Accelerating Uganda’s Development: Ending Child Marriage, Educating Girls UGANDA ECONOMIC UPDATE 10th edition Accelerating Uganda’s Development: Ending Child Marriage, Educating Girls UGANDA ECONOMIC UPDATE 10th edition contents ABBREVIATIONS AND ACRONYMS........................................................................................ i FOREWORD...........................................................................................................................ii ACKNOWLEDGEMENTS........................................................................................................ iii PART1 State of the Ugandan Economy............................................................................... 1 1.0 Recent Economic Developments.......................................................................................... 2 1.1 The real output growth remains subdued amidst external shocks and unrealized policy stimulus ................................................................................................................................ 2 1.2 A sustained easing of monetary policy has only marginally reduced commercial lending rates ......................................................................................................................... 6 1.3 The external position has improved, driven by recovery of exports...................................... 7 1.4 The fiscal policy stance has been less expansionary, in part due to under-execution of the capital budget............................................................................................................ 10 2.0 Economic Outlook................................................................................................................ 14 2.1 Higher real GDP growth expected from increased public investment and agriculture output ................................................................................................................ 14 2.2 Risks remain tilted to the downside...................................................................................... 16 2.3 Higher productivity is needed to support higher sustainable growth over the medium term........................................................................................................................ 18 PART2 The Cost of Not Investing in Girls: Marriage, Early Childbearing, and Low Educational Attainment for Girls, and their Impacts................................................................................................... 20 3.1 Why should we measure the costs of child marriage, early childbearing, and the lack of education for girls?........................................................................................................... 22 3.2. How have child marriage, early childbearing, and girls’ education changed over time?..... 23 3.3. How does poverty affect child marriage, early childbearing, and low educational attainment for girls?.............................................................................................................. 26 3.4. What types of impacts and costs are associated with child marriage, early childbearing, and low educational attainment for girls?.................................................................................... 27 3.5. What are the estimated impacts of child marriage, early childbearing, low educational attainment for girls?.............................................................................................................. 30 3.6. What are the costs associated with those impacts? The case of child marriage................. 39 3.7. How can we end child marriage, prevent early childbearing, and improve girls’ educational attainment? Lessons from international experiences........................................................... 41 3.8. Conclusion: What Have We Learned?.................................................................................. 47 References...........................................................................................................................49 Figures Figure 1: The decline in economic activity has come from all sectors of the economy, except agriculture......................................................................................................................................... 3 Figure 2: Sustained decline widening distance from long term average ......................................... 3 Figure 3: Uganda’s growth driven by consumption as investments remain subdued...................... 4 Figure 4: Uganda’s recent growth performance lower than comparators’........................................ 5 Figure 5: Uganda’s GDP growth across sectors............................................................................... 6 Figure 6: Core and headline inflation has remained within central bank target................................ 7 Figure 7: Commercial banks response to lower interest rates remains modest............................... 8 Figure 9: Current Account Deficit matching the trend in GDP growth.............................................. 8 Figure 8: The value of loans approved was increasingly lower than applications............................ 8 Figure 10: Uganda’s exports still dominated by coffee, as Middle East takes larg er share of exports............................................................................................................................. 9 Figure 11: Uganda is mainly importing machinery and equipment, largely from Asia..................... 10 Figure 12: Sustained achievement in poverty reduction recently reversed..................................... 19 Figure 13: Conceptual Framework for Measuring Impacts and Costs............................................ 28 Tables Summary Table 1: Summary of Statistically Significant Estimated Impacts by Domain.......... vii Table 1: Fiscal Operations .............................................................................................................. 11 Table 2: Uganda / Macro poverty outlook indicators (annual percent change unless indicated).... 15 Table 3: Trends in Child Marriage and Early Childbearing for Girls in Uganda (%)......................... 23 Table 4: Trends in Educational Attainment for Girls in Uganda (%)................................................. 25 Table 5: Child Marriage and Early Childbearing by Region (%)...................................................... 26 Table 6: Child Marriage, Early Childbearing and Educational Attainment for Girls by Quintile (%).26 Table 8: Impacts of Child Marriage/Early Childbearing on Educational Attainment......................... 31 Table 9: Impacts of Girls’ Educational Attainment on Child Marriage/Early Childbearing............... 32 Table 10: Impacts of Child Marriage/Early Childbearing on Fertility and Population Growth........... 33 Table 11: Impacts of Girls’ Educational Attainment on Fertility and Population Growth................... 34 Table 12: Impacts of Child Marriage/Early Childbearing on Health, Nutrition, and Violence........... 35 Table 13: Impacts of Education on Health, Nutrition, and Violence................................................. 36 Table 14: Impacts of Child Marriage/Early Childbearing on Work, Earnings, and Welfare.............. 36 Table 15: Impacts of Girls’ Educational Attainment on Work, Earnings, and Welfare...................... 37 Table 16: Impacts of Child Marriage/Early Childbearing on Women’s Agency................................ 38 Table 17: Impacts of Girls’ Educational Attainment on Women’s Agency........................................ 38 Table 18: Summary of Statistically Significant Estimated Impacts by Domain................................ 39 Table 19: Order of Magnitude of Benefits from Ending Child Marriage............................................ 41 Table A1: Key Macroeconomic Indicators....................................................................................... 52 Table A2: Growth and Structure of the Economy............................................................................. 53 Table A3: Central Government Fiscal Framework (% of GDP) ........................................................ 54 Table A4. Monetary indicators......................................................................................................... 55 Table A5.1: Balance of Payments (percent of GDP unless otherwise stated).................................. 56 Table A5.2: Balance of Payments (USUS$ millions)........................................................................ 57 Table A6. Inflation Rates (percentage changes).............................................................................. 58 © 2017 International Bank for Reconstruction and Development / International Devel- opment Association or The World Bank Group 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. 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 govern- ments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. Cover art: “Ssenga” Oil on canvas by: Prof. George Kyeyune Image courtesy of Afriart Gallery Design/Layout: Artfield Graphics Ltd, info@artfieldgraphics.com. Typesetting, graphics, and page layout: Artfield Graphics Ltd. Printed in Uganda by Artfield Graphics Ltd Additional material relating to this report can be found on the World Bank Uganda website (www.worldbank.org/uganda). ABBREVIATIONS AND ACRONYMS BoU Bank of Uganda PAPP Project Analysis and Public–Private BOP Balance of Payments Partnerships CBR Central Bank Rate PDU Procurement and Disposal Unit CCB Capital Conservation Buffer PFMA Public Finance Management Act CEM Country Economic Memorandum PFM Public Financial Management CNOOC China National Offshore Oil Corporation PIMA Public Investment Management Analysis CPI Consumer Price Index PIMAC Public and Private Infrastructure CPIA Country Policy and Institutional Assessment PMC Investment Management Centre COMESA Common Market for Eastern and Southern Africa PIMI Public Investment Management Index DAC Development Assistance Committee PIMS Public Investment Management System DHS Demographic and Health Survey PPDA Public Procurement and Disposal of Public Assets DRC Democratic Republic of Congo Authority DSA Debt Sustainability Analysis PPP Purchasing Power Parity EAC East African Community PPPs Public-Private Partnerships ELA Empowerment and Livelihoods for Adolescent Girls RAPs Resettlement Action Plans EU European Union RDP Reconstruction and Development Plan FDI Foreign Direct Investment REER Real Effective Exchange Rate FY Financial Year SBFP Sector Budget Framework papers GDP Gross Domestic Product SMEs Small and Medium-sized Enterprises HIPC Highly Indebted Poor Countries SRH Sexual and Reproductive Health HIV/AIDS Human Immunodeficiency Virus/ Acquired Immune SSA Sub-Saharan Africa Deficiency Syndrome SWA Sector Wide Approach IBP Integrated Bank of Projects SWG Sector Working Group ICT Information and Communications Technology UEU Uganda Economic Update IFC International Finance Corporation UBOS Uganda Bureau of Statistics IMF International Monetary Fund UNESCO United Nations Educational, Scientific and MDRI Multilateral Debt Relief Initiative Cultural Organization MFPED Ministry of Finance, Planning and Economic UNFPA United Nations Population Fund Development UNHS Uganda National Household Survey MDA Ministries, Departments and Agencies UNICEF United Nations Children’s Fund MoLG Ministry of Local Government UGX Uganda Shillings MTEF Medium Term Expenditure Framework UPE Universal Primary Education RWA Risk Weighted Assets UPPET Universal Post Primary Education and Training NCN Non-Concessional Borrowing USA United States of America NDP National Development Plan UNRA Uganda National Roads Authority NEER Nominal Effective Exchange Rate URA Uganda Revenue Authority NPA National Planning Authority USE Universal Secondary Education NBFP National Budget Framework Paper VAT Value Added Tax NSSF National Social Security Fund WASH Water, Sanitation and Hygiene ODA Official Development Assistance WB World Bank OECD Organization of Economic WDI World Development Indicators Cooperation and Development OPM Office of the Prime Minister i FOREWORD I n Uganda, almost six in ten people are below the age of 18 years, making its population one of the most youthful in the world. This youthful population can be an asset if it is educated and provided access to productive work. A sizable share of this human capital is, however, not fully utilized because of early marriage, early childbearing and high fertility rates that result in young children that do not attain sufficient education. As demonstrated in many countries, education raises labor productivity, which in turn supports the creation and expansion of businesses, and consequently raises output growth. As a result, incomes rise and poverty declines. In Uganda, official statistics show that the poverty rate in households headed by persons with no formal education is 2.3 times higher than the households headed by persons with at least some secondary school education. Let us not forget that the young are the most vulnerable, and are disproportionately affected by shocks that push many below the poverty line. Against this background, I am pleased to introduce the Tenth Uganda Economic Update Series, which highlights the economic costs of child marriage, early childbearing, and low educational attainment for girls. The 10th Uganda Economic Update argues that educating girls, ending child marriage, and preventing early childbearing is essential for girls to have agency, not only as future wives and mothers, but also as a productive workforce that contributes to the development of the country. Specifically, lowering population growth, the fertility rate, and the dependency ratio, will raise the productivity of the labor force, hence offering an opportunity to accelerate economic growth. Even more fundamentally, girls marrying, having children, or dropping out of school early are disempowered in ways that deprive them of their basic rights. In line with the structure of the earlier editions of Uganda Economic Update series, the report first discusses the status of macro-economy, before delving into the special topic related to child marriage, early childbearing, and low educational attainment for girls. I hope that this update is useful and will inform the debate and agenda for this important aspect of Uganda’s development. Diarietou Gaye Country Director - Eritrea, Kenya, Rwanda, and Uganda; Africa Region ii ACKNOWLEDGEMENTS The Tenth Edition of the Uganda Economic Update was prepared by a team consisting of Rachel Kaggwa Sebudde, Quentin Wodon, and Joseph Mawejje. The team is grateful to Tihomir Stucka for additional inputs on the structure and messaging of the document. Barbara Katusabe provided logistical support, while Sheila Kulubya managed the communications and dissemination strategy. The Uganda Country Team provided useful feedback during the preparation of the report. Abebe Adugna (Practice Manager, Macroeconomic and Fiscal Management), and Christina Malmberg Calvo (Country Manager) provided overall guidance on the project. The report benefitted from the insights provided by peer reviewers, including Elizabeth Ninan (Senior Education Specialist), (Senior Social Development Specialist), and Michael Geiger (Senior Economist) Close collaboration with external stakeholders was intended to ensure the relevance of the messages to policy makers and practitioners. These external collaborators included the Ministry of Gender and Social Development; the Ministry of Finance, Planning and Economic Development; and the Gender Development Partners Group in Uganda. The analysis of the impacts of child marriage, early childbearing, and low educational attainment for girls was supported through grants from the Children’s Investment Foundation Fund and the Global Partnership for Education. It builds on a study on the economic impacts of child marriage globally, conducted by the World Bank and the International Center for Research on Women with additional funding from the Bill and Melinda gates Foundation. The core team that conducted the analysis for Uganda consisted of Chata Male, Adenike Onagoruwa, Aboudrahyme Savadogo, Quentin Wodon, and Ali Yedan. Irfan Kortschak provided professional editing services. iii KEY MESSAGE Following decades of sustained economic growth in terms of increasing the level of girls’ participation during which Uganda made dramatic progress in primary education. However, many girls and young towards poverty reduction, the country has women are still unable to achieve their full potential. recently experienced a period of economic growth At present, more than a third of girls in Uganda get slow down. To return to higher rates of economic married before the age of 18, and almost three in ten growth and poverty reduction, the Government must have a child before that age. As a result, the proportion address fundamental constraints. Facing a range of of girls enrolling in secondary education is still internal and external shocks, Uganda’s economy has lower than for boys, with an even smaller proportion grown at the average annual rate of 4.5 percent over the completing lower secondary level education and with past five years, far lower than the historical average of the trend becoming even more pronounced at higher about 7.8 percent. The recent deceleration in growth levels of education. Girls who are married or have affected all sectors of the economy. With these lower children at a young age or drop out of school early are growth rates, combined with other external shocks to more likely to experience poor health, to have more the households, the remarkable progress that Uganda children over their lifetime, and to hold mainly low had made towards reducing poverty since 1992 has productivity and low paying jobs during adulthood. been reversed. In the period from 1992 to 2013, the Following an assessment of the current state of national poverty rate declined from 56 percent to 19.7 Uganda’s economy and its outlook for the future, percent. Since then, it has rebounded, increasing to a the tenth Uganda Economic Update focuses on preliminary 27 percent. For Uganda to again achieve issues related to how ending child marriage, higher rates of economic growth and poverty reduction, preventing early childbearing, and improving it must address two fundamental factors: its low levels girls’ levels of educational attainment are of productivity and the vulnerability if its people to essential for the achievement of faster growth and poverty. inclusive development. In addition to addressing Uganda’s youthful population has the potential the substantial risks and suffering faced by adolescent to achieve increased productivity and to build its girls and their children, investments in ending these resilience to shocks. Today, almost six in ten people practices can produce huge economic gains, not just in Uganda are below the age of 18 years, making for the girls, or their households, but for the country at Uganda’s population one of the most youthful in the large. world. On the one hand, this youthful population is an asset. If Uganda’s youth were provided with the appropriate education and enabled to fully participate Part 1: State of the economy in production processes at an increasingly higher During the second half of FY2016/17, Uganda’s level of productivity, this would enable Uganda to real GDP annual growth accelerated to reach the reap significant demographic dividends. On the other, rate of 5.0 percent, up from 3.0 percent achieved Uganda’s young are disproportionately affected by in the first half of FY2016/17. Despite an increase in shocks that threaten to push a significant proportion of final consumption expenditure, the real GDP growth the population below the poverty line. rate decelerated to 4 percent for FY2016/17, as a whole, Efforts to increase productivity and to reduce compared to the rate of 4.7 percent recorded in the vulnerability must be implemented with the previous year, when output growth was driven by a intention of realizing the full potential of Uganda’s strong rebound in ICT investment. With energy and large number of youths. In particular, this will fuel prices firming up, the inflation rate increased to 6.1 require the provision of universal education, percent for the period from January to October in 2017, irrespective of gender. In this area, Uganda has made compared to the figure of 5.5 percent recorded during significant progress in a number of areas, particularly the same period last year. During the same period, the iv core inflation rate decelerated to 4.7 percent, down from budget revenues totaling 14 percent of GDP. Increasingly, the figure of 6.0 percent recorded a year ago. Inflationary the Government is financing the fiscal deficit through the pressures remained subdued, despite the reduction in the domestic debt market, crowding out the private sector’s Central Bank Rate (CBR) by a total of 5.5 percentage points, access to credit. Overall, the total public debt stock increased with this rate declining to 9.5 percent in October 2017. Despite to about 39 percent of GDP by the end of FY 2016/17, with the ongoing reduction to the CBR, banks’ lending rates two-thirds of this debt stock consisting of external public declined only marginally, to around 21-22 percent. Thus, the debt. Uganda’s stock of public debt is lower than that reduction resulted in only a marginal increase in lending to of many of its regional comparators, including Kenya the private sector. (53 percent of GDP at end-2016) and Rwanda (44 percent of The current account deficit narrowed markedly GDP at end-2016). Tanzania’s public debt stock is roughly at in FY2016/17 due to the decline in government the same level as Uganda’s. consumption, the deceleration in private investment, Despite an expected softer output growth outcome and the improved performance of exports. The current in the first half of FY2017/18, based on some leading account deficit narrowed to USUS$ 1.0 billion, or 4 percent indicators, real GDP is expected to grow at the rate of of GDP, in FY2016/17, down from USUS$ 1.2 billion, or 5.1 percent for the year as a whole in FY2017/18. 4.9 percent of GDP, in FY2015/16. The narrowing of the This acceleration is expected to be driven by stronger private deficit was driven by an increase in the export of goods, with consumption as pent up demand and supply activities imports remaining roughly constant, and with a broadly materialize following a reduction in the political uncertainty unchanged service balance. Once adjusted for net FDI in the Kenyan elections, which led to the suspension inflows, the current account deficit shrunk to 2.1 percent of trade and investment activities. of GDP in FY2016/17, down from 2.3 percent of GDP in the Domestically, during the first half previous year. The current account shortfall was financed of the year, there were by long-term government project financing. With the value tensions related of long-term loans steady at around 3.3 percent of GDP to and with budget support flows amounting to 0.6 percent of GDP, financing risks stemming from volatile short-term 4.0% inflows are limited. The total value of external debt stock has stabilized at 23 percent of GDP in FY2016/17, after increasing by 6 percentage points to 22 percent of GDP in the previous year. Foreign exchange reserves remained stable at around 5.4 months imports of goods and services cover, while depreciation pressures dissipated in FY2016/17, following a cumulative depreciation of 33 percent during the period from FY2014/15 to FY2015/16. The current account The overall fiscal deficit contracted sharply during deficit of GDP, in FY2016/17 FY2016/17, going down from 4.4 percent of GDP in FY2015/16 to 3.5 percent in FY2016/17, due to a less expansionary fiscal stance, than originally budgeted. The actual fiscal deficit was almost half of the budgeted deficit of 6.2 percent of GDP, largely due to the under- execution of budgeted capital expenditures. The total value of collected tax revenues increased by another percentage point of GDP in FY2016/17 compared to a year ago, with total v the proposed constitutional amendments regarding apart from the synthesis provided here (Wodon et al., the age limit for presidential candidates. Stronger 2017). government capital expenditure, especially on key The issues of child marriage, early childbearing, infrastructure projects should help support stronger and low educational attainment for girls are closely output growth in the second half of 2017/18. If weather related, as they affect each other. Child marriage conditions remain favorable, a rebound in agricultural is likely the cause of more than half of all instances of production could support the gradual increase in early childbearing. In some cases, early childbearing the overall growth rate to 5.1 percent this year and may lead to child marriage, but this is probably 5.7 percent in FY2018/19. However, there are significant less likely. In addition, the causality between child downside risks to this growth outlook. In particular, marriage and early childbearing on the one hand, and over-reliance on the domestic debt market to finance educational attainment on the other hand, goes both the fiscal deficit will continue to inhibit private sector ways. Child marriage and early childbearing have a investment. Furthermore, increasing pressures for negative effect on educational attainment. Conversely, higher public sector wages, particularly by medical keeping girls in schools reduces the risks of child personnel, may affect the composition of expenditures marriage and early childbearing. This is why incentives and result in a higher proportion of recurrent spending, for girls to remain in school or go back to school if delaying plans for the full execution of infrastructure they dropped out appear to be effective to delay the investments. Increased domestic demand should result age at first marriage and prevent early childbearing. in a widening of the current account deficit, although It is worth noting that achieving universal secondary greater net FDI inflows would keep the adjusted current completion for girls could dramatically reduce the account deficit between 2 to 2.5 percent of GDP. prevalence of child marriage and early childbearing. On the other hand, while ending child marriage and early The Cost of Not Investing in Girls: Child Marriage, childbearing would help improve girls’ educational Early Childbearing, and Low Educational attainment, this would not be sufficient in itself to Attainment for Girls, and their Impacts ensure universal secondary completion. Ending child marriage, preventing early Child marriage, early childbearing and low childbearing, and improving educational educational attainment for girls have negative attainment for girls are essential for Uganda’s impacts on many other areas including (i) fertility development, yet investments towards those and population growth; (ii) health, nutrition, and goals remain limited. The international community violence; (iii) work and earnings; and (iv) women’s is increasingly aware of the negative impacts of child agency. Overall, the impacts of child marriage, early marriage, early childbearing, and low educational childbearing, and low educational attainment for attainment for girls on a wide range of development girls, whether taken individually or collectively, are outcomes. The Government of Uganda has adopted large. As shown in the Table below, child marriage/ a national strategy to end child marriage and early childbearing, secondary education completion teenage pregnancies. Improving girls’ education for girls, or both have a statistically significant impact is also a priority of the government. Still, more on all the indicators listed except birth registration. could be done to accelerate progress in these areas. This shows how pervasive and widespread the impacts To inspire greater commitments towards ending of a lack of opportunities for adolescent girls are. In child marriage, preventing early childbearing, and terms of the size of the impacts, the effects on fertility improving educational attainment for girls, this study and thereby population growth are very substantial in demonstrates the economic and social costs of not Uganda. Impacts on women’s earnings and intimate investing enough in these areas in Uganda. The study partner violence are also large. Some of the impacts also suggests options that could be considered for on health and nutrition are large at the margin, but investments. This study for Uganda is adapted from a not necessarily large at the national level. This is global study, and a more detailed report is available the case for under-five mortality and stunting where large negative marginal effects are observed for early vi Summary Table 1: Summary of Statistically Significant Estimated Impacts by Domain Domains and Indicators Child marriage or early Secondary education Either one of the childbearing completion two Mutual relationships Child marriage/ Early childbearing - Yes Yes Educational attainment Yes - Yes Fertility and population growth Fertility Yes Likely Yes Population growth Yes Likely Yes Modern contraceptive use No Likely Yes Health and nutrition Under-five mortality Yes No Yes Under-five stunting Yes No Yes Labor force participation No Yes Yes Demand for healthcare No Yes Yes Work and productivity Intimate partner violence Yes No Yes Women’s earnings Yes Yes Yes Household welfare Yes Yes Yes Women’s agency Decision-making ability No Yes Yes Land ownership No Yes Yes Knowledge of HIV/AIDS Yes Yes Yes Birth registration No No No Source: Wodon et al. (2017). Note: The term “likely” is used for some impacts of secondary completion because for those indicators, while the impact of secondary education completion is not statistically significant possibly due to small sample sizes, the impacts of some secondary as well as higher education are statistically significant. childbearing, but ending childbearing would not make rate of population growth, ending child marriage and a major difference at the national level simply because associated childbirths could lead to welfare benefits for relatively few children are born of mothers younger than 18. Uganda of about USUS$2.4 billion (in purchasing Other direct impacts of child marriage and early childbearing power parity terms) by 2030. Also as including on women’s decision-making, knowledge of HIV- a result of lower population AIDS and birth registrations are not always statistically growth, ending child significant, but low educational attainment for girls in those marriage areas tends to have a large impact on these outcomes. These negative impacts in turn have monetary costs, especially for fertility and population growth, earnings, US$ 2.4 and the health of the children born of young mothers. Tentative estimates of costs are provided for some of the impacts of child marriage. The estimates should not billion be considered as precise given that they depend on (1) econometric estimates of impacts that have themselves WELFARE BENEFITS by 2030 standard errors and (2) a range of assumptions for costing from lower population that could be debated. Still, they provide an order of magnitude of the potential costs of child marriage. The growth largest economic cost of child marriage is the welfare loss associated with population growth. By reducing the annual vii and early childbearing today could result in education to help delay the age at marriage and childbearing. budget savings for the Government of up to USUS$257 Some of these programs also enable girls who million by 2030 if Uganda were to achieve universal dropped out to return to school. Programs providing secondary education by then (this is an upper bound financial incentives to girls or families directly to for savings given that achieving universal secondary delay marriage may also work. education is a tall order). In addition, today, if women Implications for policy: While some of the programs who were married early had been able to avoid child work better than others to delay marriage and marriage, the gains in earnings and productivity that childbearing and to improve educational attainment would have resulted are estimated at USUS$514 million. for girls, all three categories of programs have benefits. Finally, substantial economic benefits would result By targeting different groups of girls, for example those from reductions in under-five mortality and stunting in school or with the potential to return to school, and rates, estimated together to reach USUS$275 million by those who dropped out and may not be able to return, 2030. all three categories of programs should be considered Based on a literature review of programs when implementing a strategy aiming to improve intended to improve young women’s sexual and opportunities for adolescent girls. reproductive health outcomes, delay marriage In addition, to improve educational attainment and childbearing, and improve girls’ participation for girls, basic conditions must be met, and in education, a three-pronged approach is for challenges such as gender-based violence, recommended. other interventions are needed. The focus on the • Programs to provide life skills and reproductive above three types of interventions is not meant to health knowledge: These interventions often rely be exhaustive in terms of what is needed to invest on safe space programs empowering girls through in adolescent girls in Uganda. For educational life skills training, better knowledge of sexual attainment, at the secondary level, there is a need to and reproductive health, and other skills. These build schools closer to where children (boys and girls) programs have achieved important benefits for live, or as an alternative, provide adequate modes of girls, not only in terms of knowledge acquired, but transportation to schools. Providing separate water, also through gains in self-esteem and confidence, sanitation and hygiene facilities for girls is also among others. Yet, without additional livelihood important, as is the need to reduce the risk of violence opportunities or incentives for schooling, it is not and sexual harassment in school. Finally, for specific clear that safe spaces by themselves are sufficient to challenges such as intimate partner violence or gender- delay marriage and childbearing. based violence more generally, community-based • Programs to expand economic opportunities: interventions with men, women, leaders and service Interventions that combine an emphasis on providers are also needed, with successful pilots in empowering girls, often through safe spaces, with Uganda. a focus on providing livelihood opportunities While investments to end child marriage, prevent have demonstrated some success in increasing early childbearing, and promote education for earnings for participants, as well as employment girls should not be based solely on economic and savings. In some cases, they may also improve considerations, this study demonstrates that the reproductive health outcomes and delay marriage benefits from such investments would be large. or childbearing, but not systematically so. In The primary motivation for ending child marriage, Uganda, there is evidence that such interventions preventing early childbearing, and promoting have worked well. education for girls should be to address the substantial • Programs to keep girls in school or delay risks and suffering faced by adolescent girls and their marriage: Interventions to promote education, children. The evidence of the negative impacts of these especially by reducing the opportunity and out-of- issues on a wide range of development outcomes is pocket cost of schooling, are among the most likely clear. However, in addition, these issues have large economic costs. Demonstrating the magnitude of these viii costs provides an additional justification for investments in adolescent girls in Uganda. Ending child marriage, preventing early childbearing and improving education opportunities for girls is not only the right thing to do from a moral and ethical standpoint, it is also a smart investment for Uganda’s development. Ending child marriage, Educating Girls is a smart investment ix PART State of the Ugandan Economy 1 During the second half of FY 2016/17, Uganda’s real GDP growth rate accelerated to 5.0 percent, up from 3.0 percent in the corresponding period in the previous year. However, during the first half of FY 2017/18, economic activity has been subdued, with the positive effects of sustained monetary easing, improved weather conditions, and the improved external environment partially offset by the negative impact of a number of regional and domestic developments. With the annual inflation rate remaining at around the targeted level of 5 percent, the Bank of Uganda has maintained easy monetary policies, reducing the CBR to a historical low of 9.5 percent in October 2017. By September 2017, the full impact of this stimulus policy had yet to be realized, with commercial banks’ lending rates declining only marginally to 22 percent and with the stock of private sector credit increasing only modestly by 6.5 percent compared to the level at the same point in the previous year. In FY 2016/17, Uganda’s external current deficit declined by almost a full percentage point of GDP to 4.0 percent. The strong performance of coffee exports offset the decline in the performance of tourism. The shilling remained stable, with Uganda’s foreign reserves increasing to 5.4 months of import cover. The implementation of the FY 2017/18 budget has been constrained by shortfalls in revenues, with these revenues standing at 3.2 percent below the target level for the half of FY 2017/18, and by the under-execution of the development budget. The failure to meet revenue collection targets and the pressure to increase recurrent spending may increase the overall fiscal deficit beyond the level of 6.4 percent of GDP budgeted for the year. The economy is projected to grow at the rate of 5.1 percent during FY 2017/18, with this rate expected to accelerate to 5-6 percent for the subsequent three years. However, this acceleration is predicated on the improved execution of public projects, an increased level of private investment, favorable weather conditions, increased access to credit, and improved external conditions. There are a number of significant downside risks to the growth forecast, with the most significant risks related to fiscal management; to the potential of adverse weather conditions to negatively impact agricultural output; and to regional political developments that may adversely affect private investment. To achieve higher rates of sustainable growth, Uganda must increase the productivity of its labor force. The elimination of early marriages and early child bearing and the increased participation of girls and education are factors that may provide significant support to the achievement of increased productivity growth and higher levels of sustained economic growth. Young women make a living as market vendors, Nakasero Market, Kampala (Morgan Mbabazi, 2017) 1 During FY2016/17, it is estimated that Uganda’s economy grew at the rate of 4.0 percent, a significant deceleration from historical averages, which stand at about 7-8 percent. Over the past year and a half, the economy suffered from adverse weather, unrest in South Sudan, and an uncertain global economic recovery. The fiscal deficit contracted sharply, while monetary easing had little impact in terms of reducing banks’ interest rates. In addition, structural constraints and liquidity issues in the financial market prevented commercial banks from increasing the volume of loans to the private sector. While growth is expected to accelerate in FY2017/18 as a whole to 5.1 percent, economic activity seems to have been weaker than expected in the first half of the year. The latter may have been caused by uncertainties related to Kenya’s electoral process and to political developments in Uganda, which in turn are likely to have postponed trade activities and investment. With these uncertainties reducing, economic activity is expected to pick up in the second half of the year. While significant progress has been made towards poverty reduction over the past 20 years, in recent years, the poverty rate has begun to increase, with the proportion of households living below the national poverty line rising from 19 percent in FY2012/13 to a preliminary estimate totaling 27 percent in FY2016/17. Over the past five years (2012 to 2016), Uganda’s low differential between the population growth rate economy has grown at the average annual rate of and the economic growth rate, there have been only 4.5 percent, significantly below the average annual modest improvements to average per capita incomes. rate of 7.8 percent achieved over the decade to In other words, the average annual growth rate of per 2011. On the production side, the deceleration in real capita income stands at only around 0.2 percent. Thus, growth has occurred across all sectors. In industry and with the current deceleration in economic growth, it agriculture, output growth has declined by around 50 may take more than a decade for Uganda’s average per percent, largely due to a slowdown in manufacturing capita incomes to increase from the current level of and food crops. Meanwhile, in the services sector, the US$ 706 to US$ 1000, the threshold level that defines growth rate decelerated from 8 percent to 5 percent, middle-income status. with increased growth in real estate being more 1.1 The real output growth remains subdued than offset by a decline in trade services, ICT and amidst external shocks and unrealized policy professional,1 and scientific and technical services. stimulus Throughout 2017, the global economic recovery has The recent decline in the real GDP growth rate been strengthening, driven by increased domestic means that it will take significantly longer for demand and exports, and supported by investment Uganda to obtain its goal of achieving middle stimulating growth in global trade2. In the first income status than it had envisaged. Uganda’s half of 2017, the Euro Area recorded its highest rate of average annual population growth rate is high, growth for two years, with increased growth driven standing at around 3.3 percent. With the relatively by decreased policy uncertainty, increased industrial 1 The ICT grew around 20 percent between FY 2009/10 and FY 2012/13, as activity, and more rapid credit growth, following years firms aggressively expanded services to new areas and introduced new products. This has recently slowed as these firms consolidate positions, which has resulted into exits and mergers. 2 World Bank, 2017, Global Economic Prospects. June 2017 2 Entebbe Expressway construction works (Morgan Mbabazi, 2017) Figure 1: The decline in economic activity has come from all Figure 2: Sustained decline widening distance from long term sectors of the economy, except agriculture average 4.5%   Contribu>on  to  GDP  growth(percentage  points)   4.0%   3.5%   3.0%   2.5%   2.0%   1.5%   1.0%   0.5%   0.0%   Agric   Industry   Services   Manufacturing   Construc>on   2008-­‐2011   2012-­‐2017   Source: Uganda Bureau of Statistics Source: Uganda Bureau of Statistics of accommodative monetary policy. The US economy has and more benign financial conditions. As a result, real GDP sustained improvements to its labor market, with this trend growth in the SSA region is projected to double from 1.3 likely to continue throughout 2017. China’s economic growth percent in 2016 to 2.6 percent in 2017. rate for the first half of the year stood at the robust rate of 6.9 Despite the overall increase in SSA regional growth, percent. These more favorable global trends have supported economic activity in Uganda has remained subdued, a recovery in the SSA region, with the highest growth rates largely due to the impact of political uncertainty, both recorded by large commodity-exporting economies, such as within Uganda and in neighboring countries. During the Nigeria and Angola3. The region is benefiting from a modest first half of FY2017/18,4 Uganda’s GDP growth rate is forecast recovery in commodity prices, increased demand for exports, 3 World bank, 2017, Africa Pulse, Volume 16. October 2017 4 July 2016 to December 2016 3 to have remained weaker than that of the previous of high frequency information related to Uganda’s period. With the conflict in South Sudan still not labor market. However, the increase in the volume resolved, Uganda’s exports have still not returned to of bank loans to the household and personal sector levels recorded prior to 2016. In addition, Kenya’s recent suggests that increased disposable incomes may have elections were marked by a high degree of uncertainty contributed to the growth. Overall, final consumption and volatility, conflict and unrest, particularly in the grew at the rate of 1.4 percent in FY2016/17, following period after August 2017, when election results were negative growth of -0.7 percent in the previous year. annulled and a new election was held two months Furthermore, net exports shrunk by almost a third later. As a result of these developments, exporters in FY2016/17 relative to the level recorded in the and importers in both countries adopted a ‘wait and previous year, with total imports declining by 12.5 see’ approach, which is likely to have had a negative percent, significantly outpacing the decline in total impact on the performance of Uganda’s exports. Within exports of 2.2 percent. Meanwhile, private investments Uganda, there have been tensions regarding proposed contracted by 2.5 percent, a stark contrast to the growth constitutional amendments related to the age limit of 11 percent recorded last year. This contraction was for presidential candidates. This has likely affected the result of both demand- and supply-side factors, investment sentiment of the private sector, which also including the high interest rates on loans to the private suffered due to the closure of some large business, such sector and the uncertainty in the financial sector as retail supermarkets and the posting of significant triggered by near-crisis risk levels. In particular, the losses by large banks. These uncertainties are likely to banking sector’s non-performing assets ratio reached have affected also public investment in the first half of a record high of 10 percent in December 2016, at which the year. point the Crane Bank was dissolved. The situation was further exacerbated by the disruptions in the In FY2016/17, private consumption and net exports South Sudan market for produce. In addition, private were the main drivers of real output growth. investment has been affected by the Government’s Private consumption increased by around 3 percent, high level of accumulation of domestic arrears. With more than offsetting the sharp decline in government some improvement to the execution of the public consumption, which fell by 11 percent over the year. infrastructure investment programs, government It is difficult to accurately determine the demand- investments increased by 2.8 percent, a reversal from side factors that led to this increase due to the lack of the negative growth rate of -1.5 percent recorded in FY short-term demand-side data, with a lack in particular 2014/15 and -3.6 percent recorded in FY2015/16. Overall, Figure 3: Uganda’s growth driven by consumption as investments remain subdued 6.0%   0.2%   0.2%   0.5%   5.0%   0.5%   0.2%   1.1%   0.4%   CONTRIBUTION  TO  GROWTH,  %   1.2%   4.0%   0.2%   0.1%   1.1%   0.3%   0.3%   0.8%   3.0%   0.8%     2.0%   1.0%   0.0%   -­‐0.2%   -­‐0.4%   -­‐0.4%   -­‐0.6%   -­‐0.4%   -­‐1.0%   2012/13   2013/14   2014/15   2015/16   2016/17   Private  ConsumpHon   Private  Sector  Investment   Government  ConsumpHon   Public  Sector  Investment   Net  Trade   Source: Uganda Bureau of Statistics 4 gross fixed capital formation remained stagnant at the rate of agro-processing in particular benefitting from improved 0.1 percent. regional trade. Meanwhile, the strong performance in services came from the double-digit growth rates in the ICT On the production side, the contribution of the sector and an acceleration in education services, mainly agricultural sector to overall economic value-added driven by high levels of private investment in the sector. has continued to decline. The deceleration in this With Uganda’s deceleration in output growth, its sector’s real growth rate, from 2.8 percent in FY2015/16 to performance continued to lag behind those of its peers 1.6 percent in FY2016/17, was mainly the result of the poor in the East Africa region, despite the fact that growth in performance of cash crops, livestock and forestry, with the these economies also decelerated in 2017. Growth across production of food crops remaining stagnant at 1.4 percent. the East African economies decelerated in 2017 (see Figure 5), The delayed rainy season adversely affected cash and food due to the prolonged effect of the drought that begun in 2016 crop production, with the delays resulting in drought-like and that persisted into 2017. This drought has had a negative conditions during the first half of the year. In addition, pest impact on agricultural output and thus on GDP growth attacks, particularly by the army worm, impacted production across the region. A cyclical downturn in credit growth across in major grain-producing districts. Finally, the restructuring the region’s economies has also had a dampening effect of service delivery mechanisms in the agricultural sector has on growth. As a result of these factors, Tanzania’s real GDP not resulted in the envisaged improvements to productivity. growth rate declined from 7.7 percent in the first half of 2016 The sector continues to be beset by institutional bottlenecks to 6.8 percent in the corresponding period of 2017; Kenya’s including those related to the delivery of extension services, declined from 5.8 percent to 4.8 percent; while Rwanda’s low rates of technological adoption, and the proliferation of declined from 8.2 percent to 2.9 percent, with the particularly sub-standard inputs. drastic decline in the latter due to the imposition of a tighter fiscal stance. In addition, insecurity and political tensions Whereas growth in the industrial sector continues to continued to constrain economic activity in Burundi, Central decelerate, the performance of the services sector has Africa Republic, and South Sudan. remained strong. The slowdown in the growth rate of the industrial sector, from 4.6 percent in FY2015/16 to 3.3 percent Despite the lackluster performance of Uganda’s in FY2016/17, was largely driven by the deceleration in the agricultural sector, this sector continues to be the growth of the mining and construction subsectors, with primary source of employment for the country’s the latter being caused at least in part by challenges to the workforce. Using a methodology that distinguishes ‘work’ execution of the Government’s investment program. On the from ‘employment,’5 the Uganda National Household Survey other hand, with the impact of the South Sudan conflict 2016/17 report suggests that even though a much higher receding, manufacturing activity has started to recover, with proportion of the working population is engaged in the Figure 4: Uganda’s recent growth performance lower than comparators’ 5 According to the 19th International Conference of Labour Statisticians (ICLS)held October 2013, employment and work are two different concepts, especially within developing countries. Work comprises any activity performed by persons of any sex and age to produce goods or to provide services for use by others or for own use. This concept of work is aligned to the General production boundary as defined in the System of National Accounts 2008 (SNA, 2008) and its concept of economic unit that distinguishes between: (i) Market units (i.e. corporations, quasi-corporations and household unincorporated market enterprises; (ii) Non-market units (i.e. gov- ernment and non-profit institutions serving households); and (iii) Households that produce goods or services for own final use. Yet employment comprises any activity engaged in by a working age person during a short reference period to produce goods or provide services for pay or profit Source: Uganda Bureau of Statistics and World Bank (MFMod) 5 Figure 5: Uganda’s GDP growth across sectors  12.0      10.0      9.4      8.0     Growth  rates,  %    6.0      5.6      5.1      5.2      4.7      4.0      3.8      4.0      3.6      2.0      -­‐         2009/10   2010/11   2011/12   2012/13   2013/14   2014/15   2015/16   2016/17   Agriculture   Industry   Services   GDP   Source: Uganda Bureau of Statistics agriculture sector, about 36 percent of the working continued to decline, with the exchange rate remaining population derives an income from employment in stable. this sector, with 21 percent employed in casual labor. An additional 43 percent of the working population With the decreased inflationary pressures, the is involved in subsistence agricultural production. Bank of Uganda aggressively maintained its easy Meanwhile, 52 percent of the population derives a monetary policies, with the Central Bank Rate livelihood from the services sector, with about 45 (CBR) reaching a historical low 9.5 percent in percent involved in trade. The manufacturing sector October 2017. As a result, the CBR is at a level lower provides employment to approximately 16 percent of than at any point since its inception in 2011. The Bank the working population, while the construction sector of Uganda’s sustained commitment to easy monetary provides employment to an additional 12 percent. policies is appropriate in the context of the low level With the exception of the mining sector, the bulk of demand resulting from the under-execution of of jobs within the services and industry sectors are government projects; the strength of the Ugandan concentrated in urban areas. Shilling; and the contraction in private investments described above. 1.2 A sustained easing of monetary policy has only marginally reduced commercial lending rates As has been the case in the past, the easing of the policy rate has not had a significant effect on Subdued demand and favorable weather banks’ lending rates, with a limited stimulating conditions have exerted downward pressure on effect on the economy. In the period from June 2016 Uganda’s inflation rate, which has remained at to August 2017, the weighted average commercial bank less than 6 percent over the past 18 months. In FY lending rate declined from 23.9 percent to 20.9 percent. 2016/17, the annual inflation rate stood at 5.7 percent, However, this rate later increased again to 22 percent. down from the figure of 6.6 percent recorded in the Commercial banks significantly increased their previous year. Even though the rate was lower than deposits, with their holdings of securities declining in FY2015/16, there were upward pressures on food with the declining return. In contrast, the total value prices during the second half of the year, with the rate of credit extended to the private sector increased standing at 23 percent in May 2017. During the first four modestly, by 5.7 percent, following a strong recovery months of FY 2017/18, the rate gradually decreased to during the second half of FY 2016/17. The growth of 4.8 percent, with this level significantly lower than these loans increased at an average rate of six percent. the central bank’s target of 5 percent. Inflationary With these developments, the ratio of private credit to pressures receded as the impact of drought conditions GDP declined to 13.3 percent, down from 13.7 percent in dissipated and the prices of Uganda’s imports FY 2015/16. 6 Figure 6: Core and headline inflation has remained within central bank target Source: Uganda Bureau of Statistics Apart from the impact of interest rates, bank lending percent, which is more than twice the statutory requirement has been strongly affected by the reduced appetite of 20 percent. The banking sector’s profit margins declined for foreign currency denominated loans, which by severely in FY 2016/17, but started to recover gradually in the September 2017 still contributed to about 41 percent three months leading to June 2017. of total credit. In contrast to the trends observed over the 1.3 The external position has improved, driven by recov- past three years, the rate of growth of credit denominated ery of exports in foreign currency has decelerated. Thus, the total value of foreign currency loans increased by only 2.7 percent in the six months to June 2017, a significantly lower rate than the 8.4 In FY2016/17, the current account deficit narrowed percent increase recorded for loans denominated in shillings. significantly compared to the previous year due to the Over 60 percent of foreign currency denominated bank decline in government consumption, the deceleration lending to the private sector goes to the building, mortgage, in private investment, and the improved performance construction and real estate sector and the manufacturing of exports. The current account deficit narrowed to US$ sector. Overall, lending to these sectors declined over the past 1.0 billion, or 4 percent of GDP, in FY2016/17, down from 18 months, with this decline offsetting the strong recovery in the figure of US$ 1.2 billion, or 4.9 percent of GDP, recorded credit extended for personal and household use, agriculture, in FY2015/16. The narrowing in the deficit was driven by and business services. an increase in the export of goods, with imports remaining stagnant, and a broadly unchanged service balance. Once Beginning in the second half of FY 2016/17, the banking adjusted for net FDI inflows (i.e. non-debt creating flows), system has begun to recover from a near crisis, when the current account deficit shrunk to 2.1 percent of GDP in non-performing assets (NPAs) reached almost 11 percent FY2016/17, down from the figure of 2.3 percent recorded in the of total assets. By the end of June 2017, the proportion of previous year. NPAs was 6.1 percent, lower than the figure of 10.5 percent recorded by December 2016. The closure of Crane Bank, During FY2016/17, the total value of goods exports which carried a substantial proportion of the NPAs, reduced increased to 12 percent of GDP, up from the figure of 11 the overall banking system credit risk. Nonetheless, more percent recorded in the previous year. The value of coffee stringent lending conditions were also imposed throughout exports increased by 39 percent, reaching US$ 490 million. the banking system, with banks increasing their capital Meanwhile, the value of formal non-coffee exports increased buffers in preparation for the imposition of higher standards by 12 percent, largely due to increases in the export of cotton, for capital adequacy. By June 2017, the average ratio of gold, electricity, fish and flowers. At the same time, the capital to risk-weighted assets stood at 24 percent, far higher growth rates for the export of tea, tobacco, hides and skins, than the statutory requirement of 12 percent. The ratio of simsim (sesame), and maize all declined relative to the levels liquidity assets to deposits had also increased to about 50 recorded in the previous year. According to Bank of Uganda 7 Figure 7: Commercial banks response Figure 8: The value of loans approved to lower interest rates remains modest was increasingly lower than applications Source: Bank of Uganda Source: Bank of Uganda Figure 9: Current Account Deficit matching the trend in petroleum prices, started to recover. During FY2016/17, GDP growth the average unit price of Uganda’s oil imports increased by 17 percent, offsetting in part the decline of 8 percent in the volume of oil imported during that period. Other non-oil imports declined, driven by a reduction in the import of machinery and equipment as private investment contracted. The financing of the current account deficit relies mostly on long-term government project financing7. With the value of long-term loans remaining roughly constant at around 3.3 percent of GDP and with budget support at around 0.6 percent of GDP, financing risks stemming from volatile short-term inflows remained limited. As a result, the total stock of data,6 the value of informal cross-border trade (ICBT) external debt has stabilized at 23 percent of GDP, after exports continued to grow, driven in particular by a 21 increasing from 16 percent of GDP in FY 2014/15 to 22 percent increase in the value of exports of industrial percent last year. Foreign exchange reserves remained products, with this value increasing to US$ 305 million roughly constant, at around 5.4 months of import cover. in FY 2016/17. With the huge capital inflows, the value of the In FY2016/17, the total value of goods imports shilling has remained relatively stable for the declined to 18 percent of GDP, down from the eight-month period since January 2017. After a figure of 19 percent recorded in the previous cumulative depreciation of 33 percent during the year. This was largely due to a 11 percent decline period from FY2014/15 to FY2015/16, the value of the in the Government’s import bill, resulting from the shilling depreciated further during the six months steady progress on the construction of two key public prior to January 2017. During that period, the nominal infrastructure projects (the Karuma and Isimba value of the shilling against the US dollar depreciated hydropower electric power projects). At the same time, by 7 percentage points. In the period from January to the value of private sector imports increased slightly, 7 In FY 2016/17, net portfolio inflows increased by 38 percent to a value of by 2.2 percent, as global commodity prices, especially US$ 212 million and the value of disbursed loans amounted to US$ 1,051 million in FY 2016/17, up from US$ 639 million in the previous year. This 6 Bank of Uganda, 2016, Exports of goods. http://www.bou.or.ug/bou/rates_ offset the decline in the total value of net FDI was US$ 494 million, from US$ statistics/statistics.html 630 million received in FY 2015/16. 8 Figure 10: Uganda’s exports still dominated by coffee, as Middle East takes larger share of exports Source: Uganda Bureau of Statistics September 2017, the shilling traded within a narrow range, increasing. In FY 2016/17, the share of imports from Asia (i.e. helped by reduced demand for imports and a surge in exports China, India, Japan, and Indonesia) declined to 45 percent, during the last half of FY 2016/17 and the first quarter of FY down from the figure of 51 percent recorded in the previous 2017/2018. year. An increasing proportion of Uganda’s imports come from within Africa, with the share of imports from COMESA The composition of Uganda’s export partners is increasing from 15.5 percent to 17 percent over this time changing, with new export markets in Asia and the period. While Kenya and South Africa together account for 60 Middle East opening up. Until recently, Uganda’s three percent of Uganda’s imports from within Africa, the value of biggest markets were the East African region (including imports from DR Congo increased from US$ 22 million to US$ Burundi, Democratic Republic of Congo, Kenya, Rwanda, 148 million, while those from Tanzania increased from US$ South Sudan, and Tanzania); the European Union; and the 76 million to US$ 169 million over the period. The increase Middle East. During FY 2016/17, the value of Uganda’s exports in imports from DR Congo has been mainly driven by an to the Middle East amounted to US$ 505 million, around the increase in the import of raw gold for use by a same as its exports to the European Union. The vast majority gold refinery commissioned during of the exports to the Middle East were to the United Arab FY2016/17 to process gold Emirates, which accounted for 96 percent of the total. The for re-export. value of exports to Asia reached US$ 195 million, with more 96% than 80 percent of these exports going to India, Malaysia, China, Singapore, Hong Kong, and Indonesia. Following the unrest in South Sudan, the share of exports to regional markets declined dramatically. However, since FY 2016/17, a recovery has taken place, largely due to the diversification of ICBT trade to other destinations, particularly to DR Congo and Kenya, which now accounts for 50 percent and 25 percent of the value of ICBT export trade respectively. of the exports to the As a result, the value of ICBT exports to regional markets is Middle East were to the estimated to have increased to US$ 484 million, up from US$ 390 million in FY 2015/16. United Arab Emirates While the bulk of Uganda’s imports comes from Asia, the share of imports from the Middle East and Africa is 9 Figure 11: Uganda is mainly importing machinery and equipment, largely from Asia Composition of imports Source of imports Source: Uganda Bureau of Statistics Conversely, the increase in imports from Tanzania As a result of these outflows, the overall deficit in the has been mainly driven by an increase in industrial income account increased significantly over the year, consumer goods following recent efforts by both from a value of US$ 490 million to US$ 673 million. countries to improve the transport corridor through With the value of official transfers declining by about Tanzania, hence opening up the market in Uganda. 23 percent to US$ 184 million, the improvement in secondary income was mainly driven by an increase Over the year, the performance of Uganda’s in the value of private transfers by 25 percent over the services, income and transfer accounts subsectors fiscal year, to reach US$ 1,183 million. It is possible was mixed, with the net impact of the performance that this increase was driven by the improving global of these sectors failing to support an improvement economy and by the inflow of resources resulting from in the overall current account. During this year, the the increased influx of refugees from South Sudan. value of tourism receipts (referred to as travel credit 1.4 The fiscal policy stance has been less expan- within the BOP statistics) amounted to US$ 979 million, sionary, in part due to under-execution of the capi- falling below the US$ 1 billion mark for the first time tal budget since FY2011/12. A number of factors have contributed to the decline, including the lingering impact of The overall fiscal deficit contracted sharply election-related uncertainties during the first half of during FY2016/17, declining from 4.4 percent of the year, followed by the Kasese clashes in western GDP in FY2015/16 to 3.5 percent in FY2016/17. The Uganda; by sporadic outbreaks of disease across actual fiscal deficit was almost half of the budgeted Africa, including avian flu and Ebola; and by outbreaks value of 6.2 percent of GDP, indicating a significant of conflict and unrest in a number of neighboring under-execution of the planned budget. This under- countries, including the ongoing clashes in South execution was primarily due to lower-than-budgeted Sudan and the conflict related to the election in Kenya. development spending (capital expenditures), with In terms of the income account, factors contributing actual expenditure 2.7 percentage points of GDP to outflows increased during the year, including: lower than the budgeted level. In addition, not all (i) repatriated profits associated with foreign direct expenditures were envisaged in the approved budget, investment (by 35 percent); (ii) income payments for with the Government thus being required to issue portfolio investments, mainly on account of the higher supplementary budgets to a total value of Shs 62 billion rate of interest paid on securities (by 50 percent); (iii) (equivalent to four percent of the approved budget), compensation for foreign residence employees (by 16 well in excess of the threshold limit of three percent percent); (iv) and other investments (by 38 percent). 10 Table 1: Fiscal Operations In percent of GDP FY2014/15 FY2015/16 FY2016/17 FY2017/18 FY2017/18 App. Budget Proj. Revenues and grants: 14.4 15.2 15.4 16.2 15.5 Domestic revenues 13.2 13.8 14.3 14.4 14.0 o/w Tax revenues 12.4 13.0 13.5 13.6 13.2 External Grants 1.2 1.4 1.0 1.8 1.5    Total expenditure 18.7 20.1 19.3 22.5 20.2 Recurrent 10.0 11.0 11.0 10.4 10.5 Development 6.8 7.1 7.4 9.8 8.6 Domestic Development 4.3 4.2 4.7 4.7 4.3 Externally Financed Projects 2.5 2.9 2.7 5.1 4.3 Net Lending & Investment 1.6 1.8 0.6 1.9 0.7 o/w Hydro-power Project 1.3 1.6 0.5 1.9 0.7 Other (e.g. Clearance of domestic arrears) 0.3 0.1 0.1 0.4 0.4 Float 0.1 0.4 -0.4 0.0 0.0 Overall balance -4.4 -5.3 -3.5 -6.2 -4.7 External Financing 1.2 3.0 2.9 5.4 3.7 Domestic Financing 3.2 2.3 0.7 0.9 1.1 o/w Petroleum Fund withdrawals 0.0 0.0 -0.1 -0.1 0.2 Memorandum items:     Nominal GDP (Shs billions) 76,517 82,903 91,351 100,552 101,267 Source: Ministry of Finance, Planning and Economic Development, IMF, and World Bank of the approved budget, defined by the Public Financial The FY 2017/18 budget forecasts an increase to the Management (PFM) (Amendment) Act 2015. headline fiscal deficit to more than 6.0 percent of GDP, with the increase intended to support investment in Over the year, the total value of collected tax revenues infrastructure and social services. The budget envisages continued to increase in proportion to GDP. In FY2016/17, an increase to the total value of expenditure by a full tax revenues, and domestic revenues more broadly, increased percentage point of GDP, to 22.5 percent. Of the total budget, by an additional percentage point of GDP compared to 21 percent has been allocated for the development of the value recorded in the previous year. Budget revenues transportation infrastructure, with 10.5 percent allocated to came to a value equivalent to 14 percent of GDP, despite the the energy sector. In terms of investment in social services, Uganda Revenue Authority (URA) recording a shortfall in allocations to the education sector increased by 44 percent the collection of taxes of Shs 459 billion, equivalent to 0.5 and to the health sector by 26 percent, but as share of GDP, percent of GDP. Nevertheless, the total value of domestic both have remained unchanged compared to the previous revenues has increased dramatically in proportion to GDP year. With domestic revenues projected to increase to 14.4 since FY2014/15, when the value of these revenues stood at percent of GDP, the fiscal deficit was expected to reach a 11.6 percent of GDP. 11 level in excess of 6 percent of GDP in FY 2017/18, with context of declining growth, the authorities have more than 80 percent of the fiscal deficit being funded proposed a new draft of tax incentives to stimulate through external borrowing, mainly from commercial the growth of key sectors of the economy and to sources on non-concessional terms (with the total value address employment and export bottlenecks. These of loans contracted on these terms standing at 5 percent proposals include tax relief to several companies and of GDP). Domestic financing is projected to reach 0.9 organizations in FY 2017/18. These additional tax percent of GDP, consistent with the Government’s incentives could complicate the Government’s efforts to strategy to limit its level of dependence on this achieve the targeted fiscal deficit for the year. Half-year expensive source of financing and to avoid crowding estimates of Government revenue and expenditure for out the private sector’s access to credit. FY 2017/2018 show that allocations for the provision of tax support to exempted service providers had been The Government has proposed changes to tax overspent by more than three times compared to the policy that are intended to stimulate the economy, figure envisaged in the original budget. During the first but with mixed effects on revenue mobilization. quarter of FY 2017/18, the Uganda Revenue Authority With the intention of boosting agricultural production, collected tax revenues amounting to Shs 2.86 trillion, the Government proposed in its new tax bills for FY representing a shortfall of about 3.2 percent compared 2017/18 to remove VAT on crop extension services, to the targeted level, with the shortfall largely due to animal feeds and premixes, deep cycle batteries and the lower than expected collection of international composite lanterns, irrigation equipment and works, trade taxes. sprinklers and ready-to-use drip lines. VAT on access to tourist sites and tour guides and game driving services The authorities have increasingly resorted to has also been removed and first-time investments domestic borrowing to finance the fiscal deficit. on sites more than 50 km from Kampala have been The recurrent deficits had increased Uganda’s debt granted tax relief. At the same time, a withholding tax stock to about 39 percent of GDP by the end of FY of 15 percent will be introduced on gross payments 2016/17. While about 65 percent of the total debt stock for winnings from various forms of betting on sports, consist of external debts, the nominal stock of domestic while the gaming tax rate will be reduced from 35 debt at cost has increased rapidly to reach Shs 11,486 percent to 20 percent. The authorities have also billion, or 13.2 percent of GDP, by the end of FY 2016/17. indicated that in FY2017/18, the Uganda Revenue The increased reliance on government securities will Authority will be given the mandate to collect non-tax necessitate a higher level of coordination between the revenues. This is expected to reduce leakages in the monetary and fiscal authorities to avoid crowding out system and to contribute significantly to domestic the private sector’s access to credit. The Government’s revenue mobilization efforts. At the same time, some unexpected decisions to issue securities for fiscal key tax policy proposals such as the expansion of operations undermine the conduct of the monetary the infrastructure levy to cover more goods and policy and increase its costs. intermediate inputs, were presented in the budget, but were not approved by Parliament. In addition, plans to re-impose import duty on crude palm oil were reversed. Despite its stated intention of tightening tax loopholes, the proposed granting of tax exemptions to a number of domestic companies could severely impact revenue outcomes. In the 12 Primary leaving examination time at one school in the outskirts of Kampala. (Morgan Mbabazi, 2017) 13 2. Economic Outlook Uganda’s real GDP growth is expected to increase from 4.0 percent in FY 2016/17 to 5.1 percent in FY 2017/18. With constrained private investment, the increased output growth will be driven by private consumption and public investment. Barring a recurrence of the protracted drought-like conditions that have occurred in previous years, growth in the agricultural sector should drive an increase in real GDP growth rates to around 6 percent in FY 2018/19. With stronger domestic demand, the current account deficit will widen to 5 percent in FY 2017/18 and to 6 percent in FY 2018/19. However, a large proportion of the current in private consumption and public investment will account deficit will be financed with non-debt result in a widening of the current account deficit to creating FDI inflows, while the further build-up 5.5 percent of GDP in FY2017/18, and to 6 percent of GDP of public external debt does not pose a threat in FY2018/19. to debt sustainability yet. However, there are a number of significant downside risks, with If capital expenditures are executed according lower than budgeted tax collections threatening to the approved budget, the fiscal deficit is the execution of capital expenditure and of the expected to widen to 5 percent of GDP by FY2018/19 attainment of the fiscal target and with weather- before declining to around 4.0 percent of GDP related shocks having the potential to adversely in FY2019/20. With the accelerated construction impact agricultural production, particularly given of key infrastructure projects, total expenditure is the sector’s low level of resilience. likely to increase to a level in excess of 20.2 percent of GDP in FY2017/18 and FY 2018/19. Specifically, the development of oil-related transportation 2.1 Higher real GDP growth expected from infrastructure, especially in the Albertine region, increased public investment and agriculture will help to ensure the smooth commencement of oil output production by FY2020/21. In addition to several road development projects, the Government is preparing In FY2017/18, Uganda’s output growth rate is to contract loans amounting to the value of US$ 2.9 expected to increase to 5.1 percent, up from the billion from China Exim Bank to finance the Malaba- figure of 4.0 percent recorded in FY 2016/17, with Kampala Railway, which will connect the regional the increase largely due to higher levels of private network with the existing network between Nairobi consumption and accelerated public investment. and the commercial sea port of Mombasa. The first Improvements in the macroeconomic environment, phase of the network, between Nairobi and Mombasa, weather conditions and credit availability for some was completed by the Kenyan government in June 2017. sectors have resulted in an increased private enterprise The railway line will be extended in three phases, first output growth of and increases in new orders and to Naivasha, then to Kisumu, and finally to Malaba. In employment. This loosening in the labor market may addition to its transport infrastructure development drive increased private consumption, which may also initiatives, the Government intends to complete the be supported by higher levels of agricultural output Karuma and Isimba hydroelectric power projects and due to improved weather following the end of the to construct power lines to support the distribution of two-year spell of drought-like conditions. The increase electricity, especially in the rural areas. 14 The prospect of fiscal revenues falling short of budgeted The authorities will have to balance the need to stimu- targets and thereby undermining the accelerated late the economy with the need to restrain inflationary execution of capital spending is significant. Based on an pressures. Given the currently low inflationary pressures, assessment of performance in the first quarter of FY2017/18, the authorities may further ease monetary policy to stimu- the value of fiscal revenues, excluding grants, is set to reach late output growth. However, central bank financing of the 14 percent of GDP for the year as a whole. This is close to fiscal deficit could complicate liquidity management and 0.5 percentage points of GDP below the budgeted target. undermine the credibility of monetary policy. In addition, The shortfall is largely due to the lower-than-expected with food crops still accounting for 27 percent of the national collection of taxes on international trade, particularly VAT CPI basket, supply-side factors such as persistently low pro- on imports. While external grants may reach the budgeted ductivity in the agricultural sector combined with weather level of 1.8 percent of GDP by the end of the year, so far, shocks may result in price volatility. Furthermore, if interna- performances have been below par, with disbursements tional oil prices continue to increase, the domestic prices for amounting to Shs 142 million, or less than 10 percent of the many imported goods will also likely rise, contributing to in- budgeted figure of Shs 1.7 billion. The value of disbursements flationary pressures. If monetary policy tightens in response of grants during the first quarter of FY2017/18 is equal to 1/3 to such developments, then private investment may remain of the value disbursed during the same period last year. low in response to tightened credit market conditions. Table 2: Uganda / Macro poverty outlook indicators (annual percent change unless indicated)  Real GDP growth, at constant market prices 5.2 4.7 4.0 5.1 5.7 6.0 Private Consumption 9.0 2.9 2.9 5.0 5.3 5.9  Government Consumption 4.6 -17.9 -14.9 -1.7 7.5 1.7  Gross Fixed Capital Investment -0.5 8.9 0.1 5.2 5.3 6.0  Exports, Goods and Services -2.4 3.6 -2.1 6.0 6.1 6.5  Imports, Goods and Services 4.6 -3.8 -12.5 4.0 4.8 5.0 Real GDP growth, at constant factor prices 5.4 5.1 4.1 5.1 5.6 6.0  Agriculture 2.3 2.8 1.6 2.6 3.1 3.3  Industry 7.8 4.7 3.3 5.7 5.8 6.1  Services 5.9 6.4 5.4 5.9 6.6 7.0 Inflation (GDP Deflator) 5.2 4.3 5.3 4.8 5.1 5.1  Inflation (Private Consumption Deflator) 6.3 7.8 4.6 4.8 5.0 5.0  Inflation (CPI inflation) 2.9 6.6 5.7 5.1 4.9 4.7 Current Account Balance (% of GDP) -8.7 -6.5 -5.1 -4.5 -5.5 -6.1 Financial and Capital Account (% of GDP) 6.3 5.8 4.2 5.1 6.0 7.0 Net Foreign Direct Investment (% of GDP) 2.9 2.2 1.9 1.9 1.9 2.0 Fiscal Balance (% of GDP) -4.0 -4.7 -3.7 -4.7 -5.5 -4.0 Debt (% of GDP) 32.2 35.7 38.5 40.7 44.3 45.8 Primary Balance (% of GDP) -2.5 -2.7 -1.7 -1.5 -3.1 -1.4 Sources: World Bank, Macroeconomics and Fiscal Management Global Practice, and Poverty Global Practice.  Notes: e = estimate, f = forecast, Data reported in financial years July-June   (a) Calculations based on 2009-UNHS and 2012-UNHS.  (b) Projection using point-to-point elasticity (2009-2012) with pass-through = 1 based on GDP per capita in constant LCU.  © Actual data: 2012. Nowcast: 2013 - 2017. Forecast are from 2018 to 2020 15 Box 1: Doing Business 2018: Uganda’s ranking declines According to the World Bank Doing Business 2018 report, released in October 31, 2017, Uganda’s ranking in terms of the ease of doing business indicator has slipped to 122nd place out of 190 countries, down from 115th place in the previous year. On the one hand, the report acknowledges that Uganda has made some significant improvements. In terms of the distance to frontier indicator, Uganda’s score increased to 56.94 in 2017, up from 56.52 in the previous year. This shift is based on Uganda’s improvements to its business regulations. In addition, Uganda’s implementation of reforms to facilitate trade across borders was also acknowledged. These involve measures to reduce the time to comply with export document requirements by allowing for electronic submission of documents related to processing certificates of origin and to develop and improve the Malaba One-Stop Border Post. However, the report also emphasizes that Uganda is not doing enough, as reflected through a comparison with regional comparators. As in previous years, Sub-Saharan African (SSA) countries implemented a greater number of reforms in 2016/17 than those in any other region, with a total of 83 reforms recorded in all the areas measured by Doing Business. In addition to the reform count, SSA also recorded impressive progress in terms of the distance to frontier (DTF) indicator. The region recorded the most significant increase in terms of this indicator, with the average SSA country improving by 1.23 percentage points, compared to the global average of 0.76 percentage points. Eight of the region’s economies improved their DTF score by more than three percentage points this year: Malawi (+6.33), Djibouti (+3.99), Zambia (+3.92), Nigeria (+3.85), Senegal (+3.75), Rwanda (+3.21), Liberia (+3.10), Madagascar (+3.05). Three economies in SSA were among the top ten reformers worldwide. These three countries included Malawi, Nigeria, and Zambia. Malawi and Nigeria joined this list for the first time. Meanwhile, Mauritius and Rwanda (the two African economies in the top 50), have made significant strides upward, with Mauritius ranking in the 25th place and Rwanda in the 41st place. With six reforms, Kenya implemented more reforms than any other country in the region. Rwanda, Mauritania, Nigeria and Senegal each implemented five reforms; Mauritius, Niger, Malawi implemented for reforms; and Zambia, Angola and Benin implemented three reforms. Over the medium term, Uganda’s real output weather forecasts pointing to more favorable climatic growth rate should gradually increase to conditions and better rainfalls than in the recent past. 6 percent or above. Higher growth rates will be Uganda remains vulnerable to risks associated with supported by increased private investment as FDI volatile climatic conditions and food prices, unless inflows to the extractives sector recover following the implementation of mitigation measures, including the issuance of exploration agreements, supported the development of improved irrigation systems, takes by the Government’s renewed prioritization of off. With agriculture remaining the primary source the development of oil-related infrastructure. of livelihood for more than 69 percent of Uganda’s The Government’s renewed commitment to the population, supply disruptions resulting from changes development of this infrastructure follows the to weather patterns could have significant negative agreement between Uganda and Tanzania, signed in effects on consumption and livelihoods. They could November 2017, to construct a pipeline between Hoima also complicate the management of inflation. and Tanga port to facilitate the export of oil. With these developments, the acceleration in output growth 2.2 Risks remain tilted to the downside should be supported by growth in the construction and services sectors. Furthermore, barring another In the short term, delays to the execution of protracted episode of drought-like conditions, the planned public investments and the continued real output growth rate in the agricultural sector crowding out of the private sector may constrain could reach 3 percent in FY2018/19, with longer-term real GDP growth. If the execution of capital 16 expenditure is delayed due to revenue shortages, as has Over the medium term, the prospects for improved been the case with several energy projects in the past, or human development outcomes may be undermined by a if the large investment program fails to deliver a sizable continued emphasis on infrastructure development. The growth dividend, then this may slow output growth and Government’s current strategy identifies infrastructure gaps place the public debt-to-GDP ratio on a steeper upward as the key constraint on growth and is focused on addressing trajectory. In that context, it would be even more important these gaps. This emphasis on infrastructure could lead to for the Government to contain leakages that undermine the underinvestment in human capital development. With more collection of revenues and to maintain tight control over imminent returns on investments in infrastructure than expenditure, including by limiting further tax exemptions in human capital development, global experience shows to companies, as these exemptions could further reduce that many countries may be tempted to postpone the latter, revenue. despite their critical importance for long-term development. The Government’s increased reliance on domestic Any further deterioration in the region’s social, financing threatens to crowd out private investment. The economic and political conditions could further value of the Government’s domestic debt stock has almost undermine private investment in Uganda. Protracted doubled in proportion to GDP, increasing from 7 percent of post-election unrest in Kenya could adversely impact GDP in FY2008/09 to 14 percent in FY2016/17. While this Uganda’s trade with Kenya. In the case of South Sudan, increase may assist in deepening the capital market, it is increased tensions could increase the refugee inflows into also crowding out private sector investment, with higher Uganda, with the numbers rising beyond the one million interest rates increasing the cost of borrowing for the private living in resettlements in the Northern Uganda, with sector. Although interest rates for commercial loans in local implications for social service delivery. Similarly, if tensions currency have declined somewhat since mid-2017, there are within Uganda related to the constitutional amendment are still at exorbitant levels, standing in excess of 20 percent. prolonged, this could also have a negative impact on private investment. Nurses particpate in a health awareness march (Morgan Mbabazi, 2017) 17 2.3 Higher productivity is needed to support security, with the South Sudan peace accord in 2010 higher sustainable growth over the medium term leading to increased agriculture production in the North; and a relatively long period of favorable weather The recent deceleration in Uganda’s real output conditions.1 In terms of non-monetary indicators, growth can be attributed to insufficient investment Uganda also recorded strong improvements. In the in productivity to sustain longer-term growth and period from 2001 to 2011, the infant mortality declined to enhance resilience to exogenous shocks. At the from 88 deaths per 100,000 births to 54 and the under- aggregate level, the shift in the production structure five mortality declined from 137 per 100,000 births to from agriculture to services has not been matched by 90. Over the same period, the ownership of modern a corresponding transformation in the structure of the assets such as telephones and motorcycles increased, labor force. The relative contribution of agriculture to with a concurrent decline in the use of less advanced total output has declined by more than half over the technologies, such as bicycles.2 However, since past two decades, dropping to around 25 percent of FY2012/13 poverty rates have either remained stagnant GDP over the past five years. Nonetheless, it continues or have possibly even reversed. According to the latest to be the primary source of employment for more than Uganda National Household Survey (UNHS) report, 75 percent of the labor force. Increasing productivity published in FY2016/17, the national poverty rate is in agriculture will go a long way towards supporting estimated to have increased by 7 percentage points to output growth and reducing poverty. 27 percent since the previous survey was conducted four years ago. Output growth has been driven by the proliferation of small low-productivity firms, in both the Faster poverty reduction and inclusive agricultural and services sectors. Despite development has been adversely impacted by lack households shifting from involvement in agriculture of access to basic services, regional disparities, to small-scale, low-value added retail services, and gender inequality. Only 14 percent of Ugandan output growth has been little affected by this shift households have access to adequate sanitation in production. With the majority of new firms being facilities, with this rate half of the Sub-Saharan Africa established in low value production sectors, the (SSA) regional average. In Uganda, only one in five transformation of production into manufacturing and households uses electricity for lighting, with the higher value-added products has been slow. Businesses regional average standing at one in three.3 Increasingly, that have recorded the highest increases in productivity poverty has become concentrated in the Northern have generally achieved this through the deployment of and Eastern regions, with 84 percent of Uganda’s new technology, which in part has replaced labor. poor located in these areas. Finally, a large proportion of Uganda’s poor live just above the poverty line The deceleration of output growth, particularly in and remain highly vulnerable to external shocks, agriculture has been accompanied by a reversal including shocks related to price volatility and adverse in the progress made towards poverty reduction weather incidents. During the period from 2006 to since 1992. Uganda’s trajectory in poverty reduction 2010, for every three Ugandans who rose above the can be characterized by three distinct periods: the first poverty line, two fell back below the line, illustrating runs from FY1992/93 to FY2005/06; the second from the fragility of the gains achieved.4 In FY 2012/13, 43 FY2005/06 to FY2012/13; and the third from FY2012/13 percent of Ugandan households could be categorized to the present. During the first period, the poverty as vulnerable to falling back below the poverty line, rate declined from 56 to 31 percent, largely due to with incomes above the threshold level that marks the increased productivity of food crops, high average poverty line but not exceeding twice this figure. global coffee prices, and economic transformation characterized by diversification into non-farm income. 1 World Bank, 2016, The Uganda Poverty Assessment, September 2 World Bank, 2016, The Uganda Poverty Assessment Report 2016. Farms, The second period was characterized by improved cities and good fortune: assessing poverty reduction in Uganda from 2006 to 2013. connectivity, with better infrastructure resulting in 3 World Bank, 2016, Uganda Systematic Country Diagnostic, Boosting Inclusive Growth and Accelerating Poverty Reduction, Washington DC enhanced access to markets; by improved regional 4 Economic Policy Research Centre, 2012. Sewanyana and Kasirye. 18 Figure 12: Sustained achievement in poverty reduction recently reversed Source: Uganda Bureau of Statistics The recent spell of slower real GDP growth has For Uganda to achieve higher rates of economic postponed Uganda’s goal to advance to middle growth, it must increase the productivity of its labor income status, with issues related to child marriage, force and increase its stock of capital. The approaches early childbearing and low educational attainment adopted to address obstacles to higher productivity need playing an important role in this postponement. With to ensure inclusive growth. Indeed, efforts to build both Uganda’s high population growth rate, roughly at around physical and human capital through investments in 3.3 percent with no sign of decline over the last twenty infrastructure and social services can yield good results, if years, the current economic growth rate generates only appropriately implemented. Achieving sustainable, long- modest improvements in average per capita incomes. This term improvements to productivity require a sustained focus in turn leads to limited gains in consumption per capita on the development of an inclusive education system that for households, contributing to persistently high rates of equips the workforce with the skills needed to facilitate poverty, and in some cases increases in poverty when growth these improvements. Not only do phenomena such as early in not pro-poor. Child marriage, early childbearing, and low marriage, early child bearing and failure to attain a good educational attainment for girls contribute to high fertility level of education have an adverse impact on productivity at rates for women and thereby to high population growth. the individual level, they also have an adverse impact at the In addition, they reduce the earnings potential of women. community and national level. The second part of this report Larger household sizes as well as limited employment presents an analysis of the cost of failure to invest in the opportunities for poorly educated girls both contribute to girl child and outlines a path forward to achieving inclusive household poverty, as discussed in the second part of this human capital development and thereby to achieving higher report. With the current deceleration in growth, it may take rates of inclusive growth and poverty reduction. at least another 10 years for Uganda to elevate its average per capita income from the current US$ 706 to US$ 1000 and achieve middle-income status as well as substantial poverty reduction. 19 2 PART The Cost of Not Investing in Girls: child Marriage, Early Childbearing, and Low Educational Attainment for Girls, and their Impacts Child marriage, early childbearing, and low educational attainment for girls have a wide range of negative impacts not only on the girls themselves, but also on their children, their families, and society at large. Some of the impacts with the largest economic costs relate to fertility and population growth, women’s earnings, and the health of children born of young mothers. Other impacts range from losses in women’s agency to higher risks of intimate partner violence. Ending child marriage today could generate by 2030 up to US$2.7 billion in annual benefits (in purchasing power parity terms) simply from lower population growth and a reduction in rates of under-five mortality and stunting for young children. In addition, women’s earnings today would be higher if they had been able to avoid marrying early. This loss in earnings is estimated at more than US$ 500 million. Other benefits from lower population growth would include budget savings for the government for the provision of basic services. These savings could be invested to improve the quality of the services provided. To delay the age at first marriage and childbearing, adequate laws are a first step, but interventions are also needed. Interventions that alleviate economic constraints to girls’ education tend to be the most proven. But there is also a role for interventions expanding economic opportunities for adolescent girls who dropped out of school and are not likely to be able to go back to school. Imparting adolescent girls with life skills and reproductive health knowledge is also essential, whether girls are in school or out of school. This can be done among others through safe space clubs that have proven effective for such purposes. To improve educational attainment for girls, basic conditions also must be met. At the secondary level, there is a need to build schools closer to where children (boys or girls) live. As an alternative, adequate modes of transportation to schools must be provided. Providing separate water, sanitation and hygiene facilities for girls is also important, as is the need to reduce the risk of violence and sexual harassment in school. Finally, for the broader challenge of gender-based violence and inequality, community-based interventions to work with men, women, leaders and service providers can also be beneficial. 20 Revision time for lower secondary students in a Kampala school (Morgan Mbabazi 2017) 21 3.1 Why should we measure the costs of child within the household. Fundamentally, these girls are marriage, early childbearing, and the lack of at a higher risk of being disempowered in ways that education for girls? deprive them of their basic rights. This affects not only the girls, but also their children. For example, the Despite progress over time, adolescent girls still children of young mothers face higher risks of dying have on average in Uganda lower educational by age five, being malnourished, and doing poorly attainment than boys at the secondary level. in school. The economic and social costs of child In addition, rates of child marriage and early marriage, early childbearing and low educational childbearing remain high. Educating girls and ending attainment are large first and foremost for individuals, child marriage and early childbearing is essential for but also at the level of communities and the nation. girls and young women to have agency in their life, not only as wives and mothers, but also beyond those While there is support in Uganda for ending child roles. It is also essential to enable Uganda to reach marriage, preventing early childbearing, and middle income status. Unfortunately, despite progress educating girls, investments are not sufficient for over time, more than one in three girls still get married lasting change. Ensuring that girls receive a good before the age of 18 according to the latest publicly education and do not marry at an early age are targets available Demographic and Health Survey for 2011 under the Sustainable Development Goals. Uganda (the unit level data from the 2016 survey have not yet and several other countries in sub-Saharan Africa been publicly released; therefore the analysis is based have adopted laws and national strategies to end child on the 2011 survey). In addition, almost three in ten marriage, prevent teenage pregnancies, and invest girls have their first child before the age of 18. In part in girls’ education. But while laws and strategies can because of persistently high rates of child marriage and help, they are not sufficient. Targeted interventions early childbearing, completion rates for primary and are needed to deal with economic constraints as well especially secondary education for girls remain low. as social norms and cultural traditions contributing to This means that girls, their children and the country poor outcomes for girls. This requires significant and are not empowered to reach their full potential. Given long-term financial investments and political will. In that Uganda has one of the most youthful populations other words, while there is a consensus that dropping in Sub-Saharan Africa, and one of the highest rates out of school and marrying or having children too early of young women out of the labor force, the benefits have a wide range of negative development impacts, from empowering adolescent girls should be large. actual investments to improve opportunities for girls This study, which is adapted from a global study with remain limited. a similar focus, makes the case for investments to end child marriage, prevent early childbearing, and educate A better understanding of the cost of not investing in girls in Uganda. A more detailed report on the analysis adolescent girls could help convince policy makers and is available apart from the synthesis provided here (see other stakeholders to do more. The lack of adequate `Wodon et al., 2017). investments to benefit adolescent girls may be due in Low levels of educational attainment, child By documenting the economic impacts marriage, and early childbearing affect girls’ of child marriage, early childbearing, life trajectories in profound ways. Girls who were and low educational attainment for girls, married, had children, or dropped out of school early as well as the aggregate costs associated are more likely to experience poor health, have more children over their lifetime, and earn less in adulthood. with those impacts, the study makes a This makes it more likely that their household will live case for investments in adolescent girls in poverty. Other consequences for women associated by the government, donors, and other with a lack of educational attainment, child marriage, stakeholders. and early childbearing include higher risks intimate partner violence and a lack of decision-making ability 22 part to the fact that the economic case for these investments from the 2016 DHS were not yet available for this study; has not yet been made forcefully enough. This may be see Box 2 on data sources). This share was ten points lower especially the case for child marriage and early childbearing, than the share observed among women aged 23-30 (46.3 as the evidence on the benefits of girls’ education is firmly percent), and 16 points lower than for women aged 41-49. established. Child marriage, itself the cause of most early This is encouraging, and it is hoped that this decline will childbearing, is often perceived as a social or human rights be confirmed by the 2016 DHS. At the same time, child issue, and not necessarily as an economic issue. This may be marriage rates remain high, and the same is true for early one of the reasons why sufficient targeted investments have childbearing, defined as mothers having their first child not yet been made to end the practice in many countries. By before the age of 18. In some countries, three fourths or more documenting the economic impacts of child marriage, early of all instances of early childbearing are likely to be due to childbearing, and low educational attainment for girls, the child marriage. In Uganda, as discussed below, about half of study makes a case for more sustained investments by the instances of early childbearing are a likely consequence of government and other stakeholders in those areas. child marriage. This suggests that the two phenomena are closely related, although not always so. Table 3 also provides The study starts with a basic diagnostic, proceeds with an the share of girls marrying or having their first child before analysis of impacts and costs, and ends with a discussion of age 15, with similar trends over time. policy options. More precisely, the report: (i) provides a brief analysis of trends in child marriage, early childbearing, and educational attainment for girls over time and how these Table 3: Trends in Child Marriage and Early issues are linked to poverty; (ii) analyzes the relationships Childbearing for Girls in Uganda (%) between child marriage, early childbearing, and educational Child marriage Early childbearing attainment, as well as their impacts on other development Before 18 Before Before Before 15 outcomes and selected costs associated with those impacts; years 15 years 18 years years and (iii) uses evidence from Uganda and other countries All women 46.8 14.4 37.7 8.5 to discuss policy options to end child marriage, prevent ages 18-49 early childbearing, and improve educational attainment for Age group adolescent girls. 18-22 years 36.5 7.3 28.6 5.0 23-30 years 46.3 14.5 39.2 8.4 3.2. How have child marriage, early childbearing, and 31-40 years 53.3 17.8 42.6 10.1 41-49 years 52.8 19.4 40.4 11.6 girls’ education changed over time? Wodon et al. (2017) Following global trends, child marriage and early childbearing have decreased in Uganda over time, but rates remain unacceptably high. Child marriage is defined as a formal or informal union before the age of 18. Some studies suggest that over the past three decades, the prevalence of 36.5% child marriage in developing countries may have decreased by about a dozen percentage points (Nguyen and Wodon, 2015; see also UNFPA, 2012, and UNICEF, 2014). The decline in Uganda has been slightly larger. As shown in Table 1, the share of women aged 18-2214 who married before the age of 18 was at 36.5 percent according to DHS data for 2011 (data women aged 18-22 who 14 The prevalence of child marriage has been estimated in previous reports among others by UNICEF and UNFPA for women aged 20 to 24. In this married before the age of 18 study, the analysis is carried out for women aged 18 to 22, which tracks more closely the conditions prevailing in countries at the time of the survey. Measures of child marriage could also be estimated solely among girls 18 years of age, but using a larger bracket in terms of years provides more robustness in terms of statistical results. In addition to estimating the prevalence of the practice, other useful measures can also be estimated (see Nguyen and Wodon, 2015). 23 Box 2: Data Sources for Analysis The analysis of child marriage, early childbearing, and educational attainment for girls was based on several large scale and nationally representative surveys that capture information on these outcomes. The main data sources used for the quantitative analysis are the following: (i) Demographic and Health Surveys (DHS), with the latest publicly available DHS for Uganda implemented in 2011; (ii) the Uganda National Household Survey (UNHS) for 2012-13; and (iii) the National Uganda Panel Survey for 2011-12. While new data may become available from the DHS 2016 and the UNHS 2015-16, the main messages from the analysis should remain valid. The qualitative analysis relies in large part on data collected in 2012-13 through focus group discussions and interviews in 14 districts acoss the country: Bukomansimbi and Gomba (Central 1 region), Buvuma (Central 2 region), Namayingo (East-central), Kween (Eastern), Makindye (Kampala), Lamwo (North/Mid-northern), Napak and Kaabong (Karamoja/North-east) Yumbe (West Nile), Ntoroko and Hoima (Western/Mid-western), and Kanungu (Southwest/South-western). Topics discussed in the focus group discussions include education, health and water and sanitation. While the focus group discussions were not specifically on child marriage, they provide rich insights into the constraints faced by households that may lead to child marriage and early childbearing. In addition, the qualitative analysis also draws on insights from a range of published studies. Source: Wodon et al. (2017). Uganda has also made progress towards higher educational attainment for girls. Still, fairly few girls complete their upper or even lower secondary education. Substantial efforts have been made in Uganda to improve educational attainment for both boys and girls (see Box 3). At the same time, large gaps remain. Table 4 provides data from the 2011 DHS survey on completion rates for primary, lower secondary, and upper secondary school. The estimates are provided for various age groups, starting a few years after the normal completion age to allow for late starts and repetitions . The data suggest gains in completion rates over time, but with a long way to go to achieve universal primary and secondary education. While progress has been achieved in the last few years since the implementation of the 2011 DHS, data from the UNESCO Institute of Statistics suggest that in 2015, the primary completion rate for girls was still at only 53.4 percent (52.7 percent for boys). At the lower secondary level, the completion rate for girts was at 24.9 percent, Young graduates at Aga Khan University in Kampala below that for boys at 26.8 percent. (Morgan Mbabazi, 2017) 15 For primary, the first age group is girls aged 15 to 18, with a lower comple- tion rate than the next age group. This suggests substantial late starts and repetitions leading to several years of delayed completion. 24 Table 4: Trends in Educational Attainment for Girls in Uganda (%) Primary Completion Lower Secondary Completion Upper Secondary Completion Age Group Age Group Age Group All 15 - 49 38.72 All 18 - 49 17.59 All 21 - 49 8.47 15 -18 37.35 18-20 20.36 - - 19 - 22 51.82 21-24 26.37 21 - 24 12 23 - 30 45.55 25-30 20.45 25 - 30 10.08 31 - 40 29.35 31-40 12.43 31 - 40 6.93 41 - 49 25.56 41-49 8.91 41- 49 4.4 Source: World Bank staff using latest publicly available DHS. Box 3: Efforts to Improve Educational Attainment in Uganda Uganda has a long history of promoting educational attainment for both boys and girls. The policy of providing free primary education was launched in 1997 and later became known as Universal Primary Education (UPE). Ten years later, the Government of Uganda launched a comprehensive reform program to provide universal access to post primary education and training (UPPET) to accommodate the bulge from free primary education provided under UPE. The Government created two sub-programs under UPPET: (i) Universal Secondary Education (USE) to improve transition rates from primary to secondary schools; and (ii) similar policies for vocational education (schools/community/polytechnics). Through UPPET, Uganda became the first country in sub-Saharan Africa to introduce a Universal Secondary Education Policy. Thanks to these efforts, Uganda made progress towards higher educational attainment for both boys and girls. Gender parity has been achieved at the primary education level. At entry into lower secondary (Secondary 1), intake rates are similar for boys and girls, but girls start to fall behind towards the end of lower secondary (Secondary 4). At the upper secondary level (Secondary 5-6), the gap between girls and boys is more pronounced, in part due to child marriage and early childbearing, two issues that affect girls much more than boys. 53.4% There are differences between regions in the prevalence of child marriage, early childbearing, and low educational attainment for girls. These differences can be measured finely using census data, or at a higher level of geographic aggregation using survey data. They are illustrated for child marriage and early childbearing in Table 5 at the regional level and for urban and rural areas In 2015, the primary using data from the 2011 DHS. These differences reflect heterogeneity in standards of living, opportunities for girls, completion rate for girls and gender norms across various parts of the country, and they can be used to target interventions to areas with the least favorable outcomes for adolescent girls. 25 Table 5: Child Marriage and Early Childbearing poverty should not be underestimated. In the case by Region (%) of child marriage, as noted in multiple reviews (e.g., Early UNICEF 2005; Vogelstein 2013; UNFPA 2012; UNICEF Child marriage (18-22 years) childbearing 2014; Klugman et al. 2014; Parsons et al., 2015; Wodon, (18-22 years) 2017), social and cultural norms related to gender roles All 36.5 28.6 and gender inequality contribute to the persistence of Region the practice. When overlaid with a culture that assigns Kampala 17.8 19.4 Central 1 26.9 22.9 specific gender roles to men and women, poverty and Central 2 37.4 34.5 a lack of opportunities for girls often leave few options East central 42.0 35.4 but to marry early. Higher fertility and a reduction in Eastern 49.4 38.1 earnings due to child marriage and early childbearing North 51.5 35.3 as well as low educational attainment contribute to Karamoja 59.7 33.9 West-Nile 36.5 23.8 poverty, a lack of empowerment, and the perpetuation Western 39.5 33.1 of disadvantage and gender discrimination over time. Southwest 23.9 13.8 Residence The links between poverty and child marriage, Urban 22.5 24.6 early childbearing, and low educational attainment Rural 40.6 29.8 for girls are very clear in Uganda. In Table 6, Source: Wodon et al. (2017). households are categorized in five quintiles from poorest to richest based on their level of wealth in 3.3. How does poverty affect child marriage, early the 2011 DHS. The prevalence of child marriage and childbearing, and low educational attainment for early childbearing is much higher in the bottom three girls? quintiles of wealth, after which it decreases, especially While cultural factors play an important role in the among girls from the richest quintile. A similar but persistence of child marriage, early childbearing, inverted pattern is observed for girls’ educational and low educational attainment for girls, the role of attainment, with higher shares of girls completing various levels of schooling in the higher quintiles. Table 6: Child Marriage, Early Childbearing and Educational Attainment for Girls by Quintile (%) Primary Lower sec. Upper sec. Child marriage Early childbearing completion completion completion 18-22 18-22 15-18 years 18-20 years 21-24 years Poorest 54.4 41.6 11.6 4.2 0.8 Poorer 49.9 36.6 24.1 7.3 2.3 Middle 40.5 31.3 26.2 8.0 3.5 Richer 29.9 21.8 47.2 22.1 7.9 Richest 19.9 18.9 61.7 45.4 32.0 Source: Wodon et al. (2017) Multiple factors are leading to differences in than boys simply because the benefits of educating outcomes between socio-economic groups. When girls are likely to accrue to in-laws, while the benefits poverty makes it hard for a household to send all of educating boys are more likely to benefit the family children to school, prevailing gender norms may mean of origin. A lack of formal employment opportunities that boys receive preferential treatment, at least at the for young women may also mean that less value is secondary level, while girls may be kept home from placed on secondary education for girls as parents school to take care of housework. Parents in traditional find little benefit in investing in their education. In communities may also place a lower value on girls some cases, girls may have few career choices outside 26 Conversely, child marriage, early childbearing, and low educational attainment for girls also contribute “Parents sometimes make the unrealistic to poverty and differences in wealth between groups. assumption that education means only Early marriage leads girls to have children earlier and more studying up to university, and ignore children over their lifetime. This reduces consumption per all other openings that education can capita in the household, thereby increasing the likelihood culminate into, such as technical, of being poor. Girls marrying early often must leave school, business and vocational sectors which which curtails their earnings potential as adults due to low can be as fulfilling and rewarding in life. educational attainment. These are but two of the channels Discrimination between sexes of children through which child marriage may lead to higher poverty. and predetermining the direction each One important implication of the relationship between child’s education should take is not good.” poverty and child marriage, early childbearing, and low educational attainment for girls is that most benefits from interventions in those areas would accrue to the poor. The costs associated with the impacts of child marriage, early of marriage and child rearing. A lack of meaningful social childbearing, and low educational attainment for girls are and economic alternatives may make it difficult for girls and borne by the poor. Therefore, programs and policies in these their families to envision viable alternatives to early marriage areas would be pro-poor in Uganda as in other countries. and childbearing. Structural weaknesses in the provision of education also play a role. The fact that schools are of poor 3.4. What types of impacts and costs are associated with quality, sometimes far away, or costly for families in terms child marriage, early childbearing, and low educational of both fees and lost hours of (unpaid) household work for attainment for girls? girls when they go to school may lead to de-prioritizing girls’ education and may encourage parents to marry off their A simple conceptual framework is used to measure the daughter, particularly if she is deemed to be of marriageable economic and social impacts of girls dropping out of age and the suitor is acceptable. Further, when poor families school, marrying early, or having children early. The face food insecurity, having a girl marry early gives that framework is illustrated in Figure 1, which is adapted from family one less mouth to feed. Finally, financial transactions a global study on the economic impacts of child marriage. (bride prices or bride wealth) around marriage may also lead As discussed in the previous section, the top panel in Figure parents, or in some case their daughters, to benefit from 13 recognizes that girls’ education and child marriage/ marrying these daughters early. early childbearing are closely linked. The literature and estimates from this study suggest that keeping girls in school is one of the best ways to delay marriage and early childbearing in Uganda. Conversely, marrying or having The causality between poverty and child a child early often lead girls to drop out of school. In turn, marriage, early childbearing and a lack both girls’ educational attainment and child marriage/early childbearing matter for other development outcomes. Four of educational attainment for girls runs main outcomes are considered in this study: (1) fertility and both ways. Poverty increases these risks population growth; (2) health for mothers as well as their substantially for girls, and those risks children (including under-five mortality and malnutrition contribute to the transmission of poverty for children, and the risk of exposure to intimate partner from one generation to the next. violence for mothers); (3) work (including labor force participation, land ownership, and earnings); and (4) agency (including decision-making and other impacts). While some 27 of the impacts of child marriage, early childbearing, Apart from looking at the relationships and educational attainment for girls are estimated between child marriage, early for the girls themselves, others are estimated for their childbearing, and education for girls, children. We use the term “impacts” for simplicity, this study looks at their impacts in four recognizing the difficulty of establishing causality areas: fertility, health, work, and agency. (see Box 4). In turn, various types of costs associated with those impacts are estimated. Selected monetary costs or benefits associated with the impacts of child marriage, early childbearing, and low educational attainment for girls are estimated. The framework distinguishes Figure 13: Conceptual Framework for Measuring Impacts four types of benefits from educating girls and ending and Costs child marriage and early childbearing: income gains, welfare gains, budget savings, and other benefits. Examples of benefits include (i) higher growth in GDP per capita thanks to lower population growth; (ii) higher earnings for women in adulthood thanks to higher educational attainment; (iii) higher earnings for children in adulthood thanks to a reduction in stunting; (iv) benefits associated with the valuation of children’s lives saved; and (v) reduced budget needs for the Ministry of Education thanks to lower population growth. This list is by no means exhaustive, but it includes some of the largest expected economic benefits. Finally, benefits from educating girls and ending child marriage and early childbearing at the level of individuals and households have broader implications at the national and even global level. By raising standards of living (including through higher GDP per capita thanks to lower population growth and higher earnings for women), educating girls and ending child marriage and early childbearing should reduce Source: Wodon et al. (2017). Youth enjoy welding work at a Workshop in outskirts of Kampala (Morgan Mbabazi, 2017) 28 poverty and inequality, as well as increase standards of living substantially. The economic benefits from improving outcomes for girls can be measured as flows from one year to the next or as stocks. Examples of flows include impacts on GDP per capita and annual earnings for women. An example of stock is the estimate of the wealth of a nation. The concept of a nation’s wealth differs from that of its GDP. While a country’s GDP is an aggregate measure of annual production or income, wealth is a measure of the assets of a country. A country’s assets are what enables it to produce future income or GDP. Recent research suggests that human capital accounts for two thirds of global wealth worldwide, with produced and natural capital accounting for the rest. Improving outcomes for adolescent girls would contribute to increasing Uganda’s human capital, and thereby its wealth and long-term capacity to generate future incomes. Said differently, educating girls and ending child marriage and early childbearing would help in ensuring the long- term sustainability of Uganda’s development path. In this study the focus is on measuring benefits as annual flows, although the analysis of impacts on wealth is available from the authors. It is worth noting that estimates of costs based on wealth are much larger than Youthful nursing officers at service of in Jinja referal hospital (Morgan Mbabazi, 2017) those based on annual flows of income. Box 4: What Is Meant by “Impacts” and Associated Costs? The aim of this study is to estimate the impacts of child marriage, early childbearing, and girls’ education on a wide range of development outcomes and the economic costs associated with some of these impacts. The term “impact” is used for simplicity, but one must be careful about not necessarily inferring causality. Estimates of impacts are obtained through regression analysis aiming to isolate the potential impact of child marriage, early childbearing or girls’ education on various outcomes controlling for other factors affecting those outcomes. . What is measured is a statistical association In the literature, this approach is known as “association studies” between child marriage, early childbearing or girls’ education and various outcomes. This is not necessarily an impact as could be observed with a randomized control trial. Since child marriage or early childbearing cannot be randomized, the study must rely on regression analysis to estimate impacts, but there is always a risk of bias in the resulting measures of likely impacts. Based on measures of impacts, costs associated with some of these impacts are computed. These costs are based on assumptions that could be debated, including discount rates in some cases. Therefore, cost estimates only represent an order of magnitude of potential costs, and not precise estimations. These caveats must be kept in mind. 29 3.5. What are the estimated impacts of child mar- childbearing can be complex. In some cases, having riage, early childbearing, low educational attainment one or more children before the age of 18 may be the for girls? consequence of child marriage. In other cases, the inverse may be true, as child marriage may result This section provides estimates of the economic from an early childbirth or an expected birth. For and social impacts of child marriage, early yet other girls, early childbearing may not be related childbearing, and low educational attainment to child marriage at all. Still, child marriage is more for girls. The section looks first at the relationships likely to lead to early childbearing than the reverse. in Uganda between the three issues (child marriage, Using simple assumptions and data on the timing of early childbearing, and low educational attainment for a women’s first marriage and first child (in terms of girls). Thereafter, the section looks at the impacts of the number of months separating the two events), it these issues on other development outcomes related to is possible to provide a rough estimate of the share of fertility, work, health, and agency broadly defined. early childbearing likely due to child marriage. For Uganda, estimates suggest that at the level of mothers, 3.5.1. Links between child marriage, early childbearing, more than half (52.4 percent) of early childbearing may and girls’ education be due to child marriage. At the level of children, 56.4 percent of early childbirths (children born of mothers In Uganda, child marriage is likely the cause of more younger than 18) may be due to child marriage. In than half of all instances of early childbearing. The other words, ending child marriage could reduce early relationships between child marriage and early childbearing for mothers and early childbirths for children by about half. These estimates are lower than in many other countries, suggesting that a substantial Slightly more than half of all instances proportion of cases of early childbearing may occur of early childbearing are likely due outside of formal or informal unions. Still, it appears to child marriage in Uganda. This that the majority of cases of early childbearing is proportion is lower than in many other likely due to child marriage. Ending child marriage countries. should therefore have a major positive impact towards reducing early childbearing, whether at the level of mothers or children16 . Table 7: Impacts of Child Marriage on Early Childbearing Indicators Estimated Impacts (1) Early childbearing for women (first child Child marriage is likely the cause of more than half (52.4 percent) of girls before the age of 18) having children before the age of 18 (2) Early childbirths for young children (being Child marriage is likely the cause of more than half (56.4 percent) of born of a mother younger than 18) births of children from mothers younger than 18 (3) National rates of early childbearing Ending child marriage could reduce early childbearing for girls as well as (for mothers) and early childbirths (for early childbirths for children by about half children) Source: Wodon et al. (2017) 16 At the margin, ending child marriage may entail behavioral responses which in some cases could lead to births among young mothers out of wedlock. The extent to which such behavioral responses could be observed would need to be estimated using more advanced models as opposed to simple statistics. But the simple statistics provided here suggest that even if such behavioral responses were to be observed in some cases, it is still likely that ending child marriage should lead in Uganda to a major reduction in early childbearing. It is important to note, however, that ending child marriage would not be sufficient for avoiding all early pregnancies and childbirths. Providing adolescents with access to sexuality education and adolescent-friendly reproductive health information and services are critical to ensure that adolescent girls do not face unintended pregnancies within or outside marriage. 30 level of education considered. Principals also confirm that child marriage and pregnancies are major reasons for girls Child marriage and early childbearing dropping out. The second approach consists of estimating have large negative impacts on girls’ the impact of child marriage on educational attainment educational attainment. Conversely, for girls econometrically. Estimates for Uganda suggest remaining in secondary school reduces the that child marriage has a large and statistically significant risk of marrying or having children early. impact on secondary education enrollment and completion. For example, child marriage may reduce the likelihood of completing secondary school by 12 to 23 percentage points depending on how early girls marry. This is confirmed by the The causality between child marriage/early fact that once a girl is married, statistical analysis suggests childbearing and educational attainment goes both that it is very difficult for her to remain in school. Finally, ways. First, child marriage and early childbearing have a negative effect on educational attainment. while child marriage for a mother itself may not have a direct Two approaches can be used to assess the impact of child negative impact on the education of her children, the fact marriage on educational attainment for girls. The first that child marriage may reduce the educational attainment approach consists of asking parents in household surveys of the mother leads to a negative impact on the education why their daughters dropped out of school. According prospects of her children (boys and girls). The results are to parents responding to household survey questions in summarized in Table 8. They point to a large negative impact Uganda, a pregnancy is the reason for dropping out of of child marriage on educational attainment for girls, and to secondary school for up to a third of girls depending on the some extent on the education of their children too Table 8: Impacts of Child Marriage/Early Childbearing on Educational Attainment Indicators Estimated Impacts According to parents and principals, early pregnancies and marriages are major reasons for (1) Girls dropping out of school dropping out of school Child marriage reduces the likelihood of completing secondary school by 12 to 23 (2) Educational attainment for girls percentage points Once a girl is married, statistics suggest that it is very difficult for her to remain in school, (3) Marriage vs. schooling trade-off whatever her age (4) Intergenerational effects Child marriage affects the education of the children of girls marrying early at least indirectly Sources: Wodon et al. (2017). Conversely, keeping girls in schools often helps to delay child marriage and early childbearing. The analysis for this study “We are faced with long distances to suggests that each year of secondary education leads to a primary schools. Girls on their way to reduction in the likelihood of marrying as a child of seven school meet men who entice our daughters percentage points, and a similar impact is estimated on the with money for sex. Later some get probability of having a first child before the age of 18. These pregnant and drop out of school. Also, we impacts are large as Table 9 shows. These impacts are also have no vocational school that will train confirmed by a review of the literature on interventions our girls after P7 and S4, so we see it as a that have the potential to delay marriage, given that many of the more successful interventions to delay marriage and waste of resources to educate a girl.” childbearing tend to focus on keeping girls in school or enabling them to return (as discussed later in this report) 31 Table 9: Impacts of Girls’ Educational Attainment on Child Marriage/Early Childbearing Indicators Estimated Impacts Each year of secondary education leads to a reduction in the likelihood of (1) Education’s impact on child marriage marrying as a child of seven points (2) Education’s impact on early Each year of secondary education leads to a reduction in the likelihood of early childbearing childbearing of seven points Sources: Wodon et al. (2017). However, qualitative data suggest heterogeneity Yet in other communities, support for girls’ between communities in factors leading to child education remains weaker than it is for boys. marriage, early childbearing, and low educational This contributes to child marriage and early attainment for girls. In some communities, there is childbearing. In those communities, preference for equal support for boys and girls. Wodon et al. (2016) investing in the education of boys is rooted in cultural look at the relationship between child marriage and frames considering sons as natural heirs. Parents educational attainment for girls in Uganda with data prefer educating boys because girls must inevitably from focus group discussions and key informants. get married, and the wealth they accumulate benefits In some communities, respondents express virtually in-laws. By comparison, boys normally remain within equal support for the education of boys and girls. the environment of their parents even after marriage They express the belief that “All children are equal, and help them throughout their life. There is also a so all deserve an equal right to education as vital to perception that girls are diverted from education by all children in the community for the sake of self- men at an early stage, so families would rather educate independence in the future.” Education is seen as boys who may stay longer in school: “We are faced shaping character, which motivates parents to give with long distances to primary schools. Girls on their equal attention to all children irrespective of their way to school meet men who entice our daughters with sex: “Education is good, and it is the responsibility of money for sex. Later some get pregnant and drop out all parents to ensure that both girls and boys attend of school. Also, we have no vocational school that will school. It is education that empowers a person to train our girls after P7 and S4, so we see it as a waste get to his or her destination in life. Most dreams and of resources to educate a girl.” Another factor are the ambitions are easier to fulfill and realize for people consequences of HIV/AIDS orphaning children and who are educated”. Educating boys only is considered leaving them under the care of grandparents who may “old fashioned thinking”, because while there is a risk be overwhelmed. As a result, some grandparents may that a girl will get married sooner, there is also a risk exhort girls to “kula ogende ofumbirwe,” which means that a boy, when educated, may care more about his “grow up and get married quickly”. Overall girls tend in-laws. In some communities, educating girls is also to be more vulnerable than boys in many communities. seen as a way to fortify girls against uncertainties in In Karamoja, the poorest area of the country, daughters today’s more volatile and fragile marriage institution. may also be seen as a source of wealth for parents. In Educated girls are more likely to be able to have an the past, when there was plenty of cattle, bride wealth independent life if confronted with problems in their could reach 100 cows. Although this number has now marriage. In one community, the education of girls been reduced considerably, girls are still potentially was considered important for future generations, as a source of wealth, and their socialization at home is members explained that if you take girls to school, they towards getting married as soon as possible. Puberty become better mothers for their children. also makes it difficult for girls to continue to go to 32 girls to remain in school may be the way to go, but in other communities, these interventions may not work without additional programs tackling gender-based social norms. There is heterogeneity between communities in the drivers of child 3.5.2. Impacts on fertility and population growth marriage, early childbearing, and low educational attainment for Child marriage and early childbearing have large impacts girls. The role of social norms and on fertility, and thereby on population growth, but effects gender inequality is stronger in some on modern contraceptive use are smaller. Child marriage communities than others. and early childbearing can have devastating and long- lasting consequences for girls (see Box 5). They also have implications for fertility and population growth at the national level. Total fertility is defined as the number of live births that women are expected to have over their lifetime school, as does the division of labor in the home, with girls under current conditions. The analysis for this study focuses tending to most of the household chores. on the impact of child marriage and educational attainment on total fertility. Controlling for other factors, a girl marrying This heterogeneity must be acknowledged when at 14 will have on average 19 percent more children over her designing interventions to end child marriage, prevent lifetime in Uganda than if she had married at 18 or later. early childbearing, and improve girls’ education. The If a girl marries at 17, she would still have on average 10 close links between child marriage, early childbearing, and percent more children over her lifetime. Considering the rate low educational attainment for girls in many communities of child marriage and the characteristics of girls married suggest that access to quality primary and secondary early, ending child marriage could reduce the national rate education is one of the best ways to delay marriage and of total fertility by 8 percent. This impact of child marriage childbearing. Parents know that school can sometimes lead on total fertility comes in part from the fact that women to highly valued employment. Why then do so few girls marrying early tend to also have children earlier. As a result, complete secondary school? Parents wishing to educate their according to demographic projection models, ending child daughters face an array of economic, social and institutional marriage and early childbearing could reduce population barriers, including schooling costs (out of pocket and growth in Uganda by 0.17 percentage point. By contrast, opportunity costs) and the poor quality of the education marrying early tends not to be associated with an increase or provided. Without alternatives, early marriage may be decrease in modern contraceptive use later in life, possibly attractive to parents and even some girls. This suggests that because of limited access to contraception. These impacts are in many communities, providing economic incentives for summarized in Table 10. Table 10: Impacts of Child Marriage/Early Childbearing on Fertility and Population Growth Indicators Estimated Impacts (1) Number of live births over lifetime Depending on the age at marriage, child marriage increases total fertility for women by 10% to 19% (2) National rate of total fertility Ending child marriage would reduce the national rate of total fertility by 8% (3) Use of modern contraception Marrying as a child does not have a statistically significant impact on modern contraceptive use (4) National rate of contraceptive use Ending child marriage would not affect national use of modern contraceptives (5) Population growth Ending child marriage and early childbearing could reduce population growth by 0.17 percentage point Sources: Wodon et al. (2017). 33 Box 5: Early Childbearing May Have Long-Lasting Negative impacts The Susan was 18 years old at the time she was interviewed. Her mother had died. With one sister and four brothers, she lived with her father. She started school at six years of age and dropped out last year because she became pregnant at the age of 17. She was still in primary school. She had dropped out previously, in 2008, to help her mother who was bed-ridden just before she died. At that time, she was in the third year of primary school. She now works as a casual laborer in people’s gardens, earning about 8,000 shillings a week. Payment is usually in cash, but at times in kind with sorghum or millet to bring back home. She uses her earnings to buy essential things for the home such as soap, salt, sugar, and food. The challenge she faces now is that she cannot work effectively because she is pregnant and sickly. Yet, she is still supposed to look after her siblings. In her assessment, gardening is much tougher than school, but she is emphatic that “I cannot go back to school any more. I just want to take care of my young siblings and see them through primary school, and if possible up to secondary school.” Support that could help her realize her wish of a better education for her siblings could be seed money to help her start an income generating activity, again to help her siblings complete school. Source: Wodon et al. (2016). There is also evidence of an impact of low educational gains associated with primary and secondary education attainment for girls on fertility, population growth, completion, for which effects in the regression analysis and contraceptive use. In the regression analysis, the are not statistically significant, we conclude in Table impacts of the completion of primary or secondary 11 that the evidence is mixed even if the estimates for education on total fertility as well as on modern other levels of education suggest impacts. contraceptive use are not statistically significant. However, having some secondary education or Child marriage and educational higher education is associated with a statistically attainment for girls both have impacts significant reduction in total fertility (-11 percent on fertility, and thereby population for some secondary education, and -30 percent for growth, but in Uganda data on those higher education). Furthermore, some secondary impacts are more robust for child education is also associated with an increase in modern marriage than for girls’ education. contraceptive use (increase of six points), and this is also the case for higher education (increase of 15 points). Because the focus in this study is on estimating Table 11: Impacts of Girls’ Educational Attainment on Fertility and Population Growth Indicators Estimated Impacts (1) Number of live births over lifetime Some secondary and higher education lead to reductions in fertility of -11% and -30% respectively (2) National rate of total fertility Due to a focus on estimated impacts for primary or secondary completion, no national effects are reported (3) Use of modern contraception Some secondary and higher education lead to increases in use of 6 points and 15 points respectively (4) National rate of contraceptive use Due to a focus on estimated impacts for primary or secondary completion, no national effects are reported. (5) Population growth Due to a focus on estimated impacts of primary or secondary completion, no national effects are reported Sources: Wodon et al. (2017). 34 3.5.3. Impacts on health and nutrition Through its impact on early childbearing, child marriage “Some children look younger than their affects the health and nutrition of children born of young biological age due to poor nutrition. Those mothers. It also has an impact on intimate partner violence, children between eight and nine years look but the impact on maternal mortality is less clear. The as if they are five or six years old.” literature suggests that adolescent girls have in many countries a higher level of maternal morbidity and mortality than women aged 20-24. At the same time, while avoiding pregnancy at a very young age is essential, it does not effect on the national rates of under-five stunting and follow that ending child marriage and thereby reducing mortality because only a small share of children are born early childbearing would necessarily result in a decrease of mothers younger than 18 at the time of birth. Finally, in maternal mortality ratios at the national level. This is child marriage is associated with higher risks of intimate in part because women may then have children at a later partner violence for women in Uganda, with larger effects age when rates of maternal mortality are higher. Other than in other countries. Noting that this study does not health impacts of child marriage and early childbearing address all aspects of health that may be affected by child are more clear-cut. When a child is born of a young mother, marriage and early childbearing, such as obstetric fistula or controlling for other factors, this increases the risks of under- sexually-transmitted infections (including HIV/AIDS), the five malnutrition (stunting) and mortality. In Uganda, the magnitude of the impacts on health, nutrition, and violence impacts on stunting and mortality are large at the margin. are summarized in Table 12. Ending early childbearing would however not have a large Table 12: Impacts of Child Marriage/Early Childbearing on Health, Nutrition, and Violence Indicators Estimated Impacts (1) Maternal mortality The impact that ending child marriage would have on maternal mortality ratios is not fully clear (2) Risk for children of dying by age 5 Being born of a mother younger than 18 increases the risk of under-five mortality by 4.7 percentage points (3) National rate of under-five mortality Ending all early childbirths would reduce under-five mortality from 6.58 percent to 6.31 percent nationally (4) Risk for children of being stunted Being born of a young mother has an unusually large impact on under-five stunting at 22 percentage points (5) National rate of under-five stunting Ending all early childbirths would reduce under-five stunting from 33.61 percent to 32.58 percent nationally (6) Intimate partner violence Child marriage appears to have a substantial direct impact on intimate partner violence Sources: Wodon et al. (2017). Evidence is mixed on the impact of a mothers’ educational Early childbearing has a larger impact attainment on the health of her children. The same holds on under-five mortality and stunting for the impact of a women’s educational attainment on for the children of young mothers than the risk of intimate partner violence. In the regression the mother’s educational attainment. analysis, primary and secondary education completion for In addition, child marriage has a larger mothers do not have statistically significant impacts on the impact on intimate partner violence than risks of under-five mortality and stunting for their young is the case for educational attainment. children. For stunting, there is a statistically significant reduction in risk when the mother has higher education only. Furthermore, in Uganda (but not in many other not necessarily reduce the risk of intimate partner violence. countries), higher educational attainment for women does The impacts are summarized in Table 13 35 Table 13: Impacts of Education on Health, Nutrition, and Violence Indicators Estimated Impacts Controlling for other factors, the educational attainment of the mother does not (1) Risk for children of dying by age 5 affect the risk of under-five mortality Universal primary or secondary education for girls would not necessarily lead to a (2) National rate of under-five mortality national decline in under-five mortality The effect of a mother’s education on under-five stunting is statistically significant (3) Risk for children of being stunted only for higher education (-15 points) Universal primary or secondary education for girls would not necessarily lead to a (4) National rate of under-five stunting national decline in under-five stunting Educational attainment for women does not appear to have a statistically (5) Intimate partner violence significant impact on intimate partner violence Sources: Wodon et al. (2017). 3.5.4. Impacts on labor force participation and earnings attainment, child marriage reduces women’s earnings in adulthood. Specifically, for child brides, earnings Child marriage (and indirectly early childbearing) in adulthood could increase by 15 percent with the do not affect labor force participation in a elimination of child marriage. This could increase the substantial way, but they reduce educational population’s total earnings nationally by one percent. attainment and thereby earnings and household Based on research for other countries, child marriage welfare. In Uganda, child marriage itself does not is not likely to have a direct impact on household have a statistically significant direct impact on labor consumption per capita or food adequacy after force participation for women. Through its impact on controlling for other variables. However, again through fertility and educational attainment for girls, child their impact on fertility and girls’ education, child marriage may result in a small decrease in labor force marriage and early childbearing are likely to reduce participation for women. Ending child marriage would household welfare and increase poverty. These various therefore increase labor force participation only slightly. impacts are summarized in Table 14. More importantly, through its impact on educational Table 14: Impacts of Child Marriage/Early Childbearing on Work, Earnings, and Welfare Indicators Estimated Impacts Ending child marriage could lead to a small increase in labor force participation (1) Women’s labor force participation for women through its impact on education Ending child marriage could increase earnings in adulthood for women marrying (2) Impact on earnings early by 15% through better education Ending child marriage could increase the population’s earnings and productivity (3) National impact on earnings nationally by one percent Child marriage is likely to reduce welfare and increase poverty due to its impact (4) Household welfare on women’s education and fertility. Sources: Wodon et al. (2017). Higher educational attainment increases earnings, with especially large expected benefits from Educational attainment for women universal secondary completion. Universal primary has a large impact on their expected or secondary education may lead more women to work, earnings, and more so than is the case with gains in labor force participation estimated at one for child marriage. percentage point nationally with universal primary education, and five percentage points with universal 36 secondary education. Gains in earnings would be much account for weaknesses in the demand for educated labor. In larger, as a lack of education and agency for girls affects other words, especially for universal secondary education, their productivity, including on the farm (see Box 6). Total the simulations may be overstating potential gains. Still, earnings for the labor force (men and women) could increase earnings gains could be large. The potential impacts on labor by 18 percent with universal primary education. Earnings force participation and earnings as well as household welfare could potentially double nationally with universal secondary are summarized in Table 15. education for women. These simulations however do not Table 15: Impacts of Girls’ Educational Attainment on Work, Earnings, and Welfare Indicators Estimated Impacts Universal primary and secondary education could increase labor force participation by one (1) Women’s labor force participation and five points, respectively Universal primary education could raise earnings by 18%. The impact of universal secondary (2) National impact on earnings education would be larger. Universal primary or secondary education could have large positive effects on welfare and (3) Household welfare reduce poverty substantially Sources: Wodon et al. (2017). Box 6: Women’s Empowerment and the Returns to Farming Even though women make up a large proportion of Africa’s farmers, they are, for the most part, locked out of land ownership, access to credit and productive farm inputs, support from extension services and access to markets, to name just a few factors essential for productivity. This array of challenges means that, on average, Africa’s female farmers produce less per hectare compared with men, which adversely affects their families, communities and, in the long term, their entire country. Analysis by the Africa Gender Innovation Lab at the World Bank suggests that in Uganda, plots managed by women produce on average 13 percent less (in terms of gross value of output) per acre than plots managed by other family members. Relative to men, women’s lower levels of schooling, access to extension services and application of non-labor inputs on their plots, including pesticides and organic fertilizer, widen the gender gap. The cost of the gender gap in agricultural productivity is estimated at US$ 67 million per year in Uganda. Closing this gender gap could lift more than 100,000 people out of poverty and increase annual crop output by 2.8 percent. Policies targeting the provision of technical information to women, securing equal access to and use of non-labor inputs, and supporting women’s education and training would help to alleviate gender inequality in the sector and thereby enhance productivity. Source: O’Sullivan et al. (2014) and Buehren (2015). 3.5.5. Impacts on Women’s Agency and Other Outcomes child marriage does not affect an index of women’s decision- making ability within the household directly, but it does The impacts of child marriage (and indirectly of early matter indirectly through its impact on education, as will be childbearing) on women’s decision-making and other discussed below. Child marriage is associated with a higher indicators of agency are not always large. For this study, likelihood of land ownership for women (more research is we consider measures of household decision-making, land needed to understand why). Child marriage is not associated ownership, knowledge of HIV/AIDS, and birth registrations as with a reduction in knowledge of HIV/AIDS in adulthood, elements of women’s agency. Controlling for other variables, nor does it affect directly a women’s ability to seek care for 37 herself without the permission of her husband (this “They come and get services [at the is one of the aspects of decision-making ability). Yet health center], except they may want a again, indirect effects may be at work through education [contraceptive] method which we don’t as discussed below. Finally, child marriage is also not have. Now this lady is escaping from the associated with a higher risk of not registering a child’s husband. If you tell her go to Jinja and birth in comparison to marrying at age 18 or later, after get counseled, she will not come back.” controlling for educational attainment. These various impacts are summarized in Table 16. Table 16: Impacts of Child Marriage/Early Childbearing on Women’s Agency Indicators Estimated Impacts (1) Women’s decision making ability Child marriage does not affect decision-making ability directly, but it matters indirectly through education. (2) Women’s ability to seek care Child marriage does not affect women’s ability to seek care directly, but it matters indirectly through education. (3) Women’s land ownership Child marriage is associated with a higher likelihood of land ownership (two percentage points) for women (4) Women’s knowledge of HIV/AIDS Child marriage is not associated with a reduction in adulthood in women’s knowledge of HIV/AIDS (5) Birth registration for children Child marriage is not associated with a reduction in birth registration rates for young children Sources: Wodon et al. (2017). same indicators of women’s agency are summarized in Table 17. Controlling for other variables, educational Educational attainment for women often has a larger impact on various measures attainment has a statistically significant effect of women’s agency than child marriage. on women’s decision-making ability within the household. Positive Impacts are also observed for the ability of women to seek care for themselves and their knowledge of HIV/AIDS. The impact for birth Educational attainment tends to have a larger registrations is not statistically significant. There impact on women’s agency than child marriage. is also a positive correlation between educational The impacts of girl’s education attainment on the attainment and land ownership. Table 17: Impacts of Girls’ Educational Attainment on Women’s Agency Indicators Estimated Impacts (1) Women’s decision making ability Universal secondary education could increase women’s decision-making ability within the household by 19 percent (2) Women’s ability to seek care Universal secondary education could increase women’s ability to seek care for themselves by close to one third (3) Women’s land ownership Educational attainment for women is associated with a higher likelihood of land ownership for women (4) Women’s knowledge of HIV/AIDS Universal secondary education could increase women’s knowledge of HIV/AIDS by four percent (5) Birth registration for children Educational attainment for women is not associated with an increase in birth registration rates for young children Sources: Wodon et al. (2017). 38 3.5.6. Synthesis on Estimated Impacts For all indicators except birth registrations, Overall, the negative impacts of child marriage, early either child marriage/early childbearing childbearing, and low educational attainment for girls are or secondary education completion have a large. Table 18 summarizes the main estimates of impacts statistically significant impact. This shows discussed in previous sections. First, the mutual relationships how pervasive and widespread the impacts of between child marriage, early childbearing, and low a lack of opportunities for girls are. educational attainment for girls are strong. Second, all three issues tend, in turn, to have negative impacts individually or collectively on a wide range of other outcomes. For all a statistically significant impact. Clearly, the impacts of a outcomes except birth registrations, either child marriage/ lack of opportunities for adolescent girls are pervasive and early childbearing or secondary education completion have widespread. Table 18: Summary of Statistically Significant Estimated Impacts by Domain  Domains and Indicators Child marriage Secondary education Either one of the two or early childbearing completion Mutual relationships Child marriage/ Early childbearing - Yes Yes Educational attainment Yes - Yes Fertility and population growth Fertility Yes Likely Yes Population growth Yes Likely Yes Modern contraceptive use No Likely Yes Health and nutrition Under-five mortality Yes No Yes Under-five stunting Yes No Yes Labor force participation No Yes Yes Demand for healthcare No Yes Yes Work and productivity Intimate partner violence Yes No Yes Women’s earnings Yes Yes Yes Household welfare Yes Yes Yes Women’s agency Decision-making ability No Yes Yes Land ownership No Yes Yes Knowledge of HIV/AIDS Yes Yes Yes Birth registration No No No Source: Wodon et al. (2017). Note: The term “likely” is used for some impacts of secondary completion because for those indicators, while the impact of secondary education completion is not statistically significant possibly due to small sample sizes, the impacts of some secondary as well as higher education are statistically significant. 3.6. What are the costs associated with those impacts? of child marriage, tentative estimates can be provided for The case of child marriage some of the largest impacts. The estimates in Table 19 should not be considered as precise given that they depend on (1) The impacts documented so far have large monetary econometric estimates of impacts that have themselves costs not only for girls and their children, but also for standard errors and (2) a range of assumptions for costing the country. This is illustrated in this section for child that could be debated. Still, the estimates provide an order of marriage. While it is not feasible to provide a monetary magnitude of the monetary costs of child marriage. Estimates valuation of all costs associated with the negative impacts are provided in terms of annual benefits from ending child 39 marriage in 2015. Estimates are provided for 2015 and nevertheless avoid stunting as well as survive past for 2030 as the reference year for the Sustainable age five. The benefits from such improvements Development Goals. This helps to illustrate how some are not primarily monetary. But with all necessary of the benefits of ending child marriage (and indirectly caveats, a tentative monetary value can be early childbearing) increase over time. associated with these improvements. The valuation is based on the discounted value of future wages • Welfare benefits from lower population for the children who avoid stunting, and future growth: If child marriage had been eliminated in GDP per capita for children surviving past age 2015, the welfare benefits from lower population five. The share of early childbearing likely due to growth would have been at US$95 million that year child marriage is factored in the estimations. For (in purchasing power parity or PPP terms). The stunting, the estimated benefits rise from US$43 benefits increase to US$2.4 billion by 2030. This million (PPP) in 2015 to US$81 million in 2030. For rapid increase comes first from the fact that annual under-five mortality, the benefits rise from US$104 impacts on population growth are cumulative. Each million (PPP) in 2015 to US$194 million in 2030. year, economic gains become larger because the cumulative reduction in population growth keeps The economic costs associated with growing. In addition, as standards of living (GDP the impacts of child marriage on per capita) improve, valuations also become larger. development outcomes are very large. Ending child marriage today would • Budget savings from lower population growth: generate billions of dollars of benefits Lower population growth implies that the size by 2030. of the population requiring publicly provided services is smaller, leading to budget savings. We consider the benefits of ending early childbearing for education budgets. Savings start six years • Education and earnings: The costs associated after early childbearing is ended since this is the with earnings losses for women married as time needed for fewer children to enter primary children are high. These costs are related for the school. Savings are estimated as the reduction most part to the fact that child marriage reduces in the anticipated cost of a trajectory that would girls’ educational attainment and thus earnings in achieve universal secondary education by 2030. adulthood. For adult women, the gains in earnings The benefits increase over time and could reach that could have resulted if past child marriages up to US$257 million in current US dollars by 2030 could have been avoided are estimated at US$514 if Uganda were to achieve universal secondary million in 2015. education by then. This is an upper bound estimate of potential savings as the country may not reach universal secondary education by 2030. The cost of keeping more adolescent girls in school is also not US$2.4B factored in. Still, the estimate provides an order of magnitude of potential benefits. When considering the elimination of only child marriage, benefits are a bit lower than when considering the prevention of ANNUAL welfare benefits from all early childbearing. lower population growth by • Benefits from reduced under-five mortality 2030 and stunting: While reducing early childbearing thanks to the elimination of child marriage may not reduce national rates of under-five mortality and stunting dramatically, many children would 40 Table 19: Order of Magnitude of Benefits from Ending Child Marriage Benefits Annual Benefit in 2015/2016 Annual Benefit in 2030 [Most estimates in PPP or Purchasing Power Parity] Fertility and population growth (1) Welfare cost US$95 million (PPP) US$2.4 billion (PPP) (2) Budget savings for education No benefit Up to US$257 million (current) Health, nutrition, and violence (3) Under-five mortality US$104 million (PPP) US$194 million (PPP) (4) Under-five stunting US$43 million (PPP) US$81 million (PPP) Education and earnings (5) Earnings loss for women US$514 million (PPP) Not estimated Sources: Wodon et al. (2017). Estimates for education budget savings are an upper bound and actual savings are likely to be lower. To illustrate the magnitude of the benefits from 3.7. How can we end child marriage, prevent early ending child marriage, comparisons with net Official childbearing, and improve girls’ educational attainment? Development Assistance may be useful. Net Official Lessons from international experiences Development Assistance (ODA) consists of disbursements of loans made on concessional terms (net of repayments Ending child marriage, preventing early childbearing, and of principal) and grants by official agencies. The agencies educating girls requires interventions. Beyond appropriate included are members of the Development Assistance laws, life skills and sexual reproductive health knowledge, Committee (DAC), multilateral institutions, and non-DAC economic opportunities, and incentives for schooling are countries. Net ODA includes loans with at least a fourth of needed. At the minimum, countries need to promulgate grant element. Net ODA in Uganda has amounted to six to appropriate laws, especially in the case of child marriage eight percent of Gross National Income in recent years. This as discussed in Box 7. But laws and more generally suggests that if child marriage could have been eliminated in strategies to empower girls as well as broad information 2015, this could have generated by 2030 benefits equivalent and mobilization campaigns to tackle gender inequities to one fourth of the net ODA likely to be received by Uganda, are not likely to be enough by themselves. Specific targeted simply through the welfare benefits obtained from lower interventions to empower girls are needed to ensure that population growth. When adding all the other benefits from they have appropriate life skills and reproductive health ending child marriage, gains would be larger. knowledge. Economic incentives are also needed for girls to remain in school, go back to school if they dropped out, While economic costs and benefits should not be the sole or expand their livelihood opportunities if they cannot go rationale for investments to end child marriage, prevent back to school in order to delay marriage and childbearing. early childbearing, and improve girls’ education, they While the literature on these interventions is too large to be are an important consideration. Ending child marriage, reviewed comprehensively here, subsets of this literature preventing early childbearing and improving education can be synthetized. Building on a recent review by Botea et opportunities for girls are the right thing to do from a moral al. (2017), the section focuses on three types of interventions and ethical standpoint. But in addition, these issues have for adolescent girls: (1) programs providing life skills and large economic costs. The hope is that the demonstration of reproductive health knowledge; (2) programs expanding some of these costs in this study will help generate consensus economic opportunities; and (3) programs keeping girls in for higher investments to end child marriage, prevent early school or enabling them to return to school. childbearing, and promote girls’ education in Uganda. 41 Box 7: Child Marriage Laws and their Limitations The Convention on the Rights of the Child emphasizes the need for full and informed consent for marriage. It notes that children typically do not have the ability to provide full and informed consent. This is one of the reasons why 18 is recommended as the minimum age for marriage. In Uganda, the 1995 Constitution sets the minimum age of marriage at 18 years (Article 31), but national laws have other provisions. The Marriage Act of 1904 sets the minimum age for consent at 21 years but allows written consent of the father, mother, guardian, or registrar for the marriage of minors. The Marriage and Divorce of Mohammedans Act of 1906 is silent on the minimum age for consent. Both the Hindu Marriage and Divorce Act of 1961 and the Customary Marriages (Registration) Act 1973 set the minimum age for consent at 16 years for girls and 18 years for boys. They allow marriage of minors upon consent of parents or a guardian. As discussed by Wodon, Tavares et al. (2017), such exceptions allowing girls to be married early with parental or judicial consent should be avoided. The focus on these three types of interventions stems as part of a broader target group; (2) Provide life skills from a body of evidence showing that they can have and sexual and reproductive health (SRH) knowledge, positive impacts. Each of these three types of programs economic opportunities, or education opportunities; (3) is hypothesized to potentially delay marriage/ Demonstrate results in terms of improving the health of childbearing and increase educational attainment in young women, especially for SRH, or delay marriage or different ways. They have different “theories of change” childbearing; and (4) Have been tested in a developing (see Box 8). Close to 40 interventions are reviewed country, usually in sub-Saharan Africa, but also in other by Botea et al (2017). To be included in the review, low income settings. Several of the interventions were interventions had to fulfill the following selection implemented in Uganda. criteria: (1) Target girls aged 10-19, either exclusively or A water point at a village in Mukono. Central Uganda (Morgan Mbabazi, 2017) 42 Box 8: Theories of Change for Interventions Targeting Adolescent Girls Life skills and SRH knowledge: By increasing knowledge and awareness, life skills can increase young women’s perceived risk of becoming pregnant at an early age and the desire to avoid early pregnancies (through family planning). Through these channels, life skills may lead to better health outcomes for the girls and their children. By increasing girls’ confidence and self-esteem, life skills may also increase girls’ aspirations. With increased aspirations, girls may have a greater desire to delay marriage and childbearing. Finally, life skills can increase young women’s communication and decision-making skills, leading to increased abilities to negotiate their preferences for delayed marriage and childbearing. At the same time, while life skills and SRH knowledge may empower girls, they may not be sufficient to delay marriage and childbearing if social norms curtailing agency for girls are not also addressed at the same time. Life skills together with economic opportunities: Programs increasing earnings potential for young women may increase their ability to plan marriage and childbearing decisions in three ways. First, the ability to make an economic contribution expands the role of women beyond that of sex and reproduction. This can increase their desire to limit or space childbearing. The transformation of girls from economic liabilities into assets in the eyes of their societies and families can also alleviate external pressures on girls to marry or have children early. Second, the loss in earnings associated with childrearing is an opportunity cost which may increase women’s desire to limit or space births and exercise reproductive control. Third, a young women’s increased earnings may improve her bargaining power within the household and allow her to effectively exercise reproductive control by negotiating delays in sexual debut or marriage, and negotiating the terms of sex including the use of contraceptives. Creating income-generating opportunities for women can therefore contribute to female empowerment beyond the economic realm by widening personal choice and control over SRH outcomes. Incentives for schooling or delayed marriage: In many communities, the economic, cultural, and social environment does not provide viable alternatives to marriage for adolescent girls. Once girls drop out of school, possibly because of poor quality or high cost, it may be difficult for parents not to get their daughter married. In those communities, improving the provision of quality and affordable primary and secondary education may be one of the best way to delay marriage and childbearing as parents often see schooling as a viable alternative to marriage for their daughters. Incentives and programs to keep girls in school may also lead to “tipping points” in communities whereby more and more girls remain in school and are able to delay marriage. A few interventions have also aimed to delay marriage through financial incentives conditional on not marrying early, with additional schooling often as an additional benefit. Source: Botea et al. (2017). The first category of programs emphasizes the in Africa and Latin America. The clubs have proven effective empowerment of girls by providing life skills and when implemented well. By combining socializing, fun, and reproductive health knowledge. The typical intervention access to mentors, the clubs are attractive for girls to attend. is that of a “safe space club” for adolescent girls. These clubs From there, other services are delivered. Clubs can be held in are delivery platforms for convening girls with a trusted a variety of settings, including schools or community centers. adult mentor at a specific time and place. The approach was Girls meet regularly and are able with the help of the mentors pioneered by BRAC in South Asia and the Population Council to discuss a range of issues, including those related to SRH. 43 They learn “life skills” in those meetings, including Without other incentives, safe space “soft” or socio-emotional skills such as critical thinking programs may not be sufficient to delay and problem solving, communication and negotiation marriage and childbearing or improve (for example within one’s household). One of the schooling. Still, they achieve important objectives is often to boost the girls’ self-awareness and intermediary outcomes related among self-esteem, so that they can explore and fulfill their others to aspirations and self-esteem, own aspirations. In many cases, safe space clubs are confidence, and SRH knowledge. also used to impart “hard” skills, such as basic literacy and numeracy, or basic business skills. These programs have helped improve SRH services. Impacts on the age at marriage and early knowledge and behaviours. This includes an increase childbearing tend to be larger than with life skills/SRH in girls undergoing HIV testing or counseling; an knowledge alone, but not in all cases. Given their focus increase in the use of modern contraception or other on economic opportunities, the programs often have methods of family planning; a reduction in the desire some success in increasing earnings, employment, for practicing female genital mutilation for daughters in and/or savings. Several of the programs also succeed countries where the practice is prevalent; a reduction in in increasing the use of modern contraceptives and the risk of intimate partner violence when the program SRH knowledge, which may help delay childbearing. also reaches out to men; an increase in self-esteem; In some cases, the programs also succeed in delaying and gains in specific skills taught during safe space the age at marriage and reducing teen pregnancies, as sessions, for example in the areas of financial literacy observed in Uganda (see Box 9). The message from the or basic literacy and numeracy. At the same time, review is that adding a livelihood dimension to life skills without additional interventions related to schooling and SRH knowledge programs may help delay marriage or employment and livelihoods, it is not clear that safe and childbearing, but not in all cases. The focus on spaces are sufficient to delay marriage and childbearing economic opportunities may also help in ensuring (though that may not have been a primary goal of regular participation by girls in the programs. these projects). Therefore, it is important to consider programs whereby safe spaces have been combined Interventions combining an emphasis with livelihood opportunities and incentives to remain on empowering girls, often through in school, usually with larger impacts on the age at safe space clubs, with livelihood marriage and childbearing. opportunities may improve reproductive health outcomes and delay marriage The second category of programs combine an or childbearing. This has been the case emphasis on empowering girls, often through in Uganda, but not systematically so in safe spaces, with in addition a focus on providing other countries. Since these are often the livelihood opportunities. These programs are only option available for out-of-school appropriate for girls who are not in school. For these girls, more research is needed to figure girls, building skills for income-generation may provide out what works and what doesn’t. an alternative to early marriage and childbearing. Two groups of interventions are distinguished: livelihood interventions and financial literacy/access to financial 44 Box 9: BRAC Uganda Empowerment and Livelihoods for Adolescent Girls (ELA) The ELA project in Uganda aimed to increase economic empowerment for adolescent girls in rural areas by providing life skills training, skills related to income-generation, and access to microfinance. The program has demonstrated strong positive impacts on economic, health, and agency outcomes for girls. Among other outcomes, the program (1) increased the likelihood of engaging in income-generating activities by 32 percent; (2) increased self-reported routine condom use by those sexually active by 50 percent; (3) reduced fertility rates by 26 percent; and (4) reduced reporting of unwanted sex by 76 percent. There were also reductions in teenage pregnancies and child marriage. To gather further evidence on the effectiveness of the intervention in promoting entrepreneurship, the evaluation looked at the impact of the program on the willingness to compete in an experimental setting, including for the girls’ brothers. The results suggest that programs that target adolescent girls’ empowerment such as ELA may also have spillover effects on their brothers and shift gender dynamics in the community. Sources: Bandiera et al. (2014) and Buehren et al. (2016). The third set of programs focuses on keeping girls in intervention types hold promise and multiple interventions school or enabling them to return if they dropped out. are needed to reach different profiles of girls. A few programs directly aim at delaying marriage. The literature, including a recent review by Kalamar et al. (2016), The interventions mentioned above are not meant to suggests that there are multiple intervention options available be exhaustive. For example, to improve educational to keep girls in school and delay marriage. In a few cases, attainment for girls, additional interventions are evaluations are also available for programs focusing directly needed. The three types of interventions listed above were on delaying marriage through financial incentives, often with selected because their evaluations looked at changes in SRH the additional benefit of enabling girls to remain in school. knowledge, child marriage, and/or early childbearing. In The programs providing incentives for schooling succeed the case of educational attainment, there is a much broader quite often in keeping girls in school and sometimes delay literature on what is needed to achieve gains (see Box 10). marriage and childbearing. Some of these programs enable Basic conditions need to be in place, and they matter quite a girls who dropped out of school to go back. Not all programs bit especially in low income countries such as Uganda. First, there is a need to build new schools closer to where children (boys and girls) live. In Uganda, access to lower secondary Of the three types of interventions education remains extremely low in part because there just reviewed in this study, interventions to are not enough secondary schools. Building schools closer promote education, including by reducing to populated areas means that girls do not have to walk too out-of-pocket and opportunity costs for far to go to school and parents can be more comfortable with schooling, are the most likely to help their daughters’ safety on their way to and from school. In delay marriage and childbearing. cases where schools cannot be built nearby, providing modes of transportation for girls to go to school is an option. Second, access to water, latrines and hygienic facilities are important for adolescent girls. Building and upgrading schools with succeed in all areas, but the evidence is broadly convincing separate water, sanitation and hygiene (WASH) facilities is that in comparison to the other two types of programs also an important intervention that should be pursued in reviewed above, those focusing on schooling for girls, or in Uganda. Third, an unacceptably high percentage of girls some cases on delaying marriage with financial incentives, in Uganda are at risk of violence and sexual harassment in may be more successful in indeed delaying marriage school. There is a need for specific interventions to deal with and childbearing. At the same time, however, all three these risks too. 45 Box 10: Improving Educational Attainment and Learning for Girls Because multiple reasons may contribute to gender gaps in educational attainment and learning, the types of interventions that could be implemented to reduce these gaps are multiple. Should the distance to schools be reduced, whether this is done by building new schools in remote areas or reducing travel time through modes of transportation? Should scholarships be provided to girls? Should more female teachers be hired? Should the priority be to make separate toilet blocks available for boys and girls? Should more focus be placed on understanding and changing cultural practices? Should pedagogical interventions targeting girls be implemented? The right choice between potential interventions depends on a country’s or a community’s context. But reviews of the evidence can help, and such reviews are becoming available thanks to a substantial increase in rigorous impact evaluations in recent years. One such review was published in June 2014 (Unterhalter et al., 2014). The review assessed the evidence on the impact of interventions for girls’ education focusing on (i) providing resources (including transfers) and infrastructure, (ii) changing institutions, and (iii) changing norms and including the most marginalized in education decision making. The review summarized the impact of different types of interventions on three outcomes: participation, learning, and empowerment. For each type of intervention and category of outcome, the evidence on the likelihood of impact was classified as strong, promising, limited, or needed (i.e., weak). For participation, the evidence on the impact of conditional cash transfers, information about the potential employment returns to education, and the provision of additional schools in underserved and unsafe areas was found to be strong. This was also the case for the evidence on some interventions related to teacher training, group-learning, and measures to promote girl-friendly schools, as well as learning outside the classroom, for example through tutoring. Several of these interventions (group-learning, programs for learning outside the classroom, and scholarships linked to student performance) were also found to have impacts on learning. The evidence on the impact of interventions on empowerment was generally weaker. Source: Unterhalter et al. (2014). For specific challenges, such as gender-based community-based and have multiple components to violence, additional specific interventions may work with men, women, leaders and service providers. also be needed. Uganda has high levels of intimate There is also promising evidence to recommend partner violence (IPV). International evidence suggests economic empowerment interventions for women that that prevention programs can help in reducing the are combined with gender transformative training and prevalence of IPV, especially when they address the engagement of male partners and family members. harmful social norms that are leading to gender-based Lessons can be learned from programs implemented in violence. The most successful interventions tend to be Uganda, such as SASA! (see Box 11). Box 11: Interventions to Reduce Intimate Partner Violence SASA! means “Now!” in Kiswahili. The program was developed by Raising Voices and it has been implemented in Uganda by the Center for Domestic Violence Prevention. It appears to be the first community-based violence prevention program in sub-Saharan Africa to be rigorously evaluated. The program employs multiple strategies to build a critical mass of engaged community members, leaders, and institutions, including local activism, media and advocacy, communication materials, and training. The Activist Kit that is central to SASA! community engagement and mobilization involves four phases: Start, Awareness, Support, and Action. The content evolves with each phase, with power as a central theme. Results from a randomized controlled trial suggest positive effects after three years of programming. In comparison to control communities, SASA! communities reported (i) a reduction in levels of violence against women of 52 percent; (ii) an increase in the share of women and men who believe it is acceptable for women to refuse sex of 28 percent; and (iii) an increase of 50 percent in the share of men and women who believe that physical violence against a partner is unacceptable. Source: Abramsky et al. (2014). 46 3.8. Conclusion: What Have We Learned? norms. This may be why in those areas, while ending child marriage may help, impacts may be smaller or not statistically This study has provided strong evidence of the negative significant. Still, when considering together child marriage/ impacts child marriage, early childbearing, and low early childbearing and low educational attainment for girls, educational attainment for girls. After a brief review of negative impacts are observed for almost all indicators. trends over time, and an assessment of the relationships between child marriage, early childbearing, and low The economic costs associated with these impacts educational attainment for girls, four domains of impacts are large. This has been illustrated in the case of child were considered: (i) fertility and population growth; (ii) marriage. Tentative estimates of the costs associated with the health, nutrition, and intimae partner violence; (iii) work, impacts of child marriage were provided. These are annual earnings, and productivity; and (iv) women’s agency. The estimates of the benefits from ending child marriage as of analysis indicates that child marriage, early childbearing, 2015. The estimates only provide rough orders of magnitude of and low educational attainment for girls have a wide potential costs. The largest cost is the welfare loss associated range of negative impacts on girls, their children, families, with population growth. By reducing population growth, communities, and the country at large. ending child marriage could lead to welfare benefits for Uganda of US$2.4 billion (in purchasing power parity terms) The largest monetary impacts are related to fertility and by 2030. As another benefit from lower population growth, population growth, earnings, and the health of children ending child marriage and early childbearing could result born of young mothers. These impacts are all closely in education budget savings for the government of up to related in terms of their timing in the life of adolescent girls. US$257 million by 2030 under a trajectory to achieve universal When the use of modern contraception is low, child marriage secondary education (this is an upper bound for savings given leads to early childbearing, which increases health risks for that achieving universal secondary education is a tall order). mothers and children alike. The timing of child marriages In addition, today, if women who were married early had been and early childbearing conflicts with the ability of girls to able to avoid child marriage, they could have earned US$514 remain in school, which depresses earnings in adulthood. million more. Finally, substantial economic benefits would All those effects are at work when girls are vulnerable. By result from reductions in under-five mortality and stunting contrast, impacts in other domains – from labour force rates, estimated together at US$275 million by 2030. participation to decision-making, are observed throughout a woman’s life and may depend on many other factors than In terms of policy options to end child marriage, prevent whether girls married early. For example, a lack of decision- early childbearing, and improve education opportunities making ability is the result of widespread gender inequality. for girls, a review of interventions provides insights. Child marriage, early childbearing and low educational Several lessons emerge from this review. First, interventions attainment for girls contribute to perpetuating gender that alleviate economic constraints to schooling are the inequality, but delaying marriage by a few years may not be most proven to delay marriage and childbearing. Second, sufficient to fundamentally change gender roles and social interventions providing economic opportunities have mixed There are examples of successful Ending child marriage, preventing interventions to provide better early childbearing, and improving opportunities for adolescent educational attainment for girls is not girls, including in Uganda. These only the right thing to do. It is also a interventions could be scaled up for smart economic investment. larger impacts nationally. 47 results in terms of delaying marriage and childbearing. adequate modes of transportation to schools must In some contexts, they have large impacts, while in be provided. Providing separate water, sanitation others, this is less the case. Third, life skills and SRH and hygiene facilities for girls is also important, as knowledge interventions, while valuable, may not is the need to reduce the risk of violence and sexual be sufficient to end child marriage, prevent early harassment in school. Finally, for the broader challenge childbearing, and improve education opportunities for of gender-based violence and inequality, community- girls. At the same time, all three intervention types have based interventions to work with men, women, leaders value and combining them is often needed to reach and service providers can also be beneficial. different groups of girls such as those in school or not, and those married or not. More generally, to improve educational attainment for girls, basic conditions also must be met. At the secondary level, there is a need to build schools closer to where children (boys or girls) live. As an alternative, 48 References Abramsky, T., K. Devries, L. Kiss, J. Nakuti, N. Kyegombe, E. Starmann, B. Cundill, L. Francisco, D. Kaye, T. Musuya, L. Michau, and C. Watts. 2014. Findings from the SASA! Study: A Cluster Randomized Controlled Trial to Assess the Impact of a Community Mobilization Intervention to Prevent Violence against Women and Reduce HIV Risk in Kampala, Uganda. BMC Medicine 12:122. Bandiera, O.; N. Buehren, R. Burgess, M. Goldstein, S. Gulesci, I. Rasul, and M. Sulaiman. 2014. Empowering Adolescent Girls: Evidence from a Randomized Control Trial in Uganda. Washington, D.C.: The World Bank. Botea, I., S. Chakravarty, S. Haddock, and Q. Wodon. 2017. Interventions Improving Sexual and Reproductive Health Outcomes and Delaying Child Marriage and Childbearing for Adolescent Girls. Ending Child Marriage Notes Series. Washington, D.C.: The World Bank. Buehren, N. 2015. The Cost of the Gender Gap in Agricultural Productivity in Malawi, Tanzania, and Uganda. Washington, D.C.: The World Bank Buehren, N.; M. Goldstein, K. Leonard, J. Montalvao, and K, Vasilaky. 2016. Women’s Empowerment, Sibling Rivalry, and Competitiveness: Evidence from a Lab Experiment and a Randomized Control Trial in Uganda. Policy Research Working Paper No. 7699. Washington, D.C.: The World Bank. Kalamar, A. M., S. Lee-Rife, and M. J. Hindin. 2016. Interventions to Prevent Child Marriage among Young People in Low- and Middle-Income Countries: A Systematic Review of the Published and Gray Literature, Journal of Adolescent Health 59: S16-S21. Klugman, J., L. Hanmer, S. Twigg, T. Hasan, and J. McCleary-Sills. 2014. Voice and Agency: Empowering Women and Girls for Shared Prosperity. Washington, DC: The World Bank. Nguyen, M. C., and Q. Wodon. 2015. Global and Regional Trends in Child Marriage. Review of Faith and International Affairs, 13(3): 6-11. O’Sullivan, M., A. Rao, R. Banerjee, K. Gulati, and M. Vinez. 2014. Levelling the Field: Improving Opportunities for Women Farmers in Africa. Washington, D.C.: The World Bank. Parsons, J., J. Edmeades, A. Kes, S. Petroni, M. Sexton, and Q. Wodon. 2015. Economic Impacts of Child Marriage: A Review of the Literature, Review of Faith and International Affairs, 13(3): 12-22. UNFPA. 2012. Marrying Too Young: End Child Marriage. New York: UNFPA. UNICEF. 2005. Early Marriage: A Harmful Traditional Practice, A Statistical Exploration. New York: UNICEF. UNICEF. 2014. Ending Child Marriage: Progress and Prospects. New York: UNICEF. Unterhalter, E., A. North, M. Arnot, C. Lloyd, L. Moletsane, E. Murphy-Graham, J. Parkes, and M. Saito. 2014. Girls’ Education and Gender Equality. London: Department for International Development. Vogelstein R. 2013. Ending Child Marriage: How Elevating the Status of Girls Advances U.S. Foreign Policy Objectives. New York: Council on Foreign Relations. Wodon, Q., C. Male, A. Onagoruwa, A. Savadogo, and A. Yedan. 2017. Economic Impacts of Child Marriage, Early Childbearing, and Low Educational Attainment for Girls in Uganda, Washington, D.C.: The World Bank. Wodon, Q. 2017. Early Childhood Development in the Context of the Family: The Case of Child Marriage, Journal of Human Development and Capabilities, 2016, 17(4): 590-98. Wodon, Q., M. C. Nguyen, and C. Tsimpo. 2016. Child Marriage, Education, and Agency in Uganda, Feminist Economist, 22(1): 54-79. Wodon, Q., P. Tavares, O. Fiala, A. Le Nestour, and L. Wise. 2017. Child Marriage Laws and their Limitations. Ending Child Marriage Notes Series. London and Washington, D.C.: Save the Children and The World Bank. 49 Annexes Accessing practical skills at a technicla school in Jinja, one of the secodnary cities of Uganda (Morgan Mbabazi, 2017). Table A1: Key Macroeconomic Indicators Indicator Unit measure 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 2016/17 Population Millions 31.0 31.9 35.1 37.6 38.7 39.9 41.1 42.3 GDP USD millions 20181.6 20262.3 23218.9 24663.1 27757.4 27573.2 24661.1 26282.8 Per capita GDP USD 651 635 662 656 717 691 600 621 GDP growth % 5.6 9.4 3.8 3.6 5.1 5.2 4.7 4.0 Gross Domestic Savings as % of GDP 19.7 19.5 17..7 21.7 18.8 17.1 18.2 20.1 Gross Investments as % of GDP 12.5 12.3 28.2 29.5 26.7 24.6 24.9 24.5 Inflation (period average) % 4.0 18.7 14.0 4.9 5.3 2.9 6.6 5.7 Exchange Rate (end-year) UGX/USD 2177.6 2522.7 2504.6 2591.1 2538.0 2827.7 3443.0 3528.3 External Sector                   Exports, f.o.b. Million USD 2317.3 2297.8 2667.4 2912.1 2706.3 2738.4 2687.8 3163.7 Imports - f.o.b. Million USD -4,116.8 -4,671.1 -5,241.5 -5,035.1 -5,073.5 -4,988.0 -4574.5 -4,718 Current Account Balance Million USD -1497.2 -1732.6 -2030.9 -1517.3 -2051.7 -1876.5 -1208.8 -1039.8 Balance of Payments (overall balance) Million USD -229.6 606.0 -746.6 -338.0 -378.5 352.8 -101.5 -437.6 Gross Foreign Reserves Million USD 2384.7 2044.0 2643.8 2912.3 3394.0 2895.0 2962.0 3429.0 External Debt Million USD 2343.4 2904.9 3254.1 3825.2 4300.7 4380.1 5309.2 6014.1 Foreign Direct Investment Million USD 659.7 706.4 1243.9 939.9 1087.4 884.6 629.4 493.8 Monetary Sector                   Average Deposit Rate % 2.0 2.1 3.2 3.0 3.1 3.3 3.2 3.3 Average Lending Rate % 20.7 19.8 24.6 24.8 22.1 25.2 23.7 22.6 Growth in Money Supply (M3) % 23.6 25.7 26.1 6.6 17.4 15.9 7.1 13.6 Government Finance                   Total Domestic Revenue as % of GDP 10.5 13.6 11.2 11.3 11.6 13.2 13.8 14.3 Tax Revenue as % of GDP 10.3 10.9 10.3 11.0 11.1 12.4 13.0 13.5 Non Tax Revenue as % of GDP 0.3 0.2 0.2 0.3 0.5 0.6 0.7 0.7 Grants as % of GDP 2.1 1.9 1.9 1.4 1.0 1.2 1.4 1.0 Total Expenditure and net lending as % of GDP 16.7 19.1 15.6 16.2 16.6 18.7 20.1 19.3 Recurrent Expenditure as % of GDP 10.5 12.7 9.1 9.0 9.5 10.0 11.0 11.0 Development Expenditure as % of GDP 6.1 6.1 6.1 6.5 7.0 6.8 7.1 7.4 Fiscal Balance (overall) as % of GDP -4.0 -3.6 -2.5 -3.5 -3.5 -4.4 -5.3 -3.5 Source: GOU Authorties and World Bank Staff Estimates Table A2: Growth and Structure of the Economy Economic Activity 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 2016/2017 2017/18 Proj Real GDP Growth Rates (%) 5.6 9.4 3.8 3.6 5.1 5.2 4.7 4.0 5.1 Agriculture 2.9 3.1 0.6 1.9 2.7 2.3 2.8 1.6 2.6 Industry 7.9 11.3 3.1 2.1 6.3 7.8 4.6 3.3 5.7 o/w manufacturing 4.5 7.8 2.7 -2.5 2.2 11.6 0.6 2.1 3.0 o/w construction 12.6 14.8 3.9 4.2 12.5 1.9 7.3 6.1 6.5 Services 7.0 11.7 4.0 5.4 5.4 4.8 5.9 5.7 5.9 GDP Shares (% of constant GDP) Agriculture 26.2 24.7 24.0 23.6 23.0 22.4 22.0 21.5 21.0 Industry 18.1 18.4 18.3 18.0 18.3 18.7 18.7 18.6 18.7 o/w manufacturing 8.5 8.4 8.3 7.8 7.6 8.0 7.7 7.6 7.6 o/w construction 5.8 6.1 6.1 6.1 6.5 6.3 6.5 6.6 6.6 Services 48.5 49.5 49.6 50.5 50.6 50.5 51.1 51.9 52.3 FISM and net taxes 7.2 7.3 8.1 7.9 8.1 8.4 8.2 8.0 8.0 GDP Shares by expenditure type (% of nominal GDP)  Final Consumption Expenditure 85.9 87.4 85.6 81.9 82.7 86.2 84.5 83.5 83.7 Households 76.3 74.6 77.4 73.9 74.2 76.9 77.0 75.4 75.5 Government 9.6 12.8 8.2 8.0 8.5 9.3 7.5 8.1 8.2 Gross Capital Formation 25.6 27.5 27.3 28.4 27.3 24.6 25.5 23.9 23.8 Gross fixed capital formation 25.2 27.1 26.9 27.9 26.8 24.2 25.0 23.4 23.3 Charges in inventories 0.3 0.3 0.4 0.4 0.5 0.4 0.5 0.5 0.5 Net exports -11.5 -14.9 -12.9 -10.3 -10.0 -10.8 -10.1 -7.4 -7.8 Gross domestic saving (% of GDP) 12.5 12.3 17.7 21.7 18.8 17.1 18.2 20.1 21.0 Public 2.9 3.3 2.4 3.1 2.3 2.9 2.8 3.3 4.6 Private 9.6 9.0 15.3 18.6 16.5 14.2 15.4 16.8 16.4 Source: Uganda Bureau of Statistics and World Bank Staff Estimates 54 Table A3: Central Government Fiscal Framework (% of GDP) 2017/18 App. Item 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 2016/17 Budget 2017/18 Proj                         REVENUES AND GRANTS 13.5 12.7 15.5 13.1 12.8 12.6 14.4 15.2 15.4 16.2 15.5 Revenues 11.0 10.5 13.6 11.2 11.3 11.6 13.2 13.8 14.3 14.4 14.0 URA 10.6 10.3 10.9 10.3 11.0 11.4 12.4 13.0 13.5 13.6 13.2 Non-URA 0.4 0.3 0.2 0.2 0.3 0.5 0.6 0.7 0.7 0.8 0.7 Oil Revenues 0.0 0.0 2.5 0.7 0.0 0.0 0.2 0.1 0.1 0.0 0.1 Grants 2.6 2.1 1.9 1.9 1.4 1.0 1.2 1.4 1.0 1.8 1.5 Budget Support 1.5 1.1 1.1 1.0 0.3 0.3 0.3 0.4 0.3 0.3 0.3 Project Support 1.0 1.0 0.8 0.9 1.1 0.7 0.9 1.0 0.8 1.5 1.2 EXPENDITURE AND LENDING 15.0 16.7 19.1 15.6 16.2 16.6 18.7 20.1 19.3 22.5 20.2 Current Expenditures 9.5 10.5 12.7 9.1 9.0 9.5 10.0 11.0 11.0 10.4 10.5 Wages and Salaries 3.4 3.2 3.5 3.1 3.3 3.4 3.6 3.6 3.7 3.7 3.5 Interest Payments 1.0 0.9 0.9 1.0 1.4 1.4 1.6 2.0 2.6 2.7 2.7 Other Recurr. Expenditures 5.1 6.4 8.2 5.0 4.3 4.8 4.8 5.4 4.7 4.0 4.3 Development Expenditures 4.8 6.1 6.1 6.1 6.5 7.0 6.8 7.1 7.4 9.8 8.6 Net Lending/Repayments -0.2 -0.1 -0.1 -0.1 0.6 0.0 1.6 1.8 0.6 1.9 0.7 Domestic Arrears Repaym. 0.8 0.2 0.4 0.5 0.1 0.0 0.3 0.1 0.2 0.4 0.4 OVERALL DEFICIT                       Overall Fiscal Bal. (excl. Grants) -4.0 -6.1 -5.5 -4.4 -4.9 -5.0 -5.6 -6.7 -4.5 -8.0 -5.2 Overall Fiscal Bal. (incl. Grants) -1.5 -4.0 -3.6 -2.5 -3.5 -4.0 -4.4 -5.3 -3.5 -6.2 -3.7 Financing: 1.5 4.0 3.6 2.5 3.5 4.0 4.4 5.3 3.5 6.2 3.7 External Financing (Net) 1.6 1.9 1.5 1.9 2.2 1.3 1.2 3.0 2.9 5.3 2.7 Domestic Financing (Net) 0.0 1.7 2.3 0.0 1.1 2.3 3.2 2.3 0.6 0.9 1.0 MEMORANDA ITEMS                       Nominal GDP (Shs billions) 34504 40946 47078 59420 64758 70458 76517 82903 91351 100552 101267 Source: Uganda Authorities and World Bank Staff Estimates Table A4. Monetary indicators   2008/9 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 2016/17                     Monetary Aggregates                   M3 as % of GDP 18.3 20.5 22.4 19.0 18.6 20.1 21.3 21.1 22.0 M2 as % of GDP 14.3 15.9 17.1 13.0 14.0 14.5 14.4 14.5 15.5 M3 growth rate (%) 25.0 33.2 25.7 7.2 6.6 17.4 15.9 7.1 13.6 M2 growth rate (%) 26.3 32.1 23.9 -4.2 15.7 14.1 8.8 8.9 16.0 Domestic Credit                   Total domestic credit (% of GDP) 9.2 11.9 16.0 11.8 12.5 13.8 16.7 17.2 16.2 Private sector credit (% of GDP) 10.4 11.4 14.3 12.7 12.4 12.9 14.3 13.7 13.3 Total domestic credit growth (%) 64.1 54.7 54.1 -6.5 13.4 21.9 32.3 10.8 2.6 Private sector credit growth (%) 31.3 29.8 44.1 11.6 6.0 14.1 20.3 4.3 5.7 Interest Rates Structure                   Average TB rate (period average, %) 8.4 5.3 7.6 17.2 10.3 9.9 12.0 17.8 13.2 Average lending rate (%) 20.9 20.7 19.8 24.6 24.8 22.2 21.6 24.0 22.6 Average deposit rate (%) 2.1 2.0 2.1 3.2 3.0 3.1 2.8 3.0 3.5 Source: Bank of Uganda 55 56 Table A5.1: Balance of Payments (percent of GDP unless otherwise stated)   2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 2016/17 2017/18 Proj. Current Account (incl transfers) -8.1 -9.8 -9.5 -6.3 -7.6 -6.8 -4.9 -4.0 -5.1 Exports of goods 11.5 11.3 11.4 11.7 9.7 10.1 11.1 12.4 12.3 o/w coffee 1.3 1.8 1.9 1.7 1.5 1.5 1.4 1.9 2.0 Imports of goods -20.4 -23.1 -22.6 -20.1 -18.3 -18.3 -18.9 -17.8 -18.8 o/w oil imports -2.5 -3.4 -4.1 -4.1 -3.9 -3.4 -2.6 -2.6 -2.6 Services (net) -2.1 -3.4 -1.7 -1.6 -1.2 -2.5 -2.7 -1.9 -1.7 Trade balance -8.9 -11.8 -11.1 -8.5 -8.5 -8.3 -7.8 -5.4 -6.5 Income (net) -1.7 -1.7 -2.0 -2.1 -2.2 -1.6 -1.7 -2.2 -2.5 Current transfers (net) 4.6 7.1 5.3 5.9 4.3 5.1 6.3 5.3 5.5 Capital and Financial Account 8.8 5.3 10.1 6.1 6.5 4.2 4.2 3.9 5.4 Capital account 1.0 0.8 0.8 0.1 0.3 0.4 0.5 127.0 137.0 Financial account 7.9 4.5 9.3 5.9 6.1 3.9 3.7 910.0 1385.0 o/w direct investment 3.4 3.5 5.4 3.8 3.9 3.2 2.1 667.0 897.0 o/w portfolio investment 0.2 0.0 -1.1 -0.2 0.0 -0.6 -0.7 164.0 48.0 Overall Balance 1.2 -2.9 3.3 2.1 1.1 1.5 2.9 1.6 0.4 Gross International Reserves (million USD) 2384.7 2044.0 2643.8 2912.3 3394.0 2895.0 2962.0 3380.0 3475.0 Gross international reserves in months of imports 4.4 3.2 4.3 4.5 5.2 5.0 5.4 5.5 4.9 Source: Bank of Uganda Table A5.2: Balance of Payments (USUS$ millions) Variable 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 2016/17 2017/18 Proj. Current Account (incl transfers) -1,631.0 -1,984.0 -2,219.0 -1,582.0 -2,103.0 -1,958.0 -1531.0 -1,058 -1,418.0 Exports of goods 2,317.0 2,298.0 2,660.0 2,912.0 2,706.0 2,738.0 2688.0 3169 3,387.0 o/w coffee 262.0 371.0 444.0 423.0 404.0 400.0 352.0 490 548.0 Imports of goods -4,117.0 -4,680.0 -5,241.0 -5,035.0 -5,074.0 -4,988.0 -4574.0 -4,561 -5,177.0 o/w oil imports -501.0 -679.0 -947.0 -1,028.0 -1,090.0 -933.0 -646.0 -694 -739.0 Services (net) -416.0 -691.0 -404.0 -405.0 -331.0 -561.0 -563.0 -487 -481.0 Trade balance -1,800.0 -2,382.0 -2,581.0 -2,123.0 -2,367.0 -2,250.0 -1870.0 -1,392 -1,790.0 Income (net) -335.0 -341.0 -471.0 -528.0 -610.0 -492.0 -492.0 -582 -687.0 Current transfers (net) 920.0 1,430.0 1,238.0 1,473.0 1,204.0 1,345.0 1411.0 1403 1,540.0   Capital and Financial Account 1,786.0 1,081.0 2,357.0 1,519.0 1,813.0 952.0 1060.0 1,037 1,522.0 Capital account 197.0 160.0 194.0 33.0 91.0 99.0 120.0 127 137.0 Financial account 1,589.0 921.0 2,163.0 1,486.0 1,722.0 853.0 940.0 910 1,385.0 o/w direct investment 693.0 719.0 1,244.0 940.0 1,087.0 785.0 530.0 667 897.0 o/w portfolio investment 31.3 -2.1 -264.7 -47.0 26.0 -191.0 -151.0 -164 -48.0   Overall Balance 235.0 -597.0 759.0 534.0 378.0 -353.0 101.0 427 104.0 Gross International Reserves (million USD) 2,384.7 2,044.0 2,643.8 2,912.3 3394.0 2895.0 2962.0 3380.0 3475.0 Gross international reserves in months of imports 4.4 3.2 4.3 4.5 5.2 5.0 5.4 5.5 4.9 Memoranda items GDP at current prices (USHS billions) 40,946.0 47,078.0 59,420.0 64,758.0 70458 76517 82903 91351 100552 GDP at current prices (USUS$ millions) 20,181.4 20,262.5 23,237.3 24,663.1 27,757.5 27,102.7 24,079.0 25,891.0 … Exchange rate (period average) 2,028.9 2,323.4 2,557.1 2,591.0 2,538.3 2,823.2 3,443.0 3,528.3 … Source: Bank of Uganda 57 58 Table A6. Inflation Rates (percentage changes) Item 2008/9 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 2016/17   FY FY FY FY FY FY FY FY FY CPI (average) 14.2 9.4 6.5 23.7 5.8 6.7 2.9 6.6 5.7 CPI (end of period) 10.9 4.6 4.0 15.4 6.4 3.0 4.9 5.9 6.4 Food (end of period) 27.9 16.5 9.3 12.8 -1.4 7.2 7.2 -2.1 18.1 Core Inflation (end of period) 8.9 6.7 5.7 19.5 5.8 2.9 5.0 6.8 5 Source: Uganda Bureau of Statistics For more information, please visit: www.worldbank.org/en/uganda Join the discussion on: ugandainfo@worldbank.org http://www.facebook.com/worldbankafrica http://www.twitter.com/worldbankafrica http://www.youtube.com/worldbank