint for high-performing on-the-job learning. Improving access to gen- firms: businesses that complain of a skills eral schooling and the quality of that school- shortage consistently outperform those that do ing—even at the basic level—is a priority for not (Barker and Mengistae 2013). This infor- increasing the productivity of labor in the mation suggests that while raising skills is not modern wage sector. going to solve the employment problem by Parental investment and early childhood itself, it should be part of the policy package interventions in health and education also for improving competitiveness. Even if skills influence personality traits that become rele- are currently not a binding constraint, they will vant to productivity later in life, either directly quickly become binding as the sector takes off. or through the fact that some personality traits In addition to the limited availability of will make some students more likely to suc- cognitive, soft, and technical skills among job ceed in school (Almlund et al. 2011). Recently, seekers, a lack of “managerial capital” could many governments have included soft skills in also be constraining the competitiveness of training programs for youth, and while there is African firms. As discussed in chapter 3, there some evidence of their impact, evidence on the is increasing evidence that firms in the region significance of early childhood investments is have visibly poor basic management practices. much better established. There may be considerable scope for improving Governments should also reevaluate their productivity (and hence raising employment) role in providing TVET directly. As discussed by investing in business and management skills in chapter 3, publicly provided TVET has a training, and perhaps even in individualized poor track record in Africa, where it is expen- management consulting. sive but often inefficient or irrelevant to the 226 Youth EmploYmEnt in Sub-Saharan africa labor market. An increasingly large share of better alternative for public support may be TVET in Africa is provided by the private sec- active labor market programs that have train- tor (Mingat, Ledoux, and Rakotomalala 2010; ing components and target specific groups World Bank 2012d; Atchoarena and Esquieu (Almeida, Behrman, and Robalino 2012). 2002; Kitaev 2002). Although private providers It is also critical to make formal TVET more may offer little training in some specific skills, efficient and market relevant. Several policy this lack of interest does not necessarily mean recommendations are pertinent here, but they that nascent private institutes are incapable of are based largely on the experience of devel- providing such training or that the returns to oped countries. Box 6.9 discusses potential those skills are too low. Similarly, there may reforms and how they have been undertaken be an equity-based case for establishing pub- in Africa. lic TVET institutes in poor or remote areas, where private entry is limited, but there may not be high enough returns to TVET training Address Information and Credit in remote areas. In many cases, low take-up of Constraints to Improve Access to TVET training is better addressed by improving access among Poor and Disadvantaged Groups to credit (or grants) and providing information As discussed in chapter 3, the government about training programs. In other contexts, a could step in to provide targeted financial Box 6.9 Reforming TVET systems in Africa One strategy for improving the relevance and quality of job, type and quality of employment, and students’ satisfac- skills developed through TVET programs is to grant public tion with the education. TVET institutions increased autonomy and accountabil- Many African countries have begun to reform their TVET ity for results. Institutions would have the freedom to set systems along these lines. While this is a positive step, there fees, adapt training to local needs, hire appropriate staff, is little rigorous evidence yet on whether these reforms and choose methods of instruction. Accountability can be have improved cost-effectiveness or market relevance. For increased by basing financing on performance and out- instance, Ethiopia, Ghana, and Tanzania have developed comes rather than inputs, accompanied by better measure- national qualification frameworks to set standards. Recog- ment of performance. nizing the need for better coordination in their TVET sys- Governments also have a role in providing standards, tems, some countries have tried new governance arrange- accreditation, quality assurance, and information. The TVET ments and have introduced national coordinating bodies qualifications system should be focused less on providing and national training authorities, with mixed results. Ghana, inputs and more on assuring that the training has provided for example, established the Council for TVET (or COTVET) the competencies (skills) required for the market. Lack of in 2006 as a coordinating body. A study of this reform sug- information about private TVET providers can constrain gests that COTVET’s engagement with the private sector and young people from making informed choices about training, collection of demand-side data could be improved and that and it can also dull providers’ incentives to improve their qual- insufficient coordination within the government has led to ity. Governments can play a role in collecting and disseminat- parallel agendas, plans, programs, and committees. Govern- ing this information. Instead of measuring inputs (number of ments in Africa may need to build more capacity in the rel- students, institutions, and teachers) as is currently the norm, evant agencies before they can successfully transplant com- the emphasis should be on tracking efficiency (dropout, rep- plex TVET systems from developed countries. etition, and survival rates) and outcomes (share of trainees entering wage or self-employment after graduating), which are more relevant to young people’s training decisions. For Sources: Based on Johanson and Van Adams 2004 (mixed results of governance arrangements); Darvas and Palmer 2012 (TVET in Ghana); instance, in the Netherlands, almost all graduates of higher World Bank 2012a (TVET with a business component in Uganda); Son- education institutions are surveyed a year and a half after dergaard and Murthi 2012 (reforms in Eastern Europe and Central Asia); they graduate to collect information on time taken to find a Krishnan and Shaorshadze 2013 (TVET in Ethiopia). raising productivity in africa’s modern Wage Enterprises to foster Job Growth for Youth 227 support and information that young people that African firms are particularly constrained from poor or otherwise disadvantaged back- in providing on-the-job training. World Bank grounds need to access TVET markets.21 Young enterprise surveys indicate that on average 30 women are less likely to participate in TVET percent of formal firms in Africa provide on- than young men, and they frequently focus on the-job training, which is comparable to the narrow types of training such as tailoring or standard level across low- and middle-income weaving, suggesting that direct financial sup- countries. Nor does the absence of on-the-job port or information is needed to increase their training in a firm necessarily signal a constraint. take-up of training with high returns. Targeted A recent study estimates that many firms do voucher programs that have an information not provide it because the returns to such train- component, such as the Technical and Voca- ing are close to zero for them (Almeida and tional Vouchers Program (TVVP) in Kenya, Carneiro 2009). Governments could consider can improve access to training among poor and supporting on-the-job training in specific cases disadvantaged youth. The TVVP experience where there is compelling evidence of a market shows that flexible vouchers (not restricted to failure (box 6.10), but not more generally. specific types of training providers) are more Some of the constraints to firm-specific effective than inflexible ones. Another impor- training could be addressed by broader policies tant lesson from TVVP is that participants are to improve the business climate or productivity frequently mistaken about the returns to voca- in firms. For example, if liquidity constraints tional education, with males strongly prefer- prevent firms from investing in training, then ring traditionally “male-dominated” courses policies that improve firms’ access to finance such as motor vehicle repair and women will also have an impact on on-the-job train- almost exclusively choosing traditionally ing. To the extent that firms and workers would “female-dominated” courses such as hairdress- be more likely to reach an agreement on shar- ing. TVVP provided information on the higher ing the returns from training if contracts were economic returns in male-dominated trades, enforceable, strengthening the enforceability which encouraged women to take up training of contracts and the rule of law could have a in those areas. positive impact on on-the-job training. Some- times, firms train workers in very specific or Focus Government Intervention on outdated skills because they use old technology. Building Skills That Are Portable, Not Policies that promote innovation and techno- Job Specific logical upgrading in firms could therefore Government interventions should be focused also improve the social returns to on-the-job on building more “portable” skills, as opposed training. to skills required only for a particular job or firm. More portable skills have higher social Facilitate International Linkages in value, because they are less vulnerable to Training and Technology Upgrading in changes in the demand for skills specific to cer- Specific Industries tain firms or industries. Because their current Governments can facilitate technical coop- market returns could be lower than this social eration with other countries in training and return, the likelihood of underinvestment is upgrading technology in specific industries, higher for more portable skills. Firms are also especially with other developing countries that less likely to invest in portable skills because have technical capacity in the industry. A recent workers trained in such skills could be poached example is the technical cooperation between by other firms. Ethiopia’s Ministry of Trade and Industry, the Firm-specific skills can be acquired through Ethiopian Leather and Leather Products Tech- on-the-job training, and public support may nology Institute, and the Footwear Design and not be necessary to supply them. Credit con- Development Institute of India. The Footwear straints could prevent firms from investing in Design and Development Institute provided firm-specific skills, but there is little evidence technical assistance to seven Ethiopian shoe 228 Youth EmploYmEnt in Sub-Saharan africa Box 6.10 Does government have a role in on-the-job training? To the extent that they are substitutable, on-the-job train- that did not. Ghana offered a similar voucher program to ing is by design more likely to be more market relevant than informal enterprises in the early 1990s, but it largely failed, technical or vocational education in the classroom. When apparently because of poor marketing and distribution. there is evidence that on-the-job training is being underpro- Even when there is evidence that training is underpro- vided because of credit or other market failures, government vided by some firms, governments should be cautious with subsidies could be considered. Some developing countries on-the-job training subsidy schemes and their targeting. The use subsidy schemes (in the form of levy-rebate systems) returns could vary across firms. Training might be subsidized to promote on-the-job training. For example, Peru allows in firms in which the net returns are insufficient to justify the firms to use payroll-tax proceeds for in-service training or training, or scarce public resources might be wasted on firms to purchase training at an accredited training institution. that would have provided on-the-job training even without Levy-rebate systems have a mixed record of success, how- the subsidy. It is not easy to identify which firms are con- ever. In South Africa, for example, only 35 percent of small strained in providing training. Many programs try to improve enterprises claimed the training grant to which they were targeting by using proxies, such as targeting smaller firms entitled—and even fewer large firms appear to be claiming based on the assumption that they are more likely to be them. credit constrained. These proxies are imperfect, and govern- Some countries have also used a voucher system that ments should also invest in better monitoring and evaluation subsidizes the cost of training to firms while giving them systems for such programs. A randomized pilot, for example, flexibility in choosing the training provider. Kenya’s Jua Kali could help to test whether subsidies are a substitute source Program is a good example. An evaluation of this program, of funding for on-the-job training that would have occurred which targeted small and microenterprises, found that firms in any case. using vouchers experienced markedly greater growth in rev- enue and assets. While this finding is promising, it should be Sources: Based on Johanson and Van Adams 2004; Filmer 2012 (South interpreted with caution, because the firms that participated Africa); World Bank 2011c (vouchers in Peru); Almeida, Behrman, and in the program are not necessarily comparable to those Robalino 2012 (Jua Kali). factories in design, technical upgrading, quality foreign workers may actually create additional assurance, productivity enhancement, and test- jobs. Building a base of evidence on the net ing, which improved their productivity and the employment effects of skilled immigration can overall quality of shoes. For instance, cutting at help to generate support.22 Policies for import- one of the shoe factories rose from 2,000 pairs ing skills can be designed to maximize domestic a day to 2,400, and defect rates dropped from 3 spillovers—for instance, by encouraging firms to 1 percent (Dinh et al. 2012). to train domestic workers in return for permits If certain technical skills are in demand in to hire technical skills from abroad. several countries, but the demand is too small to justify the fixed costs of building a training cen- Foster Systems That Link Employers ter in every country, governments can cooperate with Trainers to create regional hubs for technical skills train- Linking TVET systems to employers is neces- ing. “Importing” such skills could also make sary to ensure that the training is relevant to the economic sense, and in response governments market for skills. The returns to specific techni- could ease constraints on the immigration cal skills can vary substantially across markets of skilled workers. The current policy regime and over time. For this reason, policies oriented in most African countries is not conducive to toward improving the skills delivered through importing technical skills. Policy makers do TVET should build TVET systems with need to take political economy considerations employers’ involvement to ensure that employ- into account, such as concerns about domestic ers transmit information to trainers about jobs being taken by foreigners, but highly skilled which skills are in demand on a regular basis. raising productivity in africa’s modern Wage Enterprises to foster Job Growth for Youth 229 Some developing countries already produce Can the transition to work be eased by regular reports on the supply of and demand inserting training graduates in firms through for skills, such as the Labor Market Intelligence temporary internships and other programs? A Reports produced by the Philippines Technical case in point is the German-style dual system, Education and Skills Development Authority in which theory is taught in educational institu- (TESDA). TESDA’s labor market reports are tions and practical skills are acquired through based on consultations and focus group discus- an apprenticeship (or internship) in a com- sions with employer associations as well as on pany. Some evidence suggests that this system labor market surveys (TESDA 2012a, 2012b). reduces unemployment among graduates (Pio- Similar approaches are being rolled out in piunik and Ryan 2012), but it is not clear how Africa. Tanzania’s vocational and educational this system can be imported at scale into Africa, training authority has a monitoring and evalu- where the scope for internships with firms is ation system based on regular data that are col- much more limited. If adopted, training plus lected by zonal labor market analysts and mini insertion programs should be selectively tar- market surveys that track current and prospec- geted to those most in need. Latin America’s tive industry needs (United Republic of Tanza- Jóvenes en Acción model, described later, may nia and UNESCO 2012). Regular interaction be more suitable to the African context. between training centers and industry and com- Governments also support insertion merce is promoted by attaching tutors and stu- through employment or job search agencies. dents to industry to gain practical exposure and The evidence suggests that such services have by attaching industrial practitioners to training limited relevance for Africa. Formal public centers to teach specific topics. The success of placement agencies are underused. A survey this approach has yet to be determined. of job seekers in 11 African cities found that Linkages between the training system and only 7 percent were registered with an employ- industry can also help youth to make the tran- ment agency (De Vreyer and Roubaud 2013). sition to work. In Africa, matching specialized Similarly, a recent study on the Middle East and skills to jobs is not a major problem, but in low- North Africa found that many job seekers rely skilled wage employment, young people from solely on personal and family connections, and disadvantaged backgrounds have trouble con- most job placements occur outside of public necting to employers because of basic informa- employment services (Kuddo 2012). A review tional problems. A signal of these problems is of impact studies—drawn largely from devel- the ubiquity of informal networks as a source oped countries—found that these services of information about jobs. Young people with- tend to be more effective for the most educated out networks in modern firms—especially workers, when demand for labor is high (Bet- youth from poor families or rural areas, whose cherman, Olivas, and Dar 2004). family members or friends are less likely to have Comprehensive training-related programs wage jobs—will have greater difficulty finding that link students to employers through intern- such jobs. Recent research suggests that infor- ships are well-established in Latin American mal job networks also put women at a disad- countries and have begun to spread to other vantage. An experiment on hiring in Malawi regions, including Africa. The Jóvenes en found that men referred men unless specifically Acción (Youth in Action) Program in Colom- asked to refer women, and women referred ill- bia, for example, provided three months of qualified women unless specifically rewarded classroom-based training and three months for providing good-quality referrals (Beaman, of on-the-job training through internships in Keleher, and Magruder 2012). Aside from dis- firms to unemployed young people between advantaging certain groups, informal job net- the ages of 18 and 25 in the two lowest socio- works reduce the productivity of job matches economic strata of the population (Attanasio, by causing them to be based on connections Kugler, and Meghir 2011). Similarly, the Juven- rather than qualifications and by limiting the tud y Empleo (Youth and Employment) Pro- pool of potential hires that firms can tap. gram in the Dominican Republic subsidizes 230 Youth EmploYmEnt in Sub-Saharan africa training for individuals between the ages of 16 Improve Management Practices and 29 who have not completed high school, Management training can have positive with the training followed by an internship at impacts, but more evidence is needed on how a private firm (Ibarraran et al. 2012). The pro- to target it. If poor management practices are grams involve collaboration with the private as prevalent in Africa as some recent studies sector to provide internships and select the suggest, there may be considerable scope for content of training. They usually offer a pack- improving productivity by investing in busi- age of skills, complementing vocational train- ness skills training and perhaps even in indi- ing with life skills training intended to improve vidualized management consulting. Better social interaction and work-related practices management can make firms more productive and attitudes. and ultimately increase hiring. There is evi- Rigorous impact evaluations of several dence that firms underinvest in management Jóvenes en Acción programs show mixed but skills because they underestimate its value or promising results. On average, the programs are credit constrained. increase formal wage employment at the rate of Programs seeking to promote managerial about four to six additional jobs per 100 par- skills in established small and medium-size ticipants. Given the low wage employment rates firms are on the rise in developing countries, for African youth from poor backgrounds, this including programs that give intensive or impact could be large in relative terms. The pro- individualized management consulting. Some grams also lead to significantly higher earnings, have been evaluated in quasi-experimental but the impacts are larger for female participants, or randomized studies, including studies by and it is unclear if there is a significant impact the Kaizen Institute of industrial clusters in on males. Evaluations of Jóvenes en Acción pro- Dar es Salaam and Addis Ababa.24 The evalu- grams and of recent training programs from ations suggest that such programs do lead to India and Jordan suggest that a more compre- the adoption of new management practices. hensive package—such as a combination of soft The magnitude of these effects is generally skills training and on-the-job training—is more small, with some exceptions. Among Ghanaian effective (see Maitra and Mani 2012 for India; metalwork firms, for instance, the percentage Groh et al. 2012 for Jordan). of firms keeping records increased 30 percent These findings suggest that a targeted, com- (Mano et al. 2012). prehensive (training plus internship) program Evidence on whether the adoption of new could be an effective instrument for promot- management practices actually improves busi- ing productive wage employment among dis- ness performance and employment is more advantaged youth. Such programs are being mixed, with few studies finding statistically adopted in Africa but have not yet been evalu- significant impacts on sales, profits, or survi- ated, and there are serious questions about vorship. The most promising finding is from a their cost-effectiveness in Africa. Jóvenes pro- field experiment on large Indian textile firms, grams were estimated to be cost-effective in in which management practices raised produc- Latin America, in the narrow sense that their tivity 17 percent in the first year and led within average impact on earnings exceeded the pro- three years to the opening of more production gram cost per participant.23 Because the mod- plants (Bloom, Sadun, and Van Reenen 2012). ern wage sector is relatively small in Africa, These results suggest that improved manage- the cost of subsidizing internships could be ment practices can create more jobs. higher there, and impacts on earnings lower. Given the limited but promising evidence, The small size of the wage sector also makes more pilot studies in Africa would be useful, scalability a bigger concern. Considering the including studies that examine issues related to mixed evidence on what type of training works targeting and selectivity. For instance, is limited best in these programs, they should be imple- managerial capital a constraint only for larger mented on a pilot basis, and the pilots should firms, as the Indian study suggests? Are the be evaluated to find the most cost-effective returns to managerial training highest when model for Africa. training targets firms with the most ability or raising productivity in africa’s modern Wage Enterprises to foster Job Growth for Youth 231 “potential” for success, and can that potential in many countries, combined with barriers to be identified? In industries where healthy com- imports and regional trade, suppress com- petition among firms does not always occur, it petition and reduce pressure to innovate and is difficult to identify such potential by look- improve productivity, even among firms that ing at current performance. Moreover, high- produce for domestic markets. potential firms may do well even in the absence Governments in Africa can do much to rem- of government-supported managerial training. edy this situation. The most important step is Small pilot programs directed at firms with to improve the business climate so that entre- alternative profiles could be useful for under- preneurship and productive firms can grow. standing where support is most needed. Key reform objectives are to improve access to finance and infrastructure services, improve trade logistics, and ease regulatory constraints Many Options to Promote to entrepreneurship. Some efforts—such as Competitive Modern Enterprises providing physical infrastructure for public utilities and transport—require costly invest- Modern manufacturing and services firms ment in the long run. Yet many other policy have been an engine of job growth in devel- reforms are not so expensive in monetary terms oped countries and more recently in East Asia. and can deliver huge impacts by reducing dis- In the long run, the same should be true for tortions and increasing efficiency. Selective and Africa. The modern enterprise sector presently spatially targeted support to emerging clusters accounts for a small share of African employ- can make the most of limited resources by pro- ment, and it is growing unevenly across the moting agglomeration economies. Localized region. The manufacturing sector, which has interventions of this sort should assist clusters been a vibrant source of new jobs in other by providing good infrastructure and support- regions, accounts for less than 3 percent of total ive regulation, but they should never try to employment in Africa. force clusters into existence by trying to pick Because African economies are small, “winners.” domestic demand alone cannot support a A skills shortage does not seem to be a bind- thriving modern enterprise sector, where so ing constraint on modern wage employment in much depends on achieving scale and diversity. Africa, but there is much evidence that a range Trade in manufacturing (and tradable services) of skills—cognitive, behavioral, and techni- may be essential for modern firms in these sec- cal—matters to productivity in firms. African tors to realize their full potential for productive youths need those skills if they are to be truly youth employment, but presently they export productive in modern firms, particularly when very little. Labor costs are generally not lower in this sector starts to take off. Governments African countries than in their potential com- should, first of all, foster a strong foundation petitors. If they cannot compete on the strength in basic skills by improving the quality of gen- of low labor costs, the region’s modern enter- eral education. Rather than directly providing prises must compete on the strength of their TVET, they should focus on providing “public productivity—but African firms are relatively goods” such as quality assurance and informa- unproductive. tion to develop a training sector that is efficient The lack of productivity signals deeper and relevant to the market for skills. For poor problems that are not identical across the sub- and disadvantaged youths to gain access to continent. In some countries, the cost of com- training, governments should provide finan- plementary inputs to labor (electricity, over- cial support in conjunction with information land transport, and so on) is too high. In others, that enables them to obtain the right kind of bureaucratic red tape delays investors’ access to training. Programs for disadvantaged youths land or permits or complicates efforts to move that integrate training with internships show goods (inputs and output) through ports. The promise, as do programs offering managerial high costs of financial intermediation are starv- training—but their scalability and affordability ing investors of capital. Small domestic markets in Africa remain unproven. 232 Youth EmploYmEnt in Sub-Saharan africa Notes 10. Associations between these business environ- 1. Africa’s wage employment sector is only partially ment variables and plant-level and aggregate “modern,” in the sense that many wage employ- productivity in World Bank enterprise survey ees work without a formal contract. On average, data are analyzed in several papers, including about 57 percent of wage employees have a con- Dollar, Hallward-Driemeier, and Mengistae tract with their employer. Among young people, (2005); Li, Mengistae, and Xu (2011); Harri- the share with a contract is 47 percent. son, Lin, and Xu (2013). See also Hall and Jones 2. New thinking on international trade reflects the (1999); Bartelsman, Haltiwanger, and Scarpetta major impact that trade can have on productiv- (2009). ity. An increasingly large share of the trade in 11. That is, the simple average of firm produc- goods now comes from trade in different vari- tivities, not weighted by their market shares, is eties of products within the same industry; for higher in Country A. example, the United States and European coun- 12. Specifically, it is possible to decompose indus- tries export cars to each other. Such trade is ben- try productivity into the average of firm-level eficial, because it allows different varieties to be TFPs—which would be the industry’s TFP produced at an efficient scale, which would not level if all firms had equal market shares—and be possible if each variety were sold only within an “allocative efficiency” component. This is its country of origin. This kind of trade also referred to as the Olley-Pakes decomposition of allows more efficient producers to take fuller aggregate productivity (Olley and Pakes 1996). advantage of their efficiency, because a larger See also Melitz and Polanec (2012) and Foster, market—a direct consequence of lower trade Haltiwanger, and Syverson (2008) for related barriers—has a disproportionately greater posi- decompositions. The Olley-Pakes decomposi- tive impact on the profitability of more efficient tion can be described as follows: let at be the firms. As more productive firms gain market weighted average of (log) TFP of a given indus- share at the expense of less productive ones, the try in year t and let ait be the log TFP of enter- industry’s potential for generating employment prises constituting the industry with respective rises (a point addressed later in the analysis of market shares, sit , where i indexes enterprises. Africa’s manufacturing competitiveness). Then at can be written as Nt 3. Less productive firms are limited to selling in the at = at + ∑ (sit – st )(ait – at) , domestic market, as the price they would have i −1 to charge to recover transport and other costs where letters with upper bars represent un- of exporting would not be competitive. See, for weighted industry means of variables. instance, Melitz and Trefler (2012). 13. Banerjee and Duflo (2012), for instance, show 4. The caveat with these data is that the surveys that firms are credit constrained in India. Udry collect information on firms that are formally and Anagol (2006) discuss evidence on excep- registered. In most countries, all firms above a tionally high rates of return to capital in Ghana, certain size are supposed to register, but in prac- which suggests credit constraints. tice, the registration rates and quality of the reg- 14. In India, for example, labor laws prevent firms istry information vary across countries. from adjusting employment in response to 5. Some evidence indicates that Sub-Saharan external shocks. See Adhvaryu, Chari, and Africa, along with South Asia, may have the low- Sharma (2013). est labor productivity in manufacturing in the 15. Among the indicators of the potential gains developing world. See Clarke (2011). from reductions in trade costs are the large and 6. An additional reason is that African firms sub- persistent cross-border disparities in the price of stitute labor for capital by using relatively labor- goods, earnings, and rates of return to capital intensive production methods, despite the fact among neighboring countries across the region. that many African countries have higher wages These suggest substantial room for increasing than China. employment opportunities for young people 7. Moreover, foreign direct investment in Africa is simply by making it easier for countries to trade dominated by investment in extractive indus- and invest across borders, without necessarily tries (World Bank 2013a). involving large international migration of peo- 8. The technical term is total factor productivity ple (Habyarimana and Mengistae 2013). (TFP). 16. Barker and Mengistae (2013). Beck, Dermigüç- 9. As discussed later, some recent studies, notably Kunt, and Maksimovic (2005) report that firms Bloom et al. (2013), measure organizational claiming to be more constrained by access to efficiency. credit are likely to grow more slowly. raising productivity in africa’s modern Wage Enterprises to foster Job Growth for Youth 233 17. Kenya Gazette Supplement no. 52 (Legislative Almlund, Mathilde, Angela Duckworth, James Supplement no. 31) and Legal Notice no. 97, of Heckman, and Tim Kautz. 2011. “Personality, 11 July 2008. Psychology, and Economics.” In Handbook of the 18. Cadot et al. (2013) found that new exporters are Economics of Education, edited by Eric Hanushek, more likely to survive beyond the first year when Stephen Machin, and Ludger Woessmann, 1–181. Amsterdam: Elsevier. other firms are exporting the same product to similar destinations. They argue that informa- Atchoarena, David, and Paul Esquieu, eds. 2002. tion spillovers account for the finding. Private Technical and Vocational Education in 19. Collier and Venables (2007). See also Monga Sub-Saharan Africa: Provision, Patterns, and (2012) and Lin and Monga (2011) on the instru- Policy Issues. IIEP/Prg.DA/01.300. Paris: Interna- tional Institute for Educational Planning. mentality of localization to that end in the con- text of industrial policy, which the framework Attanasio, Orazio P., Adriana D. Kugler, and Costas of Collier and Venables does not necessarily Meghir. 2011. “Subsidizing Vocational Training assume. for Disadvantaged Youth in Developing Coun- tries: Evidence from a Randomized Trial.” Ameri- 20. The issue of localizing interventions in support can Economic Journal: Applied Economics 3 (3): of clusters is often intertwined with the issue of 188–220. industrial policy (that is, policy targeted to spe- cific types of industries or firms). But the two Banerjee, Abhijit, and Esther Duflo. 2005. “Growth are quite distinct. Interventions to localize clus- Theory through the Lens of Development Eco- nomics.” In Handbook of Economic Growth, Vol. ters do not rely on identifying specific firms or 1A, edited by Philippe Aghion and Steven Dur- industries that should be clustered. For the cur- lauf, ch. 7, 473–552. Amsterdam: Elsevier. rent debate on industrial policy, as distinct from the spatial targeting of measures for improving ———. 2012. “Do Firms Want to Borrow More? Testing Credit Constraints Using a Directed the investment climate, some key references are Lending Program.” Massachusetts Institute of Lin (2011); Lin and Monga (2011); Pack and Technology, Department of Economics, Boston, Saggi (2006); Rodrik (2004). MA. 21. A survey of disadvantaged youth in Malawi found that only 13 percent had ever received Barker, M., and Taye Mengistae. 2013. “Business Environment, Market Distortions, and Employ- technical or vocational training (Hicks et al. ment in Africa: An Analysis of Business Survey 2011). Those who had not received any training Data on Manufacturing Industries and Services.” attributed it to high costs and limited access and Background paper for the Africa Region Report information. on Youth Employment in Africa, World Bank, 22. That evidence base does not yet exist for Africa. Washington, DC. 23. This assessment is based on particular assump- Bartelsman, Eric, John Haltiwanger, and Stefano tions about how long the earnings impact lasts. Scarpetta. 2009. “Cross-Country Differences in See Attanasio, Kugler, and Meghir (2011). Productivity: The Role of Allocation and Selec- 24. McKenzie and Woodruff (2013) review these tion.” IZA Discussion Paper 4578, Institute for evaluations of management and business prac- the Study of Labor, Bonn. tice training programs. Mano et al. (2012), Beaman, Lori, Niall Keleher, and Jeremy Magruder. Sonobe, Suzuki, and Otsuka (2011), and Kar- 2012. “Do Job Networks Disadvantage Women? lan, Knight, and Udry (2012) are Africa-specific Evidence from a Recruitment Experiment in evaluations. Malawi.” Working Paper, Northwestern Univer- sity, Department of Economics, Chicago. References Beck, Thorsten, Asli Dermigüç-Kunt, and Vojislav Adhvaryu, Achyuta, A. V. Chari, and Siddharth Maksimovic. 2005. “Financial and Legal Con- Sharma. 2013. “Firing Costs and Flexibility: Evi- straints to Growth: Does Firm Size Matter?” dence from Firms’ Employment Responses to Journal of Finance 60 (1): 37–77. Shocks in India.” Review of Economics and Statis- Bertrand, Marianne, Antoinette Schoar, and David tics 95 (3): 725–40. Thesmar. 2007. “Banking Deregulation and Almeida, Rita, Jere Behrman, and David Robalino, Industry Structure: Evidence from the French eds. 2012. The Right Skills for the Job? Rethinking Banking Reforms of 1985.” Journal of Finance Effective Training Policies for Workers. Washing- 62 (2): 597–628. ton, DC: World Bank. Betcherman, Gordon, Karina Olivas, and Amit Dar. Almeida, Rita, and Pedro Carneiro. 2009. “The 2004. “Active Labor Market Programs: New Evi- Return to the Firm Investment in Human Capi- dence from Evaluations with Particular Attention tal.” Labor Economics 16 (1): 97–106. to Developing and Transition Countries.” Social 234 Youth EmploYmEnt in Sub-Saharan africa Protection Discussion Paper 04012, World Bank, Africa. Paris: Agence Française de Développe- Washington, DC. ment; Washington, DC: World Bank. Biggs, Tyler, Manju Shah, and Pradeep Srivastava. Dinh, Hinh T., Dimitris A. Mavridis, and Hoa B. 1995. “Technological Capabilities and Learn- Nguyen. 2012. “The Binding Constraints on the ing in African Enterprises.” Technical Paper 288, Growth of Firms in Developing Countries.” In World Bank, Washington, DC. Performance of Manufacturing Firms in Africa: An Bloom, Nicholas, Benn Eifert, Aprajit Mahajan, Empirical Analysis, edited by Hinh T. Dinh and David McKenzie, and John Roberts. 2013. “Does George R. G. Clarke, 87–138. Washington, DC: Management Matter? Evidence from India.” World Bank. Quarterly Journal of Economics 128 (1): 1–51. Dinh, Hinh T., Vincent Palmade, Vandana Chandra, Bloom, Nicholas, Raffaella Sadun, and John Van and Frances Cossar. 2012. Light Manufacturing in Reenen. 2012. “The Organization of Firms across Africa: Targeted Policies to Enhance Private Invest- Countries.” Quarterly Journal of Economics 127 ment and Create Jobs. Washington, DC: World (4): 1663–705. Bank. Briceño-Garmendia, Cecilia, and Vivien Foster, eds. Djankov, Simeon. 2009. “The Regulation of Entry: 2010. Africa’s Infrastructure: A Time for Transfor- A Survey.” World Bank Research Observer 24 (2): mation. Washington, DC: World Bank. 183–203. Cadot, Olivier, Leonardo Iacovone, Martha Denisse Djankov, Simeon, Caroline Freund, and Cong S. Pierola, and Ferdinand Rauch. 2013. “Success Pham. 2010. “Trading on Time.” Review of Eco- and Failure of African Exporters.” Journal of nomics and Statistics 92 (1): 166–73. Development Economics 101 (C): 284–96. Dollar, David, Mary Hallward-Driemeier, and Taye Chari, Amalavoyal V. 2011. “Identifying the Aggre- Mengistae. 2005. “Investment Climate and Firm gate Productivity Effects of Entry and Size Performance in Developing Economies.” Eco- Restrictions: An Empirical Analysis of License nomic Development and Cultural Change 54 (1): Reform in India.” American Economic Journal: 1–31. Economic Policy 3 (2): 66–96. Dutz, Mark, Sergio Kannebley, Maira Scarpelli, and Cingano, Federico, and Paolo Pinotti. 2012. “Trust, Siddharth Sharma. 2012. “Measuring Intangible Firm Organization, and the Structure of Produc- Capital in an Emerging Market Economy: An tion.” Working Paper 2012-133, Bocconi Univer- Application to Brazil.” Policy Research Working sity, Paolo Baffi Centre on Central Banking and Paper 6142, World Bank, Washington, DC. Financial Regulation. Eifert, Benn, Alan Gelb, and Vijaya Ramachandran. Clarke, George R. G. 2011. “Manufacturing Firms 2008. “The Cost of Doing Business in Africa: in Africa.” In Performance of Manufacturing Evidence from Enterprise Survey Data.” World Firms in Africa: An Empirical Analysis, edited by Development 36 (9): 1531–46. Hinh T. Dinh and George R. G. Clarke, 47–86. Fafchamps, Marcel, and Simon Quinn. 2012. Washington, DC: World Bank. “Results of Sample Surveys of Firms.” In Per- Collier, Paul, and Anthony J. Venables. 2007. formance of Manufacturing Firms in Africa: An “Rethinking Trade Preferences: How Africa Can Empirical Analysis, edited by Hinh T. Dinh and Diversify Its Exports.” World Economy 30 (8): George R. G. Clarke, 139–211. Washington, DC: 1326–45. World Bank. ———. 2008. “Trade and Economic Performance: Fafchamps, Marcel, and Mans Soderbom. 2006. Does Africa’s Fragmentation Matter?” Paper pre- “Wages and Labor Management in African Man- sented at the Annual World Bank Conference on ufacturing.” Journal of Human Resources 41 (2): Development Economics, Cape Town, June 9–11. 346–79. Corrado, Carol, Charles Hulten, and Daniel Sichel. Farole, Thomas. 2011. “Special Economic Zones in 2005. “Measuring Capital and Technology: An Africa: Comparing Performance and Learning Expanded Framework.” In Measuring Capital in from Global Experiences.” World Bank, Washing- the New Economy: Studies in Income and Wealth, ton, DC. Vol. 65, edited by Carol Corrado, John Haltiwan- Filmer, Deon. 2012. “Challenges and Options for ger, and Daniel Sichel, 11–45. Chicago: Univer- Technical and Post-Basic Education in South sity of Chicago Press. Africa.” Policy Note, World Bank, Africa Region, Darvas, Peter, and Robert Palmer. 2012. “Technical Human Development Unit, Washington, DC. and Vocational Skills Development in Ghana.” Fisman, Ray, and Jakob Svensson. 2007. “Are Cor- World Bank, Washington, DC. ruption and Taxation Really Harmful to Growth? De Vreyer, Philippe, and François Roubaud, eds. Firm-Level Evidence.” Journal of Development 2013. Urban Labour Markets in Sub-Saharan Economics 83 (1): 63–75. raising productivity in africa’s modern Wage Enterprises to foster Job Growth for Youth 235 Fitzmaurice, Mike, and Olivier Hartmann. 2013. Skills, Employability, and Training for Disad- “Border-Crossing Monitoring along the North- vantaged Youth: Evidence from a Randomized ern Corridor.” SSATP Working Paper 96, World Evaluation Design.” IZA Discussion Paper 6617, Bank, Washington, DC. Institute for the Study of Labor, Bonn. Foster, Lucia, John C. Haltiwanger, and Chad Syver- IFC (International Finance Corporation). 2013. son. 2008. “Reallocation, Firm Turnover, and IFC Job Study: Assessing Private Sector Contri- Efficiency: Selection on Productivity or Prof- butions to Job Creation and Poverty Reduction. itability?” American Economic Review 98 (1): Washington, DC: IFC. 394–425. Johanson, Richard K., and Arvil Van Adams. 2004. Fujita, Masahisa, Paul Krugman, and Anthony J. Skills Development in Sub-Saharan Africa. Wash- Venables. 1999. The Spatial Economy: Cities, ington, DC: World Bank. Regions, and International Trade. Cambridge, Kalemli-Ozcan, Sebnem, Laura Alfaro, and Vadym MA: MIT Press. Volosovych. 2008. “Why Doesn’t Capital Flow Gelb, Alan, Christian Meyer, and Vijaya Ramach- from Rich to Poor Countries? An Empirical andran. 2013. “Does Poor Mean Cheap? A Com- Investigation.” Review of Economics and Statistics parative Look at Africa’s Industrial Labor Costs.” 90 (2): 347–68. CGD Working Paper 325, Center for Global Karlan, Dean, Ryan Knight, and Christopher Udry. Development, Washington, DC. 2012. “Hoping to Win, Expected to Lose: Theory Groh, Matthew, Nandini Krishnan, David McKen- and Lessons on Micro Enterprise Development.” zie, and Tara Vishwanath. 2012. “Soft Skills or Working Paper, Yale University, New Haven, CT. Hard Cash? The Impact of Training and Wage Kitaev, Igor, with contributions from T. Coleman, Subsidy Program on Female Youth Employment J. Glover, and B. Kaluba. 2002. “Synthesis of Main in Jordan.” Policy Research Working Paper 6141, Findings from Two Studies on Private Technical- World Bank, Washington, DC. Vocational Education and Training in Ghana and Habyarimana, James, and Taye Mengistae. 2013. Zambia (Phase II).” Draft IIEP/Prg.DA/02.365, “Labor Market Integration in Africa: Evidence International Institute for Educational Planning, from Trading Blocks.” Background paper for the Paris. Africa Region Report on Youth Employment in Klapper, Leora, Luc Laeven, and Raghuram Rajan. Africa, World Bank, Washington, DC. 2006. “Entry Regulation as a Barrier to Entrepre- Hall, Robert E., and Charles I. Jones. 1999. “Why neurship.” Journal of Financial Economics 82 (3): Do Some Countries Produce So Much More 591–629. Output per Worker Than Others?” Quarterly Krishna, Pravin, and Devashish Mitra. 1998. “Trade Journal of Economics 114 (1): 83–116. Liberalization, Market Discipline, and Productiv- Hao, Janet, and Charles R. Hulten. 2011. “Intan- ity Growth: New Evidence from India.” Journal of gible Capital and the Sources of Recent Chinese Development Economics 56 (2): 447–62. Economic Growth.” Paper prepared for the Con- Krishnan, Pramila, and Irina Shaorshadze. 2013. ference on Research in Income and Wealth, Pro- “Technical and Vocational Education and Train- ductivity Program Workshop, Summer Institute ing in Ethiopia.” Working Paper, International 2011, National Bureau of Economic Research, Growth Centre, London, February. Cambridge, MA, July 19. Kuddo, Arvo. 2012. “Public Employment, Ser- Harrison, Ann E., Justin Yifu Lin, and L. Colin Xu. 2013. “Africa’s (Dis)advantage: The Curse of vices, and Activation Policies.” Social Protection Party Monopoly.” NBER Working Paper 18683, and Labor Discussion Paper 1215, World Bank, National Bureau of Economic Research, Cam- Washington, DC. bridge, MA. Li, Wei, Taye Mengistae, and Lixin Colin Xu. 2011. Hicks, Joan Hamory, Michael Kremer, Isaac Mbiti, “Diagnosing Development Bottlenecks: China and Edward Miguel. 2011. “Vocational Education and India.” Oxford Bulletin of Economics and Sta- Voucher Delivery and Labor Market Returns: A tistics 73 (6): 722–52. Randomized Evaluation among Kenyan Youth.” Lin, Justin Yifu. 2011. “New Structural Econom- Report for Spanish Impact Evaluation Fund ics: A Framework for Rethinking Development.” Phase II. World Bank Research Observer 26 (2): 193–221. Hsieh, Chang-Tai, and Peter J. Klenow. 2009. “Mis- Lin, Justin Yifu, and Célestin Monga. 2011. “Growth allocation and Manufacturing TFP in China and Identification and Facilitation: The Role of the India.” Quarterly Journal of Economics 124 (6): State in the Dynamics of Structural Change.” 1403–48. Development Policy Review 29 (3): 259–310. Ibarraran, Pablo, Laura Ripani, Bibiana Tapoada, Maitra, Pushkar, and Subha Mani. 2012. “Learn- Juan Miguel Villa, and Brigida Garcia. 2012. “Life ing and Earning: Evidence from a Randomized 236 Youth EmploYmEnt in Sub-Saharan africa Evaluation in India.” Department of Econom- Rodrik, Dani. 2004. “Industrial Policy for the ics Discussion Paper 44/12, Monash University, Twenty-First Century.” CEPR Discussion Paper Melbourne. 4767, Centre for Economic Policy Research, Mano, Yukichi, Alhassan Iddrisu, Yutaka Yoshino, London. and Tetsushi Sonobe. 2012. “How Can Micro and Rotunno, Lorenzo, Pierre-Louis Vézina, and Zheng Small Enterprises in Sub-Saharan Africa Become Wang. 2012. “The Rise and Fall of (Chinese) More Productive? The Impacts of Experimental African Apparel Exports.” CSAE Working Paper Basic Managerial Training.” World Development 2012-12, Centre for the Study of Africa Econo- 40 (3): 458–68. mies, University of Oxford. McKenzie, David, and Christoper M. Woodruff. Sachs, Jeffrey D., and Andrew M. Warner. 2001. 2013. “What Are We Learning from Business “The Curse of Natural Resources.” European Training and Entrepreneurship Evaluations Economic Review 45 (4-6): 827–38. around the Developing World?” World Bank Saleman, Yannick, and Luke S. Jordan. 2013. “The Research Observer. doi: 10.1093/wbro/lkt007. Implementation of Industrial Parks: Some Melitz, Marc J., and Sašo Polanec. 2012. “Dynamic Lessons Learned in India.” World Bank, Wash- Olley-Pakes Productivity Decomposition with ington, DC. Entry and Exit.” NBER Working Paper 18182, Sondergaard, Lars, and Mamta Murthi, with Dina National Bureau of Economic Research, Cam- Abu-Ghaida, Christian Bodewig, and Jan Rut- bridge, MA. kowski. 2012. “Skills, Not Just Diplomas: Manag- Melitz, Marc J., and Daniel Trefler. 2012. “Gains ing for Results in Education Systems in Eastern from Trade When Firms Matter.” Journal of Europe and Central Asia.” World Bank, Washing- Economic Perspectives 26 (2): 91–118. ton, DC. Mingat, Alain, Blandine Ledoux, and Ramahatra Sonobe, Tetsushi, Aya Suzuki, and Keijiro Otsuka. Rakotomalala. 2010. “Developing Post-Primary 2011. “Kaizen for Managerial Skills Improvement Education in Sub-Saharan Africa: Assessing the in Small and Medium Enterprises: An Impact Financial Sustainability of Alternative Pathways.” Evaluation Study.” World Bank, Washington, World Bank, Washington, DC. DC. http://siteresources.worldbank.org/DEC/ Monga, Célestin. 2012. “Winning the Jackpot-Jobs Resources/FinalVolumeIV.pdf. Dividend in a Multi-Polar World.” Paper pre- Svensson, Jakob. 2003. “Who Must Pay Bribes and sented at the International Economics Associa- How Much? Evidence from a Cross-Section of tion–World Bank Roundtable, “Industrial Policy Firms.” Quarterly Journal of Economics 118 (1): in Africa,” Pretoria, South Africa, July 3–4. 207–30. OECD (Organisation for Economic Co-operation Syverson, Chad. 2011. “What Determines Produc- and Development). 2010. Measuring Innovation: tivity?” Journal of Economic Literature 49 (2): A New Perspective. Paris: OECD. 326–65. Olley, Steven, and Ariel Pakes. 1996. “The Dynam- Tanzania, United Republic of, and UNESCO ics of Productivity in the Telecommunications (United Nations Educational, Scientific, and Cul- Equipment Industry.” Econometrica 64 (6): tural Organization). 2012. “Management Issues 1263–97. in Higher, Technical, and Vocational Education Ozier, Owen. 2013. “The Impact of Secondary and Training in Tanzania.” In Education Sector Schooling in Kenya: A Regression Discontinuity Analysis: Beyond Primary Education, the Quest for Analysis.” Paper presented at the Annual Meeting Balanced and Efficient Policy Choices for Human of the Population Association of America, New Development and Economic Growth, ch. 8, 307– Orleans, April 11–13. 32. Dar es Salaam: Government of Tanzania and UNESCO. Pack, Howard, and Kamal Saggi. 2006. “The Case for Industrial Policy: A Critical Survey.” World TESDA (Technical Education and Skills Develop- Bank Research Observer 21 (2): 267–97. ment Authority). 2012a. “Metals and Engineer- ing Sector: Supply and Demand in Computer Piopiunik, Marc, and Paul Ryan. 2012. “Improv- Numerical Control (CNC) Machine Operation ing the Transition between Education/Training and Other Related TVET Qualification.” Labor and the Labour Market: What Can We Learn Market Intelligence Report LMIR-ST-PO 03-05- from Various National Approaches?” EENEE 12, TESDA, the Philippines. Analytical Report 13, European Expert Net- work on Economics of Education, http://www. ———. 2012b. “Technical Education and Skills eenee.de/portal/page/portal/EENEEView/_ Development Monitoring and Evaluation Sys- generische_page_eenee?content=eenee-home. tem: Towards Increasing Efficiency, Effective- htm&language=us. ness, and Relevance of TVET in the Philippines.” raising productivity in africa’s modern Wage Enterprises to foster Job Growth for Youth 237 Presentation made at the Third International ———. 2011c. “Strengthening Skills and Employ- Congress on Technical and Vocational Education ability in Peru: Final Report.” World Bank, and Training, Building Skills for Work and Life, Washington, DC. Shanghai. http://www.unesco.org/education/ ———. 2012a. “Analysis of Uganda’s System of TVET2012/roundtable/3/M-DelaRama.pdf. Business, Technical-Vocational Education and Udry, Christopher R., and Santosh Anagol. 2006. Training (BTVET).” World Bank, Washington, “The Return to Capital in Ghana.” Discussion DC. Paper 932. Yale University, Economic Growth ———. 2012b. Doing Business 2013: Smarter Regu- Center, New Haven, CT. SSRN: http://ssrn.com/ lations for Small and Medium-Size Enterprises. abstract=893023. Washington, DC: World Bank. Wood, Adrian, and Kersti Berge. 1997. “Export- ———. 2012c. “Financing Small- and Medium- ing Manufactures: Human Resources, Natural Sized Enterprises in Rwanda.” World Bank, Resources, and Trade Policy.” Journal of Develop- Washington, DC. ment Studies 34 (1): 35–59. ———. 2012d. World Development Report 2013: Wood, Adrian, and Jörg Mayer. 2001. “Africa’s Jobs. New York: Oxford University Press. Export Structure in a Comparative Perspective.” Cambridge Journal of Economics 25 (3): 369–94. ———. 2013a. Africa’s Pulse 7 (April). World Bank, Washington, DC. World Bank. 2004. “An Assessment of the Invest- ment Climate in Zambia.” World Bank, Washing- ———. 2013b. “Tanzania Economic Update, ton, DC. Opening the Gates: How the Modernization of the Port of Dar es Salaam Can Transform ———. 2008. Finance for All? Policies and Pitfalls in Tanzania.” Report 77729, World Bank, Washing- Expanding Access. Policy Research Report 41972, ton, DC. World Bank, Washington, DC. World Economic Forum. 2012. The Global Com- ———. 2009. “Zambia: Second Investment Cli- petitiveness Report 2012–2013: Full Data Edition. mate Assessment; Business Environment Issues Geneva: World Economic Forum. in Diversifying Growth.” Vols. 1 and 2. World Bank, Washington, DC. Xu, Lixin Colin. 2011. “The Effects of Business Environments on Development: A Survey of ———. 2010. “The Business Environment in New Firm-Level Evidence.” World Bank Research Southern Africa: Issues in Trade and Market Observer 26 (2): 310–40. Integration.” Vols. s 1 and 2. World Bank, Washington, DC. Zeng, Douglas. 2006. “Knowledge, Technology, and Cluster-based Growth in Africa: Findings ———. 2011a. “Botswana: Skills for Economic from Eleven Case Studies of Enterprise Clusters Diversification and Economic Growth.” World in Africa.” Knowledge for Development Bank, Washington, DC. (K4D) Program, World Bank Institute, ———. 2011b. Doing Business 2012: Doing Business Washington, DC. in a More Transparent World. Washington, DC: World Bank. Note: This chapter draws extensively, especially in the section presenting and discussing Africa’s wages, unit labor costs and competitiveness, on two background papers: Trapeznikova, Ija. 2012a. “Determinants of Earnings Differentials and Wage Trends in Africa’s Modern Wage Sector.” Background paper for the Africa Region Report on Youth Employment in Africa, World Bank, Washington, DC. ______. 2012b. “Competitiveness and Job Creation: The Case of the Manufacturing and Service Sectors in Africa.” Background paper for the Africa Region Report on Youth Employment in Africa, World Bank, Washington, DC. (Reference to these papers was inadvertently ommitted in a previous version of this document) FOCUS Youth Unemployment Financial the Inclusion andin Transition South Africa:to 3 NOTE 4 Sustainable Different Livelihoods for Configuration, Young People Different Approaches S outh Africa’s labor market—unlike most production—at the expense of low-skilled others in Sub-Saharan Africa—has a high workers and of industries that use low-skilled share of wage employment off the farm, labor intensively (Mengistae 2011). While very little employment in household enterprises South Africa’s industry has become more com- (HEs), and an even smaller share of employ- petitive over the last decade, it remains less ment in agriculture (4.5 percent). South Africa competitive than that of many trading part- also stands apart from other African countries ners. Stringent labor market regulations are in having a high unemployment rate, estimated typically put forward as the primary cause of at 25 percent (based on a narrow definition of high unemployment. Yet close scrutiny of the unemployment) in 2012. evidence suggests that the story is not just High unemployment is nothing new in about labor market rigidities. Deficiencies in South Africa, which had an unemployment rate economic growth, coupled with the rapidly of 39 percent in 2005, compared to 17 percent growing working-age population, are major in Africa’s other upper-middle-income coun- contributors to high unemployment in South tries. At 51 percent, unemployment among Africa. those ages 15–24 was twice the national aver- By many standards, South Africa has age in 2012. South Africa’s high rate of youth rigid labor market laws and high barriers to unemployment reflects the high rate of unem- entry in the HE sector. The minimum wage ployment throughout the country, which is imposed by the regulations is high, so the worsened by circumstances specific to youth. ratio of minimum to median wage is 1.52 for Especially unique is the combination of unskilled workers and can go as high as 2.66 high unemployment and a very small HE sec- in some sectors (Bhorat, Kanbur, and Mayet tor. The ratio of HE employment to unem- 2011). High unionization and the use of ployment in South Africa was estimated to be wage bargaining councils raise wages further: 0.30–0.48 in 2012, which is much smaller than conservative estimates show that union jobs the ratio in middle-income countries of Asia enjoy a wage premium of at least 17 percent. and Latin America (11 and 7, respectively) or As a result, the labor market is segmented. other countries in Sub-Saharan Africa (4.7). Well-paying nonfarm wage jobs occupy the In South Africa, residual labor is not absorbed upper end of the spectrum, and a pool of the by the informal sector, as it is in other African unemployed, queuing for wage jobs, occupies countries or even in middle-income countries the lower end. High barriers to entry in the elsewhere in the world. In South Africa, larger HE sector—created mainly by stringent local domestic firms occupy the space that HEs government by-laws, limited access to capital, occupy in many other African countries. and poor entrepreneurial skills—ensure that very few of those who fail to get a wage job join the nonfarm HE sector. Consequently, Conventional Perceptions of the a high level of open unemployment coexists Causes of Unemployment with a small informal sector that cannot be explained by high reservation wages among High wage levels and high real unit labor costs the unemployed (Kingdon and Knight 2004; played a part in dampening the demand for Mitra 2010). labor and encouraging industry to shift into Yet the contribution of labor market rigidi- 238 more capital-intensive modes and sectors of ties to unemployment appears exaggerated. focus note 4 239 An employment wage elasticity of –0.7 sug- not only from the structure of growth but also gests that minimum wage laws and the wage from its slow pace in relation to demographic premiums exacted by unions and bargaining trends. Most of the growth in employment councils do reduce employment, but the size between 1995 and 2005 was in skilled employ- of the elasticity also implies that wage rigidi- ment, which grew 22 percent, while unskilled ties only partially explain why unemployment employment declined 13 percent (figure F4.1, is so high. It is important to note that viola- panel a). Demand for unskilled labor is weak tion of minimum wage requirements is high at a time when the unskilled labor force is in South Africa, reaching 50 percent in 2007 expanding significantly. Owing to this pattern (Bhorat, Kanbur, and Mayet 2011). Violations of skills-biased growth, a great scarcity of skills are highest in sectors where minimum wages exists alongside high unemployment among are higher, suggesting that minimum wages are the unskilled. only moderately binding. Revised estimates In the decade between 1995 and 2005, prior of the unionization premium are also much to the global financial crisis, the employment smaller than previous estimates (Bhorat, Goga, elasticity of growth (ranging from 0.6 to 0.8) and van der Westhuizen 2012). Indeed, recent and growth of gross domestic product (GDP) estimates indicate that wage rigidities account were comparable in South Africa and other for only 1 percentage point of unemployment middle-income countries, except for China and in South Africa (Magruder 2012). India. Population growth was not comparable, however. Rapid population growth reduced growth per capita in South Africa compared The Real Unemployment Problem: to other middle-income countries (figure F4.1, Slow Growth panel b). South Africa’s labor force grew 46 per- cent between 1995 and 2005, outstripping the To a larger extent, South Africa’s unemploy- 29 percent growth in employment. Growth in ment problem is a growth problem, arising the labor force was propelled by an increase in Figure F4.1 Patterns in economic and employment growth in South Africa a. Employment growth by skill category, 1995–2005 b. GDP growth in South Africa and other countries 25 Argentina 20 Brazil Botswana 15 Chile 10 China Percent 5 India Indonesia 0 Malaysia –5 Poland –10 Turkey South Africa –15 Skilled Semi-skilled Unskilled Middle-income countries Upper-middle-income countries 0 2 4 6 8 10 12 Percent Per capita growth GDP growth Sources: World Bank 2011; Bhorat, Kanbur, and Mayet 2011. 240 Youth EmploYmEnt in Sub-Saharan africa labor force participation as well as population Improve the Targeting and Design of growth. Labor force participation rose 11 per- Wage Subsidies centage points between 1997 and 2005, mostly Subsidies are expected to stimulate labor because post-apartheid norms enabled greater demand by mitigating the impact of high mini- involvement in economic activity, especially mum wages, union premiums, or high reser- among women. At the same time, the share vation wages. Subsidies may not have much of of the working-age population expanded by 6 an effect on demand for labor in South Africa, percentage points between 1990 and 2010. The however, because it is difficult to substitute high unemployment that accompanied growth unskilled for skilled labor (Go et al. 2010). in the working-age population is one indica- Subsidies may serve only to move those who tion that South Africa’s demographic dividend are currently employed into other jobs, and did not materialize. their effects may well be temporary. High vio- lation of the minimum wage law in sectors with the highest minimum wage ratios will further Inefficient Job Search and reduce the impact of wage subsidies. Matching Processes Wage subsidies might be more effective if they target sectors where the response will For young people in particular, the effects be higher and are designed to reduce the dis- of South Africa’s oversubscribed market for placement of workers from one job to another. unskilled labor are amplified by frictions in job Although preliminary evidence from an impact search and matching. The quality of unskilled evaluation in South Africa showed that the wage labor is extremely varied, and unskilled work- subsidy program increased the probability of ers’ low level of education offers little indica- getting and keeping employment, a negligible tion of their real ability to prospective employ- proportion of employers claimed the subsidy. ers. Firms also face high hiring and firing costs This finding suggests that the real impact of owing to tight regulations. To minimize hiring the subsidy was that it improved the efficiency costs, firms tend to hire unskilled labor through of search and matching instead of reducing word of mouth and other nonpublic advertise- the real wage that the employer would have to ments; to screen for productivity, they rely on pay. The general finding is that although youth referrals. For young people, with their limited wage subsidies may have a positive impact, experience, inability to signal their real apti- their efficacy is conditioned by other labor tude, poor connections, and poor information market conditions, such as the flexibility of the on job openings, the job search can become labor market. A youth wage subsidy is just one very long and demoralizing, and the age at instrument—of many—to tackle youth unem- which they obtain their first wage job will con- ployment. If wage rigidities prove to be the tinue to increase. primary cause of unemployment, eliminating labor rigidities would be a longer-term solution to the problem. If overall labor demand is low, Policies to Reduce Unemployment then it will need to increase. Various policies have been pursued and pro- Provide Incentives for Labor-Intensive posed to reduce unemployment in South Growth Africa. Popular proposals on the demand Public works programs have not significantly side include wage subsidies, labor regulation improved employment in South Africa, based reform, and public works; on the supply side, on evidence from the Expanded Public Works they include reforms in education and train- Program (McCord 2007). A better alternative ing systems, as well as support for entrepre- is to implement national policies that not only neurship. What might be done to improve the increase the pace of economic growth but also potential for such strategies to work—espe- create incentives for labor-intensive firms to cially for youth? form and to locate near the pool of labor— focus note 4 241 for example, by establishing special economic licensing may work in the opposite direction by zones (as discussed in chapter 6) or offering increasing the administrative burden for HEs fiscal incentives favorable to labor-intensive (as well as small and medium enterprises). industries. References Reform Education and Training Systems Bhorat, Haroon, Sumayya Goga, and Carlene van der Westhuizen. 2012. “Institutional Wage South Africa’s long-term employment issues Effects: Revisiting Union and Bargaining Coun- cannot be addressed without reforms in edu- cil Wage Premia in South Africa.” South African cation and training systems. Such reforms can Journal of Economics 80 (3): 400–14. help to improve the productivity of the labor Bhorat, Haroon, Ravi Kanbur, and Natasha Mayet. pool, narrow the wide gap between the skills 2011. “The Determinants of Minimum Wage that the labor market demands and workers Violation in South Africa.” Working Paper WP can supply, and increase the signaling ability 2011-05, Cornell University, Charles H. Dyson School of Applied Economics and Management, of education, which, in turn, will improve the Ithaca, NY. search and matching efficiency of the labor Botha, M., G. H. Nieman, and J. J. van Vuuren. market. 2006. “Evaluating the Women Entrepreneurship The technical and vocational education Training Programme: A South African Study.” and training (TVET) system is plagued by low International Indigenous Journal of Entrepreneur- enrollment, low completion, and low prob- ship, Advancement, Strategy, and Education 2 (1): ability of finding employment, but it could be n.p. http://www.indigenousjournal.com/read. html. improved by reforms that strengthen basic edu- cation (an area where South Africa performs Filmer, Deon. 2012. “Challenges and Options for Technical and Post-Basic Education in South poorly). South Africa would also benefit from Africa.” Policy Note, World Bank, Africa Region, restructuring the further education and train- Human Development Unit, Washington, DC. ing (FET) system to permit part-time learning Go, Delfin S., Marna Kearney, Vijdan Korman, (which would make FET more flexible and Sherman Robinson, and Karen Thierfelder. 2010. accessible) and to introduce apprenticeships “Wage Subsidy and Labour Market Flexibility in and closer engagement with industry (which South Africa.” Journal of Development Studies 46 would make FET more relevant to the work- (9): 1481–502. place). Sector education and training authori- Kingdon, Geeta Gandhi, and John Knight. 2004. “Unemployment in South Africa: The Nature of ties should have a simplified bureaucracy, move the Beast.” World Development 32 (3): 391–408. to demand-driven financing, and delegate Magruder, Jeremy R. 2012. “High Unemployment training to the most competent institutions yet Few Small Firms: The Role of Centralized using a performance-based financing approach Bargaining in South Africa.” American Economic (see discussion in Filmer 2012). Journal: Applied Economics 4 (3): 138–66. McCord, Anna. 2007. “Training within the South Revise By-Laws and Offer Focused African National Public Works Programme.” In Human Resources Development Review 2008: Training to Support Entrepreneurial Education, Employment, and Skills in South Activities Africa, edited by Andre Kraak and Karen Press. South Africa should make it easier for young Cape Town: HSRC Press. people to become self-employed by intro- Mengistae, Taye. 2011. “Are South African Wages ducing by-laws that are compatible with Too High or Growing Too Fast? A Compari- self-employment and by offering financial son of Manufacturing Pay and Productivity in Selected Middle-Income Economies.” World support. Although international evidence is Bank, Washington, DC. generally mixed, the evidence on the impact Mitra, Sophie. 2010. “Disability Cash Transfers in of the Women Entrepreneurship Programme the Context of Poverty and Unemployment: in South Africa suggests that well-structured The Case of South Africa.” World Development 38 entrepreneurship training would enhance suc- (12): 1692–709. cess (Botha, Nieman, and van Vuuren 2006). World Bank. 2011. World Development Indicators. Current proposals to strengthen business Washington, DC: World Bank. Chapter 7 Conclusion: Building an Effective Youth Employment Policy The challenge of youth employment in Sub- and demographic profile. For that reason, this Saharan Africa is not amenable to simple solu- report has emphasized short- and medium-term tions. It reflects the intricate web of challenges strategies that will address current challenges and opportunities that links all countries in a while laying a foundation for the employment globalized world. For Africa now, the key employ- agenda to succeed in the long run. ment issue is that productivity—and therefore earnings—in many jobs is low, while aspira- tions, especially among young people, are high A Programmatic Approach and rising. Despite progress in many countries, most of Africa’s young people will not have an Unemployment, which is typically low in easy or structured transition to a sustainable low- and lower-middle-income countries, is livelihood, one of the core aspects of adulthood. just the tip of the iceberg. Most Africans can- All stakeholders have a role to play in supporting not afford to be unemployed. They work, but this transition. they earn little for that work. The challenge of Most African countries endorse the medium- youth employment requires a comprehensive to long-term agenda to increase employment approach designed to relieve the constraints in the nonfarm modern wage sector—in other that prevent young people and the private sec- words, to bring about a structural transforma- tor from seizing opportunities and increasing tion of their economies. If Africa follows Asia’s productivity in agriculture, household enter- model, this effort will rely primarily on expand- prises, and the modern wage sector. Govern- ing export-oriented manufacturing, with its high ments need to adopt a holistic view of how to requirements for labor. But even as governments address the situation—they need to “own” the embark on policies to foster such a transforma- whole problem. tion, they increasingly recognize that the process There is a common tendency to believe that will take time, given Africa’s small industrial base government-provided training in technical 243 244 Youth EmploYmEnt in Sub-Saharan africa and vocational skills will suffice to unlock the Rising to the youth employment challenge potential of youth. However, such action, by requires two kinds of action. Governments itself, will not address the more fundamental must act now for the short term, by addressing problems that prevent young people from find- the immediate constraints that hold youth back ing better opportunities. The gap in technical from finding higher-productivity work. They skills is only one facet of the youth employment also must act now for the medium term, by tak- challenge. A comprehensive approach to tack- ing decisions and actions now so that the chal- ling the challenge will address constraints on lenge of youth employment is easier to address both the human capital as well as the business in the future (table 7.1). environment sides of the problem. As an overall principle, it is important to recognize that youths will find their opportuni- “Do Now for Now”: Address the ties in the private sector, whether in agriculture, household enterprises, or the modern wage Constraints Facing Households sector. Governments should focus on provid- and Firms ing public goods and services that support pro- ductivity in both household economic activity Priorities to Get Youth into Productive and enterprise economic activity. Government Household Economic Activities should be an enabler for job creation. Both farm and nonfarm household enter- A key enabling action is to ensure that each prises suffer from financial exclusion. Lifting cohort of youth entering the workforce has a barriers to credit, financing, and savings vehi- solid foundation of skills acquired through cles will support greater productivity—and basic education. In addition to getting children growth—in the agricultural and nonagricul- to attend school, a critical priority is to ensure tural sectors. Improving access to credit and that substantial learning occurs while they are financing is important for financial inclusion, in school. Given the growing evidence that but the key is to facilitate saving. Saving can social and behavioral skills are important for be fostered through local savings groups and productivity, education systems could consider mobile money (especially attractive to youth), instilling such skills in combination with the as well as the more established banking sector. other, more traditional, cognitive skills. Mechanisms for saving, which help to build The defining feature of youth is that it is a the habit of saving, will increase the funds time of transitions, but youth in Africa are not available for investing and enable mobilizing finding well-marked pathways to stable and additional financing. productive livelihoods. While efforts should help all youth to make the transition, specific groups will face particular challenges along the Priorities for Agriculture way. For example, increasing young women’s Policies need to enable rural youth to access empowerment by increasing their agency— land. Approaches could include increasing their control over their own destinies—has been the security of tenure, ownership, and sales shown to reduce negative outcomes such as to encourage fluid land markets to develop. early pregnancy and exposure to human immu- Strategies will involve reducing the cost of land nodeficiency virus/acquired immunodeficiency transfer, promoting rental markets, encourag- syndrome (HIV/AIDS) and to have positive ing youth-sensitive land redistribution, and impacts on their earnings. And new approaches promoting youth-focused administration of should be tried. The new communication tech- communal or government land. nologies that are spreading throughout Africa New approaches to agricultural extension (and that hold so much appeal for youth) offer promise to overcome previous shortcomings. innovative options for government to reach To expand the repertoire of agricultural skills youth with messages, to help them to shape and increase access to information, these high- their employment aspirations, and to help quality, demand-driven extension programs youth learn about employment opportunities. should target youth. conclusion: building an Effective Youth Employment policy 245 Table 7.1 Priority actions to take now to address the youth employment challenge “Do now for now”: “Do now for later”: Area for intervention Actions to affect the current cohort of youth Actions to affect future cohorts of youth agriculture 1. Enable rental markets for land 1. Establish effective land registration and transaction systems 2. pilot intergenerational land transfer programs 2. Scale up intergenerational land transfer programs, based on lessons from pilot programs 3. Support high-quality, demand-driven extension services (covering information as well as skills) 3. mainstream youth into smart interventions aimed at increasing productivity (producer organizations, livestock development, 4. link agricultural credit to extension services irrigation, and others) 4. build skills through rapid improvements in education systems in rural areas agriculture and household 5. promote rural village savings and loan associations and self-help enterprises groups 6. Enable financial inclusion for households 7. use safety net programs as a platform to deliver interventions to disadvantaged youth household enterprises 8. Develop a national strategy for household enterprises that reflects 5. build foundational skills through rapid improvements in education the voice of their ownship and youth systems 9. Ensure access to workspace and infrastructure for household 6. address infrastructure needs of household enterprises in urban enterprises through improved urban policy development planning 10. leverage nongovernmental organizations to deliver interventions that support disadvantaged youth to enter the sector by addressing multiple constraints (building a range of skills together, building skills along with providing access to finance) modern wage sector 11. reduce the cost of infrastructure services by addressing quality 7. increase the quantity of infrastructure services and efficiency 8. Expand regional markets for products 12. address logistics bottlenecks 9. build foundational skills through rapid improvements in education 13. reduce corruption and the cost of business start-up systems 14. reform technical and vocational education and training and pursue 10. improve access to credit through financial sector reform public-private partnerships for delivering demand-driven training cross-cutting areas 15. increase awareness of opportunities and pathways to self- 11. promote early child development and nutrition to build a stronger employment, especially for young women foundation for skills development 16. consider second-chance education for basic skills 12. build behavioral skills (consider reforms within the school system) 13. reduce fertility rates to lower the size of future youth cohorts (through more girls’ education, improved maternal and child health, increased access to family planning) 14. build better employment data and a stronger evidence base to identify country constraints, priorities, and opportunities Smart interventions to raise productiv- voice of those working in the HE sector to be ity, such as effective producer organizations, heard. When the needs of HE owners cannot improved livestock production systems, or be expressed and channeled, it is not surpris- sustainable irrigation and water management, ing that policy makers and politicians neglect should mainstream youth. Projects should nei- to cater to those needs. ther treat youth as an enclave nor ignore them. To ensure that HEs can be established and operate legally and productively, urban policies Priorities for Household Enterprises should provide adequate locations where HEs A first step in most countries is to develop a can work and sell their products. In addition, national strategy for household enterprises smart urban planning should provide access to (HEs), because at present they are ignored, if essential support services for HEs, such as light- not suppressed. Such a strategy will involve ing, water, sanitation, and effective security. working with local governments as a part of Because a very large share of Africa’s new urbanization and local economic development nonfarm employment opportunities will need efforts. In addition to developing plans, the to come from individual self-employment, strategy should explicitly develop a way for the projects should encourage youth to navigate a 246 Youth EmploYmEnt in Sub-Saharan africa successful transition to starting up and operat- transportation and other infrastructure, cus- ing an HE. Depending on the individual and toms, and border control. sector, this assistance may include support for Another priority is to develop and sustain a developing a range of skills (behavioral, life, policy environment that enables the most pro- technical, and business skills) as well as access ductive firms to enter and grow. Here, the vital to finance for starting a business. Integrated steps include improving access to finance and approaches that address multiple constraints land, removing barriers to entry, and reducing simultaneously are promising. the cost of corruption. Governments should leverage nongovern- The poor quality and scarcity of infrastruc- mental organizations (NGOs) and steer them ture services clearly inhibit competitiveness, to support young people who are well placed to but the best immediate solution is not neces- enter the HE sector. Such support can involve sarily to provide more infrastructure. Strategies NGOs in delivering integrated approaches to must pay attention to the cost and quality of removing productivity constraints, with a focus infrastructure services. Better services—elec- on poorer youth. Working through NGOs tricity, roads, or ports, for example—might be is attractive, because often they can be more provided with existing infrastructure coupled flexible, community driven, and responsive with better policy and organization. Incentives to clients than other service providers. Even matter for turning the available infrastructure so, government financing for NGOs to deliver into effective infrastructure services. integrated programs for youth should be per- Throughout Africa, the private sector formance based, with payment linked to the provides a wide range of vocational train- successful delivery of services that demonstrate ing opportunities. The public sector should beneficial impact. provide post-school training programs only Safety net programs (such as those that to carefully targeted groups. When such pro- transfer cash or livestock assets) are proving grams are undertaken, public-private partner- to be powerful tools for poor areas and poor ships help to ensure that training is aligned people to improve their standard of living with business needs. Directly integrating firms and support productive investments. These into the training is an additional way to facili- programs can support youth in making the tate the transition to employment. Programs transition into productive activity (within should not simply displace adult workers and or outside agriculture) through integrated substitute young workers. approaches, such as programs combining public works or transfers with life skills train- ing, especially if the overall business environ- “Do Now for Later”: Take Action ment is conducive. Monitoring and careful for Medium-Term Payoffs evaluation are essential to identify the most cost-effective approaches. As governments identify and implement poli- cies to make their young and growing work- Priorities to Create More and Better force more productive in the near term, they Wage and Salary Jobs in Firms must also start to take action on policies that If the overall number of jobs in the modern will yield payoffs in the medium term. First and wage sector grows, youth will benefit. Because foremost, many countries will need to reduce export-oriented manufacturing has the greatest fertility rates more rapidly. Countries that have potential to generate employment in the wage larger working-age populations in relation to and salary sector, the first priority is to remove their non-working-age populations will experi- binding constraints to export competitiveness. ence the most rapid gains in productivity and Governments will need to address elements investment. of the business environment that reduce pro- Countries that move urgently to improve ductivity at the factory level and well beyond their education systems will also gain a sub- the factory gates, such as constraints related to stantial advantage. Given the size of the cohorts conclusion: building an Effective Youth Employment policy 247 coming into Africa’s education systems, no True regional integration that removes the country can afford to retain an education limitations of Africa’s relatively small domestic system that fails to instill the skills to pursue markets for goods and services will require a a stable livelihood. Underperforming educa- series of efforts that reinforce one another. And tion systems not only are inefficient but also the payoff from some of these reforms may also condemn large parts of another generation to take time to materialize, as investors and entre- poverty. At the same time, children must be preneurs wait to be certain that reforms have ready to learn when they come to school. Steps taken hold before they step in. But changes to build the foundation for effective educa- should start today. tion must be taken in early childhood—such At its core, the challenge of youth employ- as steps to improve the nutrition and cogni- ment is closely aligned with the challenge of tive development of children under age five. inclusive growth, defined not only as growth Deficiencies in skills have restricted growth in in which the poorest segments of society share, productivity in the past, and they will remain a but also as growth in which young people’s constraint if changes are not made. vitality is harnessed and rewarded. For Afri- Addressing business environment con- can countries meeting this challenge, the ben- straints to the creation of more and better efits will build progressively on each other. jobs in the modern wage sector is a long-term The demographic dividend will start to yield endeavor that will require sustained effort. results, and Africa’s prosperity will grow and Reforming land or financial markets takes time. be shared. Appendix A Note on Data Standardized and Harmonized actual underlying population. The last survey Household and Labor Force Data from each country was used to create overall aggregates. Many of the figures used in this report are In addition to these standardized and har- based on data from standardized and har- monized data sets for the Sub-Saharan Africa monized household and labor force surveys. region, the authors selectively analyzed data The data presented in these figures are derived from the World Bank Research Department’s from the authors’ analysis of the World Bank International Income Distribution Database Africa Region’s standardized and harmonized (I2D2). This is a compilation by the World data sets. This standardization was carried Bank Research Department of household and out under the Survey-Based Harmonized labor force surveys from around the world. Indicators Program (SHIP), augmented with similarly standardized and harmonized labor force surveys (SHIP-LF). These are listed in Demographic and Health Survey table A.1. Data The labor force participation variables were standardized and harmonized across these sur- Certain figures are based on analysis of Demo- veys so that the data could be aggregated in a graphic and Health Survey data. These data are meaningful way. When aggregated, population available from www.measuredhs.com. Coun- weights were used so the data from each survey tries and years used in the analysis are pre- contribute to the aggregate in proportion to the sented in table A.2. 249 250 Youth EMploYMEnt in Sub-SAhArAn AfricA Table A.1 Standardized and Harmonized Household and Labor Force Surveys used in report analysis Country Type of survey Year Last year cameroon Ship 2001 Ship 2007 x comoros Ship 2004 x côte d’ivoire Ship 2002 Ship 2008 x Ethiopia Ship-lf 2005 x Ghana Ship 1991 Ship 1998 Ship 2005 x Kenya Ship 2005 x Malawi Ship 2004 x Mozambique Ship 2003 Ship 2008 x niger Ship 2005 x nigeria Ship 2004 Ship 2010 x rwanda Ship 2005 Ship 2010 x São tomé and príncipe Ship 2000 Ship 2010 x Senegal Ship 2001 Ship 2005 x Sierra leone Ship 2003 x tanzania Ship-lf 2006 x uganda Ship 2005 Ship 2010 x Zambia Ship 2010 x A note on Data 251 Table A.2 Demographic and Health Survey Data Sets used in report analysis Transitionsa Education Country Year Female Male profileb benin 2006 x x x burkina faso 2010 x x x burundi 2010 x x x central African republic 1994 x x cameroon 2011 x x x chad 2004 x x x congo, rep. 2005 x x côte d’ivoire 2005 x x congo, Dem. rep. 2007 x x Ethiopia 2011 x x x Gabon 2000 x x x Ghana 2008 x x x Guinea 2005 x x Kenya 2008 x x x lesotho 2009 x x x liberia 2007 x x x Madagascar 2008 x x x Malawi 2010 x x Mali 2006 x x x Mauritania 2000 x x Mozambique 2003 x x x namibia 2006 x x x niger 2006 x x x nigeria 2008 x x x rwanda 2010 x x x São tomé and príncipe 2008 x x Senegal 2010 x x x Sierra leone 2008 x x x South Africa 1998 x Swaziland 2006 x x x tanzania 2010 x x x togo 1998 x uganda 2011 x x x Zambia 2007 x x x Zimbabwe 2010 x x x a. Chapter 2. b. Chapter 3. ECO-AUDIT Environmental Benefits Statement The World Bank is committed to preserving Saved: endangered forests and natural resources. • 10 trees The Office of the Publisher has chosen to • 5 million Btu of print Youth Employment in Sub-Saharan total energy Africa on recycled paper with 50 percent • 879 lbs of net postconsumer fiber in accordance with the greenhouse gases recommended standards for paper usage set • 4,767 gal. of waste by the Green Press Initiative, a nonprofit pro- water gram supporting publishers in using fiber • 320 lb. of solid that is not sourced from endangered forests. waste For more information, visit www.greenpress initiative.org. High fertility and declining mortality rates have led to a very young population in most Sub-Saharan African countries. The region’s labor force is expected to increase by 11 million people per year over the next 10 years. Most of this increase will be new entrants seeking their first job. While the younger generation is better educated than their parents, they often lack the means to translate that education into productive employment. Today, most work is in nonwage jobs on farms and in household enterprises. Even if greater economic activity were to create the conditions for robust growth and economic transformation, the private modern wage sector in low- and lower-middle-income countries could not absorb all the applicants. This report focuses on how to improve the quality of all jobs and to meet the aspirations of youth. It emphasizes that building a strong foundation for human capital development can play an important role in boosting earnings, and it argues that a balanced approach focused on building skills, raising productivity, and increasing the demand for labor is necessary. Youth Employment in Sub-Saharan Africa notes that many youth employment challenges are problems of employment in general. However, youth is a time of transition, and young people face particular constraints to accessing productive work. The report brings together original analysis of household and labor force surveys; it reviews the experience of a number of promising interventions across the continent; it draws from qualitative studies in several countries; and it surveys the most up-to-date evidence from rigorous evaluations of policies and programs. From this information base, the report provides guidance to policy makers on how to intervene along two dimensions—human capital and the business environment—and in three priority areas—agriculture, household enterprises, and the modern wage sector. The ultimate goals are to increase productivity, improve livelihoods, and multiply opportunities for young people. ISBN 978-1-4648-0107-5 90000 9 781464 801075 SKU 210107