GROWTH, TRADE, AND TRANSFORMATION A Country Economic Memorandum for Uganda GROWTH, TRADE, AND TRANSFORMATION A Country Economic Memorandum for Uganda © 2022 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; Internet: www. worldbank.org Some rights reserved 1 2 3 4 23 22 21 20 This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not nec- essarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions Attribution — Please cite the work as follows: World Bank. 2022. Growth, Trade, and Transformation: A Country Economic Memorandum for Uganda. Washington, DC: World Bank. Translations — If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations — If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole re- sponsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content — The World Bank does not necessarily own each compo- nent of the content contained within the work. The World Bank therefore does not warrant that the use of any third-party-owned individual component or part con- tained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to reuse a component of the work, it is your responsibility to determine whether permission is needed for that reuse and to obtain permission from the copyright owner. Exam- ples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to World Bank Publica- tions, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; e-mail: pubrights@worldbank.org. Cover and interior art: Design and Development Minds. Contents Acknowledgements............................................................................................................................................... 8 10 Key Messages......................................................................................................................................................9 EXECUTIVE SUMMARY .......................................................................................................................................13 INTRODUCTION...................................................................................................................................................... 39 Growth: Not Fast Enough...........................................................................................................................40 Demography: Slowing, But Still Growing Too Fast..................................................................... 47 Nature: Increasingly Fragile.........................................................................................................................51 The Outline of This Study.......................................................................................................................... 52 CHAPTER 1. GROWTH AND TRANSFORMATION IN UGANDA...................................................59 What Has Driven Growth in Uganda?............................................................................................... 60 Why Does Structural Transformation (Still) Matter for Uganda?...................................64 What Is Holding Back Rapid Transformation? .............................................................................73 Synopsis................................................................................................................................................................89 CHAPTER 2. THE ROLE OF INTERNATIONAL TRADE......................................................................95 The State of Uganda’s International Trade.....................................................................................96 The Firm-Level Drivers of Uganda’s Exports................................................................................99 The Determinants of Uganda’s External Trade...........................................................................107 Trade’s Effect on Transformation and Inequality .....................................................................115 Synopsis ............................................................................................................................................................... 121 Spotlight A. Uganda’s Booming Dairy Sector..............................................................................123 Spotlight B. Uganda’s Changing Fisheries.....................................................................................126 Spotlight C. Ugandan Coffee: High Quality at a Discount..................................................129 CHAPTER 3. FUTURE CHALLENGES, OPPORTUNITIES, AND POLICIES.............................135 The Future of Transformation in Uganda...................................................................................... 136 Table of contents Agricultural Productivity............................................................................................................................139 Regional Integration and Trade............................................................................................................ 145 Tourism.................................................................................................................................................................150 Digital Transformation.................................................................................................................................153 Hydrocarbons...................................................................................................................................................159 3 Climate Change.............................................................................................................................................. 168 Synopsis............................................................................................................................................................... 172 APPENDICES...........................................................................................................................................................175 Appendix A. A Simple Calculus of Migration Decisions....................................................... 177 Appendix B. The Structural Transformation Model................................................................180 Growth, trade, and transformation 4 List Of Boxes, Figures, Maps, And Tables Boxes Box 1. Employment statistics in Uganda....................................................................................... 62 Box 2. What is structural transformation and why does it matter?........................... 67 Box 3. The role of frictions and land access in explaining Uganda’s transformation. ...............................................................................................................................85 Box 4. Hydrocarbons: curse or blessing?.....................................................................................167 Figures Figure 1. The analytical outline of the report...................................................................................14 Figure 2. The drivers of growth in value added per worker, by sector.......................... 16 Figure 3. Why are Ugandans not as productive as the Vietnamese?.............................17 Figure 4. Factors that may affect spatial or sectoral transitions.......................................18 Figure 5. The role of frictions and land access in transformation (simulated)......20 Figure 6. Exporter characteristics in Uganda and its peers................................................ 22 Figure 7. Trade policies of Uganda and comparison countries, 2018––all products..................................................................................................................... 24 Figure 8. Aggregated Logistics Performance Index (LPI) scores (2012–2018)........ 25 Figure 9. The impact of lower food prices on structural transformation (10-year).............................................................................................................................................27 Figure 10. Distributional impact of a 10 percent price decline in all agricultural products............................................................................................................................................. 28 Figure 11. Medium-term (10-year) projections: baseline vs. policy outcomes.......30 Figure 12. Income trends in Uganda.........................................................................................................41 Figure 13. Expenditure side decomposition (constant 2010 USD series)....................44 Figure 14. Real dynamics in economic sectors (constant 2010 USD series).............45 Table of contents The spatial distribution of the Ugandan population............................................48 Figure 15 . The age distribution of the Ugandan population...................................................49 Figure 16 . Figure 17. Land cover change in Uganda, 1990–2015................................................................ 52 Figure 18. The analytical outline of the report.................................................................................54 Figure 19. The drivers of growth in value added per worker................................................... 61 5 Figure 20. Estimates for the employment share of agriculture............................................63 Figure 21. Benchmarking sector structures: Uganda vs. all countries.............................66 A simple model of structural transformation based Figure 22. on the Lewis Model..................................................................................................................... 67 Why are Ugandans not as productive as the Vietnamese?...........................68 Figure 23. Employment sector of household heads, percent.................................................71 Figure 24. Figure 25. Internal migrants in Uganda by area of origin and destination.....................72 Figure 26. Factors that may affect spatial or sectoral transitions..................................... 76 Figure 27. Spatial transitions: change in consumption by adult equivalent, 2010–2016.........................................................................................................................................77 Simulation results........................................................................................................................86 Figure 28. Uganda’s exports: dynamics, composition, and destinations...................... 97 Figure 29. Figure 30. Uganda’s imports: dynamics, composition, and sources................................98 Figure 31. Uganda’s exporting firms (2020 unless otherwise noted)..............................101 Exporter entry, exit, and survival.....................................................................................105 Figure 32. Exporter size and domestic value added.................................................................106 Figure 33. Figure 34. Trade policies of Uganda and comparison countries, 2018––all products...................................................................................................................108 Figure 35. Unilateral deviations from the CET...............................................................................110 Figure 36. Trade facilitation and logistics indicators................................................................. 112 Figure 37. The impact of lower food prices on structural transformation (10-year)............................................................................................................................................114 Figure 38. Distributional impact of a 10-price decline in all agricultural products.116 Growth, trade, and transformation Figure 39. Price changes by sector, structural transformation, and inequality...... 120 Figure 40. Uganda’s growing exports of dairy.................................................................................124 Figure 41. Uganda’s formal and informal fish exports................................................................ 127 Figure 42. Medium-term (10-year) projections: baseline vs. policy outcomes......137 Figure 43. Allocation of public expenditure in support of the agriculture sector... 141 Figure 44. Use of improved inputs in Uganda..................................................................................143 Figure 45. The estimated impact of the AfCFTA by country and sector in Uganda...............................................................................................................147 Figure 46. International visitor arrivals in Uganda..........................................................................151 Figure 47. Uganda’s Travel and Tourism Competitiveness Index rankings..................152 6 Figure 48. Major oil fields and refinery and pipeline positions .........................................160 Figure 49. Oil production and fiscal projections............................................................................161 Figure 50. Capex and government financing needs projections...................................... 164 Figure 51. Is there a resource curse in Uganda already?....................................................... 166 Predicted climate-driven migration patterns.......................................................... 171 Figure 52. Figure A.1. Initial income and the gain from migration...............................................................179 Tables Table 1. Trade policies of Uganda by broad industry groups........................................... 23 Table 2. Migration flows evaluated at destination.....................................................................73 Table 3. Migration flows evaluated at source................................................................................74 Table 4. Migration is considerably less prevalent in Uganda compared to other countries...............................................................................................75 Table 5. Sectoral transitions (all workers including self-employed, age 15+)........ 79 Table 6. Sectoral transitions (only wage workers, age 15+)..................................................81 Table 7. Marginal effects on the probability of having a head of household internal migrant................................................................................................... 83 Table 8. Cross-country regressions for export competitiveness indicators....... 102 Table 9. Cross-country regressions for exporter dynamics indicators.................. 103 Table 10. Trade policies of Uganda by broad industry groups.........................................108 Table 11. Digital Evolution Index components............................................................................. 155 Table B-1. Value of parameters................................................................................................................. 185 Table of contents 7 Acknowledgements This study was prepared by a multidisciplinary team led by Harun Onder (Senior Economist, TTL-Lead Author) and Richard Walker (Senior Economist, Co-TTL) un- der the guidance of Vivek Suri (Practice Manager). Tihomir Stucka (Senior Econo- mist) was the TTL in the earlier stages. Contributing authors by field are as follows: • Macroeconomic analysis: Harun Onder, Tihomir Stucka, and Richard Walker • Firm-level trade analysis: Hiau Looi Kee (Lead Economist) and Alejandro Forero Rojas (Trade Expert) • Trade logistics and regional integration: Jean-Christophe Maur (Senior Economist), Aleksandar Stojanov (Junior Professional Officer), and Silas Kanamugire (Trade Expert) • Migration analysis: Aziz Atamanov (Senior Economist) and Harun Onder • Structural transformation model: Erhan Artuc (Senior Economist), Isambert Leunga (Trade Expert), and Harun Onder • Discrete choice model for distributional analysis: Erhan Artuc • Hydrocarbon sector: David John Santley (Senior Energy Specialist) and Su- sana Moreira (Senior Gas Specialist) • Digital and ICT sector: Qursum Qasim (Private Sector Specialist) • Tourism: Liliya Repa (Tourism Expert) The team is grateful for guidance and support from Keith Hansen (Country Direc- tor), Asad Alam (EFI Regional Director), R. Mukami Kariuki (Country Manager), Anto- nio Nucifora (ETIRI Practice Manager), Allen Dennis (EFI Program Leader), and Philip Schuler (Lead Economist). The team benefited from excellent comments from peer reviewers Luc Christiaensen (Program Manager, HSPJB), Nora Dihel (Senior Trade Economist, ESAMU), Dino Merotto (Lead Economist and Growth Cluster Leader, Growth, trade, and transformation HSPJB), and Bob Rijkers (Senior Economist, DECTI). Esther Ampumuza (Team As- sistant, AEMUG) and Karima Laouali Ladjo (Program Assistant EAEM1) provided excellent operational and administrative support. The team would also like to acknowledge numerous government officials for pro- viding invaluable comments and suggestions. The climate change analysis draws from a background note prepared by Marco Boggero (STC) and Kanta Rigaud (Lead Environmental Specialist) for the World Bank SCD Update, and the sectoral and spatial transition analysis benefited from inputs from Carolina Mejia-Mantilla (Se- nior Economist). This study was partially financed by Umbrella Facility for Trade. Editorial review and typesetting services provided by JPD Systems and Design and Development Minds. 8 10 Key Messages To support Uganda’s policy agenda for economic growth and transformation, this report analyzes: (i) Uganda’s structural transformation and growth experience in recent years, (ii) the role of international trade in shaping the country’s economic trends, and (iii) future risks and opportunities. The key findings are as follows: The patterns of growth and transformation in Uganda 1. Uganda has made impressive progress since the 1990s. With policy re- forms unlocking productivity growth, and with favorable external condi- tions in the 2000s, per capita income increased 3-fold and life expectan- cy by 50 percent in the last three decades. 2. Uganda’s growth failed to foster rapid structural transformation and even- tually stalled in the 2010s. This has left about two thirds of the country’s labor force in agriculture, producing less than a quarter of the country’s GDP. Without a rapid transformation, Uganda’s high population growth led to one of the fastest deforestation rates in the world. 3. The slow pace of transformation was driven by: (i) a weak pull from non-agricultural sectors; (ii) major frictions in the economy including high job search (lacking digital facilitation) and mobility (low rural connectivi- ty) costs; (iii) rural credit constraints and market imperfections (e.g., land) that render upfront investments (e.g., human capital) unaffordable. 4. Going forward, Uganda faces two structural fault lines: (i) generating enough jobs for its rapidly growing labor force, which will grow 2.5-fold in the next three decades, and (ii) halting the depletion of woodlands and bushlands, which have been converted to small-scale farms. The role of trade in facilitating growth and transformation 5. International trade will shape Uganda’s future growth and transformation for two main reasons: (i) with a relatively small economy, Uganda needs access to other markets to grow; (ii) alignment with international markets can help reduce the effects of domestic frictions on transformation. 6. To boost trade, Uganda needs to address: (i) high trading costs driven by logistics and infrastructure problems, (ii) trade policy obstacles driven by protectionist reflexes (e.g., restrictive import regulations and tariffs) and; (iii) frequent problems in regional integration frameworks. 7. Lowering food prices, for example by reducing trade costs (better facil- Key message itation) and tariffs, can boost growth and transformation and reduce in- equality by helping the poorest Ugandans. Simulations in this report show that a 10 percent reduction in food prices can boost the welfare of rural wage workers, who are less likely to own land, by 2 percent. 9 Future risks and opportunities 8. Without a breakthrough, Uganda’s transformation is projected to remain modest in the medium-term. However, policies aimed at boosting pro- ductivity and reducing frictions can help. With 2 percentage points higher TFP growth in manufacturing, annually, Uganda’s GDP can gain an addi- tional 20 percent in a decade, and with lower frictions, 12 percent. 9. A breakthrough can come from removing the prevailing constraints on the economy, which can address import restrictions, costly trade logis- tics, limited rural access to infrastructure and information, and market imperfections like credit constraints and low liquidity of assets like land. 10. Uganda’s future development will also be shaped by developments in: (i) agricultural productivity, (ii) regional integration and trade (e.g., the Af- CFTA), (iii) tourism, (iv) digital transformation, (v) hydrocarbons, and (vi) climate change. In each of these areas, adopting the right policies is cru- cial for exploiting opportunities and mitigating the risks that can derail the country’s development trajectory. Growth, trade, and transformation 10 Executive summary Overview 11 Executive Summary U ganda has made significant progress in the last three decades; most im- portantly, Ugandans now live much longer and better lives than before. Since 1990, the life expectancy of a Ugandan baby has increased by almost half, from 43.8 years to 63.4 years. At the same time, per capita Gross Domestic Product (GDP) has registered more than a 3-fold increase, from USD 248 to USD 817. With higher incomes, the country’s poverty rate1 decreased from 64.4 percent to 41.3 percent. Overall, Ugandans have managed to increase significantly the longev- ity and quality of their lives in a relatively short period of time. This progress was driven by rapid economic growth and enabled by good pol- icies and favorable conditions. Between 2000 and 2011, Uganda’s real GDP grew by an impressive average of 7.9 percent annually. This was made possible by three factors: (i) the Government’s reform agenda, including macro-stabilization reforms to establish fiscal discipline and bring inflation under control, price liberalization in coffee, consumer products and the country’s exchange rate regime, and the privat- ization of several inefficient state-owned enterprises; (ii) a ‘peace-dividend’ that materialized as the conflict with the Lord’s Resistance Army subsided; and (iii) an emergence of favorable external conditions, including the creation of South Sudan as a trading partner and a destination for international aid sourced from Uganda, and the liberalization of regional agricultural trade through the East African Com- munity (EAC). Since 2011, however, growth has slowed down, exposing major economic fault lines in the long run. The period after 2011 provided a striking contrast to the previ- ous decade. With slowing reform momentum, the reversal of external factors, and a number of exogenous shocks like droughts, growth slowed to 4.3 percent per year, barely surpassing the stubbornly high population growth rate of over 3 percent per year. This provides the first major fault line of the Ugandan economy: in the next 3 decades, population growth will remain high, projected to double and the working-age population to increase 2.5-fold. Thus, better growth performance is essential to sustain the improvements in Ugandans’ lives. The second fault line is about natural assets: with increasing rural population pressure, large swathes of the country’s forests and bushland have so far been repurposed for agriculture, mostly to support low productivity, small-scale farming. Since 1990, this process has led to a 69 percent decrease in woodlands and 41 percent decrease in bushlands. Going forward, this shock-absorption mechanism (using natural capital to offset missing economic opportunities for an ever-growing population) will become increasingly Executive summary costly, irreversibly damaging Uganda’s natural assets, and less effective, failing to provide real economic opportunities. It is, therefore, crucial to unlock new sources of growth in Uganda. 13 Figure 1. The analytical outline of the report Spatial & Sectorial Analysis of Discrete choice sectorial analysis & exporter model-inequality transition Fiscal model dynamics analysis analysis for oil Chapter 1 Chapter 2 Chapter 3 Drivers of growth The role of Future and stuctural international opportunities transformation trade and policies Structural transformation model To support Uganda’s search for new growth, this report analyzes: (i) Uganda’s structural transformation, (ii) the role of international trade, and (iii) future op- portunities. The analysis first considers the drivers of economic growth in Uganda, with a special emphasis on structural transformation—that is, the reallocation of labor towards more productive activities (Figure 1). The analysis shows that, for Uganda, where nearly two thirds of the workforce is occupied in agriculture, which Growth, trade, and transformation produces less than a quarter of the GDP, a more productive employment of la- bor would inevitably become a central element of a future growth strategy. Next, the analysis turns to analyzing the role of international trade in promoting such transformation. For many economies that suffer from small domestic markets and distortions, trade can help allocate productive factors to more efficient uses. Fi- nally, the report takes a forward-looking approach to explore future opportunities and policies in Uganda, focusing specifically on five key areas: regional integration and trade, hydrocarbons, tourism, digital transformation, and climate change and the environment. In each of these areas, the discussion identifies potential gains for growth and transformation, possible risks and pitfalls, and suggests policies to maximize gains while managing those risks. The report utilizes a large number of data sources, statistical methods, and economic models to derive systematic and transparent conclusions. These in- clude: an empirical analysis of sectoral and spatial transition patterns among Ugan- 14 dans using multiple Ugandan household surveys; a simulation-based analysis of structural transformation in Uganda using a general equilibrium model of structural transformation; an analysis on the firm-level drivers of Uganda’s international trade using customs data; an analysis of the role of trade policy on inequality in Uganda using a discrete choice model drawing from household surveys; and a fiscal analy- sis of the hydrocarbon sector using a sector-based model. The technical details of these approaches are provided in Appendix B. Why does transformation matter in Uganda? In Uganda, economic growth has primarily been driven by within-sector pro- ductivity changes, and structural transformation has remained weak. In the 2000s, all major sectors, including agriculture, registered significant within-sector productivity growth, which drove economic growth (Figure 2). Labor reallocation towards sectors with higher productivity was a contributing factor, but remained modest. In comparison, within-sector productivity slowed down significantly after 2011, especially after 2015 and in agriculture. This, together with unfavorable labor allocation effects, has broken the growth momentum in Ugandan productivity. Al- though the slowdown in productivity growth within industry and services was sig- nificant, especially in the second half of the decade, they still registered positive within-sector productivity growth. Unlike those, a decreasing within-sector pro- ductivity in agriculture pulled total labor productivity down by 1 percentage point in 2010–15 and about 0.2 percentage points in 2015–20. To achieve a middle-income status, Uganda will need to overcome its slow pace of economic transformation. The majority of the per capita income gap between Uganda and its middle-income peers (e.g., Vietnam) can be explained by two fac- tors (Figure 3): first, low productivity in agriculture (which explains 34.7 percent of the gap with Vietnam); and, second, the failure to reallocate larger shares of its labor force to more productive activities outside of subsistence agriculture (45.3 percent of the gap with Vietnam). Moreover, a low measurement of agricultural productivity can be intertwined with rural underemployment; e.g., measurement problems, like production for own consumption and actual hours at the farm being lower than that implied by overall employment numbers, can bring down produc- tivity indicators using national accounts, as shown by Christiaensen and Demery (2017). Nevertheless, about two-thirds of the country’s workforce remain in agri- culture, producing less than a quarter of total value added, which is likely the most prominent stumbling block for Uganda’s development. Executive summary 15 Figure 2. The drivers of growth in value added per worker, by sector In Uganda, growth has largely been driven by within-sector productivity dynamics. This was shared by all three sectors in the 2000s, but changed after that as agriculture registered a reversal in productivity growth. Agriculture 2000 - 2005 2005 - 2010 2010 - 2015 2015 - 2019 Industry 2000 - 2005 2005 - 2010 2010 - 2015 2015 - 2019 Services 2000 - 2005 2005 - 2010 2010 - 2015 2015 - 2019 -2 -1 0 1 2 3 Annualized growth in value added per worker in % Within-sector Across-sector Dynamic Source: World Bank Group Jobs and Structural Change Tool (accessed in September 2021), which uses WDI data for labor allocations (based on ILO modeled estimates). For a discussion on differences across various labor market estimates in Uganda, see Box 1. What is holding back transformation? The slow pace of structural transformation in Uganda is paralleled by sectoral Growth, trade, and transformation and spatial transition patterns. People often move as a response to major differ- ences in living conditions between different places (Figure 4). Those can be driv- en by wages, living costs (prices), amenities like schools and electricity, and other factors like age, gender, religion, and ethnicity. Unmeasured benefits like the pro- duction of food for own consumption (prevalent in Uganda’s subsistence agricul- ture) are also important. The gain from the movement, inclusive of all those factors, should exceed its costs, including direct mobility costs like travel tickets and others like costs from job search or liquidating/acquiring assets (home or land). Moreover, individuals should be able to afford any frontloaded costs, e.g., education/training for future jobs, which may not be possible if they are credit constrained (no access to credit and/or inability to liquidate assets). For many in Uganda, these conditions are not fulfilled, as discussed below. 16 Figure 3. Why are Ugandans not as productive as the Vietnamese? Vietnamese workers are twice as productive as Ugandan workers. More than three quarters of this gap is driven by Uganda’s excessive labor share and low productivity in agriculture. The productivity gap in manufacturing does not play a major role. Drivers of the productivity gap 100 Lower productivity in manufacturing 3.6 90 16.5 Lower productivity in services Share of the difference (%) 80 70 Lower productivity 60 34.7 in agriculture 50 40 30 Differences in 20 45.3 labor allocation 10 0 Source: WDI and UNU-WIDER Economic Transformation Database, World Bank staff calculations. Manufacturing Notes: WDI Source: in this Economic and UNU-WIDER figure includes mining andDatabase, Transformation construction in ISIC-4 World classification Bank staff for Notes: simplicity. calculations. Manufacturing in this figure includes mining and construction in ISIC-4 classification for simplicity. The pull generated by modern sectors has remained weak. Higher productivity and wages in industry and services are often the main drivers of rural-urban tran- sition. Household surveys show that, between 2010 and 2016, rural-urban migrants in Uganda did not attain additional consumption levels relative to those who stayed in rural areas (on average, a 1.5 percentage point loss was associated with the tran- sition). With no major changes in work hours, this points to a weak pull factor from the urban side. An interesting exception is the group of agricultural wage workers, whose hourly wages increased by an additional 76.2 percent (compared to those who stayed in agriculture) after transitioning to industry jobs. In comparison, for self-employed Ugandans in subsistence agriculture (who are more likely to pro- duce for own consumption), the ex-ante gain from transitioning out of agriculture does not seem to be large enough. Ugandans also face acute constraints in traditional sectors/locations. Even when moving out of subsistence agriculture may be gainful, it may not be afford- Executive summary able due to credit constraints. In Uganda, this problem looms large. An empirical analysis of the Uganda National Household Surveys show that outmigration is low in agriculture intensive areas, especially where traditional land tenure prevails. These, together with the fact that access to financial credit is severely limited in rural ar- eas, imply that Ugandans cannot borrow or liquidate their assets (if any) to pay for a livelihood transition. 17 Figure 4. Factors that may affect spatial or sectoral transitions What could slow down structural transformation? Destination (pull) issues Insufficient infrastructure, unfavorable institutions, and high trade costs can lead to low investments and sluggish growth in productivity and wages. Transaction costs Transportation costs, job search cost and uncertainty, and social fragmentation can slow down labor flows despite significant Source (push) issues wage gaps across regions. Illiquid assets, credit constraints, and unrecorded benefits (e.g., own consumption) from farm can reduce incentives to migrate. Source: World Bank. Additionally, the empirical analysis shows that areas with lower provision of public services like electricity, roads, and education have exhibited lower outmigration, and improvements in these services within the last 5 years have accelerated out- migration. These findings point to major frictions (reinforced by credit constraints) Growth, trade, and transformation in the form of skill barriers, high job search costs (lacking digital facilitation), and high mobility costs (low quality transportation), which restrict spatial and sectoral transitions. The binding constraints for labor outflow from traditional sectors have intensi- fied in recent decades. Figure 5 provides a number of simulations that help ana- lyze the role of binding constraints (frictions) in Uganda’s structural transformation, quantitatively. Panel a shows the actual and modeled dynamics of agriculture’s la- bor share in Uganda. These rely on the actual wage ratios between agriculture and manufacturing in panel b, which decrease from 35 percent to 13 percent, over time. These panels reveal two important insights. First, the wage gap between agricul- ture and manufacturing has nearly doubled while the share of labor employed in agriculture remained large, with mainly workers from better off households leaving the sector.— that is, frictions constraining labor outflow have intensified. Second, 18 the actual and modeled estimates match (panel a) in a relatively simple model. Therefore, the wage gap pattern observed in the data (driven by the frictions in the economy) by itself explains well the observed labor dynamics. Counterfactual analyses confirm the intensified role of frictions, and the ex- pansion of agricultural land, in slowing transformation in Uganda. The last two panels in Figure 5 provide two counterfactual exercises. In panel c, the wage ratios are kept at the initial 35 percent rate (instead of decreasing to 13 percent), which, according to the model, causes the labor share of agriculture to decrease to 27 percent (instead of 64 percent in the baseline model). This shows that intensifying frictions have played a significant role in preventing transformation. In panel d, rural population intensity is held constant at its 1990 value by allowing agricultural land to expand. This increases the simulated labor share of agriculture from 69 percent to 74 percent and shows how agricultural land expansion can slow structural trans- formation. Public policies play an important role in shaping the factors that limit transfor- mation. Despite making progress in recent decades, Uganda still has major gaps in public service provision, especially in rural areas. The country has one of the lowest electricity access rates in the world and its transportation infrastructure suffers from underinvestment and lack of maintenance. Together, these have limited busi- ness opportunities and investments, especially in agro-processing and tourism, weakening the pull factor in modern sectors. Similarly, Uganda’s chronically low ed- ucation outcomes, together with severely lagging rural financial development, have rendered spatial and sectoral transitions unaffordable. The COVID-19 pandemic has likely worsened these outcomes further. Schools have been closed for longer than any other country in the world––as result, the majority of 15 million children have remained out of school for most of the past two years. Finally, for the last 5 years, there has been an upward trend in political violence, risk, and uncertainty, which threaten to further constrain investments and, thus, reduce gains from sec- toral and spatial transitions. What role can international trade play? The future of growth and transformation in Uganda will largely rely on interna- tional trade. There are at least two major reasons why international trade is essen- tial for Uganda’s future development. First, with its relatively small economic size and low income, Uganda’s growth needs to benefit from access to other markets. Executive summary Good access to other markets can help the country import essential inputs that cannot be produced domestically. Instead, Ugandans can then focus their labor and capital in areas where the country has comparative advantages and, with glob- al market opportunities, also reap benefits from economies of scale. Second, as described above, economic transformation in Uganda has been slowed by sizeable distortions and frictions. Further removal of trade barriers can help either eliminate them or restrain their consequences for growth and transformation. However, these effects are not realized automatically, and typically require an active policy stance. 19 Figure 5. The role of frictions and land access in transformation (simulated) a. Labor share of agriculture 1.0 0.8 0.6 Share 0.4 0.2 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 b. Wage ratios (wa/wm): actual vs. predicted 1.0 0.8 0.6 Ratio 0.4 0.2 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 c. Counterfactual labor share of agriculture: No change in wages 1.0 0.8 0.6 Share 0.4 0.2 0 Growth, trade, and transformation 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 d. Counterfactual labor share of agriculture: more land access 1.0 0.8 0.6 Share 0.4 0.2 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Actual Model Predicted Counterfactual Source: Artuc, Leunga, and Onder (2022) for this report. Notes: Simulations employ an open economy general equilibrium model for structural transformation with three sectors (agriculture, manufacturing, and services), where frictions are modeled using exog- enous wage ratio dynamics from actual data. The model is calibrated to Uganda by using national account statistics. 20 Uganda’s international trade performance has paralleled its GDP patterns: an exceptionally dynamic growth episode in the 2000s followed by stagnation in the 2010s. From 2000 to 2012, the exports of Uganda grew 7-fold. Services did especially well (10-fold growth); in fact, tourism has become the top foreign-ex- change generator (32 percent of all export revenues in 2019). On the merchandise side, agriculture is the most prominent export sector: coffee (8.5 percent of all exports), dairy (3.2 percent), fish (2.8 percent), tobacco (1.4 percent), and cocoa (1.2 percent). The agricultural exports are largely segmented by destination: almost all coffee is exported to advanced economies in Europe and North America, dairy goes to Kenya, and fish is largely sold in Asian markets. The firm-level analysis in this report has identified both strengths and weak- nesses underlying Uganda’s trade performance so far. A firm-level diagnostic of Ugandan exporters’ size, entry, exit, and trade patterns (Figure 6) reveals important observations: • High levels of churn: A dynamic firm entry pattern in Uganda shows that an open business environment and low entry costs encourage firms to partici- pate in exporting. On average, 45 percent of all exporters are newcomers and each year 47 percent stop exporting. This level of churn is higher than in most countries, except in Tanzania. • Difficulties in accessing imported inputs: There are symptoms of major distortions in import markets. If exporting firms cannot import inputs, they fail to maintain competitiveness in international markets. Only larger firms who both export and import survive. This latter group exported about 6 times more compared to the former group in 2020. Ugandan exporters also have too high domestic value added compared to other countries. While this may sound good, it shows that Ugandan exporters face problems in access- ing imported inputs/materials. What is limiting the country’s trade potential? Uganda faces natural obstacles for trade like a landlocked location, but there are important man-made impediments too, and those can be addressed to im- prove trade. As a landlocked country, with the nearest maritime port about a thou- sand kilometers away from the country’s border crossings, Uganda relies heavily on the economic relationship with its immediate neighbors. About 96 percent of Uganda’s transit comes from the port of Mombasa in Kenya through the Northern Executive summary Corridor. However, there remain important factors that limit Uganda’s trade po- tential, broadly categorized into three groups: (i) a trade policy that is influenced by protectionist reflexes, (ii) increasing frictions in regional integration frameworks, and (iii) high trading costs driven by logistics and infrastructure problems. As dis- cussed below, these factors have far-reaching consequences. 21 Figure 6. Exporter characteristics in Uganda and its peers Exporter Entry Rate TZA 0,52 KEN 0,45 UGA 0,45 CMR 0,43 ETH 0,30 MUS 0,30 COL 0,30 ZAF 0,25 0 0,10 0,20 0,30 0,40 0,50 0,60 Entry rate Median Exporter Size ETH 145 COL 65 UGA 33 CMR 30 ZAF 24 KEN 19 TZA 14 MUS 14 0 20 40 60 80 100 120 140 160 Median exporter size (Thousand USD) Growth, trade, and transformation Exporter Exit Rate TZA 0,50 UGA 0,47 CMR 0,43 KEN 0,43 MUS 0,32 ETH 0,31 COL 0,30 ZAF 0,24 0 0,10 0,20 0,30 0,40 0,50 0,60 Exit rate 22 Average Size of Exporters Exports (Million USD) 2,5 2,0 1,5 1,0 0,5 0 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Exporter-Importers Exporters Only Source: Authors’ calculations based on data used for the Exporter Dynamics Database and Word Integrated Trade Solutions. Note: Panels a–c show averages for 2010–2020 for Uganda, Kenya, Mauritius, and South Africa; 2010–2015 for Cameroon; 2000– 2017 for others. The horizontal axes denote shares (e.g., 0.3 mean 30 percent) in the top panels. Panel d shows Uganda series only. Table 1. Trade policies of Uganda by broad industry groups AVE - Non-Tariff Measures Tariff Measures (NTMs) Products Simple Weighted Products Simple Weighted with average average with average average tariff>0 tariffs tariffs NTMs NTMs NTMs Agricultural raw 24.7 3.9 0.4 100.0 1.0 0.2 materials All Food 82.9 24.1 22.5 91.8 5.3 0.9 Manufactures 57.9 5.8 0.6 100.0 0.5 0.0 Uganda’s trade policies, especially its tariffs against food imports, are quite restrictive. Average tariffs in Uganda are among the highest compared to other countries (Figure 7). Although Uganda does not stand out in its tariff coverage (79.4 percent of imports), the average simple (22.8 percent) and import-weighted (22.3 percent) tariffs in Uganda are the highest in the group. Examples of high tariffs include rice (75 percent), corn (60 percent), and dairy products (60 percent). In Executive summary comparison, Vietnam, an agricultural success case, has less than 40 percent of products facing positive tariffs and its import-weighted average tariff is only 4.2 percent. Uganda’s non-tariff measures are not large in ad valorem equivalent (AVE) terms; however, import permits and license requirements impose sizeable hurdles. Overall, the food tariffs (Table 1) elevate the domestic prices of food, which is paid by Ugandan consumers, and other import restrictions reduce Ugandan producers’ access to inputs, in turn, reducing exports. 23 Figure 7. Trade policies of Uganda and comparison countries, 2018––all products a. Tariff Measures b. Non-Tariff Measures Share Of Products With Tariffs Share Of Products With NTMs 100 100 90 90 80 80 70 70 Percent Percent 60 60 98.9 98.0 96.6 97.9 50 50 92.6 92.3 91.6 86.1 82.2 82.2 79.4 82.1 40 70.6 40 71.8 30 30 20 38.7 20 4 25.3 10 10 0 0 CIV VNM SEN GHA NGA CIV UGA KEN VNM MAR NGA GHA UGA MAR KEN SEN Simple Average Tariff Simple Average AVE-NTM 25 7 6 20 5 15 Percent Percent 4 22.8 6.4 6.4 3 10 5.0 5.0 5.1 4.4 15.4 14.4 2 13.6 12.6 12.5 3.2 11.4 5 7.3 1 1.4 0 0 CIV VNM CIV KEN GHA SEN UGA NGA VNM MAR NGA GHA UGA MAR KEN SEN Growth, trade, and transformation Weighted Average Tariff Weighted Average AVE-NTM 25 4 4 20 3 3 Percent Percent 15 2 3.8 22.3 10 2 12.9 1 1.8 11.1 5 8.0 1.5 1.3 7.7 6.9 6.6 0.9 0.9 1 0.8 4.2 0.6 0 0 CIV VNM CIV KEN GHA SEN NGA UGA MAR VNM NGA GHA UGA MAR KEN SEN Notes: Notes: World World Bank Bank estimates estimates based based on on UN-COMTRADE. UN-COMTRADE AVE-NTM Kee = Ad and Nicita valorem (2021). equivalent AVE-NTM = Adnon-tariff measures. non-tariff mea- valorem equivalent sures. 24 Figure 8. Aggregated Logistics Performance Index (LPI) scores (2012–2018) Uganda’s LPI scores perform well in customs and timeliness, but lag in infrastructure, tracking & tracing, and logistics competence. 3,6 Customs 3,4 3,2 Uganda 3 Vietnam Timeliness 2,8 Infrastructure 2,6 Rwanda Ghana 2,4 Tanzania 2,2 Ethiopia 2 Kenya International Tracking & tracing shipments Logistics competence Source: World Bank Logistics Performance Indicators, https://lpi.worldbank.org/. Regional integration in East Africa has created opportunities for Ugandan pro- ducers; however, implementation problems have increased. Uganda relies on regional partners for connectivity and market access, but disputes about country of origin and unilateral deviations from the Common External Tariff (CET) framework of the EAC have been growing in recent years. The EAC CET has given a boost to Uganda’s regional exports (like in the dairy sector) by reducing tariffs within the region and erecting ones against imports from elsewhere. However, the recent dis- putes with Rwanda and Kenya, which led to a blockade against Ugandan exports Executive summary including maize, sugar, and dairy with the latter, are reminders that EAC members should target global competitiveness rather than protection-driven development. Uganda’s trade logistics performance lag in three categories: infrastructure, lo- gistics competence, and tracking and tracing (Figure 8). The first shortcoming (infrastructure) reflects an over-reliance on road transport, leading to inefficiencies. The latter two dimensions reflect logistics market issues as Ugandan freight and logistics operators cannot compete with larger and more sophisticated neighbors. 25 With large imbalances between export and import volumes, the relatively small size of some exports like coffee, and the country’s price-taker status, Ugandan export- ers pay large markups on transport from Uganda to Mombasa Port. In the long run, regional infrastructure cooperation will be a decisive factor in Uganda’s trade facilitation, but targeted interventions around key sectors can already help. A modal shift from road to rail or rail/lake transportation can help reduce costs but requires significant cooperation across borders. Such cooperation is also essential in improving border processes, e.g., to elimi- nate unnecessary check points and ad hoc fees. Unilaterally, authorities in Uganda can support strategic sectors by enhancing local infrastructure (e.g., rural roads for better collection and electricity access for cold storage in the dairy sector) and lowering the regulatory burden (e.g., streamlining inspection procedures for coffee). How do food prices affect transformation & inequality? How much do protectionist trade policies affect growth, structural transfor- mation, and the overall well-being of Ugandans in reality? The discussion so far has identified a high number of trade frictions that can potentially impose adverse effects on Ugandans’ welfare. To quantify this effect, simulations in this report con- sider a hypothetical 10 percent reduction in food prices, driven by better trade logistics or lower tariffs over the next decade, by utilizing the general equilibrium model for structural transformation described before. This helps to analyze the consequences of such a reduction on GDP and structural transformation. However, the benefits of these effects are not shared equally by all Ugandans. To quantify such distributional effects, a second set of simulations are then used. Results show that lower food prices can accelerate growth and structural transformation in Uganda significantly. Figure 9 shows the impact of a 10 percent reduction in food prices, which can be achieved by lowering trade frictions, on labor shares and value-added dynamics across different sectors in Uganda. The base- Growth, trade, and transformation line figures show business as usual outcomes. Overall, the food price reduction increases GDP growth by an additional 10 percentage points in the next decade compared to the business as usual scenario. This is done purely by reallocating 4.6 percentage points more labor from agriculture to other sectors, where labor pro- ductivity is higher in relative terms, compared to the baseline. As a result, the labor share of agriculture falls to 58.5 percent, solely because lower prices in agriculture makes working in other sectors more attractive. Lower food prices increase welfare across all income groups, but especially for those in the lower and higher income percentiles. Figure 10 shows the simulated effects of a 10 percent reduction in food prices across income groups in Uganda. The first panel shows the share of food items in consumption expenditures and incomes of Ugandans according to their income level. The poorest Ugandans, e.g., wage work- ers in rural areas who do not own land, spend a higher share of their consumption expenditures on food and they rely less on food for their incomes in a direct manner. 26 Figure 9. The impact of lower food prices on structural transfor- mation (10-year) In the business as usual scenario, the labor share of agriculture decreases from 64.7 percent to 63.1 percent in a decade. With 10 percent lower agricultural prices, it decreases to 58.5 percent. Labor share projections 70 64.7 63.1 58.5 60 50 Percent 40 30 24.3 21.5 22.0 20 17.1 13.8 14.9 10 0 Agriculture Manufacturing Services As a result, GDP grows faster with lower food prices. Agricultural value added shrinks more (1.7 points relative to initial GDP), but that is small compared to the additional gains in manufacturing (6.8 points) and services (4.3 points). Value added (as a share of base-year GDP) 80 71.6 69.3 70 64.8 65.0 60 50 42.0 Percent 40 36.0 30 21.9 20 16.1 14.4 10 0 Agriculture Manufacturing Services Executive summary Base year Business as usual projections 10% lower food prices Source: World Bank Logistics Performance Indicators, https://lpi.worldbank.org/. Source: Artuc, Leunga, and Onder (2022) for this report. Simulations employ the open economy general equilibrium model de- scribed before. 27 Figure 10. Distributional impact of a 10 percent price decline in all agricultur- al products Consumption and Income Share - Scenario: All Agriculture 50 40 30 Share 20 10 0 -10 0 10 20 30 40 50 60 70 80 90 100 Income percentile Consumption share Income share Welfare Change - Scenario: All Agriculture 3.5 3 % change in welfare 2.5 2 1.5 1 0.5 0 -0.5 -1 0 10 20 30 40 50 60 70 80 90 100 Income percentile Short-run Long-run Growth, trade, and transformation Source: World Bank calculations based on Artuc, Porto, and Rijkers (2021a and 2021b). High income Ugandans also rely less on food production for their income, but for different reasons. Food consumption patterns also differ between low and high in- come Ugandans. For example, while the latter consume more wheat and rice, the former consume more corn and other cereals in relative terms. Food consumption patterns also differ between low and high income Ugandans. For example, while the latter consume more wheat and rice, the former consume more corn and oth- er cereals in relative terms. The second panel shows the welfare effects. Overall, lowest-income households are more likely to benefit from this shock as they are more likely to be net food consumers and they spend a large share of their income on food. Middle-income households benefit the least, nonetheless positively, since their consumption-production gap is smaller than other income groups. 28 What are the future opportunities in Uganda? Without a breakthrough from recent trends (business as usual scenario), Ugan- da’s transformation is projected to remain modest in the medium-term. Figure 11 shows medium-term (10 years) simulations using the structural transformation model. In the baseline, where all exogenous factors have the same trend as in the last decade, the labor share of agriculture decreases by only 1.6 percentage points (first panel). This outflow, combined with no productivity growth in the sector, leads to a 0.6 decrease in the sector’s value added (second panel), while other sectors keep growing as their momentum from the last decade is preserved. Policies aimed at boosting productivity and reducing frictions can put the country on a more rapid growth and transformation path. With 2 percentage points higher total factor productivity (TFP) growth in manufacturing, Ugandan GDP can gain an additional 20 percent in the next decade compared to the baseline projections. Part of this is driven by the labor outflow from agriculture: the sector’s labor share decreases by 10.3 percentage points instead of a modest 1.6 percent- age points in the baseline. Similarly, with lower frictions, labor reallocation acceler- ates, and the economy grows by 12 percentage points more despite no additional gains in sectoral TFPs. To facilitate such a breakthrough, the authorities can adopt a two-pronged ap- proach: remove prevailing constraints and exploit more targeted opportunities. The analysis has so far identified several key constraints, including high food tar- iffs, costly trade logistics, limited rural access to infrastructure and information, and market imperfections like credit constraints and low liquidity of assets like land. Addressing these constraints can help both increase productivities in each sector and reduce frictions against sectoral and spatial transitions. A broad strategy to address these problems comprises the first element of a two-pronged approach. The second element involves proactively identifying and exploiting opportunities that can transform the economy. The remainder of this summary focuses on six candidates: agricultural productivity, regional integration and trade, especially the African Continental Free Trade Area (AfCFTA), tourism, digital transformation, hy- drocarbons, and climate change. (i) Boosting productivity in agriculture Opportunities: unleashing the muted productivity in agriculture can drive growth, transformation, and job creation in Uganda. Despite agriculture’s important role Executive summary in Uganda’s economy, the sector’s output is far below potential, growing only at 2 percent, annually, compared to the 3 to 5 percent in other EAC members, in recent years. Total factor productivity growth has largely been absent from Ugandan ag- riculture. Farmers’ limited market participation, low uptake of improved agricultural inputs, and limited adoption of technology are some of the main causes of stagnat- ing yields. These, in turn, reflect high logistics costs, low adoption of digital technol- ogies, and insufficient public priority for maintaining innovation – for instance in the area of improved and resilient seeds – and effective extension services in the sector. 29 In fact, Ugandan agriculture has one of the lowest adoption levels of improved seeds, inputs and mechanization. If these problems are addressed, Uganda’s agri-food system can enhance employment opportunities, both skilled and unskilled, for the country’s predominantly young population, the majority of whom live in rural areas. Main risks: increasing rural population density (reducing farm size and incen- tives to commercialize) and climate shocks (suppressing productivity) threat- en the sector. In the absence of a rapid structural transformation, and with over- lapping land tenure systems inhibiting the efficiency of the land market, the share of small household farms with less than 2 hectares of land rose from 75 to 83 per- cent between 2006 and 2016. As a result, the average net land operated in Uganda fell from 1.7 to 1.2 hectares per household, thus reversing the trend toward larger farm holdings, which are more likely to commercialize due to economies of scale and ease of adopting modern technologies. Ugandan farming is also exposed to increasing climate variability and weather shocks, with the country being the most vulnerable to climate change among regional peers on the ND-GAIN index. More than 95 percent of cropland is rainfed and based on subsistence farming, making it especially vulnerable to weather variability and climate hazards. In recent years, this lack of resilience has resulted in huge losses in livestock and crops. For exam- ple, due to drought and pests such as the armyworm in 2016, output plummeted. With increasing climate variability in the future, widespread food insecurity can become increasingly more likely. Policies: numerous broad-based and targeted solutions are within reach. With high transaction costs in transportation and access to information, and a burden- some institutional setup for public policies, the Ugandan farming remains uncoor- dinated and underinvested. To boost the sector’s productivity and resilience, au- thorities should (i) encourage the expansion of vertical integration practices like contract farming and outgrower schemes, (ii) reduce transportation costs through better logistics and competition in agricultural intermediation to enhance the profit margins of farmers engaged in domestic and international trade transaction costs, (iii) facilitate the adoption of better inputs like quality seeds, digital technologies, and climate-smart agriculture practices including sustainable land management Growth, trade, and transformation techniques. (ii) Enhancing regional integration and trade Potential opportunities: income and efficiency gains driven by reforms. The re- forms envisaged by the AfCFTA, including on tariffs, non-tariff measures (NTMs), and trade facilitation measures, are expected to boost Uganda’s GDP by about 3.3 percent by 2035. Food processing and agriculture are the biggest winners with gains of 7.9 percent (about USD 1 billion) and 5.1 percent (about USD 1.6 billion), re- spectively. Natural resources and overall manufacturing register losses of 10.4 per- cent and 0.8 percent, respectively. With additional cooperation in investment and competition policy, and intellectual property rights, Uganda’s total income gains can increase by up to 5.7 percent (the AfCFTA deep scenario). 30 Figure 11. Medium-term (10-year) projections: baseline vs. policy out- comes In the baseline, the labor share of agriculture decreases by 1.6 percentage points (pp) in 10 years. With 2 pp higher manufacturing TFP growth annually, it decreases by 10.3 pp, and with lower frictions by 15 pp. Labor share projections 70 64.7 63.1 60 54.4 49.7 50 Percent 40 29.1 30 26.9 21.1 21.5 22.0 18.8 20 14.9 13.8 10 0 Agriculture Manufacturing Services In the baseline, the value added of agriculture decreases by 0.6 pp in 10 years. With 2 pp higher manufacturing TFP growth annually, it decreases by 3.5 pp, and with lower frictions by 3.0 pp. Value added (as a share of base-year GDP) 120 107.7 105.0 102.1 100 95.9 85.7 86.0 80 Percent 60 42.0 36.0 40 21.9 21.3 18.9 18.4 20 0 Agriculture Manufacturing Services Base year Baseline Higher manufacturing TFP Lower frictions Executive summary Main risks: coordination challenges with the EAC and the loss of protec- tion-driven competitiveness. Synchronizing with the EAC (especially the CET), and maintaining cooperation in this process, will require considerable effort. The AfCFTA has only been ratified by three of the six EAC members, and without broad- 31 er ratification the integrity of the CET may come under pressure. The other prob- lem is the potential impact of liberalization on protected sectors like dairy. Ugan- da’s booming dairy sector has largely relied on the high CET in the EAC, and may lose competitiveness, without this protection in an integrated African market, to other producers like South Africa and Egypt. Ugandan authorities should take pol- icy actions (see below) to align domestic businesses with the market conditions and increased competition that will eventually materialize. Policies: four key actions to enhance regional cooperation and prepare busi- nesses for open competition. (a) Reset cooperation within the EAC by reduc- ing the stay of applications and the CET and programming further liberalizations; (b) Prepare producers for the implementation of the AfCFTA with a transparent roadmap of transition, including better provision of public services, regulations, and infrastructure to reduce costs; (c) Soften import substitution policies and scale up efforts to remove barriers to trade and mobility, including regulatory obsta- cles against imports; (d) Expend more effort in building international branding and quality standards, and enhance and communicate conservation efforts to differen- tiate Ugandan products in competitive markets. (iii) Tapping into the untapped potential of tourism Opportunities: potential to diversify products and attract global visitors. Al- though tourism is already the leading sector in Uganda generating foreign exchange, the demand is largely regional, leaving global travelers untapped. The country has a large number of attractions, including East African savanna and tropical rainforest, and houses rich wildlife stocks, including more than half of all mountain gorillas in the world and numerous bird species, some of which are native to Uganda. Finally, the indigenous cultures provide an additional opportunity. With such attractions, Uganda can offer diversified services that cover cultural heritage, health and well- ness, ecotourism, and religious tourism, in addition to general tourism attractions. Risks: a lingering pandemic; environmental hazards from oil development; fail- ure to strengthen infrastructure, improve hygiene, and enhance tourism man- Growth, trade, and transformation agement. The impact of the COVID-19 pandemic on tourism has been debilitating. International travel came to a standstill on March 21, 2020 when Uganda officially closed all of its borders. A prolonged downturn can have persistent effects on the sector. Potential hazards from oil production and transportation can also irrevers- ibly damage the country’s biodiversity and other natural capital. Finally, Uganda ranks low in infrastructure quality, even lower than Sub-Saharan Africa (SSA) aver- ages, mostly due to insufficient air and ground transportation options; poor hygiene of facilities; and inadequate tourism management, including human resources that sometimes fail to meet customer expectations. Policies: a comprehensive strategy for the sector. (i) Develop coherent strategies and flexible action plans for tourism development; (ii) Improve national-subnation- al coordination by empowering local communities and fully sharing benefits from tourism locally; (iii) Implement a digital transformation of tourism by modernizing regulatory frameworks, strengthening capacity, and expanding accessible digital 32 infrastructure; (iv) Place a greater focus on the environmental and sociocultural pil- lars of sustainability; (v) Collect and release tourism-related data in a timely man- ner to help individual destinations develop appropriate niches; and (vi) Identify tourism-related tradeoffs from other economic activities like deforestation and oil infrastructure for explicit use in medium-term planning and policies. (iv) Embracing digital transformation Opportunities: digital transformation can foster trade and structural trans- formation. Some of the key obstacles to trade and structural transformation in Uganda can be overcome with better digitalization. Most prominent are high job search costs and low access to finance. Enhancing digital services in labor market transactions (e.g., online labor supply, demand, and matching mechanisms) and accessibility of such services can help reduce the information asymmetry and the uncertainty surrounding spatial and sectoral transitions. Similarly, improvements in digital financial services (DFS) can help Uganda’s poor and underbanked consum- ers, such as women and rural dwellers, have access to the formal financial system, thereby reducing the credit constraints that slow their spatial and sectoral tran- sitions. Finally, with e-government services for trade facilitation and value chain digitalization, transaction costs for trade can be reduced substantially. Risks: comprehensive challenges in access and usage, some of which are policy driven. Uganda ranks 81 out of 90 economies on key Digital Intelligence Index met- rics, including infrastructure access, digital inclusion, institutional readiness and level of innovation in the digital economy. Compared to its regional peers, Uganda ranks particularly low on access and fulfillment infrastructure (supply conditions) and innovation even as it exhibits average demand conditions (digital access and inclusion). Digital taxation is a key constraint to broad market development of the digital economy, and digitalization of export value chains in particular. Similarly, the tax on mobile money withdrawals is regressive and constrains market development. Policies: a combination of regulatory and market development policies. Au- thorities should invest in soft infrastructure (policies, institutional capacity, digital literacy programs) to capitalize on Uganda’s advantages. The AfCFTA provides op- portunities for regional market development, but capitalizing on these will require continued investments in expanding digital infrastructure, including in network ac- cess broadly and especially for underserved populations, women, youth, and rural residents. Effective strategies to lower the costs of digital devices and services, including the removal or reduction of taxes on enabling services, are also needed. Executive summary (v) Utilizing hydrocarbons for development Opportunities: fiscal space for boosting development. Uganda’s hydrocarbon reserves are neither large nor of the best quality, and there remain many challenges before oil can flow (estimated to start in 2025). However, if successful, the country’s fiscal revenues from oil production can reach USD 3.3 billion per year during peak production (2030), which would likely correspond to about 4.9 percent of GDP and 33 about a quarter of fiscal revenues from other sources at that time. If used well, the additional revenues from hydrocarbons can help close Uganda’s infrastructure gap and help build the country’s human capital, both much needed. Main risks: fiscal exposure driven by delays or even cancellation, environmental costs, and the possibility of resource curse. The Ugandan Government has sig- nificant capital expenditure (capex) obligations (peaking at an estimated USD 350 million in 2 years), which tightens an already narrow fiscal space, before revenues can be collected. A potential delay in production, or a complete shelving of the project, can add significantly to the country’s fiscal burden. On the environmen- tal side, the oil reserves are concentrated in an area that has among the great- est biodiversity on the planet. Besides ethical concerns, environmental hazards can directly reduce the country’s tourism and agricultural potential. The windfall revenues can also lead to a reallocation of productive resources to non-tradable sectors, thereby reducing growth potential. Finally, oil revenues can also reinforce historical grievances across social groups/regions and destabilize social cohesion. Key policies: many ways to fail, only one way to succeed. Capitalizing on the opportunities provided by oil is not easy. First, the authorities will need to get the fiscal management right to achieve stabilization and efficiency objectives. Second, policies will need to target conditions that can reinforce Dutch disease dynamics in the economy. These include among other trade barriers (i.e., difficulties in exporting final products or importing intermediary inputs) non-inclusive prospects for eco- nomic participation and institutional weaknesses that can facilitate rent-seeking. Third, policies should respond effectively to concerns about graft, environmental impact, lagging regions, and disadvantaged groups, to preempt grievances and so- cial unrest. In the final analysis, there is only one way to make the most of the oil, and that is to get these policies right. (vi) Climate change adaptation Channels: higher temperatures and variability pose broad-based challeng- es to economic transformation. Uganda’s climate has already changed––and Growth, trade, and transformation will change further. There is an uptick in extreme precipitation events, particularly during the shorter rainy season. Precipitation between rainy seasons will decrease along with annual rainfall. The average annual temperature has increased by 1.3 de- grees C since 1960, with more ‘hot’ days and ‘hot’ nights per year. With increasing global average temperatures, these trends in Uganda are likely to continue and in- tensify, with most changes likely to take place between now and the mid-century. Key challenges: direct effects for agriculture and water availability and sup- ply, and a wide spectrum of indirect effects like disease and migration. Many crops, including a key export crop like coffee, are directly facing climate-induced yield losses (up to 75 percent in the Arabica variety). Reduced water availability and watershed recharge is likely to stress fisheries, and, as a result, its significant growth and trade potential. Direct losses are estimated to reach USD 1.5 billion per year by 2050. Indirect effects include a greater incidence of disease and pests that 34 comes with climate extremes. In 2020, after a series of cyclones, Uganda was hit by the most critical desert locust outbreak in 25 years. These events will become more frequent. Finally, climate events can also aggravate migration or conflict. The World Bank (2018) identified the broader Lake Victoria region as a climate migra- tion hotspot, with between 7 percent (optimistic case) and 11 percent (pessimistic case) of the population becoming internal climate migrants by 2050. Policies: a broad-based approach is needed. (i) A climate-smart agriculture strategy should focus on accessing drought and flood prone varieties, employ- ing digital technologies for weather and market advisories and better monitoring of trends in pests, diseases, and market conditions. Uganda’s trade potential has been thwarted in the recent past by its inability to monitor and control pest and diseases. (ii) Globally, it is estimated that for every dollar invested in climate-smart infrastructure four dollars is gained in benefit. These include both extreme-weather resilient infrastructure in cities and micro-grid applications like solar-powered ag- ricultural appliances suitable for smallholder farmers. (iii) An adaptive social pro- tection system can help reduce the persistent effects of transitory shocks and variability. (iv) Finally, to implement all these policies, Ugandan authorities should have better access to data and information on disaster risk and natural resource management. Executive summary 35 Growth, trade, and transformation 36 Introduction Overview Growth, trade, and transformation 38 INTRODUCTION U ganda has achieved a lot in the last three decades, with Ugandans now living longer and better lives. Since 1990, the life expectancy of a Ugandan baby has increased from 43.8 years to 63.4 years. This is about 20 years of additional life, exceeding the earlier one by almost a half. But that is not the only achievement. Ugandans now also live better lives in each one of those years. Over the same three decades, per capita Gross Domestic Product (GDP) registered a 3.2-fold increase, from USD 248 to USD 817. With higher incomes, the country’s poverty rate (headcount ratio, USD1.90 a day, 2011 PPP) decreased from 64.4 per- cent to 41.3 percent. As in other countries, economic growth has been the main instrument for im- proving living conditions in Uganda. Economic growth has been the main driver of poverty reduction in all developing countries. According to Pritchett (2019), nearly all variation in poverty rates across countries in recent years is explained by dif- ferences in median consumption levels (with a very strong correlation of 0.994). Thus, reducing poverty in an economy is almost always about increasing the medi- an consumption through economic growth. Uganda has not been an exception to this trend. The country’s strong growth performance in the 2000s helped reduce poverty dramatically, and when growth slowed down in the 2010s, poverty reduc- tion also stalled. Despite the commendable progress, however, there remains room for improve- ment in the country. Following the rapid growth phase in the 2000s, the rate of progress in the country has slowed in recent years. Going forward, to reach a mid- dle-income status, lift its population out of poverty, and generate enough jobs for one of the fastest growing populations in the world, the Ugandan economy will need to grow rapidly, sustainably, and broadly (i.e., in a shared manner). This report aims to shed light on this process first by analyzing the economic dynamics of the country. This includes an investigation of the drivers of growth and structural transformation in Uganda. Next, it focuses on the role of international trade as a po- tential mechanism to unlock some of the current challenges faced by the Ugandan economy (i.e., shallow domestic market and distortions leading to misallocation of resources). Finally, it considers potential opportunities and future policies required to take advantage of them. This introductory chapter sets the stage for the main analysis by focusing on broad trends in the country. The discussion in this chapter summarizes three broad trends underlying recent developments in Uganda that are likely to shape Introduction the scale and sustainability of its future progress: economic growth, population growth, and changes in the country’s natural capital. It also provides an outline of the main analysis in the rest of the report. 39 Growth: Not Fast Enough By the 2000s, the Ugandan economy was well positioned for the ‘golden de- cade’ of growth in developing countries around the world. Uganda capitalized on three growth-conducive factors in the late 1990s and early 2000s: (i) the peace dividend that materialized as the conflict with the Lord’s Resistance Army subsid- ed; (ii) the emergence of favorable external conditions, which included the cre- ation of South Sudan as a newly independent trading partner, the opening up of the East African Community (EAC) to free regional agricultural trade, and the lo- cal procurement of food aid by the World Food Program (WFP) for its operations in South Sudan; and (iii) the Government’s reform agenda, which unleashed the economy’s ability to take advantage of emerging opportunities. Macro-stabilizing reforms helped establish fiscal discipline and brought inflation under control. Mar- kets were empowered through price liberalization in coffee, consumer products, and the country’s exchange rate regime. Many inefficient state-owned enterprises were privatized (62 firms) or liquidated (31 firms) by 1999. With key stars aligned, the economy soared between 2000 and 2011. Uganda’s real GDP grew by an impressive 7.9 percent per year on average. In parallel, the country was in the midst of a population boom, which grew by about 3.4 percent annually. Nevertheless, even in per capita terms the Ugandan economic growth was impressive at 4.3 percent on average per year. With such rapid economic growth over a decade, the poverty rate was almost halved (measured at USD 1.9 a day), decreasing from 68 percent to 36 percent, and 3.2 million Ugandans were lifted out of poverty. By 2011, the country registered a record USD 850 per capita in national income — only 17 percent away from the lower middle-income (LMIC) status (USD 1,026 per capita in that year, Figure 12). Since 2011, however, the reform momentum has weakened, and the economy has slowed down. The period after 2011 provided a striking contrast to the pre- vious decade. GDP growth slowed down to an average of 4.3 percent per year, barely surpassing the stubbornly high population growth rate of over 3 percent per Growth, trade, and transformation year on average for the same time frame. It is important to note that most devel- oping countries exhibited slower average yearly growth rates in this decade than the previous one. However, such a decrease was much slower elsewhere, about 1.3 percentage points in Middle Income Countries and 1.2 percentage points in Low In- come Countries. In contrast, the large reduction in Uganda (3.7 percentage points) reflects the role played by domestic factors like the slowing reform momentum and environmental shocks, as well as the reversal in supportive external factors. With robust population growth and weakening economic performance, the poverty headcount reversed between 2012 and 2016. About 5 million Ugandans fell into poverty, and Uganda’s poverty rate climbed to 41.3 percent (using USD 1.90 a day in 2011 PPP terms). 40 Figure 12. Income trends in Uganda Uganda had a rapid growth episode in the 2000s, resulting in a convergence with the lower middle-income cutoff. But this momentum was lost in the 2010s. Gross National Income (Atlas Method) 1100 1000 900 800 Current USD 700 600 500 400 300 200 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Uganda IDA cutoff: Lower-Middle Income The gap between Uganda’s per capita income and the average for IDA countries has been widening in the last decade. GDP per capita 3300 Constant 2017 USD (PPP) 2800 2300 1800 1300 800 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Uganda IDA only Source: WDI and World Bank calculations. Notes: The dotted line in the first panel shows trend line (polynomial). Introduction The ‘lost decade’ of the 2010s left a widening gap between Uganda and other developing economies. The stagnation after 2011 resulted in a widening income gap compared to LMICs and even International Development Association (IDA)-only countries. In 2015, the National Planning Authority set the LMIC status as the policy 41 target. By 2019, however, Uganda’s national income per capita was at three-quar- ters of the LMIC cutoff level (an 8 percentage points regression compared to 2011). Moreover, Uganda has not managed to catch up with IDA-only countries. In 2011, Uganda’s per capita GDP was only 22 percent lower than the IDA average. By 2019, this gap had increased to 35 percent. On the expenditure side, the last two decades have witnessed a gradual de- crease in the role consumption played in GDP. In the first half of the 2000s, con- sumption constituted more than 90 percent of GDP (Figure 13). However, its growth rate was surpassed by that of investments and exports in the next two decades, with the exception of the 2011–15 period. As a result, consumption’s share of GDP had decreased to about 80 percent by 2020. With severe droughts in Uganda in 2016–17, which led to a sharp fall in private consumption growth, the downward trend in consumption’s share of GDP accelerated. While consumption growth gen- erated more than 80 percent of net GDP growth between 2011 and 2015, it only contributed about half of this between 2016 and 2020. Exports have grown consistently over the last two decades, exhibiting an er- ror-correcting pattern. Over the last two decades, Ugandan exports of merchan- dise goods and services have grown by 11.2 percent annually, in real terms. This upward yearly trend was largely consistent except for two major episodes of de- viation that were then immediately followed by a return to the trend. The first was in 2008, where real exports grew by 85 percent, largely driven by coffee but also shared across the board. This hike was followed by two years of negative growth of around 9 percent, which brought exports back to its long-term trend. The second deviation was a slump between 2014 and 2016, where exports remained stagnant in real terms. A 34 percent hike in 2017, largely driven by services like tourism, again brought them back to the long-term trend. By 2020, exports constituted about 15 percent of GDP, a 3-percentage point increase compared to two decades ago. Imports have mimicked consumption, but exhibited more pronounced dynam- ics. Between 2000 and 2020, imports grew by 6 percent in real terms, annually. The peak of this pattern was between 2006 and 2010, where the average annual growth Growth, trade, and transformation rate reached 13.3 percent per year, and in parallel to the boom in consumption. Re- cent years, however, witnessed a suppressed pattern, where imports were reduced by 0.4 percentage points annually since their peak in 2015. This was largely driven by contractions in 2016 and 2020, which averaged at a negative 9 percent growth per year. Investments have remained robust thanks to public initiatives, but with sub- dued private investments, crowding out has become a concern. Gross-fixed capital formation increased by about 7.9 percent since 2000, annually. Between 2011 and 2015, the investment growth slowed down to about 4.7 percent per year, which increased to about 7 percent since then. Behind this rebound, however, lies a concerning shift. In the last 5 years, private investments grew by only 3 percent, compared to an average of 5.3 percent over the previous 5 years. This slowdown is driven by higher economic uncertainty, governance issues, and high borrowing costs due to extensive government financing of the public investment push. This 42 push included the completion of the Karuma and Isimba dams, electrification and IT backbone infrastructure, and the Entebbe–Kampala Expressway. Other large in- vestments in electrification, transport (such as the SGR railway connection to Ken- ya and oil well access roads), and the oil pipeline to Tanzania are either in the works or being planned. Difficulties in access to credit have further constrained private investments. Lack of competition, high interest rates, limited credit information, and weak prop- erty rights limit the access of Ugandan firms to finance. Interest rates in banking remain high, at 19 percent on average in FY19/20 (whereas the Central Bank Rate was 9.3 percent) and have remained above those in regional neighbors in the last decade, putting Ugandan firms at a disadvantage. Although the banking system is well capitalized and profitable, competition among banks is limited, the sector exhibits a lack of innovation, and its services fail to reach large segments of the market. Less than 25 percent of bank lending goes to small and medium enterpris- es (SMEs). Lending at longer maturities is constrained as banks rely primarily on short-term deposits (90 percent of their funding base) and larger firms, and SMEs often do not meet standards for long maturities. Two credit bureaus have started to reduce information asymmetry and facilitate access to finance. A law to allow movable assets as collateral was passed recently. However, legal uncertainty over property rights and lengthy proceedings to recover collateral continue to weigh on banks’ credit risks. On the supply side, services have so far provided the largest value added and contribution to output growth. Between 2000 and 2020, services constituted a stable share of GDP, hovering around 44 percent (Figure 14), with an increase in the early 2000s thanks to an episode of rapid growth, followed by a slump in the after- math of the Global Financial Crisis. More recently, services value added grew by 4.8 percent annually between 2016 and 2020, contributing more than half of the GDP growth among the three sectors. Travel and tourism exports of Uganda generated about 32 percent of all export receipts in 2019, surpassing the commodity exports like gold (22.5 percent) and coffee (7.5 percent). Although services trade grew rapidly, the country has much more potential in this area. Uganda is located strategically to serve as a hub for several tradable services. The country provides a major trade gateway for significant captive mar- kets in South Sudan and eastern Democratic Republic of Congo (DRC). There is much potential to add value to transit trade via transport, logistics, and other ser- vices like freight forwarding and e-commerce. But this requires structural solutions to binding constraints such as access to finance, digital infrastructure, regulatory weaknesses, and weak management and technical skills. In addition, inadequate physical infrastructure (including road and highway systems, electrical grids, and telecommunications) and a lack of market information and skilled personnel pro- Introduction vide further challenges. 43 Figure 13. Expenditure side decomposition (constant 2010 USD series) Investments and exports have performed better than consumption over the last two decades… a. Growth in expenditure 20 18 17.2 Growth rate (CAGR) 16 14 13.2 13.3 12 10.9 10.2 9.5 10 8.1 8 7.0 5.5 6.1 5.6 6 4.8 4.3 4.7 4.1 4 2 - 0.4 0 -2 Consumption Investment Exports Imports 2001-05 2006-10 2011-15 2015-20 ... leading to a secular decline in consumption’s share in GDP… b. Share in GDP 140 120 12.2 17.8 14.6 15.3 100 20.4 23.1 25.9 24.5 80 Percent 60 91.6 87.7 83.2 40 80.1 20 0 -24.3 -28.7 -25.6 -20.5 -20 -40 2001-2005 2006-2010 2011-2015 2016-2020 Consumption Investments Exports Imports Growth, trade, and transformation … and rebalancing contributions to growth in the last 5 years. c. Contribution to GDP growth 12 Percentage points 10 1.7 8 2.7 0.1 1.5 6 0.7 1.6 1.9 1.1 4 1.8 6.8 2 4.4 4.7 3.6 0 -1.2 -1.4 -2 -2.9 -4 2001-2005 2006-2010 2011-2015 2016-2020 Consumption Investments Exports Imports 44 Figure 14. Real dynamics in economic sectors (constant 2010 USD series) Between 2000 and 2020, agriculture grew by 3 percent annually. In comparison, manufac- turing grew by 5.7 percent and services by 6.1 percent. a. Sector dynamics 60 Services 50 2000 Share in GDP (%) 2020 40 Agriculture 30 2000 2020 20 Manufacturing 2020 10 2000 0 0 5 10 15 20 Value added (constant 2010 USD, billions) Agriculture’s growth was lower than that of manufacturing and services in every 5-year period. b. Real Growth in GDP 10 9 8.9 Growth rate (CAGR) 8 7.3 7 6.6 6.1 6 5.3 5 4.8 4.1 4.0 4.0 4.3 4 3 2.1 2 1.6 1 0 Agriculture Manufacturing Services 2001-05 2006-10 2011-15 2015-20 The largest sector in the economy, services, also contributed the most to the GDP growth. c. Contribution to GDP growth 8 7 Percentage points 6 5 3.7 4 3 2.6 1.1 1.8 1.9 2 1.2 0.7 1 2.1 0.9 0.6 0.6 1.1 Introduction 0 2001-2005 2006-2010 2011-2015 2016-2020 Agriculture Manufacturing Services Source: WDI and World Bank calculations. The black lines in the first panel show 2-year moving averages. 45 With supply chain problems and inefficiencies in input markets, agriculture fell behind a rapidly growing population. Between 2000 and 2020, agriculture grew by 3 percent in real terms, annually, while Uganda’s population grew faster at 3.4 percent. In fact, the growth rate in agriculture fell short of those in services and manufacturing in every 5-year period over the last 20 years, especially between 2006 and 2015. As a result, agriculture has made relatively modest contributions to growth. In recent years, the share of the largest segment in agriculture, the pro- duction of food crops, has declined by 2 percentage points in overall GDP to 12.5 percent despite higher recent growth (3.4 percent in 2015–2020 vs.1.7 percent in 2010–2014). Cash crops fared better. The production of coffee, tea, cotton, and tobacco rebounded, with an average growth rate of 6 percent during 2015–2019 compared to –1 percent in 2010–14. Agriculture has been hit partly by volatile weather patterns, but man-made fac- tors have been instrumental too. With increasing population pressure, agricultural land in Uganda increased by 15 percent between 2000 and 2020. However, only 7,000 hectares (1.2 percent of the estimated irrigation potential of 600,000 hect- ares) of cultivated land is under formal irrigation, largely focusing on cash crops. Most small-scale food crop farms lack improved farming practices and technol- ogies (e.g., irrigation, enhanced inputs) to manage weather variability. Addition- ally, with ineffective government interventions like the distribution of free inputs under the Operation Wealth Creation (OWC) program, problems with counterfeit and adulterated fertilizers and chemicals, and inadequate input market regulations, including seed certification, economic performance in agriculture remained sup- pressed. As a result, agriculture has made modest contributions to growth and provided unreliable supply for agro-processing. Manufacturing grew rapidly in the 2000s, but could not maintain its momentum in recent years. The sector grew rapidly at 6.1 percent between 2001 and 2005 and 7.3 percent per year between 2006 and 2010, annually. However, with slowing re- forms, this growth decelerated significantly to an average of 5.3 percent between 2011 and 2015 and 4 percent since then. Agro-processing represents the majority of Uganda’s manufacturing output, with food processing alone accounting for 40 Growth, trade, and transformation percent, led by the sugar, coffee, and tea subsectors. Manufacturing also includes construction materials (steel fabrications, bricks, and cement), and chemicals (in- dustrial and consumption) for the domestic and neighboring country markets. More recently, the COVID-19 pandemic has driven a sharp deceleration in eco- nomic activity. Growth in the Ugandan economy slowed down from 6.4 percent in 2019 to 3 percent in 2020, which was driven by sharp contractions in manufactur- ing and services. Although growth rebounded strongly to over 13 percent in Q4 of FY21, with a more severe second COVID-19 wave and related lockdown measures, the recovery into early FY22 has likely stalled. At 3.4 percent, growth in FY21 contin- ued to be below pre-COVID projections of over 6 percent. 46 Demography: Slowing, But Still Growing Too Fast In Uganda, nothing has grown faster than the population in the last few decades. Uganda’s population was estimated at 23.7 million in 2000. By 2020, it was estimat- ed to be nearly double that number, 45.7 million. In annual terms, this constituted one of the fastest growth rates in the world (3.4 percent, annually), and it is driven primarily by a widening wedge between stubbornly high fertility rates and rapidly shrinking mortality rates. Between 2000 and 2020, most mortality rates were halved in the country, but the fertility rate is yet to catch up with this downward trend. Infant mortality rates decreased from 87.4 cases per 1000 live births to 33.4 cases, and the crude mortality (death) rates for the whole population fell from 14.4 to 6.7 per 1000 per- sons. As a result, the life expectancy of a Ugandan newborn increased dramatically from 46.2 years to 63.4 years in just two decades—an excellent gain. The problem is that fertility rates have not fully caught up with this trend. Despite a sizable de- crease in fertility, from 6.9 to 4.8 births per woman, the Ugandan fertility rate re- mains one of the highest in the world. The growth of the Ugandan population took place broadly across the country, and the population density increased in urban and rural areas alike. Figure 15 shows the spatial progression of demographic growth between 2000 and 2020 in three panels. While the population expansion in major urban centers like the Kampala-Jinja-Mbale axis is obvious, the growth was broad-based with significant expansion in the Northern region and in southern Uganda as well. Although the sparse population is not marked on the map, population growth is not limited only to urban centers in various parts of the country. Worldwide Governance Indicators (WGI) statistics show that the rural population in Uganda grew from 20.2 million to 34.3 million over the same period. Thus, despite an increase in arable land (about 17.5 percent), the rural population density increased from 3.8 persons per hectare of arable land to 4.9 persons per hectare in the same period. This increase is shown by the expansion of the green-shaded areas (higher density) against the yellow shaded areas (lower density) in Figure 15. With delayed fertility transition, the Ugandan population is projected to grow significantly in the decades to come. The country’s population is estimated to double again by 2050. However, according to UN Population Division estimates, in the next few decades, the delayed fertility transition will gain momentum, with fertility decreasing faster than mortality. Between 2020 and 2050, the fertility rate is expected to gradually slow from 4.8 births per woman to 2.7 births per woman. In the same period, the mortality rate will decrease from 6.7 to 5.1 per 1000 persons. With slower reductions in mortality, life expectancy at birth will only increase by Introduction about 7 years, arriving to 70.3 years in 2050. 47 Figure 15 . The spatial distribution of the Ugandan population In 2000, Uganda’s total population was 23.6 million, and the rural population density was 3.8 persons per hectare of arable land. N High Low 100 km In 2010, the total population increased to 32.4 million and the rural population density to 3.9 persons per hectare of arable land. N High Low 100 km Growth, trade, and transformation In 2020, the total population increased to 45.7 million and the rural population density to 4.9 persons per hectare of arable land N High Low 100 km Source: WorldPop (country total UNPD adjusted). 48 Figure 16 . The age distribution of the Ugandan population In 2000, the working-age population (15–64) in Uganda was 48 percent of the population; the young age population (0–14) was at 50 percent. Population, 2000 10 800 9 Euclidian Distance = 14.0 700 8 Uganda, Millions 600 World, Millions 7 6 500 5 400 4 300 3 200 2 1 100 0 0 0-4 5-9 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85-89 90-94 95-99 100+ Age Group Uganda = 23.7 million World (RHS) = 6.1 billion In 2020, the share of working-age population (52 percent) exceeded that of the young age population (46 percent). Population, 2020 10 800 9 Euclidian Distance = 15.1 700 8 Uganda, Millions 600 World, Millions 7 6 500 5 400 4 300 3 200 2 1 100 0 0 0-4 5-9 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85-89 90-94 95-99 100+ Age Group Uganda = 45.7 million World (RHS) = 7.8 billion By 2050, 65 percent of the population will be at the working-age, and 31 percent at the young age (much closer to the world distribution of population by age). Population, 2050 10 800 Euclidian Distance = 9.9 9 700 8 600 Uganda, Millions 7 World, Millions 6 500 5 400 4 300 3 200 2 1 100 0 0 0-4 5-9 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85-89 90-94 95-99 100+ Introduction Age Group Uganda = 89.5 million World (RHS) = 9.7 billion Source : UN Population Division estimates. Notes: Euclidian distance measures the gap between the two series (sum of squared gaps). 49 Demographic transitions come with opportunities but also with important challenges. A recent World Bank (2021) study suggests that Uganda is entering the early stage of its demographic transition. With the right conditions and policies, the country can capitalize on this transition. In demographic transitions where fertility and mortality decrease, the share of economically active people in the population would increase (first demographic dividend). This may be particularly prominent if lower fertility translates into higher female labor force participation. With lower de- pendency ratios (e.g., fewer children and elderly per worker), Ugandans can invest more in education and healthcare, boosting human capital and income. In addi- tion, with higher income, national savings can increase, which, in turn are invested in productive ways, further boosting economic prospects (second demographic dividend). Uganda’s young age dependency ratio is projected to decrease in the coming decades and this can provide opportunities. In 2000, the working-age population (ages 15–64) in Uganda was about 48 percent of the population (Figure 16), while youth and children (ages 0–14) were about 49.7 percent. By 2020, the working-age population increased to 52 percent, and the younger ages fell to 46 percent. This trend is projected to accelerate going forward: by 2050, about two thirds of the population (65 percent) is projected to be in the working-age group and slightly less than a third (31 percent) in the younger age groups. Thus, other things being equal, the burden of economically inactive people on the economically active one is scheduled to decrease, creating space for better human capital and physical capital accumulation under the right conditions. However, without appropriate policies in place, the transition can lead to signif- icant economic distress. With a rapidly increasing working-age population comes the need for economic opportunities. By 2050, the working-age population will be about 2.5-fold larger than it is now, and recuperating a demographic dividend will depend on finding productive jobs and economic opportunities for them. The next section discusses in detail the expansion of Uganda’s arable land at the expense of the country’s natural capital stock since the 1990s, which absorbed some of the population pressure. Given the fragile and exhaustible nature of natural assets, this Growth, trade, and transformation will not be an option going forward for much longer. Moreover, the current eco- nomic framework, where most workers are engaged in subsistence agriculture or in low-productivity self-employment activities, cannot generate sufficient jobs for the forthcoming working-age population bulge in Uganda. The economic opportu- nities for the country’s booming working-age population will need to come from policies that eliminate obstacles to private business enterprises; support Ugan- dans with better education, healthcare, and infrastructure; and provide an inclusive, transparent, stable, and predictable business environment for all. 50 Nature: Increasingly Fragile Uganda has one of the best natural assets in the world. Uganda has a total area of 241,551 square kilometers; about 83 percent of this area is land and 17 percent open water and swamps. According to the latest Wealth Accounting and the Valuation of Ecosystem Services (WAVES) report (UBOS 2020), in 1990 the largest land cover class was small-scale farmlands, with 35 percent of national land cover equivalent to 8.4 million hectares (Figure 17). Grasslands had the second-largest cover, with 5.1 million hectares. Woodlands had the third-largest cover, with nearly 4 million hectares (16 percent), while open water was at 3.7 million hectares or 15 percent of national cover. Uganda’s natural assets are the backbone of a strong nature-based tourism industry and are some of the most biodiverse in Africa, including 54 per- cent of all mountain gorillas in the world and 50 percent of Africa’s bird species (some endemic to the country). The country’s forests provide global public goods, by sequestering carbon; regional public goods, by protecting watersheds critical for livelihoods and jobs; and private economic benefits, by providing over 80 per- cent of domestic energy through provision of wood fuels. Uganda’s population boom and the expansion of economic activity in recent decades have taken their toll on the country’s natural environment. With in- creasing pressure from rapid population growth, including in rural areas, and eco- nomic growth, Uganda’s natural assets face an increasing risk of depletion. Since 1990, large swaths of the country’s forests and bushlands have been repurposed for agriculture. In 2015 there was a woodland area of 1.2 million hectares (Figure 17), 69 percent less than the 4 million hectares in 1990. In contrast, in the same period the small-scale farmlands expanded by 22 percent from 8.4 million hectares to 10.3 million hectares. The grassland covers almost recovered from a 30 percent de- crease between 1990 and 2010 to just 0.4 percent less than the 1990 cover in 2015. But this was largely at the expense of bushlands, possibly because the grasslands are used as foraging area for livestock and the increasing wildlife numbers. Overall, bushlands decreased by 41 percent while wetlands decreased by 10 percent be- tween 2010 and 2015. The future sustainability of Uganda’s economic growth will require a trans- formation in natural capital management. According to the World Bank (2020), Uganda’s continuing loss of natural assets is among the highest in Africa. Liquidat- ing natural capital for consumption and the purchase of other assets is simply not viable in the medium run, for several reasons, including that these are bounded resources and soon the rapid depletion rate will begin limiting economic growth it- self (i.e., through negative impacts on tourism). Thus, Uganda will need to transform from a heavy reliance on the expansion of low-productivity small-scale agriculture to creating value added through the accumulation of physical, human, and intangi- Introduction ble (i.e., know-how) capital. 51 Figure 17. Land cover change in Uganda, 1990–2015 Growth, trade, and transformation Source: Uganda Bureau of Statistics––UBOS (2020). The Outline of This Study The study will first analyze the drivers of growth and structural transformation in Uganda. Chapter 1 focuses on the drivers of economic growth (i.e., productivi- ty growth by sectors and the reallocation of labor across them) over the last two decades, which differ significantly between the 2000s and 2010s. This is done by using national accounts statistics. Next, the analysis turns to investigating the fac- 52 tors that affect the country’s structural transformation patterns, which fall short of desired levels. To do this, the study employs: • An empirical analysis of sectoral and spatial transition patterns among Ugandans, by using survey data from Uganda National Household Surveys (UNHS) and Uganda National Panel Survey (UNPS), including inquiries about the factors that may affect such transitions as well, • A simulation-based analysis of structural transformation in Uganda, by using a general equilibrium model of structural transformation, which is calibrated by using national accounts-based statistics (among others) to dissect the specific role played by various factors (i.e., productivity growth, land, prices, and adjustment costs) in driving growth and labor reallocations across sectors. Next, in Chapter 2 the study focuses on the role of international trade in growth and transformation of Uganda. For many economies, which suffer from small domestic markets and distortions, trade can help allocate productive factors to their most efficient uses. Uganda is not an exception. With this in mind, the second chapter provides a summary of Uganda’s export and import performance over the last two decades. Then it provides analyses that comprise: • Firm-level drivers of Uganda’s international trade by using customs data from the Exporter Dynamics Database (EDD) to analyze the size, survival, and domestic value added of the Ugandan exporters, which also distinguishes between pure exporters and exporters who also import; • The role of trade policy in structural transformation of Uganda, by using the structural transformation model developed in the previous section, to simulate the effects of a hypothetical decrease in food prices (driven by tar- iff reductions) on growth and labor allocations across sectors; • The role of trade policy in inequality in Uganda, by using a discrete choice (partial equilibrium) model drawing from household surveys to characterize the income and consumption share of agricultural goods by income deciles, and simulate the income-differentiated impact of a decrease in food prices (driven by tariff reductions). Finally, in Chapter 3 the analysis takes a forward-looking approach to discuss future opportunities and policies in Uganda. In light of the analytical findings re- garding the dynamics of the Ugandan economy, the analysis then considers a few key issues that are likely to provide opportunities. These include the following: • Regional integration and trade, with emphasis on the advent of the AfCFTA and its implications for Uganda; Introduction • Hydrocarbons, developments to date, and projections for oil production and fiscal revenues using a fiscal model of the sector; • Tourism, challenges faced by the sector and several proposed policy actions to overcome those challenges; 53 Figure 18. The analytical outline of the report Spatial & Sectorial Analysis of Discrete choice sectorial analysis & exporter model-inequality transition Fiscal model dynamics analysis analysis for oil Chapter 1 Chapter 2 Chapter 3 Drivers of growth The role of Future and stuctural international opportunities transformation trade and policies Structural transformation model Source: World Bank. • Digital transformation, how better digital connectivity can boost transfor- mation in Uganda where physical connectivity is costly; • Climate change and environment, both the challenges and the opportuni- Growth, trade, and transformation ties in Uganda, whose natural assets have value beyond the country’s bor- ders. The report is organized to present content with increasing technical complex- ity and time commitment. The key message at the beginning provides a bird’s- eye view of the main results. The executive summary provides a systematic and non-technical overview of the analysis and the main findings of the study. The main body of the report provides a detailed exposition of the main policy problems, analysis, and results, with some discussion of the analytical methodologies and technical aspects of the arguments as needed. Finally, more technical details about underlying estimations, modeling assumptions and structures are provided in the appendices. 54 References Pritchett, Lant. 2019. “There is only one poverty strategy: (broad-based) growth.” https://lantpritchett.org/there-is-only-one-poverty-strategy-broad-based- growth-part-i/. Uganda Bureau of Statistics. 2020. Towards ecosystem accounts for Uganda. Kam- pala: UBOS. World Bank. 2020. “Uganda Natural Capital Accounting, Environment and Climate Change Programmatic ASA.” Report No. AUS0001830. Washington, DC: World Bank Group. World Bank Open-Knowledge Repository. 2021. Tackling the Demographic Chal- lenge in Uganda. www.openknowledge.worldbank.org. Washington, DC: World Bank Group. Introduction 55 Growth, trade, and transformation 56 Chapter 1. Growth and transformation in Uganda CHAPTER 1. GROWTH AND TRANSFORMATION IN UGANDA A fter rapidly growing in the 2000s, a slowing economy in recent years has revealed Uganda’s Gordian knot: about two thirds of its workforce pro- duces less than a quarter of its GDP. The discussion in the introduction showed that, with several conducive factors like post-conflict recovery, favorable external conditions, and a dynamic reform momentum, the Ugandan economy grew rapidly in 2000s. However, as this initial momentum phased out afterward, it has become evident that the country is in need of new sources of growth. The population is still growing rapidly, natural assets remain fragile and bounded, and almost two thirds of the country’s workforce are employed in relatively low produc- tive activities in agriculture, producing less than a quarter of all value added in the country. Therefore, it is only natural that the country should seek mechanisms to reallocate its workforce toward more productive activities outside agriculture. To support this pursuit, this chapter aims to provide the analytical underpinnings of future policies that promote structural transformation in Uganda. This chapter analyzes the nature of Uganda’s recent growth experience with a Chapter 1. Growth and transformation in uganda special emphasis on the drivers of structural transformation. The analysis first focuses on the nature of productivity growth in the Ugandan economy between 2000 and 2020 (or 2019, when data limitations or concerns about COVID-19-driv- en factors distorting the structural analysis prevailed). This includes a discussion on the role of labor reallocation toward more productive activities in explaining the productivity growth––or lack thereof––across sectors, using national accounts data. As discussed in detail below, this labor allocation problem lies at the core of many developing countries’ income gaps with the rest of the world, along with low productivity in agriculture. Next, the chapter analyzes the factors that may have hindered a more rapid transformation in the Ugandan economy, by using micro data from household surveys, including spatial and sectoral transitions (i.e., internal migration and job switching across sectors). Where data does not shed sufficient light, the analysis employs a general equilibrium structural transformation model, calibrated with national accounts statistics, to assess the likely mechanisms be- hind the observed trends. 59 What Has Driven Growth in Uganda? Economic growth is ultimately driven by productivity gains in an economy. The analysis so far has documented the contributions of different sectors to GDP growth in Uganda from an accounting perspective, i.e., changes in the economic activity were decomposed by corresponding sectors. However, the sector-specific contributions themselves can be conceived as the results of two underlying mech- anisms: (i) changes in per-worker productivity in a given sector, and (ii) realloca- tion of workers across sectors. Such a decomposition can help analyze changes in Ugandans’ productivity, which is the essential driver of Ugandans’ income. Between 2000 and 2010, within-sector productivity growth and the reallo- cation of labor across sectors reinforced each other in promoting growth in Uganda. Figure 2 shows the dynamics of total labor productivity growth in Uganda, using a canonical decomposition method. Specifically, this method distinguishes between (i) productivity growth stemming from labor productivity growth within sectors (labeled ‘within-sector’); (ii) productivity gain in the form of labor moving from low productivity growth sectors to high productivity sectors (labeled ‘across- sector’); and (iii) productivity gain from labor moving to sectors where labor pro- ductivity increases (labeled ‘dynamic’). The across-sector factor is positive when workers move to sectors with higher productivity regardless of its growth, and the dynamic factor is positive only if workers are moving to sectors with positive pro- ductivity growth. In Uganda, within-sector productivity growth contributed 2.6 per- centage points to labor productivity and across-sector labor allocation added 0.3 percentage points, annually, in the first half of the 2000s. Within-sector productiv- ity growth accelerated in the second half of the decade, reaching 4.3 percentage points, while across-sector dynamics remained about the same. Thus, productivity growth was largely driven by within-sector productivity growth and not by struc- tural transformation. All three major sectors exhibited noticeable growth in within-sector produc- tivities. Productivity growth in services (within-sector) generated half of the ag- Growth, trade, and transformation gregate productivity gain in the first half of the 2000s.2 In the second half of the decade, the same performance was exhibited by the industry sector, where with- in-sector productivity growth more than doubled. In agriculture, within-sector productivity growth was relatively more modest but nonetheless positive, adding about 0.6 percentage points per year to aggregate productivity over the decade. Labor reallocation from agriculture, and to a limited extent manufacturing in the first half, toward services drove the main across-sector dynamics throughout the 2000s. However, these reallocation effects remained limited compared to produc- tivity dynamics within each sector. 60 Figure 19. The drivers of growth in value added per worker In Uganda, growth has largely been driven by within-sector productivity dynamics… a. Value added per worker, aggregate 2000 - 2005 2005 - 2010 2010 - 2015 2015 - 2019 -2 0 2 4 6 Growth in value added per worker in % Within-sector Across-sector Dynamic …which were shared by all three sectors in the 2000s, but which changed after that as agriculture registered a reversal in productivity growth. b. Value added per worker, by sector Agriculture 2000 - 2005 Chapter 1. Growth and transformation in uganda 2005 - 2010 2010 - 2015 2015 - 2019 Industry 2000 - 2005 2005 - 2010 2010 - 2015 2015 - 2019 Services 2000 - 2005 2005 - 2010 2010 - 2015 2015 - 2019 -2 -1 0 1 2 3 Annualized growth in value added per worker in % Within-sector Across-sector Dynamic Source: World Bank Group Jobs and Structural Change Tool (accessed in September 2021), which uses WDI data for labor Source: World allocations Bank Group (based on ILO Jobs Structural Change and estimates). modeled Tool (accessed For a discussion in September on differences 2021), across which various uses labor WDI data market for labor estimates alloca- in Uganda, tions 1. on ILO modeled estimates). For discussion on differences across various labor market estimates in Uganda, see Box 1. (based see Box 61 Box 1. Employment statistics in Uganda In Uganda, several surveys cover labor market statistics. Annual labor force sur- veys (LFS) have the most comprehensive labor module based on the 7-day recall, and were conducted in 2012, 2016/17, 2017/18, and 2018/19. The Uganda National Household Survey (UNHS) is the main source of monetary poverty measure in Uganda and, besides a detailed consumption module, it also com- prises a labor module similar to the LFS. It was conducted in 2012/13, 2016/17, and 2019/20 in the last decade. The Uganda National Panel Survey (UNPS) uses a slightly different labor module and provides longitudinal coverage. It was conducted in 2010/11, 2011/12 and 2013/14, 2015/16, 2017/18, 2018/19, and 2019/20. In addition, the International Labor Organization (ILO) provides mod- eled estimates for key labor market indicators, which are also reported in World Development Indicators (WDI). The presence of multiple sources, with differences in methodologies, creates challenges. Most notably, indicators are sometimes not comparable between surveys or over time, and such comparisons can lead to different estimates. For example, while the UNPS showed a 6-percentage-point increase in em- ployment ratio (the share of employed people in the working-age population) between 2013 and 2019, the UNHS showed a 14-point reduction between 2012 and 2019, widening the gap between employment levels across the two sur- veys to about 26 points (up from 6 points earlier), whereas the ILO estimates have shown a stable trend. More importantly, the sectoral employment shares in agriculture also exhibit differences in the second half of the last decade. In 2019, for example, ILO and UNHS estimate it at 72 percent. The UNHS has likely picked COVID-19 effects, which pushed people to agriculture, and changes in the labor module ques- tionnaire and omission of the second quarter during the data collection has also played a role due to seasonality, whereas ILO estimates show the extrap- olation of the previous trend. In comparison, the last LFS showed 64 percent in Growth, trade, and transformation 2018 and the UNPS showed 61 percent in 2019. For the purposes of this study, following the World Bank Jobs and Structural Change tool, ILO estimates are used. This approach has facilitated comparisons across countries and provid- ed long annual series, which are used to calibrate simulations performed later in the report. 62 Box 1. (continue) Figure 20. Estimates for the employment share of agriculture Estimates for the employment share of agriculture diverge over the last half decade. While ILO and UNHS estimates show around 72 percent in 2019 (the latter likely reflecting COVID-19 effects partially captured in the sample), UNPS shows 61 percent, and LFS show 64 percent (in 2018). Employment Share of Agriculture 74 73 % of working/employed population 72 72 72 72 72 72 72 72 72 70 70 70 68 68 67 67 66 66 66 66 64 64 62 62 61 60 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 WDI (modelled), 15+ UNPS (microdata), 14-64 UNHS (microdata), 14-64 LFS (UBOS publication), 14-64 Chapter 1. Growth and transformation in uganda Source: UBOS, WDI and WB staff calculations using UNHS and UNPS data. After 2010 within-sector productivity has gradually slowed down and labor re- allocation has reversed some of the productivity gains. Within-sector produc- tivity growth in the Ugandan economy decreased to 3.7 percent between 2010 and 2015 and to a meager 1.1 percent after that (Figure 19). Labor allocation across sectors did worse, especially in the first half of the decade. The static and dynam- ic reallocation effects together reduced total labor productivity growth by about 2.5 percentage points annually between 2010 and 2015, and about 0.1 percentage points after that. Together, the slowdown in within-sector productivity growth and the reversal of labor allocation effects have broken the momentum in Ugandan pro- ductivity. The within-sector productivity slowdown was shared by all three major sectors, but that in agriculture stands out. The slowdown in productivity growth within in- dustry and services was significant, especially in the second half of the decade. Nevertheless, they registered positive growth in within-sector productivity. Un- like these, a decreasing within-sector productivity in agriculture pulled total labor 63 productivity down by 1 percentage point in 2010–2015 and about 0.2 percentage points in 2015–2020. Despite a decreasing productivity in agriculture, both indus- try and services lost employment shares to the agricultural sector, which reduced total labor productivity growth by 0.8 percentage points in the former case, and 2.1 percentage points in the latter, in the first half of the decade. That is, given lack of opportunities elsewhere and low entry barriers in agriculture, the latter became the absorbing sector for new labor entrants despite low productivity and earnings. These effects slowed by the second half of the decade, but this adverse trend has not been reversed yet. Low agricultural productivity stems from two sources: low average technical efficiency at the farm level and persistent underemployment of the rural popu- lation. The national accounts-based indicators above show relatively low and stag- nant agricultural productivity patterns in Uganda. However, micro studies that rely on surveys show that at least a portion of the productivity gap between agriculture and other sectors stems from measurement problems. For example, Christiaensen and Demery (2017) argue that accounting for production for own consumption, and measuring output per hour of work rather than the stock of labor (hours of work per person are greater in non-agricultural sectors), can account for more than half of the productivity gap between agriculture and other sectors. Therefore, the idea here is that the across-sector productivity gaps observed in output per worker per year reflect gaps in employment levels rather than gaps in returns per hour worked. Nevertheless, the low hours themselves may be driven by lower farm productivi- ties, and even accounting for those factors still leaves Uganda with the fact that a large chunk of the country’s workforce remains occupied in near-subsistence level agricultural activities. Improving productivity across the board and allocating labor toward higher productivity tasks will need to be an essential characteristic of Uganda’s future development. We turn to this issue next. Why Does Structural Transformation (Still) Matter for Uganda? Growth, trade, and transformation To achieve middle-income living standards, Uganda will need to promote high- er productivity growth within sectors and allocate its labor force to more pro- ductive uses. Figure 21 shows value-added shares in GDP, employment shares, and value added per worker by sector in Uganda, compared to all country-year observations since 1991. For each indicator, the blue boxes show the interquartile range (i.e., values between the first and the fourth quarters of the distribution lie within the boxes), with the dark blue line inside the box showing the median value. Uganda’s 2019 values are marked with a circle. The value-added share of sectors in Uganda were typically close to the global median (Panel a). However, the share of agricultural employment in Uganda was one of the highest in the world and that in services was one of the lowest. As a result, although the per worker productivity in Ugandan services and manufacturing remained higher than the global median, that 64 in agriculture was significantly lower than other countries, putting a significant drag on economic growth in the country. The slow speed of Uganda’s structural transformation significantly drives its income differences with other countries. Box 2 summarizes the idea of structural transformation for promoting economic development. Economists have long ar- gued that a large share of income shortfall in low-income countries is driven by the fact that agricultural productivity is very low in those countries, and they allocate too large a share of their workforce to agriculture. To elaborate on this point, Box 2 also compares Uganda with Vietnam, a success story in agriculture-driven devel- opment. About 35 percent of this difference is because Vietnamese workers are more productive in agriculture (75 percent more productive). In comparison, pro- ductivity differences are much lower in manufacturing (22 percent) and services (32 percent). The most prominent factor, however, is the differences in labor allo- cations across sectors, which explains about 45 percent of the total productivity gap between the two countries. Note that sector dynamics across countries are driven by numerous coun- try-specific factors. It is sometimes suggested that when agricultural productivity increases, the sector will shed labor. This is not necessarily true for all countries. For example, in a small open economy where domestic prices of tradable goods are fixed by international prices, an increase in agricultural productivity does not lead to a reduction in agricultural prices or wages. Thus, other things being equal, there is no reason for labor to move out (Matsuyama 2019). For Uganda, there are other concerns, too. An important one is the significant overlap between spatial transitions (e.g., migration) and sectoral transitions (e.g., moving out of agriculture), which may further hinder labor reallocation, and thus needs to be investigated next. Chapter 1. Growth and transformation in uganda 65 Figure 21. Benchmarking sector structures: Uganda vs. all countries The share of agriculture in total value added in Uganda in 2019 (circle) was at the median level (blue line inside the box) when compared to other country-years. a. Total value added 8 Share in total value added 6 4 2 0 agriculture industry services However, the employment share of agriculture in Uganda was one of the highest in the world, and that in services one of the lowest. b. Share in total employment 8 Share in total employment 6 4 2 0 agriculture industry services Labor productivity in agriculture is significantly lower in Uganda compared to other coun- tries. Growth, trade, and transformation c. Value added per worker in PPP 2017 constant prices (log scale) 11 Per worker value added, 10 9 8 7 econ.-wide agriculture industry services UGA - 2019 Source: Data from World Bank Group Jobs and Structural Change Tool (accessed in September 2021), covering all country-years since 1991. 66 Box 2. What is structural transformation and why does it matter? Development is essentially a never-ending process of accumulating produc- tive factors and finding better ways to organize them for productive purposes. In fact, even factor accumulation responds to organizational quality at both micro and macro levels. For example, firms can gain access to more finances and attract better workers if they organize their operations better. Similarly, economies can incentivize higher savings and human capital accumulation if they effectively allocate these resources to areas where they are most needed and thus most productive. As such, all countries are developing, always. For example, advanced economies are currently being transformed through the digital economy, which boosts trade in previously non-tradable services. Structural transformation is the terminology used to describe the reallocation of productive factors (often labor) across sectors like agriculture, manufactur- ing, and services. In earlier analyses, e.g., the Lewis model (Figure 22), invest- ments increase the marginal product of labor in the modern sector (. Thus, as labor responds to higher wages () by migrating, the modern sector gradually absorbs the subsistence workers over time (. This process can stabilize when a diminishing labor force pushes up the marginal product (and the wages—) of subsistence workers sufficiently high. Figure 22. A simple model of structural transformation based on the Lewis Model Chapter 1. Growth and transformation in uganda Traditional models of structural transformation emphasize the role of investments in boosting the marginal productivity of labor in modern sectors. This, in turn, widens the wage gap between sectors, and attracts workers from the subsistence sector. Source: Basu (2003). 67 Box 2. (continued). This concept has remained highly popular among economists working on de- velopment challenges. The theoretical clarity of the idea, and its intuitiveness, may have contributed to such popularity. However, it also shines empirically. More recent studies like Caselli (2005) and Restuccia, Yang, and Zhu (2008) suggest that the bulk of the income differences across countries can be ex- plained by two simple facts. First, low-income countries are much less pro- ductive in agriculture relative to advanced economies. Second, low-income countries allocate a much higher share of their labor force to agriculture than more advanced economies. Thus, by conjuncture, if developing countries were able to reallocate their la- bor to other sectors and increase labor productivity in agriculture, they would close much of the income gap with advanced economies. To test the validity of this idea for Uganda, Figure 23 decomposes the difference between average labor productivity in Uganda with that of Vietnam, a LMIC who implemented a big push successfully to become a LMIC. The share of Vietnamese agriculture in the economy is commensurate with other countries in its income group. Figure 23. Why are Ugandans not as productive as the Vietnamese? Vietnamese workers are twice as productive as Ugandan worketrs. More than three- quarters of this gap is driven by Uganda’s excessive labor share and low productivity in agriculture. Productivity gap in manufacturing does not play a major role. Drivers Of The Productivity Gap 100 Lower productivity in manufacturing 3.6 90 Lower productivity Share of the difference (%) 16.5 in services Growth, trade, and transformation 80 70 Lower productivity 60 34.7 in agriculture 50 40 30 Differences in 20 45.3 labor allocation 10 0 Source: WDI and UNU-WIDER Economic Transformation Database, World Bank staff calculations. Notes: Manufacturing in this figure includes mining and construction in ISIC-4 classification for simplicity. 68 Box 2. (continued). In aggregate, the average Vietnamese worker is 2.1 times more productive than the average Ugandan worker. However, when broken down, there are major vari- ations across sectors. For instance, a Vietnamese worker is only 22 percent more productive than her Ugandan counterpart in manufacturing, and only 32 percent in services. In contrast, a Vietnamese worker in agriculture is about 1.75-fold more productive than her Ugandan counterpart. Overall, only 3.6 percent of the gap between the Vietnamese and the Ugandan workers’ productivity is driven by the productivity gap between the Ugandan and Vietnamese manufacturing sector. The bulk of the gap is accounted for by the differences in labor allocation—Uganda’s labor share in agriculture is 1.75- fold greater than that of Vietnam, which explains the 45.3 percent in the gap. The second most important factor is the productivity gap in agriculture, which explains the 34.7 percent in the gap. However, any inference based on this exercise should consider two important caveats. First, this exercise only shows how one country’s productivity differs from another’s, roughly. It does not show why. Sector-specific productivity and labor allocations are mutually dependent. For example, productivity changes may affect labor allocations, and vice versa. However, they can also be de- termined by other (confounding) factors like trade, infrastructure, and several institutional factors, which are considered in more detail below. Second, aggre- gate labor productivity measures can miss important nuances. For example, Christiaensen et al. (2018) show that, at the micro level (i.e., output per hour of Chapter 1. Growth and transformation in uganda labor), productivity gaps between sectors are not as large in Africa. Therefore, a low agricultural productivity may reflect factor allocation problems at the sec- tor level rather than problems within a production unit. This can be generalized: manufacturing productivity can suffer if supply chain problems limit otherwise productive firms. Similarly, with large overhead costs firm productivity may be low even with productive individual plants. Overall, sector-based productivity differences and labor allocations can pro- vide a natural entry to analyzing growth in low-income countries, when consid- ered with reasonable caution. The Role of Migration in Job Transitions With few non-farm jobs in rural areas, moving out of agriculture and internal migration go hand in hand in Uganda. Ugandan administrative statistics do not directly measure job transitions (e.g., moving from one sector to another) and mi- gration at the national scale for recent years. However, survey data can be used to shed some light on the matter. For example, the UNHS 2016/17 captures the internal 69 migration experience of about 15,000 Ugandan households in the previous 5 years and their most recent occupation. Using this information, Figure 24 compares the current occupations of Ugandan heads of households who migrated and those who did not. In urban areas, both rural-urban and urban-urban migrants were less likely to work in agriculture compared to non-migrants. Interestingly, the same was also true in rural areas: both rural-rural and urban-rural migrants were less likely to work in agriculture than rural non-migrants. The drivers of such patterns cannot be de- termined by using only survey data: for example, it is not clear if migration is largely undertaken by non-agricultural workers (e.g., more educated) or if Ugandans are more likely to find non-agricultural jobs once migrated. However, notwithstanding the causality issue, these trends highlight the importance of internal migration for structural transformation in Uganda. Despite its importance for transformation, rural to urban migration is low in Uganda by many standards. According to UNHS 2016/17, only about 9 percent of household heads reportedly migrated between 2012 and 2017.3 Figure 25 shows the distribution of these households and population with migrant heads according to their migration patterns. Surprisingly, the migration from rural areas to urban areas is much less than expected. Rural-urban migration comprises only 16 percent of all migrants—the smallest group by this classification. Those who moved between rural areas or between urban areas were roughly between a third and a quarter of all migrants. Interestingly, movements from urban to rural areas were similarly large and exceeded the rural-urban migration, and urban areas supplied 55 percent of all internal migrants. Similar patterns also featured in other surveys. For example, the School to Work Transition Surveys (SWTS) shows that only 23 percent of youth migrants moved from rural to urban areas (Boutin 2016). Internal migration patterns exhibit heterogeneity across regions and between rural and urban areas. Table 2 and Table 3 show migration flows by regions, disag- gregated by rural and urban areas, and evaluated as a share of migrants at desti- nation (former) and at source (latter). The Central region, which is the most urban- ized and wealthiest region in Uganda, was the top destination region, absorbing 41 percent of all internal migrants. The poorer and more rural Northern region was in Growth, trade, and transformation second place, absorbing 25 percent of all internal migrants, followed by the West- ern and Eastern regions. Internal migration is largely localized. Internal migrants from rural areas within each region were more likely to stay in rural areas of the same region compared to urban migrants who were less likely to stay in urban areas. About 85 percent of all internal migrants from the Northern region migrated within the Northern region. In contrast, only half of internal migrants from the Eastern region migrated within that region, while 31 percent moved to the Central region, probably due to its relative proximity and better road connections to get there. About 69 percent of all internal migrants in the Central region came from somewhere else within the Central region. For the urban areas of the Central region, more than 70 percent of migrants were either from the same urban areas or Kampala. In fact, Kampala was the key source of migrants to urban areas of the Central region, while most migrants to Kampala came from other urban areas within the Central region. 70 Figure 24. Employment sector of household heads, percent In urban areas of Uganda, those who did not migrate within the last 5 years had a greater tendency to work in agriculture (21 percent) than migrants––only 14 percent of rural to urban migrants and 7 percent of urban to urban migrants worked in agriculture. Urban Areas 79 86 93 21 14 7 Urban Non-Migrant Rural To Urban Urban To Urban Migrant Agriculture Non-Agriculture The same pattern holds even in rural areas. About 69 percent of non-migrants in rural areas worked in agriculture. In comparison, 37 percent of urban to rural migrants and 58 percent of rural to rural migrants held agricultural jobs. Chapter 1. Growth and transformation in uganda Rural Areas 31 42 63 69 58 37 Rural, Non-Migrant Urban To Rural Rural To Rural Migrant Agriculture Non-Agriculture Source: World Bank estimations using UNHS 2016/17. 71 Figure 25. Internal migrants in Uganda by area of origin and destination Rural to urban migrants (black bars) comprise only a small share of all internal migrants in Uganda (according to UNHS data). Migrantion Statistics 28 33 27 29 29 26 16 12 Households with migrant head Population with migrant head Rural to urban Urban to urban Urban to rural Rural to rural Source: World Bank estimations using UNHS 2016/17. Kampala is not only a destination for migrants, but it also serves as a prominent source for reverse migration. For rural areas in the Eastern and Western regions, 45 percent and 47 percent of all migration took place within the region, respective- ly. However, in urban areas of the same regions, the sources of internal migration were more diversified. Interestingly, Kampala was an important source of migrants to all regions. This trend is largely explained by migrants returning to rural areas after living temporarily in Kampala. Rural-urban migration in Uganda is low even compared to other countries. Ta- ble 4 compares migration patterns across several SSA countries, estimated by the World Bank (2021). Results are presented by sizes of destination cities in each coun- try, using various data sources. Uganda clearly exhibits the lowest share of migrants Growth, trade, and transformation in every urban category, with only 13 percent of the urban population comprising migrants in Uganda as compared to an average 33 percent in other countries. 72 Table 2. Migration flows evaluated at destination DESTINATION Central Eastern Northern Western destinations Rural Urban Kampala Rural Urban Rural Urban Rural Urban All Rural 22 7 13 10 6 6 2 10 5 10 Central Urban 16 26 41 7 13 6 2 1 4 13 Kampala 26 47 0 15 7 12 22 23 18 22 Rural 5 7 12 45 45 5 15 3 0 12 Eastern Urban ORIGIN 11 5 7 17 24 4 3 0 3 7 Rural 2 2 6 1 0 31 14 2 1 9 Northern Urban 1 0 3 1 2 21 38 1 0 7 Rural 11 3 11 5 0 13 2 47 39 15 Western Urban 6 3 8 0 1 3 2 13 27 6 All sources 100 100 100 100 100 100 100 100 100 100 Chapter 1. Growth and transformation in uganda Source: World Bank estimations using UNHS 2016/17. Notes: Table shows the share of migrants at their destination (columns) by their origin (rows). For example, 6 percent of all migrants in rural areas in the Central region come from the urban areas in the Western region (last cell before the totals in the first column). Numbers are population weighted. The last column (all destinations) shows the share of migrants from specific locations of origin at all destinations, and the last row (all sources) shows the sum of all migrant shares from all sources at a specific destination. Shaded cells refer to internal migration within a geographic unit. What Is Holding Back Rapid Transformation? Spatial or sectoral transitions are either promoted or hindered by destination (pull) issues, source (push) issues, and transaction costs (frictions). People of- ten move when differences between living conditions at destination and source locations/sectors are large enough (Figure 4). Those conditions include wage gaps between locations/sectors as well as differences in living costs (prices), amenities or publicly provided services like schools and electricity, and other economic, social, and cultural factors like age, gender, religion, and ethnicity. Benefits that are not cap- tured by wage information are also important; these include production of food not for marketing purposes but for domestic consumption (auto-consumption), which plays an important role in Ugandan subsistence agriculture as described below. Overall, the net gap between the two alternatives needs to be large enough, as the transition (e.g., migration or job switching) itself is a costly and often risky endeavor. 73 Table 3. Migration flows evaluated at source DESTINATION Central Eastern Northern Western destinations Rural Urban Kampala Rural Urban Rural Urban Rural Urban All Rural 37 13 8 11 2 12 1 13 3 100 Central Urban 22 36 19 6 3 10 1 1 2 100 Kampala 20 37 0 8 1 11 5 13 5 100 Rural 7 10 6 44 13 9 7 4 0 100 Eastern Urban ORIGIN 27 12 6 28 12 10 2 0 3 100 Rural 5 5 4 1 0 74 8 2 1 100 Northern Urban 2 1 3 1 1 63 27 2 0 100 Rural 12 3 4 4 0 18 1 41 17 100 Western Urban 16 10 8 1 1 9 1 26 29 100 All sources 17 18 6 11 3 20 5 13 6 100 Source: World Bank estimations using UNHS 2016/17. Notes: Table shows the share of migrants from a source (row) by their des- tination (columns). For example, 16 percent of all migrants from the urban areas in the Western region went to rural areas in the Central region (last cell before the totals in the first column). Numbers are population weighted. The last column (all destinations) shows the sum of all migrants from a specific source, and the last raw (all sources) shows the share of all migrants in Uganda at a specific destination. Shaded cells refer to internal migration within a geographic unit. Growth, trade, and transformation For example, finding a job at the destination may be difficult and not guaranteed or transportation costs may be prohibitively high. Therefore, decisions are made by comparing the living conditions between destination and source locations, net of transaction costs. The rest of this chapter analyzes the factors driving Uganda’s structural trans- formation, or lack thereof. In what follows, we follow the classification above (real income differentials, transition costs and frictions, and the factors that can affect these two) to analyze the likely drivers of the slow transformation in the Ugandan economy. Although data constraints impede a precise estimation of bottlenecks in Uganda, the effects of such bottlenecks can be observed indirectly. Systemat- ic estimations of various bottlenecks for growth and structural transformation are difficult even in data-rich countries—a well-known observation in the growth diag- nostics literature (Hausmann, Klinger, and Wagner 2008). 74 Table 4. Migration is considerably less prevalent in Uganda compared to other countries Working-age Small Large Small Cities: Big cities: population Towns: Towns: Total 100k–1000k (>1 million) (15–64 year old) 0–20k 20k–100k Migrant share of urban population Ethiopia1 (2013) 0.46 0.45 0.43 0.25 0.40 Tanzania 1 (2010) 0.20 0.18 0.36 0.53 0.32 Uganda2 (2016) 0.09 0.12 0.17 0.16 0.13 Ghana (2010) 0.26 0.23 0.25 0.40 0.31 Kenya (2009) 0.33 0.29 0.37 0.60 0.47 Mali (2009) 0.28 0.23 0.26 0.42 0.35 Average 0.27 0.25 0.31 0.39 0.33 Share of rural-urban migrants among all migrants in urban centers Ethiopia 69 57 47 54 58 Tanzania 1 72 86 72 77 77 Chapter 1. Growth and transformation in uganda Uganda2 38 54 50 55 47 Average 60 66 56 62 61 Source: Estimations by World Bank (2021). Notes: In this table, unless specified otherwise, a person is considered a migrant if she moved to an area less than 10 years ago. 1) migrants are considered people who moved into a zone (Ethiopia) or district (Tanzania) that is not their birth district less than 10 years ago; 2) migrants are considered people who moved to a district less than 5 years ago; 3) Khartoum is not included. Data sources: Ethiopia (Labor force survey); Tanzania and Uganda (Living standard measurement surveys); Ghana, Kenya, and Mali (Censuses). In Uganda, where data is more limited, this task becomes even more daunting. Nonetheless, while bottlenecks may not be observable, their symptoms often are. Thus, the analysis follows a two-pronged approach: • Statistical analysis, which focuses on the symptoms of possible bottle- necks by using data on sectoral and spatial transitions from several rounds of household surveys (UNHS and UNPS), which provide some ability to track migration and job-switching patterns among participants. 75 Figure 26. Factors that may affect spatial or sectoral transitions What could slow down structural transformation? Destination (pull) issues Insufficient infrastructure, unfavorable institutions, and high trade costs can lead to low investments and sluggish growth in productivity and wages. Transaction costs Transportation costs, job search cost and uncertainty, and social fragmentation can slow down labor flows despite significant Source (push) issues wage gaps across regions. Illiquid assets, credit constraints, and unrecorded benefits (e.g., own consumption) from farm can reduce incentives to migrate. Source: World Bank. • Simulation analysis, which focuses on the nature and the implications of factors behind transformation dynamics that are not directly observable in the data. For example, the role of frictions in spatial and sectoral transitions is one of them. To explore these, the analysis develops a novel general equi- Growth, trade, and transformation librium model, which is calibrated by using national accounts data, and run to dissect the specific roles played by different factors. Although the issue at hand is highly complex, and gaining causal inference is difficult, this approach helps better understand the drivers of Uganda’s slow transformation. A comprehensive overview of the factors leading to the slow structural transformation, especially that with causal inference, is obstructed by the complexity of the problem and data problems. However, the empirical and simulation-based approach employed here helps make progress in better under- standing these factors (for technical details, please see the appendix at the end of the report). 76 Figure 27. Spatial transitions: change in consumption by adult equivalent, 2010–2016 Between 2010 and 2016, rural to urban migration remained low proportionally (4.5 percent). a. Transition Rates (Percent) 15.3 15.8 14.7 22.2 18.6 11.9 19.3 11 Urban to rural Stay urban 84.7 84.2 85.3 77.8 81.4 88.1 80.7 89 Rural to urban 4.5 4.8 4.2 3.8 6 4.3 3.5 9.8 Stay rural 95.5 95.2 95.8 96.2 94 95.7 96.5 90.2 All Male Female Age: 15-24 Age: 25-34 Age: 35+ Primary Higher education education Among those who migrated from rural to urban areas, work hours did not change signifi- cantly (54.9 percent) compared to who stayed in rural areas (55 percent). b. Changes in Work Hours (Percent) 30 39.3 25.4 35 18.8 86.6 13.9 22,5 Urban to rural Stay urban 65.5 40.9 37.1 35.6 38.6 30.9 34.1 29 Rural to urban 54.9 61.4 47.6 117.7 42.7 40.6 87 37.3 Chapter 1. Growth and transformation in uganda 93.8 73 Stay rural 55 65.2 44.6 47.4 39.8 47.2 All Male Female Age: 15-24 Age: 25-34 Age: 35+ Primary Higher education education The change in consumption among who migrated from rural to urban areas were about 1.5 percentage points lower than those who stayed in rural areas. c. Consumption Index (Pre-transition = 100) Urban to rural 107.7 109.3 105.9 109.6 103.7 109.2 111.1 96.7 Stay urban 106.5 104.2 108.9 100.4 104.7 108.6 107.4 103.3 Rural to urban 107.9 106.1 109.9 93.6 122.7 106.8 103.2 107.6 Stay rural 109.4 108.3 110.4 109.1 111.5 104.6 107 106.9 All Male Female Age: 15-24 Age: 25-34 Age: 35+ Low Better education education Source: Garlati et al. (2018) based on UNPS data. 77 Real income gains from transitions The first plausible explanation for a slow spatial/sectoral transition would be a relatively small size of the gap between real incomes across locations/sectors. In standard models of structural transformation for a small and open economy, investments in modern tradable sectors increase the productivity of labor, push wages up, and pull workers from subsistence sectors. This process is stabilized when a diminishing labor force pushes up the marginal product (and the wages) of subsistence workers and the wages are equalized across sectors (net of mobility costs) again. Thus, a slow productivity growth in modern sectors (possibly driven by low investments) can slow down transformation by failing to generate sufficient differences in living standards between modern and subsistence economies. For Uganda, Merotto (2020) pointed to a decline in the demand for off-farm labor rel- ative to its supply, suggesting a weak pull from modern sectors. On average, rural-urban transitions have not led to an increase in migrants’ consumption, but results vary by age, education, and gender. A previous World Bank study (Garlati et al. 2018) analyzed the spatial and sectoral transitions of Ugandans by using five rounds of the UNPS between 2009 and 2016. In line with earlier findings, the study detected a proportionally lower incidence of rural-urban migration (5 percent of all respondent-years) than urban-rural migration (15 per- cent of all respondent-years) in Uganda (the first panel in Figure 27). Comparing rural-urban migrants and those who stayed in rural areas, the study identified no significant differences in how total work hours changed (the second panel in Figure 27). However, the change in rural-urban migrants’ consumption was 1.5 percentage points lower than that of those who stayed in rural areas (the third panel in Figure 27), and this pattern was common for both males and females. In contrast, urban-rural transitions have increased the consumption of mi- grants except for females. Those who moved from urban areas to rural areas re- duced work hours by about 7 percentage points relative to the ones who stayed in urban areas. Despite this, the urban-rural migrants increased their consumption by 1.2 percentage points relative to the ones who stayed in urban areas. This was, Growth, trade, and transformation however, driven purely by males, whose relative increase was about 5.1 percentage points. Further investigation shows that the effect is largely driven by older males (above age 35—not shown in the figure). By contrast, consumption among female urban-rural migrants registered a loss of 3 percentage points relative to non-mi- grants in urban areas, and this was again driven by females age 35 or older. Education seems to play an important role in driving spatial transition and its impact on consumption. Ugandans with primary or higher education were about 5.3 percentage points more likely to migrate from rural areas to urban areas com- pared to those with lower or no education. Upon migration, their work hours in- creased less, but their consumption increased by an additional 4.4 percentage points, compared to the latter group. This urban bias of the educated was also ev- ident in the urban-rural migration. More educated Ugandans were 8.3 percentage points less likely to migrate from urban areas to rural areas than less educated ones. 78 Table 5. Sectoral transitions (all workers including self-employed, age 15+) Education All workers Primary or less Primary or higher To  Agriculture Agriculture Agriculture Services Services Services Industry Industry Industry From  Agriculture 90.6 2.6 6.8 91 2.6 6.4 79.9 4.5 15.6 Transition Industry 28.6 51.6 19.8 40 44.4 15.6 13.5 63.7 22.8 Services 19.2 4 76.8 27.1 4.9 68 12.2 3.3 84.5 Agriculture 49.4 225.5 217.3 44.2 162.2 180 60.2 456.9 312.5 Hours Industry -14.6 19.6 49.5 -21.1 23.5 75.2 -14.4 4.2 30.2 Services -14.5 66 23.1 -15.1 121.3 20.7 -26.5 17.3 21.5 Agriculture 10.3 5.6 11.1 7.6 4.9 8.6 8.2 -8.6 7.8 Consumption Industry 3.9 8.9 6.5 1.3 10.4 -6.9 6 9.4 14.8 Chapter 1. Growth and transformation in uganda Services 0 6.4 6.4 2.1 10.1 5.2 -2.3 7.2 3.8 Source: Garlati et al. (2018) based on UNPS data. Upon migration, more educated migrants’ work hours did not increase as much compared to the latter group. Moreover, they faced a loss as their consumption decreased, not only relative to the less educated group, but also in nominal terms (a 3.3 percent decrease), implying a direct welfare loss. These effects of education were common to both males and females. As for sectoral transitions, survey data exhibits a dramatic labor churning in industry. About half of all industrial workers (including the self-employed) tran- sitioned to services or agriculture over the 5 rounds of the survey covered in the analysis (Table 5). In comparison, only 10 percent of agricultural workers and about a quarter of service sectors workers switched sectors. Workers leaving industry were spread across agriculture and services (60 to 40 percent, respectively). Most workers leaving agriculture switched to services (about three quarters), and most workers leaving services switched to agriculture (about three quarters). These trends were similar between male and female workers. 79 Moving out of agriculture increased work hours significantly, but not consump- tion, relative to those who stayed. On average, workers switching from agricul- ture to industry or services increased their weekly work hours at a main job more than 4-fold compared to those who remained in agriculture. However, in terms of consumption, those who switched to services gained an additional 0.8 percent- age points while those who switched to industry lost about 4.7 percentage points, compared to the workers who stayed in agriculture. By contrast, workers switching from industry and services to agriculture reduced both their work hours and con- sumption compared to those who remained in the respective sectors. Education played a key role in driving sectoral transitions, too. Moving out of agriculture was twice as likely among Ugandans with primary or higher education (20 percent) as those with less education (about 9 percent). Moreover, the more educated group increased their work hours by much more than the less educated group, who had already greatly increased their hours in comparison to those who remained in agriculture. However, these increases in work hours did not translate to consumption gains, which were smaller for the more educated group. This may be because the less educated group had relatively lower consumption in agriculture; thus the new incomes in services and industry provided a larger gain proportionally. The suppression of gains from moving out of agriculture was largely driven by the self-employed; wage workers faced different conditions. The sectoral tran- sition trends we have discussed so far were concerned with all workers, including both self-employed and wage workers, and painted a grim picture for moving out of agriculture. However, a more disaggregated look shows a different picture. For the wage workers, transition trends are better aligned with structural transformation dynamics. First, wage workers more frequently transitioned out of agriculture. About half of all wage workers moved out of agriculture over the five rounds of the UNPS surveys. In comparison, only a quarter of wage workers in industry, and about 10 percent of those in services, did the same. Moreover, the transition from industry and services to agriculture was much smaller for wage workers compared to workers as a whole. Growth, trade, and transformation Only 7.1 percent of industry wage workers and 4.6 percent of services wage work- ers transitioned into agriculture. We should, however, note that the share of wage workers in agriculture is quite small. Thus, despite their high transition rate, and with deep inertia among the self-employed who comprise the vast majority, more than 90 percent of all agricultural workers remained in agriculture, as discussed above. Second, wage workers who moved from agriculture to industry observed real gains in hourly wages. The hourly wages of wage workers who moved from ag- riculture to industry increased by an additional 76.2 percent compared to those who stayed in agriculture. The same was not true for those who moved to ser- vices, whose hourly wage rise was about 43 percentage points lower than those who stayed in agriculture. Interestingly, however, wage workers who moved from industry and services to agriculture also observed a higher increase in their wages compared to their peers who stayed in their respective sectors. 80 Table 6. Sectoral transitions (only wage workers, age 15+) Education All wage workers Primary or less Primary or higher Agriculture Agriculture Agriculture Services Services Services Industry Industry Industry To  From  Agriculture 53.3 17.0 29.7 61.5 15.4 23.1 18.5 26.9 54.7 Transition Industry 7.1 72.4 20.5 13.8 68.2 18.1 0.8 78.6 20.6 Services 4.6 6.0 89.5 9.6 13.1 77.3 2.2 2.9 94.9 Agriculture 50.5 126.2 7.0 74.3 183.2 6.2 -14.4 98.7 -18.9 Hourly wages Industry 99.9 56.3 65.6 118.9 36.6 87.5 -12.9 63.3 54.1 Services 80.0 42.8 65.8 106.6 41.7 83.9 22.9 39.1 62.4 Source: Garlati et al. (2018) based on UNPS data. Education was a primary driver of sectoral transitions and shaped how transi- Chapter 1. Growth and transformation in uganda tions changed hourly wages. The agricultural wage workers with primary or higher education transitioned at a much higher rate than those with lower education (81.5 percent vs. 38.5 percent). While only 18.5 percent of more educated wage workers stayed in agriculture, about 79 percent in industry (10 percentage points higher than lower educated) and 95 percent in services (about 15 percentage points high- er than lower educated) stayed. The higher educated wage workers who switched to industry observed a significant gain in their hourly wages; albeit not as much as the lower educated doing the same. By contrast, in all other cases for the high- er-educated workers, transitioning to a different sector reduced hourly wages compared to those who stayed. The dynamics were different for lower educated wage workers. In their case, switching from industry and services to agriculture in- creased hourly wages relative to those who did not switch. Overall, these transition patterns do not reveal persistent gaps in income pat- terns across locations that are large enough to drive rapid structural transfor- mation. In general, spatial transitions from rural to urban areas and sectoral tran- sitions from agriculture to other services have been slow and have not provided significant gains in consumption or wages relative to no-transition cases. One im- portant exception to this trend is the case of wage workers. These comprised a small group in agriculture and gained significantly higher hourly wages when they 81 moved to industry (not to services) in relative terms. But the same is also true for wage workers who moved to agriculture from services or industry. Higher education accelerated rural-urban transitions and movements out of agriculture, and even increased the hourly wages in the latter case; however, it decreased consumption in the case of rural-urban transitions. Together, these trends show that Ugandans preferred to remain in rural subsistence agriculture if they were self-employed and less educated, which comprised the vast majority. Non-income factors and frictions Besides direct prospects for income and consumption, transitions like ru- ral-urban migration can be driven by location-specific amenities as well. The decision to migrate can be affected by the characteristics of locations where mi- grants resided before migration and the locations that migrants aimed to reach. To explain which of those factors mattered, we use the community questionnaire from the UNHS 2016/17. This survey provides retrospective questions on migration histo- ry and also community-level information regarding amenities like roads and elec- tricity for a nationally representative sample. All location-specific information is aggregated at district level using population size in each community. To control for confounding factors, the analysis relies on two multivariate regression models; us- ing, first, factors at districts of origins (push factors) and, second, those at destina- tion (pull factors). Both models used the same household level variables (Table 7). In line with the previous results, more educated Ugandans were more likely to be migrants. Table 7 shows the marginal effects of various sociodemographic factors on the probability of having a household head be an internal migrant, controlling for location-specific factors. Female, better educated, and younger household heads with lower dependency ratios were more likely to migrate. These results remain un- changed when location-specific characteristics are controlled at the current loca- tion or the location of origin (pre-migration). Post-secondary education (including university) had one of the largest effects on the probability of having a household Growth, trade, and transformation head identified as a migrant. Lower amenities seem to have restricted outmigration from districts with mod- est means, and basic improvements accelerated it. The first column in the bot- tom panel of Table 7 shows the marginal effects of district-specific factors on the likelihood of household heads migrating elsewhere and originating from that district by the time of the survey. Overall, households with a migrant head were less likely to come from districts where agriculture was the main activity or land tenure was cus- tomary, or where electricity had not been improved in the last 5 years. Urbaniza- tion increased the outmigration. Surprisingly, however, improvements in electricity access at the original location seemed to have increased outmigration. This could be potentially related to the enhanced capacity of households in these districts to overcome initial migration costs. Similarly, with improved electricity access, search costs and information asymmetry may be lowered through better communications within migration networks. 82 Table 7. Marginal effects on the probability of having a head of house- hold internal migrant Community Community Variables variables at district variables at current of origin district Head of household is male -0.0291 *** -0.0366 *** Head of household age -0.0029 *** -0.0029 *** Head of household education (base = no education) Primary incomplete 0.0127 0.0134 * Primary complete 0.0239 *** 0.0313 *** Household level variables Secondary incomplete 0.0174 ** 0.0322 *** Secondary complete 0.0371 *** 0.0711 *** Post-secondary but not university 0.0470 *** 0.0713 *** University 0.0416 *** 0.0766 *** Dependency ratio -0.0007 *** -0.0008 *** Marital status (base=married monogamous) Married polygamous -0.0119 * -0.0195 *** Divorced/ Separated -0.0082 -0.0087 Chapter 1. Growth and transformation in uganda Widow/ Widower -0.0318 *** -0.0385 *** Never married 0.0041 0.0114 Community had flood or drought -0.0001 0.0008 *** in the last 5 years Community had new employment Community level variables -0.0001 0.0004 ** opportunities in the last 5 years Community had new road in the 0.0002 0.0000 last 5 years Community had improved 0.0004 ** 0.0002 electricity in the last 5 years Most common land tenure in the -0.0002 * 0.0000 community is customary Main economic activity in the -0.0005 *** -0.0005 ** community is farming Urban dummy 0.0169 *** 0.0170 ***   Observations 15,553 15,553 Notes: Estimations reflect multilevel mixed-effects logistic regression with the binary dependent variable defined as migrant (those who lived in more than one district in the last 5 years) and non-migrant (everybody else). Household weighted estimates using UNHS 2016/17. ***, **, and * indicate significance at 1%, 5%, 10% confidence levels, respectively. 83 The negative effects of the district of origin factors on outmigration likely point to the poor’s inability to move. Inferior conditions at the origin are often expected to induce faster outmigration. This is known as the push effect. However, migration itself is a risky and costly endeavor, and the poorest members of the society often refrain from undertaking such an investment for several reasons. For example, they may simply not be able to afford it or the risks and costs associated with migration may be too high. Even when there are large gains to be made, if the poor’s current income is small and they are credit constrained (e.g., cannot borrow to overcome the income bottleneck), they will not migrate. In such circumstances, an improve- ment in amenities like electricity can relax these constraints sufficiently to enable some residents to migrate (see Appendix A for a more formal elaboration of this point). Note that the same improvement in a relatively better-off location (with higher income level) can have the opposite effect: by improving the living condi- tions there, it may reduce incentives to move away. The negative effects of agricultural intensity and prevalence of customary land tenure on outmigration reinforce the credit constraints argument. In areas with customary land tenure, land––the most important wealth of the people––is a rela- tively illiquid asset and is relatively hard to monetize. Considering also the relative- ly low financial penetration, this means that the residents in those areas will face greater opportunity costs associated with outmigration, and thus be less willing to do so. In comparison, the pull factors at destination worked as expected. Districts with larger roles of agriculture in the local economy were less likely to attract mi- grants. The districts that experienced larger employment opportunities during the last 5 years were more likely to attract migrants. Box 3 on the next page provides estimates of the magnitude of the effects that frictions and land access dynamics might have played in Uganda’s structural transformation history over the last few decades. The role of public services Growth, trade, and transformation Public policies can shape the gains from sectoral and spatial transitions and also affect the frictions that slow down such transitions. Private economic activ- ity takes place in conducive environments facilitated by public policies like securi- ty, property rights, the provision of large-scale public goods and services like infra- structure, and the broader set of institutions that help resolve common goods and coordination problems. Problems in these areas can slow down structural trans- formation significantly through multiple channels. With inadequate infrastructure, for example, firms may find it unprofitable to invest in modern sectors, reducing potential income gains from sectoral and spatial transitions. In addition, inadequate infrastructure may render mobility costs for such transitions prohibitively high. Fi- nally, infrastructure also has direct consumption value; i.e., people consider ameni- ties when they vote with their feet. Thus, both the pull factors and the push factors are highly sensitive to public policies. In Uganda, infrastructure problems abound and have surely slowed down growth and transformation. 84 Box 3. The role of frictions and land access in explaining Uganda’s transformation. To assess the role of frictions in Uganda’s structural transformation quanti- tatively, Artuc, Leunga, and Onder (2022) developed a simulation framework for this report. The model includes a 3-sector small open economy with labor market frictions. Agriculture and industry goods are tradable, while services are assumed to be non-tradable, for simplicity. The model can track structural dynamics through uneven sectoral productivity growth (the Baumol effect), or through an income effect where non-homothetic preferences can shift at- tention and resources to manufacturing and services (especially in a closed economy). The model is calibrated using national accounts data for employment, produc- tivity, prices, and wages for the period 1990–2018. Specifically, the wage ratios between agriculture, manufacturing, and services sectors observed in the data are exogenously imposed and the model’s primitives are calibrated to fit the labor allocations across sectors. Despite the simplicity of the model and the many potential market failures and distortions in Uganda’s economy (e.g., agri- cultural production for auto-consumption and the prevalence of a customary land tenure system, which reduces the liquidity of land), the model’s fit to data is good once the wage dynamics are exogenously controlled for. A more de- tailed discussion of the model’s technical characteristics and fit is provided in Appendix B. The main results (panels a and b in Figure 28) show the modeled results of Chapter 1. Growth and transformation in uganda agriculture’s share in labor and value added over the period of analysis. The model, despite its simplicity, captures well the trend decrease in the sector’s labor share and value-added share. Panel c in the same figure shows the actual wage trends that are used to generate these baseline estimates. To shed light on the relative importance of frictions in explaining structural transformation patterns in Uganda, panel c shows the wage ratios between agriculture and manufacturing (). The actual data (from the Groningen Growth and Development Centre) show a downward trend throughout the period of analysis, decreasing from 35 percent to 13 percent. In the absence of unob- servable frictions or transition costs, we would expect the wage gaps across sectors to close in real terms. However, a large and widening wedge between agricultural and manufacturing sectors point to major obstacles against such convergence. But how large is this effect? Panel d provides a counterfactual exercise, where the wage ratios are kept at the initial 35-percent rate (instead of decreasing to 13 percent). As a result, the labor share of agriculture decreases to about 27 percent (instead of 64 percent in the baseline model). Thus, we conclude that the unobservable frictions (which could be any factor that are discussed in this chapter) play a major role in slowing Uganda’s transformation. 85 Box 3 (continued) Figure 28. Simulation results a. Labor share of agriculture b. Baseline: Value added share of agriculture 1.0 1.0 0.8 0.8 0.6 0.6 Share Share 0.4 0.4 0.2 0.2 0 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 c. Wage ratios (wa/wm): d. Counterfactual labor share of actual vs. predicted agriculture: No change in wages 1.0 1.0 0.8 0.8 0.6 0.6 Share Ratio 0.4 0.4 0.2 0.2 0 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 e. Counterfactual labor share of f. Counterfactual value added agriculture: more land access share of agriculture: more land access Growth, trade, and transformation 1.0 1.0 0.8 0.8 0.6 0.6 Share Share 0.4 0.4 0.2 0.2 0 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Actual Model Predicted Counterfactual Source: Artuc, Leunga, and Onder (2022). 86 Box 3 (continued) In panels e and f, we extend the counterfactual analysis to estimate how land access affects transformation. As discussed in the introduction of this report, with rapid population increase, rural population intensity in Uganda increased significantly: from 3.1 persons per acre of arable land in 1990 to 4.7 by 2018. In the counterfactual analysis, we hold the population intensity constant in its 1990 value to show how expansion of land access as a result of population pressure can further slow the country’s transformation. Results show that in the counterfactual case, other things being equal, the labor share of agriculture increases from 69 percent to 74 percent, and its value-added share decreases from 55 percent to only 39 percent. Overall, the simulation exercise here shows that both transition frictions and land expansion can lead to significant reductions in Uganda’s structur- al transformation. Uganda has one of the lowest electricity access rates in the world. In 2017, the power grid supplied only about 22 percent of the urban population and about 12 percent of the rural population. As a result, the average Ugandan consumed about 215 kilowatt hours, falling far behind the African average of 552 kilowatt hours. Au- thorities have tried to increase power generation capacity through investments in hydro and thermal power plants, construction of new transmission lines, and the Chapter 1. Growth and transformation in uganda enactment of energy-efficiency measures. As a result, the installed capacity of the country increased from 595 MW in 2012 to 938 MW (75 percent of which is generated by hydropower), and the population with access to electricity rose to 28 percent by 2019. However, even with the slower economic growth in recent years, demand for electricity has been growing by about 10 percent per year (WTO, 2019). Thus, the country will continue to face challenges that constrain growth, until such time as the generation capacity can actually be delivered to consumers, which will then hopefully meet the existing demand and encourage higher levels of demand. Despite improvements, transport infrastructure has also been insufficient, es- pecially in rural areas. According to a Diagnostic Study of the Uganda National Roads Authority (UNRA) Transformation (World Bank, 2019), the country has made significant progress in paving roads as part of National Development Plans (NDP) over the last decade. As a result, the share of total paved (national) roads increased from 16 percent to 22 percent between 2010 and 2018. However, despite this no- ticeable progress, which brought the total paved roads to 4,551 kilometers, the im- plementation fell short of NDP targets by a third and, with the more recent slow- down during the pandemic, the gap between the implemented upgrades and the NDP II target widened (6,000 kilometers in 2020). Moreover, the total roads in the country (paved and unpaved) have remained flat in the last decade. In some local surveys, only 12 percent of participants thought their road network was in good 87 condition, with the rest rating it as fair or poor (Muhwezi, Basoona, and Acai 2021). Similarly, more than four fifths of participants indicated that roads deteriorated significantly in less than two years following construction as a result of weather conditions and poor maintenance. Together, the problems in electricity and road infrastructure are reported to limit business opportunities and investments especially in agro-processing and tourism. The inadequate transport and storage infrastructure (including cold storage), poorly maintained roads, and scarce opportunities for railway, air trans- port, and shipping facilities add to transport costs in Uganda, and erode the com- petitiveness of Ugandan producers in external markets. As a result, investment in sectors that rely on these services like agro-processing remains limited. In addition, the low infrastructure quality reduces tourism potential as well. For example, it is reported that driving on certain sections to Queen Elizabeth National Park can of- ten take more than 2 hours to travel 72 kilometers on an unpaved road, more than 4 hours if weather conditions change, or become unnavigable altogether if mudslides occur under severe weather conditions. Overall, these factors have likely restricted business opportunities and investments that could boost productivity and wages and attract workers out of subsistence agriculture. Education is another area where underachieving public policies may have weakened the structural transformation process. According to the Uganda Eco- nomic Update (World Bank, 2019), Uganda is ranked among the countries in the lowest quartile of the Human Capital Index (HCI) distribution. The Uganda HCI is lower than the average for the SSA region, and below what would be predicted by its income level, largely due to its low education outcomes. A child born today in Uganda is expected to complete only 7 years of education by age 18, compared to a regional average of 8.1. With low levels of learning achievement in Uganda, this is only equivalent to 4.5 years of learning, with 2.5 years considered as lost due to poor quality of education. Uganda’s score on this component is the lowest among comparator countries and below the SSA average. The country currently shows extremely low gross enrollment rates (GER) at pre-primary and secondary levels. In primary education, the near-universal access veils very modest completion rates. Growth, trade, and transformation With the COVID-19 pandemic, schools have been closed for longer than any other country in the world––as result, the majority of 15 million children have been out of school for the most part of two years. Furthermore, plans are unclear regarding what is needed to reopen schools effectively and catch up with all the learning lost. As discussed earlier, more education is one of the main correlates of spatial and sectoral transition; thus, transforming Uganda’s economic composition will entail a major improvement in the country’s education outcomes. Political violence, risk, and uncertainty suppress transformation in different ways. Uganda has historically experienced periodic flare-ups of political violence around elections and other political activities. Security services are often reported to use routine excessive force to stop peaceful protests and demonstrations (WTO, 2019). Although such incidences have not translated into major conflicts or pre- vented the implementation of public projects in the last decade, political tensions 88 have remained high since the 2021 general elections. Such levels of uncertainty and risk can suppress investment, as expected returns on investment will be reduced, and diminish gains from sectoral and spatial transitions. Synopsis This chapter focused on Uganda’s economic performance and investigated the drivers of economic growth and transformation over the last two decades. The analysis relied on national accounts-based statistics as well as micro data from surveys to identify these drivers. Uganda achieved a rapid economic growth and transformation phase in the 2000s, but lost momentum after 2011. The rapid growth episode was driven by three conducive factors: a peace dividend, after the elimination of internal conflict; external factors like the establishment of EAC and the emergence of South Sudan as a major export destination and international aid sourced from Uganda; and the government’s business-friendly reforms, including those that brought fiscal disci- pline and liberalized prices in key sectors. Since 2011, however, the reform drive has weakened, and growth has lost momentum. Economic growth has primarily been driven by within-sector productivity changes, while structural transformation has remained weak. In the 2000s, all major sectors, including agriculture, registered significant within-sector produc- tivity growth, which drove the economic growth. Labor reallocation toward sectors with higher productivity was a contributing factor, but remained modest. In com- Chapter 1. Growth and transformation in uganda parison, within-sector productivity slowed down significantly after 2011, especially after 2015 and in agriculture. This, together with unfavorable labor allocation effects, have broken the growth momentum in Ugandan productivity. To achieve a middle-income status, Uganda will need to overcome its slow pace of economic transformation. The majority of the income gap between Uganda and its middle-income peers (e.g., Vietnam) can be explained by two factors: first, a low productivity in agriculture and, second, the failure to reallocate larger shares of its labor force to more productive activities outside of subsistence agriculture. Moreover, the low productivity in agriculture itself is largely driven by rural under- employment. Thus, the stalling structural transformation is likely the most promi- nent stumbling block for Uganda’s development. What is holding transformation back in Uganda? The microanalysis in this chap- ter showed that three factors have been instrumental in hindering structural trans- formation, specifically: • A weak pull by modern sectors: Higher productivity and wages in industry and services are often the main drivers of rural-urban transition. In Uganda, however, rural-urban migrants have not benefited from significant increases in their consumption relative to others, signaling a weak pull factor. An inter- 89 esting exception is agricultural wage workers, whose consumption increased when they transitioned to services or industry. For self-employed Ugandans in subsistence agriculture, the weak pull factor may be driven by the pres- ence of an unaccounted portion of production used for own consumption. • Acute constraints in traditional sectors: Even when moving out of subsis- tence agriculture is clearly gainful, such a transition can be unaffordable with upfront costs. In Uganda, this problem looms large. Outmigration is slow in areas with traditional land tenure, and access to credit is severely limited in rural areas. Together, these imply that Ugandans cannot borrow or liquidate their assets (if any) to pay for a livelihood transition. • Major frictions in between: In Uganda, areas with lower provision of public services like electricity, roads, and education have exhibited lower outmigra- tion, and improvements in these services have accelerated it. These obser- vations point to major frictions in the form of skill barriers (reinforced with credit constraints), high job search costs (lacking digital facilitation), and high mobility costs (low quality transportation), which restrict spatial and sectoral transitions. Simulations in this chapter suggest that these frictions have played a significant role in slowing Uganda’s transformation. Public policies shape the factors that limit transformation. Despite making prog- ress in recent decades, Uganda still has major gaps in public service provision. The country has one of the lowest electricity access rates in the world and its trans- portation infrastructure suffers from underinvestment and lack of maintenance. Together, these have limited business opportunities and investments, especially in agro-processing and tourism, weakening the pull factor in modern sectors. Sim- ilarly, Uganda’s chronically low education outcomes, together with severely lagging financial development, have rendered spatial and sectoral transitions unaffordable. Finally, an upward trend over the last 5 years in political violence, risk, and uncer- tainty threaten to further constrain investments, and thus reduce gains from sec- toral and spatial transitions. Growth, trade, and transformation References Basu, Kaushik. 2003. Analytical Development Economics: The Less Developed Economy Revisited. Cambridge, MA: MIT Press. Beaman, Lori, Harun Onder, and Stefanie Harun. 2021. “When do Refugees Return Home? Evidence from Syrian Displacement in Mashreq.” Policy Research Working Paper Series 9688. Washington, DC: World Bank. Boutin, Delphine. 2016. “Migration Experience and Access to a First Job in Uganda.” Discussion Paper No. 10119. Bonn: Institute of Labor Economics (IZA). Caselli, Francesco. 2005. Accounting for Cross-Country Income Differences. In Handbook of Economic Growth Vol. 1, edited by Philippe Aghion and Steven N. Du- 90 ralumin. 679–741. Amsterdam and London: North-Holland Publishing Company. Christiaensen, Luc, and Nancy Lozano Gracia. 2021. “Migrants, Markets and May- ors––Rising Above the Employment Challenge in Africa’s Secondary Cities.” Jobs Working Papers No. 63. Washington, DC: World Bank Group. Christiaensen, Luc, and Lionel Demery, eds. 2017. “Agriculture in Africa––Telling Myths from Facts: A Synthesis.” Washington, DC: World Bank Group. Dennis, Allen, Taye Mengistae, Yutaka Yoshino, and Albert Zeufack. 2016. “Sources of Productivity Growth in Uganda: The Role of Interindustry and Intra-industry Mis- allocation in the 2000s.” Policy Research Working Paper 7909. Washington, DC: World Bank. Garlati, P., C. Mejia-Mantilla, Dino Merotto, and M. Webber. 2018. “Towards a Better Labor Market in Uganda: An Analysis of Labor Markets Transitions between 2009 and 2016.” Mimeo. Washington, DC: Mimeo, World Bank Group. Hausmann, Richard, and Bailey Klinger, and Rodrigo Wagner. 2008. “Doing Growth Diagnostics in Practice: a ‘Mindbook’.” CID Working Paper Series No. 177. Cambridge, MA: Center for International Development, Harvard University. Lagakos, David. 2020. “Urban-Rural Gaps in the Developing World: Does Internal Migration Offer Opportunities?” Journal of Economic Perspectives 34(3): 174–92. Matsuyama, Kiminori. 2019. “Engel’s Law in the Global Economy: Demand‐Induced Patterns of Structural Change, Innovation, and Trade. Econometrica 87(2): 497–528. Merotto, Dino. 2020. Uganda: Jobs Strategy for Inclusive Growth. Jobs Working Pa- pers No. 19. Washington, DC: World Bank Group. http://documents.worldbank.org/ Chapter 1. Growth and transformation in uganda curated/en/693101582561426416/Uganda-Jobs-Strategy-for-Inclusive-Growth. Muhwezi, Lawrence, Naomi Kamwaka Basoona, and Joseph Acai. 2021. “An Assess- ment of Rural Road Networks Management in Uganda as a Driver of Economic De- velopment: A Case Study of Kabarole District.” International Journal of Construc- tion Engineering and Management 10(1):8–16. Pritchett, Lant. 2019. “There Is Only One Poverty Strategy: (Broad Based) Growth.” https://lantpritchett.org/there-is-only-one-poverty-strategy-broad-based- growth-part-i/. Restuccia, Diego, Dennis Tao Yang, and Xiaodong Zhu. 2008. “Agriculture and Ag- gregate Productivity: A Quantitative Cross-Country Analysis.” Journal of Monetary Economics 55(2): 234–50. World Trade Organization. 2019. Trade Policy Review: East African Community (EAC). Geneva: World Trade Organization. 91 Growth, trade, and transformation 92 Chapter 2. The role of International Trade CHAPTER 2. THE ROLE OF INTERNATIONAL TRADE I nternational trade is a fundamental determinant of growth and structural transformation. International trade transforms economies in three tradition- al channels: (i) scale effect, where specialization along comparative advantag- es can increase overall economic activity in each country; (ii) composition effect, where the sector structure of the economy is realigned by international prices (i.e., a structural transformation); and (iii) technique effect, where access to new tech- nologies and inputs transforms production processes (Onder, 2012). In develop- ing countries, where market frictions and distortions abound, there may be other mechanisms at play, too. Atkin and Donaldson (2021) list two of those: (i) a distor- tion revenue effect, where trade changes the consequences of domestic distor- tions; and (ii) a direct effect on distortions, where trade changes the magnitude of prevailing distortions themselves. As discussed in the previous chapter, in Uganda such distortions play an important role in shaping transformation. Uganda’s special conditions reinforce the importance of trade in its future growth and transformation. With its relatively small economic size and low in- come, Uganda’s growth needs to benefit from access to other markets. The country faces natural barriers like a landlocked location and frequent political instability in neighboring countries. These make Uganda’s international trade more difficult, but they are largely beyond the reach of policy actions. However, there are also man- made impediments against trade, including distortionary policies. Those restrict access to imported intermediate inputs, which in turn reduces export potential Chapter 2. The role of international trade and competitiveness. They also distort domestic prices, which can slow down the country’s economic transformation. Addressing these challenges is essential to promote growth and transformation in the country. This chapter analyzes Uganda’s international trade with an emphasis on its performance, challenges, and role in transforming Uganda. The first section summarizes Uganda’s trade patterns. The second section then discusses firm-lev- el drivers of the aggregate patterns. Next, the third section focuses on the factors that affect trade, with a specific emphasis on trade policy, regional integration, and trade logistics. Finally, the fourth section analyzes how trade policies shape Ugan- da’s structural transformation and inequality by extending the structural transfor- mation model presented in the previous section and using a new model that fo- cuses on household income and consumption patterns in agricultural goods, using micro survey data. 95 The State of Uganda’s International Trade After growing exponentially in the 2000s, Uganda’s exports have stalled in the last decade. From 2000 to 2012, the exports of Uganda grew 7-fold (Panel A in Figure 29). The growth momentum was particularly remarkable for services which grew 10-fold, boosting the services’ share in total exports from a third to a half in the same period. In comparison, agricultural exports grew about 4-fold and their share in total exports decreased from a half to less than a third. However, starting from 2012, the tide has changed dramatically for most sectors. By 2019, total ex- port revenues in Uganda were 2 percent less than in 2012. This was in part driven by services exports, which contracted by 5 percent in the same period, and in part by agriculture which fell by 23 percent. It is important to note that these are only formal statistics and there may be sizeable (up to 60 percent of formal exports) underreporting of small cross-border exports (Rauschendorfer, Stojanov, and de Melo 2021); however, detecting a precise trend for those is not possible. Tourism is Uganda’s most important export, followed by cash crops like cof- fee. Uganda’s tourism receipts increased 8-fold between 2000 and 2019, making the sector the top foreign exchange generator (32 percent of all export revenues in 2019). On the merchandise side, agriculture is the largest exporting sector. It includes a variety of traditional and new exported products with sizeable export weights. In 2019, the most prominent ones included coffee (8.5 percent of all ex- ports), dairy (3.2 percent), fish (2.8 percent), tobacco (1.4 percent), and cocoa (1.2 percent). In earlier years, maize and beans were prominent, too. There is also small- er but significant export of flowers, vegetables, cotton, and tea (Panel B in Figure 29). More complex exports like those in chemicals and textiles remain very limited (less than 3 percent combined). Uganda’s merchandise exports are increasingly oriented towards Asia, includ- ing the Middle East, and regional markets in Africa. In 2000, more than half of all Ugandan’s exports were sent to European markets. However, while exports to European countries remained roughly stable at around USD 500 million since the Growth, trade, and transformation global financial crisis, exports to other markets continued to grow. For example, ex- ports to Africa increased from about USD 50 million in 2000 to about USD 1.5 billion in 2018, before plunging to USD 500 million in 2019, in part as a result of regional trade disputes. Exports to Asia, however, did not plunge. From 2000 to 2019, they grew continuously from only USD 40 million to about USD 1.2 billion. A large share of the growth in recent years comes from a hike in gold exports to the United Arab Emirates, which increased from about USD 60 million in 2015 to USD 940 million in 2019. Gold is imported into Uganda through formal and informal channels from neighboring countries and 99 percent of which is re-exported to the UAE. Uganda does not produce or further process gold. Therefore, Uganda’s gold exports have low potential for domestic value addition and job creation. 96 Figure 29. Uganda’s exports: dynamics, composition, and destinations Ugandan exports increased 7-fold between 2000 and 2012, but then remained nearly flat afterwards. An export hike in 2018 was quickly undone by 2019. a. Export dynamics (2000–2019) Services constituted about half of all exports, with tourism registering a third. Agriculture constituted about a fifth of exports, with coffee, milk, and fish among the prominent sectors. b. Exports decomposition in 2019 Chapter 2. The role of international trade The most important export destinations are regional partners like Kenya and others like United Arab Emirates, Italy, Netherlands, and Germany. c. Export destinations in 2019 (merchandise only) Source: Atlas of Economics Complexity by Harvard University using UNCOMTRADE data. 97 Figure 30. Uganda’s imports: dynamics, composition, and sources Ugandan imports increased 5.5-fold between 2000 and 2012 and only by an additional 7 percent until 2019. a. Import dynamics (2000-2019) Services constituted about 40 percent of all imports, with transportation services alone accounting for a quarter. b. Import decomposition in 2019 The most important import sources are regional partners like Kenya and others like China and India. Growth, trade, and transformation c. Import sources in 2019 (merchandise only) Source: Atlas of Economics Complexity by Harvard University using UNCOMTRADE data. 98 Uganda’s merchandise imports comprise a diversified basket of complex goods that are not produced domestically. In 2019, about a quarter of imports comprised transportation services, which is due to the country’s geographic po- sition (Figure 30). Among the merchandise imports were products with relatively high technological complexity, including chemicals (14 percent of all imports), ma- chinery (8.7 percent), vehicles (7.7 percent), and electronics (5.8 percent). There were also significant agricultural imports (11 percent), which mainly comprised ce- reals and processed agricultural products like paper, oils, and sugar. Overall, while Uganda’s imports grew about 5.5-fold between 2000 and 2019, their composition remained largely stable over time. As for import sources, although all regions exhib- ited growth, imports from Asia grew faster. Its share in Uganda’s imports grew from 22 percent in 2000 to 51 percent in 2019, of which about 35 percentage points came from China and India. In comparison, Africa’s share decreased from 45 per- cent to 25 percent, and that of Europe decreased from 28 percent to 17 percent in the same time frame. There is a considerable amount of small-scale cross-border trade (SSCBT) in Uganda. SSCBTs account for about 30 percent of total regional exports (formal plus unrecorded exports since 2013). In some years, SSCB exports reach up to 50 percent of total exports with the Democratic Republic of Congo (DRC), South Su- dan, and Tanzania. Imports are also significant – for example, 97 percent of Ugan- dan imports from the DRC through the Bunagana border are small scale trade. SSCBT provides livelihoods for many border dwellers and plays a role in regional peace and security. Most cross-border traders are female and part of poor border communities with cross-border trade constituting their sole source of incomes. Ugandan authorities collect data at border crossings at different times of the year, which provides policymakers with important information on the potential of trade in the region, particularly on how trade affects the most vulnerable communities and the role women play in this. Chapter 2. The role of international trade The Firm-Level Drivers of Uganda’s Exports Analyzing the activity patterns of Ugandan exporters and/or importers can help to better understand the drivers of Uganda’s exports. To do this, this sec- tion employs the World Bank’s Exporter Dynamics Database (EDD), for the period 2011–2020. This comprises highly disaggregated firm-level information, including exports, size, and entry/exist dynamics, using customs transactions for 74 coun- tries. Figure 31 shows summary statistics of exporters in Uganda (number, size, con- centration, number of products, and number of destinations) benchmarked against comparators (Cameroon, Colombia, Ethiopia, Kenya, Mauritius, Tanzania, and South Africa). 99 Exporters: number, size, and concentration Compared to her peers, and others, Uganda has fewer exporters. Following the country’s broader trends, the number of exporters in Uganda increased significant- ly in the 2000s, from 242 in 2000 to 1691 at peak in 2011, and has decreased after- wards, reaching 889 in 2020 (Panel A in Figure 31). By 2020, Uganda had 32 export- ing firms per million of persons in the country, which was substantially lower than in Tanzania (48), Kenya (129), and South Africa (698). However, it may be tempting to think that these differences are likely to be driven by differences in income levels, the size of the economies, and the nature of the dominant sectors. To remove such doubts, Table 8 uses a panel of country-year-product level in the 2000–20 period with all 74 countries in the EDD, where the dependent variables are different indica- tors of export performance. The regressions control for the size (GDP) and the stage of development (GDP per capita) of each country. In addition, they use year-prod- uct fixed effects to control for common commodity price shocks or technology shocks during the sample period.4 Each regression includes an indicator variable identifying the observations for Uganda, which shows how Uganda performs rela- tive to the other countries. The first column shows that relative to countries with similar size, income, and export-basket, Uganda has fewer exporters. Ugandan exporters are larger in comparison to those in other countries. The median total value of exports per Ugandan firm was USD 33,000 over the 2011–20 period (Panel C in Figure 31). This is greater than key comparators like South Africa (USD 24,000), Kenya (USD 19,000), and Tanzania (USD 14,000). Moreover, the aver- age value of exports in Uganda is much higher at USD 1.3 million, which shows that the distribution of exporter sizes is dominated by a few large exporters in Uganda. The columns 2 and 3 of Table 8 also show that, within a product category in a given year, and after controlling for income and size, Uganda’s exporters are significantly larger than in the other 74 countries, both in terms of average and median sizes. The analysis also shows that, in the last 5 years, Ugandan exporters grew regardless of the initial size. However, larger firms grew faster. As a result, Ugandan exports have become more concentrated over time. This concentration is verified by the Her- Growth, trade, and transformation findahl index of exporter shares (Column 4 in Table 8). Within a product category, Uganda’s Herfindahl index is 4 points higher than that in other countries even after controlling for country size and level of development. Ugandan exporters’ product and destination diversification are comparable to their regional peers but higher than the predicted values. On average, Ugandan firms export 3.9 products in 2.4 countries (Panels E and F in Figure 31). These values are comparable to those of Tanzania (4 products to 2.2 countries) but lower than those of Kenya (6.4 products to 2.6 countries). However, when income and size are controlled for, and product and destination numbers are considered within each product line (columns 7 and 8 in Table 8), Ugandan exporters appear to be more diverse, in terms of the number of products they export, as well as the number of countries they export to. 100 Figure 31. Uganda’s exporting firms (2020 unless otherwise noted) Number of Exporters Over Time Number of Exporters MUS 1584 1800 1691 1600 ZAF 698 1405 1400 1383 1356 Number of Exporters COL 160 1200 1214 KEN 129 1011 958 1000 1008 918 846 1016 782 968 TZA 48 800 736 856 885 889 600 704 CMR 39 400 509 UGA 32 200 ETH 19 242 0 0 400 800 1200 1600 2000 2007 20 1 0 2004 2014 2006 2009 2008 2016 2019 2018 2005 2015 2003 2013 2017 2001 2002 2012 2020 2011 Number of Exporters (Per Million People) Median Exporter Size Herfindhal Index of Exporters ETH 145 TZA 0,06 COL 65 MUS 0,04 UGA 33 CMR 0,03 CMR 30 UGA 0,02 ZAF 24 ZAF 0,01 KEN 19 ETH 0,01 TZA 14 COL 0,01 MUS 14 KEN 0,01 Chapter 2. The role of international trade 0 20 40 60 80 100 120 140 160 0 0,01 0,02 0,03 0,04 0,05 0,06 0,07 Median exporter size (Thousand USD) Index Value Average Number of Destinations Average Number of Products Per Exporter Per Exporter ZAF 18,7 COL 3,2 MUS 7,9 ZAF 3,2 KEN 6,4 CMR 3,0 COL 5,0 ETH 2,6 CMR 5,0 KEN 2,6 TZA 4,0 UGA 2,4 UGA 3,9 MUS 2,4 ETH 3,2 TZA 2,2 0 2 4 6 8 10 12 14 16 18 20 0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 Number of Products (HS-6 Digit) Number of Destinations Source: Authors’ calculations based on data used for the Exporter Dynamics Database. Notes: Products are defined at the HS 6-digit level. 101 Table 8. Cross-country regressions for export competitiveness indica- tors Hirschman Index Destinations per Share of top 5% Share of top 1% Log Number of Log Number of Log Number of Exporter Size Exporter Size Products per Log Average Herfindahl- Log Median Exporters Exporters Exporters Exporter Exporter Variables Uganda -0.365*** 0.0453** 0.245*** 0.0418*** 0.127 0.0336*** 0.0342*** 0.0572*** Indicator (0.00757) (0.0227) (0.0212) (0.00267) (0.101) (0.0117) (0.00768) (0.00329) 0.596*** 0.531*** 0.195*** -0.0714*** 0.0141*** 0.0388*** 0.0223*** 0.0583*** Log GDP (0.000908) (0.00154) (0.00142) (0.000190) (0.000807) (0.000321) (0.000876) (0.000283) Log GDP per 0.0700*** 0.178*** 0.000610 -0.00282*** 0.0633*** 0.0321*** 0.0621*** 0.0833*** Capita (0.000996) (0.00184) (0.00171) (0.000231) (0.000631) (0.000321) (0.00105) (0.000408) -13.71*** -5.073*** 3.139*** 2.316*** -0.515*** -0.700*** -0.631*** -1.833*** Constant (0.0184) (0.0363) (0.0333) (0.00437) (0.0244) (0.00885) (0.0182) (0.00676) Observations 1,419,306 1,060,805 1,034,118 1,060,805 130,227 325,640 60,719 1,060,805 R-squared 0.508 0.402 0.368 0.264 0.662 0.371 0.566 0.268 Time X HS Yes Yes Yes Yes Yes Yes HS 2-digits Yes 6-Dig FE Cluster S.E. No No No No No No No No Source: World Bank based on data for the Exporter Dynamics Database. Notes: Robust standard errors in brackets. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% confidence levels, respectively. All countries in the Exporter Dynamics Database with available data in the 2000–2020 period enter the regression. Within an industry group in a given year, the average number of exported products (6-digits HS codes) per exporter in Uganda is significantly higher than in compara- tor countries with similar size and stage of development. Growth, trade, and transformation Overall, exporters in Uganda comprise fewer large firms in cash crops (coffee, beans, cut flowers, and fish) and many small-scale firms in other areas. The large exporters exhibit more diverse product offering and more market destinations, and they are likely to be competitive internationally. This could indicate higher produc- tivity of Ugandan exporters. Next, the analysis will consider the dynamics of firm entry/exit and how market conditions have influenced such trends in Uganda. Exporters’ entry and exit Compared to her regional peers and other countries, Ugandan exporters exhib- its high entry rates and even higher exit rates. Figure 32 compares the entry and exit rates of Ugandan exporters with that of regional peers. On average each year, 45 percent of Ugandan firms that export did not do so in the previous year, where as 47 percent of Ugandan firms that were exporting in the previous year stopped exporting. 102 Table 9. Cross-country regressions for exporter dynamics indicators Hirschman Index Destinations per Share of top 5% Share of top 1% Log Number of Log Number of Log Number of Exporter Size Exporter Size Products per Log Average Herfindahl- Log Median Exporters Exporters Exporters Exporter Exporter Variables Uganda -0.365*** 0.0453** 0.245*** 0.0418*** 0.127 0.0336*** 0.0342*** 0.0572*** Indicator (0.00757) (0.0227) (0.0212) (0.00267) (0.101) (0.0117) (0.00768) (0.00329) 0.596*** 0.531*** 0.195*** -0.0714*** 0.0141*** 0.0388*** 0.0223*** 0.0583*** Log GDP (0.000908) (0.00154) (0.00142) (0.000190) (0.000807) (0.000321) (0.000876) (0.000283) Log GDP per 0.0700*** 0.178*** 0.000610 -0.00282*** 0.0633*** 0.0321*** 0.0621*** 0.0833*** Capita (0.000996) (0.00184) (0.00171) (0.000231) (0.000631) (0.000321) (0.00105) (0.000408) -13.71*** -5.073*** 3.139*** 2.316*** -0.515*** -0.700*** -0.631*** -1.833*** Constant (0.0184) (0.0363) (0.0333) (0.00437) (0.0244) (0.00885) (0.0182) (0.00676) Observations 1,419,306 1,060,805 1,034,118 1,060,805 130,227 325,640 60,719 1,060,805 R-squared 0.508 0.402 0.368 0.264 0.662 0.371 0.566 0.268 Time X HS Yes Yes Yes Yes Yes Yes HS 2-digits Yes 6-Dig FE Cluster S.E. No No No No No No No No Source: World Bank calculations based on data for the Exporter Dynamics Database. Notes: Robust standard errors in brackets. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% confidence levels, respectively.All countries in the Exporter Dynamics Database with available data in the 2000–2020 period enter the regression. TEV= total export value. Chapter 2. The role of international trade A higher exit rate also explains the decreasing number of exporters in Uganda as shown above. This level of churning is higher than those in countries like Ethiopia, Mauritius, or South Africa, comparable to those in Kenya, but lower than those in Tanzania. Next, Table 9 repeats this exercise by using cross country regressions for product categories, while controlling for income and size. This analysis shows that, in a given year, exporter entry and exit rates are significantly higher in Uganda than in other countries with similar size and level of development. Ugandan exporters are less successful in surviving the first couple of years in the export market. Survival rates of new exporters are significantly lower in Uganda when compared to its regional peers; on average 30 percent of new exporters in Uganda each year remain active in export markets the following year, and after the third year of entry only 12 percent still export. These are the second lowest 1-year and 3-year exporter survival rates among the regional peers (Panels c and d in Figure 32). Even after accounting for levels of development and year-product fixed effects, Uganda’s low exporter survival rates prevail (Table 9). 103 Uganda’s exports tend to survive longer in the EAC, the European Union (EU), and North America. Panel e in Figure 32 shows that Ugandan exporters exporting to the EU have the highest survival rate after the first year, at 44.5 percent, followed by the US and Canada, and Asian countries. The survival rate of firms in the EAC region after the first year is 38.5 percent, and others have less than 20 percent. This gap is likely driven by the concentration of cash crop-oriented exporters to the European and North American markets. Finally, Uganda’s large and medium-sized exporters have a relatively high survival rate (Panel f), in contrast to small and micro exporters. This is similar to other countries, where 1 in 4 exporters disappear after 5 years. Overall, the analysis here shows that Uganda has relatively low fixed costs to enter and exit the export markets, but competitiveness is a challenge for new- comers. Once accounted for the country size, level of development, and export products basket, the newcomers in Uganda are less likely to survive the competi- tive international markets for more than a few years.5 With high exit rates, larger and more productive firms remain in the market. As discussed in the next section, most of these surviving firms are large firms that both import and export. Smaller export- ing firms are competitively challenged as a result of limited access to imported in- puts (possibly because of import licensing requirements) or financing constraints. Imports-for-exports and domestic value added In Uganda, the most successful exporters are those who also import. In 2020, around 40 percent of exporters in Uganda were also importers. These exporter-im- porter firms accounted for 93 percent of the country’s overall export value in 2020, up from 87 percent in 2011. The decline in the number of exporters since 2011 affected both the pure exporters and the exporter-importers. These statistics show that export- er-importers are not only larger, but also grow faster than pure exporters. To con- firm these observation, Panels a and b in Figure 33 show the average and median size of firms that only export and those that export and import as well. In 2020, the average exporter-importer was almost 6 times larger than the average pure Growth, trade, and transformation exporter firm, up from 3.5 times in 2011. In addition, exporter-importers’ size in- creased 84% between 2011 and 2020. The comparative evolution of size by type of firm points to an increasing concentration of exports in larger firms, which are exporter-importers. Ugandan exporters have high domestic value-added ratios, which reflect the intensity of basic (agricultural) exports and low integration with global value chains (GVCs). Estimations following Kee and Tang (2016), show that Uganda has an average Domestic Value-Added Ratio (DVAR) of 0.90, which shows that most exports of Uganda have very low imports content. This is primarily driven by the fact that Uganda’s exports heavily comprise primary sector products (e.g., agri- culture), which typically exhibit high DVARs across countries. Furthermore, it also shows that Ugandan exporters are not integrated with GVCs, which often involves export products that use imported materials. Countries that are highly integrated with GVCs have lower DVAR, such as China (0.66) and Mexico (0.60). 104 Figure 32. Exporter entry, exit, and survival Exporter Entry Rate Exporter Exit Rate TZA 0,5 TZA 0,5 KEN 0,5 UGA 0,5 UGA 0,4 CMR 0,4 CMR 0,4 KEN 0,4 ETH 0,3 MUS 0,3 MUS 0,3 ETH 0,3 COL 0,3 COL 0,3 ZAF 0,2 ZAF 0,2 0 0,1 0,2 0,3 0,4 0,5 0,6 0 0,1 0,2 0,3 0,4 0,5 0,6 Exporter Entry Rate Exporter Exit Rate 1-Year Survival 3-Year Survival ETH 0,5 KEN 0,3 KEN 0,5 ETH 0,2 COL 0,5 COL 0,2 ZAF 0,4 ZAF 0,2 MUS 0,4 CMR 0,1 CMR 0,3 UGA 0,1 UGA 0,3 MUS 0,1 TZA 0,3 TZA 0,1 0 0,1 0,2 0,3 0,4 0,5 0,6 0 0,1 0,1 0,2 0,2 0,3 0,3 1-Year Survival Rate 3-Year Survival Rate Source: World Bank calculations based on data used for the Exporter Dynamics Database and Word Integrated Trade Solutions. Notes: Panels a-d show averages for 2010–2020 for Uganda, Kenya, Mauritius, and South Africa; 2010–2015 for Cameroon; 2000– Chapter 2. The role of international trade 2017 for others. Firms size by annual export value: Micro USD 10 million. Horizontal axes in Panels a-4 show the rate (e.g., 0.3 shows a 30 percent change) This two-sided problem (concentration in primary exports and low integration with GVCs) is also evident in sector-specific DVAR estimates. Panel c in Figure 33 presents the sectoral DVAR of Uganda and a few selected comparison countries. The DVAR of Uganda is among the highest in most of the sectors, including agricul- ture, chemicals, textiles/apparel, machinery, and glass/cement sectors. This sug- gests that Uganda is not integrated with GVCs, as most of its exports, even indus- trial products, do not rely on imported materials. Separately, a destination-based estimation of DVARs shows that Uganda’s exports have the highest DVAR in the EU and the US. Both markets import mainly agriculture products, such as coffee and fish fillets from Uganda. Thus, the high DVAR problem in Uganda is driven by both concentration of exports in primary commodities and a low degree of integration with GVCs in other sectors. 105 Figure 33. Exporter size and domestic value added Average Size of Exporter Median Size of Exporter 2,5 2,5 2,12 2,12 2,12 1,97 1,95 1,97 2,0 2,0 Exports, Million USD Exports, Million USD 1,70 2,12 1,70 1,95 1,47 1,47 1,41 1,5 1,41 1,5 1,15 1,46 1,15 1,46 1,40 1,40 0,99 0,93 0,99 0,93 1,0 1,0 0,88 0,88 0,44 0,46 0,44 0,46 0,5 0,37 0,5 0,37 0,56 0,56 0,40 0,40 0,32 0,40 0,40 0,32 0 0 2020 2020 2014 2014 2015 2016 2018 2019 2015 2016 2018 2019 2013 2013 2012 2012 2017 2017 2011 2011 Exporters Only Exporter-Importers Agriculture Chemicals Foodstufss/Beverages Footwear/Headgear 1,2 0,9 1,2 0,9 0,9 0,8 0,8 0,8 0,8 1,0 1,0 1,0 0,9 0,9 0,9 1,0 1,0 0,7 0,9 0,7 0,8 0,9 0,7 0,6 0,8 0,9 0,6 0,6 0,8 0,8 0,6 0,6 0,6 0,8 0,7 0,6 0,5 0,8 0,6 0,5 DVAR Ratio DVAR Ratio DVAR Ratio DVAR Ratio 0,6 0,7 0,5 0,4 0,5 0,5 0,6 0,6 0,4 0,4 0,4 0,4 0,4 0,3 0,4 0,3 0,2 0,2 0,2 0,2 0,1 0,1 0 0 0 0 MUS CMR ETH TZA COL ZAF UGA KEN ETH MUS COL CMR TZA UGA KEN ZAF KEN MUS COL UGA ZAF TZA ETH CMR UGA TZA KEN MUS CMR ZAF COL ETH Glass/Cement Machinery Metals Mineral 1,0 0,9 0,8 0,8 1,0 1,2 0,9 0,8 0,8 0,8 0,9 0,9 0,8 0,8 0,9 0,8 0,9 1,0 0,7 0,8 1,0 0,9 0,8 0,8 0,8 0,8 0,7 0,7 0,7 0,7 0,6 0,7 0,7 0,8 0,8 0,8 0,7 0,6 0,7 0,6 0,8 0,6 0,6 0,6 0,8 DVAR Ratio DVAR Ratio DVAR Ratio DVAR Ratio 0,6 0,6 0,5 0,6 0,6 0,5 0,5 0,6 0,4 0,4 0,4 0,3 0,4 0,3 0,3 0,2 0,2 0,2 0,2 0,1 0,1 0,1 0 0 0 0 ZAF COL TZA CMR MUS KEN UGA ETH COL CMR ETH MUS KEN ZAF TZA UGA KEN MUS UGA ZAF CMR TZA COL ETH UGA MUS KEN COL TZA ETH CMR ZAF Miscallenous Optical Equipment Paper/Pulp Plastic/Rubber 1,0 0,9 0,9 1,0 1,0 0,9 0,9 0,9 0,9 0,8 0,8 0,9 Growth, trade, and transformation 0,9 0,8 0,9 0,9 0,9 0,8 0,7 0,7 0,7 0,8 0,8 0,8 0,8 0,8 0,8 0,7 0,8 0,7 0,7 0,7 0,7 0,7 0,7 0,7 0,6 0,7 0,6 0,6 0,5 DVAR Ratio DVAR Ratio DVAR Ratio DVAR Ratio 0,6 0,6 0,6 0,5 0,5 0,5 0,5 0,5 0,5 0,5 0,5 0,5 0,5 0,4 0,4 0,4 0,4 0,3 0,3 0,3 0,3 0,2 0,2 0,2 0,2 0,1 0,1 0,1 0,1 0 0 0 0 UGA COL CMR MUS KEN TZA ETH ZAF MUS COL TZA UGA KEN ZAF CMR COL ETH CMR KEN UGA MUS TZA ZAF KEN MUS COL UGA ZAF TZA ETH CMR Raw hides/Leather Textiles/Apparel Transport Wood 1,0 0,9 0,9 0,9 1,2 1,0 1,2 0,9 0,9 0,9 0,9 1,0 1,0 0,8 0,8 1,0 0,9 0,8 1,0 0,9 0,8 0,7 0,9 0,9 0,9 0,8 0,7 0,9 0,6 0,7 0,8 0,7 0,7 0,8 0,8 0,8 0,7 0,7 0,8 0,6 0,6 DVAR Ratio DVAR Ratio DVAR Ratio DVAR Ratio 0,6 0,6 0,6 0,6 0,6 0,5 0,6 0,5 0,6 0,4 0,3 0,4 0,4 0,4 0,3 0,3 0,2 0,2 0,2 0,2 0,1 0,1 0 0 0 0 MUS CMR KEN UGA COL ZAF ETH TZA COL MUS TZA ETH ZAF KEN UGA CMR KEN COL CMR TZA UGA MUS ZAF ETH UGA COL KEN MUS ZAF TZA CMR Source: World Bank calculations based on Exporter Dynamics Database. 106 Overall, the firm-level analysis has identified both strengths and weaknesses in Uganda’s trade performance. On the positive side, dynamic firm entry patterns point to a business environment with low entry cost, which encourages firms to participate in exporting. However, firms who fail to secure imported inputs seem to struggle in maintaining competitiveness in international markets, and larger firms who survive seem to be both exporters and importers. These observations are also confirmed by the high domestic value added of Ugandan exporters, which also highlight problems in import penetration in the country. The analysis will next dis- cuss potential conditions and policy-driven obstacles that may have generated such trends. The Determinants of Uganda’s External Trade Trade policy measures and logistics conditions shape Uganda’s export pat- terns to a large extent. Being a landlocked country, Uganda relies heavily on its economic relationship with its immediate neighbors. Regional markets have been important drivers of Uganda’s export growth. About 96 percent of Uganda’s transit comes from the port of Mombasa through the Northern corridor, and some of its exports are exclusively oriented towards regional markets. For example, more than 90 percent of all dairy exports by Uganda go to EAC partners. Similarly, small-scale cross-border trade is estimated to be more than 60 percent of total formal trade in goods in Uganda. Thus, the country’s prospects for exports depends crucially on logistics and trade policies within the region and between the region and the rest of the world. Trade policy and regional integration Uganda is a member of the EAC, and its tariff policy is largely shaped by the Chapter 2. The role of international trade rules and regulations of the union. These include the Common External Tariff (CET) system, for imports originating from outside the union; an 80-percent discount in such CET, for imports originating from other countries in the Common Market for Eastern and Southern Africa (COMESA); and duty-free trade within the EAC. In ad- dition to tariffs, Uganda also employs several non-tariff measures at the border, which are not subject to EAC regulations. These include requirements for import permits and import licenses. Licenses are valid for six months from issuance, and are renewable. According to the World Trade Organization (WTO, 2019), examples include registration and approval from the Agricultural Chemicals Board (ACB) for agri-chemical imports and from the National Drug Authority (NDA) for imports of human and animal medicine. 107 Table 10. Trade policies of Uganda by broad industry groups Tariff Measures AVE Non-Tariff Measures (NTMs) Products Simple Weighted Products Simple Weighted with average average with average average tariff>0 tariffs tariffs NTMs NTMs NTMs Agricultural 24.7 3.9 0.4 100.0 1.0 0.2 raw materials All Food 82.9 24.1 22.5 91.8 5.3 0.9 Manufactures 57.9 5.8 0.6 100.0 0.5 0.0 Notes: World Bank estimates based on UN-COMTRADE. AVE = Ad valorem equivalent. Figure 34. Trade policies of Uganda and comparison countries, 2018––all products a. Tariff Measures b. Non-Tariff Measures Share Of Products With Tariffs Share Of Products With NTMS 100 100 90 90 80 80 70 70 Percent Percent 60 60 98.9 98.0 96.6 97.9 50 50 92.6 92.3 91.6 86.1 82.2 82.2 79.4 82.1 40 40 70.6 71.8 30 30 4 38.7 20 20 25.3 10 10 Growth, trade, and transformation 0 0 CIV VNM VNM MAR NGA GHA UGA KEN SEN NGA CIV GHA MAR UGA KEN SEN Simple Average Tariff Simple Average AVE-NTM 25 7 6 20 5 15 Percent Percent 4 22.8 6.4 6.4 3 10 5.0 5.0 5.1 4.4 15.4 14.4 2 13.6 12.6 12.5 3.2 11.4 5 7.3 1 1.4 0 0 CIV VNM VNM MAR NGA GHA UGA NGA KEN SEN GHA CIV UGA MAR KEN SEN 108 Weighted Average Tariff Weighted Average AVE-NTM 25 4 4 20 3 3 Percent Percent 15 2 3.8 22.3 10 2 12.9 1 1.8 11.1 5 8.0 1.5 1.3 7.7 6.9 6.6 0.9 0.9 1 0.8 4.2 0.6 0 0 CIV VNM CIV KEN GHA SEN NGA UGA MAR VNM NGA GHA UGA MAR KEN SEN Source: World Bank estimates based on UN-COMTRADE. Notes: AVE-NTMs = Ad valorem equivalent non-tariff measures. Uganda’s trade policies, especially its protections against food imports, are quite restrictive. Table 1 summarizes the tariff and non-tariff measures (NTMs) in key sectors, where NTMs are reported in ad valorem tariff equivalence (AVE) terms. About 92 percent of all imported food products face NTMs at the border. The simple average tariffs and AVEs of NTMs in food products are 24 per- cent and 22.5 percent, respectively. Examples of very high tariffs include rice (at 75 percent), corn (60 percent), and dairy products (60 percent). In contrast, oth- er product groups, including agricultural raw materials and manufactures, faced considerably lower tariffs and NTMs at lower-middle single digits. In recent years, import duties and value-added tax (VAT) collected at the border yielded about 40 percent of all fiscal revenues. Average tariffs in Uganda are among the highest compared to other countries. Chapter 2. The role of international trade To benchmark Uganda’s trade policies, Figure 34 compares its tariff and NTMs with those in Côte d’Ivoire (CIV), Ghana (GHA), Kenya (KEN), Morocco (MAR), Nigeria (NGA), Senegal (SEN), and Vietnam (VNM). Although Uganda does not stand out in its tariff coverage, its average simple and import-weighted tariffs are the highest in the group. In comparison, the aspirational examples in the group, Morocco and Vietnam, have significantly lower tariff barriers. For example, Vietnam, an agricultur- al success case, has less than 40 percent of products facing positive tariffs and its import weighted average tariff is only 4 percent. Protectionist policies are a double-edged sword: while they facilitate import substitution in certain sectors, they also reduce opportunities in other sectors, and hurt consumers. The import protection provided by the EAC CET framework has created an EAC market for dairy producers in Uganda, and led to a rapid ex- pansion of within-EAC exports since mid-2000s (see the spotlight on dairy at the end of this chapter), which may be challenged under the AfCFTA (see Chapter 3). 109 Figure 35. Unilateral deviations from the CET Unilateral deviations from the EAC CET have increased significantly in recent years. Deviations from the CET Source: Rauchendorfer and Twum (2020). However, as shown above, successful Ugandan exporters are also importers, and tariffs and stringent license requirements can hurt Ugandan exporters by reducing access to intermediate inputs. Moreover, they also increase the domestic prices for consumers, reducing their welfare, and increasing inequality, as discussed below. Trade frictions around the EAC CETs have deepened in recent years. The EAC members have increasingly employed deviations from the CET. Rauschendorfer and Twum (2020) show that in 2019/20 Uganda was, with Kenya, the largest source Growth, trade, and transformation of deviations from the CET (Figure 35). While stay of applications only apply to im- ports from outside the EAC, their frequency indicates strong protectionist motives. For example, since 2019, in response to a diplomatic dispute, Rwanda has closed its border with Uganda, effectively stopping all exports from Uganda. The slack was picked up by Tanzanian exports. Kenya has also blocked Ugandan products, including maize, sugar, and dairy, citing concerns about origins of these products and about phytosanitary standards in Uganda. These frictions serve as a reminder that Uganda should target global compet- itiveness rather than protection-driven development. There is a clear tension between Uganda’s ambitions to harness regional integration for greater exports and the country’s attempts to shield its domestic market for local firms. For exam- ple, the ‘Buy Uganda Build Uganda’ policy places a strong emphasis on enhancing local content in domestic production and procurement, but it also triggers retalia- 110 tory policies by Uganda’s neighbors. Instead of taking more protectionist stances, which endanger further disputes, the EAC members, including Uganda, should aim to boost the competitiveness of their producers outside of EAC CET protection. This requires the elimination of any barriers against imports and input access, a conducive business environment, and improvements in public infrastructure, which is discussed next. Trade logistics and infrastructure Uganda’s trade facilitation is an important aspect of the regional cooperation agenda. Several trade facilitation schemes that are implemented through the EAC, like the Single Customs Territory (SCT), have significantly reduced transaction costs in Uganda. For example, in March 2019, the EAC Regional Electronic Cargo Tracking System (RETCS) for transit trade (implemented by Uganda, Kenya, and Rwanda, and planned by DRC and South Sudan) reduced the time for cargo movement between Mombasa to Malaba from 2.35 days to 1.75 days. Uganda has also implemented one-stop border posts with Kenya (2), Rwanda (3), Tanzania (1), DRC (1), and South Sudan (1). Uganda’s efforts to facilitate trade have improved its logistics, but there is room for further progress. Uganda’s Logistics Performance Index (LPI) results are com- parable to those of Rwanda and Tanzania, but they fall behind those of Kenya (Fig- ure 8). Uganda’s performance lags comparatively in three categories: infrastructure, logistics competence, and tracking and tracing. The first shortcoming (infrastruc- ture) reflects an over-reliance on road transport. The latter two dimensions (logis- tics competence, and tracking and tracing) reflect a logistics market that is rel- atively underdeveloped, as Ugandan freight and logistics operators have trouble competing with larger and more sophisticated neighbors. Chapter 2. The role of international trade Improving information technology and modernizing/reducing controls are es- sential for efficient border management in a high-volume environment. With the support of TradeMark East Africa (TMEA), Uganda is making progress in the implementation of the Uganda Electronic Single Window (UESW). This helps to de- crease document submission and better connect government agencies. According to the Uganda Revenue Authority (URA), the system has reduced the clearance time of goods from 21 days to 4 days for imports and 2 days for exports. A more recent assessment shows that the Uganda National Bureau of Standards clearance is now issued in 2 hours as opposed to 2 days. Under the Authorized Economic Op- erator (AEO) scheme, beneficiaries now represent 26 percent of customs revenues, and the time to clear cargo under AEO has reduced from 4 days to 12 hours. Further progress in border processes rely on (i) increased cooperation with neighbors and (ii) improved infrastructure. Despite some progress, operators still report frequent customs and police check points on transit routes and ad hoc fees by local authorities at border crossings. The RETCS for transit suffers from a lack of electronic tracking devices and differing practices among countries in the region. 111 Figure 36. Trade facilitation and logistics indicators Uganda’s LPI scores perform well in customs and timeliness, but lag in infrastructure, tracking & tracing, and logistics competence. 3,6 Customs 3,4 3,2 Uganda 3 Vietnam Timeliness 2,8 Infrastructure 2,6 Rwanda Ghana 2,4 Tanzania 2,2 Ethiopia 2 Kenya International Tracking & tracing shipments Logistics competence Source: World Bank Logistics Performance Indicators, https://lpi.worldbank.org/. Similarly, differences in the implementation of the regional AEO scheme hamper the smooth running of the program. Additional support to the EAC secretariat and the corridor organization (the CCTTFA) could help iron out some of these issues. More progress in e-government can further help trade flows by reducing costs. Growth, trade, and transformation These policies should be accompanied by adequate investments in IT infrastruc- ture and Internet access at borders, both of which are sometimes deficient and thus impair the effective functioning of regional schemes such as the RETCS. There is scope for investments in transport and logistics infrastructure to sup- port the development of regional economic linkages and value chains. Although landlocked, Uganda is a key transit zone for South Sudan and eastern DRC, both relatively insecure areas. Uganda could offer services that are difficult to access on the other side of the border, including bonded warehouses (e.g., custom duties are paid only when the goods are removed from the warehouse) and value-added services like consolidation, and light assembly or processing. The Gulu logistics hub project (under development) provides a good example in this regard. A road-intensive transportation system suffers from several inefficiencies. With large imbalances between import and export volumes, there is a lot of excess 112 capacity on the leg from Uganda to Kenya, which in the long-term is internalized as additional freight costs. Moreover, trucks are stopped before getting sanitary and phytosanitary (SPS) certificates, and less than container loading (LCL) is not autho- rized by the Uganda Coffee Development Authority (UCDA), increasing costs and delays for smaller batches of merchandise like coffee and SME exports. Uganda is a price-taker in logistics, which comes with disadvantages. Smaller traders in partic- ular do not have much bargaining power with multinational logistics companies and lack control over the supply chain from the moment their goods are loaded onto trucks in Uganda. Price markups seem high on transport from Uganda to Momba- sa’s port. For instance, trading free on board (FOB) in Mombasa fetches a premium of USD 1,650–1,850 for a 20ft container of coffee (Mugenyi 2020) relative to the price paid free on truck (FOT) in Uganda. A modal shift from road to rail or rail/lake transportation can help reduce costs, but requires significant cooperation across borders. On Lake Victoria, which ac- counts for the bulk of formal water transport services, infrastructure is generally in poor condition along the Ugandan waterway network and the two main transit routes to Tanzania and Kenya (Port Bell/Jinja-Mwanza and Port Bell/Jinja-Kisumu, respectively). New lake transport operators seem interested in entering the market with one planning to put two thousand-ton ships into service. Port Bell benefited from the dredging of the entrance channel and the construction of a new oil jetty, which enables it to also handle transit petroleum products and other cargo. How- ever, it faces significant encroachment, and transport onward into Kampala is unre- liable and encumbered by the urban sprawl. A project has been in place to replace it with a new port in Bukasa, with its own challenges, on the outskirts of Kampala. In the long run, regional infrastructure cooperation will be the decisive factor in Uganda’s shift towards more cost-effective modes of transportation. Extending the standard gauge railway from Naivasha (Kenya) to Kampala (over 500 km) re- quires a considerable investment, with most of the line falling on the Kenyan side Chapter 2. The role of international trade of the border. In the short- and medium-term, the existing capacity of the current meter gauge railway can be relied upon, but this will require better rehabilitation and maintenance, as well as an expansion of the rolling stock and institutional capacity. With Tanzania, securing the revival of the rail/water multimodal route through Port Bell and Mwanza can provide an alternative to the road-only Northern Corridor. However, this would also require significant investments on the Tanzanian side. Targeted interventions around key competitive export value chains can help. The discussion above emphasizes the need to prepare Uganda’s emerging sec- tors for global competition. The dairy industry is an important example. Beyond the tariff protection it has received, it relies on efficient collection systems, entailing investments into milk collection centers (MCCs). While this analysis has not con- ducted feasibility surveys, the scope for a more comprehensive approach to set- ting up MCCs in the Central region is evident. This includes assessing the needs of local infrastructure: community access roads need improvement, and despite ef- forts, the MCCs still face electricity problems. There may also be needs for further government services (e.g., phytosanitary, veterinary services) in the target areas. 113 Figure 37. The impact of lower food prices on structural transfor- mation (10-year) In the business as usual scenario, the labor share of agriculture decreases from 64.7 percent to 63.1 percent in a decade. With 10 percent lower agricultural prices, it decreases to 58.5 percent. Labor share projections 70 64.7 63.1 60 58.5 50 Percent 40 30 24.3 21.5 22.0 20 14.9 17.1 13.8 10 0 Agriculture Manufacturing Services As a result, GDP grows faster with lower food prices. Agricultural value added shrinks more (1.7 points relative to initial GDP), but that is small compared to the additional gains in manufacturing (6.8 points) and services (4.3 points). Value added (as a share of base-year GDP) 80 71.6 69.3 70 64.8 65.0 60 50 42.0 Percent 40 36.0 30 21.9 20 16.1 14.4 10 0 Agriculture Manufacturing Services In the business as usual scenario, the value added share of agriculture decreases by 10.9 percentage points (relative to current GDP) in the next decade. With lower food prices it Growth, trade, and transformation decreases an additional 1.7 percentage points. Value added (as a share of current GDP) 70 60 50 44.4 46.1 44.6 44.6 42.0 Percent 40 36.0 30 21.9 20 11.0 9.3 10 0 Agriculture Manufacturing Services Base year Business as usual projections 10% lower food prices Source: Artuc, Leunga, and Onder (2022). 114 Trade facilitation efforts should also address regulatory burden. Ugandan coffee exporters are subject to numerous (36 steps) regulations, including UCDA export licenses, performance bonds, and strict quality and SPS requirements, among oth- ers (Mugenyi). While strict controls are necessary for brand-building, these mea- sures can be more efficient. For instance, inspections by the UCDA and the Ministry of Agriculture (MAAIF) could be co-located in regional offices closer to production sites. In fact, with only two available laboratories for conformity tests, the costs are often prohibitive (EU-EAC MARKUP, 2020). UCDA grading is also not meeting all needs of exporters, as some grades demanded by buyers are not graded by UCDA (Mugenyi). Coffee exports, while having been streamlined, do not fully benefit from trade facilitation measures: only operators working with large freight companies benefit from the AEO scheme, physical documents continue to be required, and mobile payments are not accepted. Trade’s Effect on Transformation and Inequality The patterns of Uganda’s international trade, and the underlying factors that shape these patterns, have important implications for the country’s structural transformation. Productivity growth drives structural transformation in completely opposite ways in closed and open economies. The traditional view is that pro- ductivity growth makes agriculture shed labor. But other things being equal, this is only true for closed economies (Matsuyama 2009), where production and house- hold choices (e.g., consumption or investment) must clear the market. In this case, when a sector grows faster than others, labor moves out of it because households maintain the composition of their consumption basket (Ngai and Pissarides 2007). Similarly, if sectors have identical productivity growth, but households consume proportionally more manufactured goods or services as they become richer (i.e., Chapter 2. The role of international trade non-homothetic preferences), labor can still move out of agriculture (Kongsamut, Rebelo, and Xie 2001). In contrast, for small open economies, the link between con- sumption and production is broken, and productivity growth in, say, agriculture, can attract labor rather than shedding it. The next two sections will investigate in more detail the role of trade and trade policy for Uganda’s structural transformation. Specifically, the analysis will first consider how changing international prices, e.g., as a result of lower import tariffs in Uganda, can affect the country’s structural transformation. This will be based on the structural transformation model developed in the previous chapter (see Ap- pendix B for technical details). Next, it will consider the distributional effects of these price adjustments with and without structural transformation. This will be done by using a partial equilibrium model of households’ decisions on labor supply and agricultural land allocation, relying on detailed household survey data on in- come, employment, and consumption for realism. 115 Figure 38. Distributional impact of a 10-price decline in all agricultural products Consumption and Income Share - Scenario: All Agriculture 50 40 30 Share 20 10 0 0 10 20 30 40 50 60 70 80 90 100 Income percentile Consumption share Income share Welfare Change - Scenario: All Agriculture 3.5 % change in welfare 3 2.5 2 1.5 1 0.5 0 -0.5 0 10 20 30 40 50 60 70 80 90 100 Income percentile Short-run Long-run Short-Run Welfare Change Composition - Scenario: All Agriculture 6 % change in welfare 5 4 3 2 1 0 -1 -2 -3 0 10 20 30 40 50 60 70 80 90 100 Growth, trade, and transformation Income percentile Consumption Labor income Land income Total Long-Run Welfare Change Composition - Scenario: All Agriculture 6 % change in welfare 5 4 3 2 1 0 -1 -2 -3 0 10 20 30 40 50 60 70 80 90 100 Income percentile Consumption Labor income Land income Total Source: World Bank calculations based on Artuc, Porto and Rijkers (2021a and 2021b). 116 How does trade policy affect structural change? To quantify the role of trade policy in Uganda’s structural transformation, the analysis focuses on price dynamics of agricultural goods in the next decade. The earlier analysis in this chapter showed the disproportionately high tariffs on food imports in Uganda. In order to analyze how these affect the country’s sec- toral labor and output composition, the analysis here runs a hypothetical scenar- io through the lens of the structural change model developed in Artuc, Leunga, and Onder (2022) for this report. Specifically, this scenario fixes the growth rate of exogenous variables at their historical trends (10-year) for the next decade, and compares the effects of a 10 percent decrease in international prices of food (as a result of tariff reductions, removal of trade frictions, or other factors). Figure 9 shows the results. The baseline projections show a moderate TFP-driven transformation in Ugan- da within the next decade. In the business as usual scenario, all exogenous vari- ables (e.g., sectoral TFP, international prices, and mobility costs across sectors) trend at the same rate as in the last decade. As a result, the labor share of agricul- ture decreases from 64.7 percent to 63.1 percent by the end of the next decade (the first panel in Figure 9), while those of manufacturing and services increase by about 1.1 and 0.5 percentage points, respectively. These trends are largely driven by faster productivity growth in manufacturing and services, a continued produc- tivity suppression in agriculture, and high mobility costs between sectors. As a re- sult, agriculture’s value-added shrinks by 5.8 percent of the initial GDP, and those of manufacture and services increase by 28.8 and 23.0 percent, respectively (the second panel). Lower food prices can accelerate growth and transformation. With lower food prices, more labor moves out of agriculture (4.6 percentage points more than that in the business as usual scenario). As a result, manufacturing and services employ Chapter 2. The role of international trade more workers (2.2 percentage points and 2.3 percentage points, respectively). As a result of this reallocation of labor from relatively low productivity activities to higher ones, the aggregate GDP grows by an additional 10 percentage points in the next decade. The value added in agriculture shrinks by 1.7 percentage points more with lower food prices (14.4 percentage points of initial GDP with lower food prices vs. 16.1 points in business as usual). However, with the gains in manufacturing (an additional growth equivalent to 6.8 percent of initial GDP) and services (4.3 ser- vices), that loss is small. In the end, a 10 percent reduction in food prices boosts GDP growth and accelerates the sectoral transformation of the economy. How does trade policy affect inequality in uganda? Besides affecting the scale and sectoral composition of the economy, trade policy also has important distributional implications in Uganda. The discussion in the previous chapter showed that, among the factors that limit rural-urban and sectoral transitions in Uganda, land ownership structure and small-scale agricul- tural production for self-consumption played a key role. Together with tight credit 117 constraints and sizeable frictions in liquidity of assets, like the customary land ten- ure, these factors can shape the effects of trade policies on different segments of the Ugandan society. For example, how import tariffs affect Ugandans’ income and welfare can vary by the consumption and production characteristics of different households. Whereas a reduction in food prices can increase the welfare of a net consumer of such goods, it can decrease the welfare of those who are net produc- ers. Given these nuances, this section sheds light on income-differentiated effects of trade policies. For a nuanced assessment of distributional implications, the analysis here em- ploys a partial equilibrium model drawing from Ugandan household surveys. This approach helps to study households’ decisions on labor supply and agricul- tural land allocation to analyze the impact of exogenous price shifts on detailed income inequality. The model features 53 agricultural commodities and 10 sectors, including agriculture, where households’ production and expenditure shares for each product are identified following Artuc, Porto, and Rijkers (2021a and 2021b). The analysis separates the short-term effects from long-term effects. The former approach blocks structural transformation channels by fixing land and labor allo- cations in the short-run. This can be because intersectoral mobility costs are pro- hibitively high. In comparison, in the long-term analysis, households can respond to shocks by changing their employment and land allocation decisions in a costly manner. This adjustment is calibrated by using factor response elasticities from Sotelo (2020) and Lagakos and Waugh (2013). Finally, following from the previous section, the analysis first considers the effects of a 10 percent reduction in the price of all food imports. The results are presented in Figure 10. Next, it repeats the same exercise for key sectors (cash crops, dairy, and cereals), individually. Those results are presented in Figure 39. In Uganda, the poorest households are the most likely to be the net consum- ers of agricultural goods and the low- and middle-income households are the least likely. The first panel in Figure 10 shows the household income percentiles on the horizontal axis, and the share of households’ income and consumption asso- ciated with agricultural products on the vertical axis. The poorest households are Growth, trade, and transformation spending a much higher share of their income on food compared to other house- holds. Low- and middle-income households seem to spend the smallest share of their income on food, possibly due to self-consumption, as they are more likely to engage in farming. This group of households has the highest shares of income as- sociated with agriculture, which increases until the 70th percentile and decreases afterwards. Food consumption patterns also differ between low and high income Ugandans. For example, while the latter consume more wheat and rice, the former consume more corn and other cereals in relative terms. An across-the-board reduction in food prices increases welfare across all in- come groups, but especially for those in the lower and higher income percen- tiles. The second panel in Figure 10 shows the short-run and long-run welfare im- pact of a 10 percent reduction in all food prices. Low-income households are more likely to benefit from this shock as they are more likely to be net food consumers and they spend a large share of their income on food. Middle-income households 118 benefit the least since their consumption-production gap is smaller than other in- come groups. Interestingly, however, in the long-run, low-income households will be more likely to benefit relative to the short-run. To explain this pattern, the welfare effect needs to be further decomposed, which is done in the third and fourth pan- els in the same figure. The inability to switch sectors magnifies the wage suppression effect of lower agricultural prices. The third panel in Figure 10 decomposes the short-run effects. The negative effects of the agricultural price decline on low-income households are caused by reduction in wages of households at the bottom of the income distribu- tion, who are more likely to provide labor to the agriculture sector. Higher income households are more likely to own land and less likely to provide agricultural labor. When factors reallocate in the long-run (the fourth panel), workers move away from agriculture, hence the marginal product of labor increases after the initial decline as the number of workers decline and wages rebound partially. However, it is import- ant to note that these effects can vary across different products within the broader category of agricultural goods, which is analyzed next. The income and consumption shares of Ugandans vary significantly across dif- ferent agricultural products and income groups. The left-hand side panels in Fig- ure 39 show the income and consumption shared for three product groups: cash crops, dairy, and cereal. Ugandans, regardless of their income level, do not spend a significant share of their budget for consuming cash crops like coffee, tea, and tobacco (first panel). However, middle-income Ugandans rely on these crops for income more than other income groups. In the dairy sector, both the income share and the consumption share of the sector increase monotonically with income. As for cereals, the income shares of Ugandans exhibit a familiar inverted-U shape, with middle-income Ugandans relying more on cereal production for income. However, households in the lowest income percentiles spend a significantly larger share of their consumption budget on cereals, and they do not rely on cereals for income. Chapter 2. The role of international trade The welfare effect of reducing the price of a single good is relatively small, but it varies across income groups nonetheless. The panels on the right-hand side of Figure 39 show the welfare effects of a 10-percent reduction in prices of cash crops, dairy, and cereals, individually. With negligible domestic consumption of cash crops, income effect dominates, and almost all income groups receive negative welfare shocks when the prices of cash crops decrease. This effect is somewhat more pronounced for households in the middle-income group who earn a larger share of their income from these products. The welfare impact of a dairy price reduction is generally positive, increasing income; i.e., the households at the higher end of the income distribution are more likely to gain from such price reduction. This is largely because low-income Ugandans do not (i.e., cannot) consume much dairy. Finally, a reduction in cereal prices has a large positive effect on low-income Ugandans who spend a large share of their consumption budget on cereals. In all these three cases, individual sectors constitute a small share of the larger agricul- tural sector. Thus, the differences between short-term (without adjustment) and long-term (with adjustment) effects are small. 119 Figure 39. Price changes by sector, structural transformation, and inequality Consumption and Income Share - Welfare Change - Scenario: Scenario: Cash Crops Cash Crops 6 0.2 0.1 5 % change in welfare 0 4 -0.1 3 -0.2 Share -0.3 2 -0.4 1 -0.5 0 -0.6 -0.7 -1 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 Income percentile Income percentile Consumption and Income Share - Welfare Change - Scenario: Scenario: Dairy Dairy 2.5 0.25 0.2 2 % change in welfare 0.15 1.5 0.1 Share 1 0.05 0 0.5 -0.05 0 -0.1 -0.15 -0.5 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 Income percentile Income percentile Consumption and Income Share - Welfare Change - Scenario: Scenario: Cereals Cereals Growth, trade, and transformation 18 2 16 14 1.5 % change in welfare 12 1 10 Share 8 0.5 6 4 0 2 0 -0.5 -2 -1 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 Income percentile Income percentile Consumption share Income share Short-run Long-run Source: World Bank calculations based on Artuc, Porto, and Rijkers (2021a and 2021b). 120 Overall, the analysis here highlights significant but differentiated patterns of the relationship between trade policy, structural transformation, and inequali- ty in Uganda. Lower agricultural prices can be achieved by reducing trade frictions and import tariffs, and they are generally pro-poor. Low-income Ugandans broadly benefit from lower prices as the impact on their consumption costs exceed other channels, like lower wages in agriculture. In cash crops like coffee, tea and tobacco, domestic consumption is very low, thus, low prices hurt all income groups. However, in that case, the trade policy tools controlled by Uganda, like reducing trade costs, are designed to increase the net prices faced by Ugandan producers rather than decrease them. Therefore, even when the concerns about access to intermediate inputs are set aside, improving trade distortions at the border is still highly aligned with the inclusive growth and structural transformation agendas in Uganda. Synopsis This chapter analyzed Uganda’s international trade. The analysis first focused on the country’s trade patterns in recent years, then investigated the firm-level drivers of those trends by using detailed firm-level customs data that revealed exporter size, and entry and exit characteristics. Next, it considered factors that affect firm-level and aggregate trade flows, with a special emphasis on trade pol- icy, regional integration, and trade logistics and infrastructure. Finally, the analysis employed two models to investigate how Uganda’s international trade policy af- fects its structural transformation and inequality. This was done by using the aggre- gate structural transformation model developed in the previous chapter and a new model that incorporates the differentiated income and consumption relationship of households with 53 different agricultural products using a discrete choice model drawing from micro survey data. The key findings are as follows. Chapter 2. The role of international trade There are both strengths and weaknesses in Uganda’s trade performance. The firm-level analysis in this chapter revealed the following observations: • On the strengths side, a dynamic firm entry pattern shows that, with a fa- vorable business environment, entry costs are low, which encourages firms to participate in exporting. • On the weaknesses side, there are symptoms of major distortions in import markets. Those firms who cannot import inputs fail to maintain competitive- ness in international markets, but larger firms who both export and import survive. This is also confirmed by higher than expected domestic value-add- ed rates of Ugandan exporters. Uganda faces important natural obstacles against trade (e.g., landlocked loca- tion), but there are important man-made impediments too, and those should be addressed. The analysis in this chapter identified important factors that limit Uganda’s trade potential: 121 • Trade policy: Uganda’s trade policies, especially its protections against food imports, are quite restrictive. Average tariffs in Uganda are among the high- est compared to other countries. While they facilitate import substitution in certain sectors, they also reduce opportunities in other sectors, and hurt consumers. • Regional integration: Uganda relies on regional partners for connectivity and market access, but disputes about country of origin and deviations regarding the EAC CET have been growing in recent years. These frictions are a good reminder that Uganda should target global competitiveness rather than pro- tection-driven development. • Logistics and infrastructure: Uganda’s Logistics Performance Index (LPI) re- sults lag in three categories: infrastructure, logistics competence, and track- ing and tracing. The first shortcoming (infrastructure) reflects an over-reli- ance of road transport. The latter two dimensions reflect an underdeveloped logistics market as Ugandan freight and logistics operators have trouble competing with larger and more sophisticated neighbors. Protectionist policies and trade frictions in agriculture slow down growth and structural transformation, and they hurt the low-income Ugandans. Simulations in this chapter showed that a 10 percent reduction in food prices, which can be achieved by tariff reduction and lower trading frictions, can increase GDP growth by an additional 10 percentage points in a decade by reallocating labor from low productivity agriculture to manufacturing and services. Lower prices boost the welfare of the poorest Ugandans who rely less on agriculture for direct income (e.g., wage workers in rural areas who do not own land) and spend a large share of their expenditures for food. Growth, trade, and transformation 122 Spotlight A. Uganda’s Booming Dairy Sector Chapter 2. The role of international trade Source: The Market Project (2020). Uganda’s dairy sector has boomed over the past decade. With strong domestic and external demand, Uganda’s annual milk production increased from 1.4 billion liters of milk in 2006 to 2.2 billion litters in 2018 (van Campenhout et al. 2019). Domestic demand was driven by population growth and an increase in per capita consumption—the average Ugandan doubled her milk consumption over the pre- ceding decade. But exports played the predominant role. In fact, dairy products have become one of Uganda’s fastest-growing export commodities, increasing 7-fold, from USD 10 million to USD 70 million, between 2010 and 2020 (Figure 40). These exports mostly consist of milk and milk powder, which make up about two- thirds and a third of total dairy exports, respectively. More sophisticated products, like butter, are also increasingly finding their ways into foreign markets. In terms of their share in the country’s export basket, dairy exports have become comparable to long-established products like tea or cut flowers. 123 Figure 40. Uganda’s growing exports of dairy With the enactment of the EAC CET, Uganda dairy exports took off. Dairy Exports 80 70 Exports (USD, Millions) 60 50 40 30 20 Establishment of the 10 EAC Customs Union 0 2000 2005 2010 2015 2020 EAC RoW Source: Constructed from UN Comtrade data (2021). Notes:: Data Source for 2019from Constructed UN Comtrade is missing. data (2021). EAC = Kenya, Notes: Tanzania, DataBurundi, Rwanda, for 2019and is missing. EAC = Dairy South Sudan. Kenya, Tanzania, includes Rwanda, milk, Burundi, and milk powder, South Sudan. butter, Dairy cheeses, includes curd, milk, milk powder, butter, cheeses, curd, and eggs. and eggs. The Ugandan milk economy is roughly subdivided into two separate milk sheds with important differences. In the southwestern milk shed, low prices for milk have attracted FDI, with firms competing internationally. As a consequence, this area has become the main supplier for Uganda’s exports of dairy products and has pushed the other main milk shed, the central milk shed, towards producing for and supply- ing the domestic market. The export-oriented southwestern milk shed performs considerably better across a number of criteria. First is the number and density of milk-collection centers. These facilities are crucial parts of the dairy value chain, Growth, trade, and transformation as they provide cooling services prior to further transport. With significantly more milk-collection centers, the average distance to a collection center in the south- western shed is about half that in the central shed. Second is the productivity. The southwestern shed has more heavily improved breeds than the central shed and milk output per cow is thus higher. The final criterion concerns financing. Milk farmers in the southwest borrow significantly more (almost twice) to invest in their businesses. Uganda’s exports of dairy products are dominated by a handful of large export- ers and highly regionalized, targeting mainly Kenya Before the COVID-19 pan- demic, the top eight exporters accounted for 95 percent of Uganda’s exports of dairy products. Moreover, individual exporters also specialized in markets. Ugan- da’s largest dairy exporter served a total of nine different destination markets, with more than 90 percent of its export value being shipped to five different markets 124 (Egypt, Kenya, Oman, Japan, and Ethiopia). All other firms in the top eight served a maximum of three countries, with most specializing in a single country of des- tination. Overall, in 2018, Kenya alone imported 84 percent of Uganda’s total ex- ports of dairy, while the other four EAC members (Burundi, South Sudan, Tanzania, and Rwanda) accounted for another 11 percent. While 2020 witnessed an uptick in exports to other countries, the share of non-EAC countries in total dairy exports remain at about 15 percent. The deep regional concentration of exports is largely driven by the extremely high preference margins in the EAC. Dairy products, including milk and milk pow- der, are considered sensitive Items in the EAC CET and protected by a maximum 60 percent ad valorem tariff when imported from outside the union.6 By compari- son, the average tariff for dairy products stood at around 22 percent ad valorem in 2006 (before the establishment of the EAC). With the introduction of the CET, the average tariff on dairy products increased to about 46 percent (Frazer, 2012), and Uganda began exporting dairy products after the introduction of this preferential access in 2007. The lack of market diversification makes Uganda’s dairy exports vulnerable to shocks. In recent years, Kenya has increasingly blocked Ugandan exports of dairy products (as well as sugar and maize), on the basis of country of origin concerns. Allegations suggested that these products were not actually produced in the coun- try but imported from third states and then re-exported tariff-free under the rules of the EAC customs union.7 Going forward, Uganda’s dairy sector can face addition- al challenges if Kenya and the other EAC members decrease external protection in the advent of the AfCFTA or if Kenya implements bilateral agreements with the United States (US) or the United Kingdom (UK). Chapter 2. The role of international trade 125 Spotlight B. Uganda’s Changing Fisheries Source: https://www.worldbank.org/en/programs/africa-program-for-fisheries Growth, trade, and transformation Fisheries contribute significantly to Uganda’s agricultural exports and, like the broader agriculture sector, comprise formal commercial and informal tradi- tional subsectors. Bordering large lakes, Uganda is the seventh-biggest fish ex- porter on the African continent (after Mauritania, Namibia, Senegal, and South Afri- ca in SSA). According to the Ministry of Agriculture, Animal Industries and Fisheries, the sector contributes to the livelihoods of 1.5 million Ugandans (4 percent of the population). Fisheries exports are segmented between modern (largely formal) and traditional (largely informal) subsectors, which export different species, to different destinations, and with different levels of processing. Modern fish exports target overseas markets and comprise higher domestic value addition. Nile perch (L. niloticus) and Nile Tilapia (O. niloticus) are formally exported to European and Asian markets. These are processed by a small number of large commercial enterprises who organize the bulk (more than four fifths) of Uganda’s fish exports. 126 Figure 41. Uganda’s formal and informal fish exports Source: Rauschendorfer, Stojanov, and de Melo (2021). Notes: The Bank of Uganda‘s informal cross-border survey has not been conducted since the outbreak of COVID-19 in March 2020; hence we do not report values for after 2019. The formal fish processors mostly export products that have already undergone considerable value addition locally, mainly chilled, or frozen value-added products like fillets, loins, or headed and gutted fish, as well as canned fish. In 2019, the corre- sponding industry association reported employing 32,000 workers directly (NEMA, 2021). Traditional fish exports serve regional markets, often informally. Mukene (R. ar- gentea), and Muziri (N. bredoi) are exported (largely) informally to Uganda’s EAC partners. These informal cross-border exports of fish constitute an important source of livelihood for small-scale traders. However, with lower capital intensity Chapter 2. The role of international trade and use of cooling facilities, informal traders face high post-catch losses (30 per- cent), which reduce earnings. Over the past decade, overfishing has led to a declining stock of fish and re- stricted further growth of fish exports. In the early 2000s, Uganda’s fish exports boomed like dairy exports. Between 2000 and 2006, fish exports increased more than 4-fold, which was largely driven by a major expansion in exports to Europe- an markets. After that, however, the exports of fish from Uganda remained largely stagnant. This loss of momentum has in part been attributed to overfishing and illegal fishing. In Uganda, the bulk of national catches are conducted in Lake Albert (43 percent), followed by Lake Victoria (40 percent) and Lake Kyoga (12 percent). Lake Victoria and Lake Kyoga were considered overfished, while Lake Albert was nearly overfished in 2018 (NEMA, 2021). 2018 saw the lowest catch value in 15 years, with 345,000 tons (NEMA, 2021). Currently, out of 21 fish processors, only 11 are operational. The fisheries sector has adapted by introducing aquaculture. The sector’s re- sponse to decline in catches includes a successful introduction of aquaculture, 127 which now accounts for 20 percent of national production. With this success, Uganda has become the third-largest aquaculture producer in Africa (Adeleke et al.); with support from the government and donors, production grew from 1,500 metric tons of fish in 2005 to 120,000 metric tons by 2018. Aquaculture practice is currently concentrated in the Central region and the main species farmed are North African catfish and Nile tilapia. The sector comprises approximately 20,000 small-scale fish farmers, while commercial, large-scale, fish farmers are also slowly developing an interest (NEMA, 2021). To maintain the growth of the fisheries exports, more regulation is needed. Most importantly, overfishing must be addressed at regional and domestic levels. The Government of Uganda responded to the threat by passing the Fish (Amendment) Act of 2011, requiring all boats, including all people involved in fishing activities, to be registered. A new Fisheries and Aquaculture Bill remains under consideration by Parliament. The bill constitutes a comprehensive revision of an earlier act which only considered capture fish, and would also address issues like post-harvest han- dling, transportation, fisheries research, as well as surveillance and monitoring of fisheries. In addition to regulatory issues, several challenges need to be addressed to improve the quality of fish exports. Uganda relies on high-value markets in the EU under the Everything But Arms agreement, but this opportunity is often threat- ened by an inability to meet stringent quality and safety requirements. In the past, Uganda has been subject to fish import bans by the EU in 2002, 2015, and 2019. To avoid similar problems, the authorities should encourage best farming practices, access to sustainable quality inputs (feed), biodiversity control, and maintaining food safety alerts. These will require both better provision of public services (e.g., better electricity and transportation for the cold supply chain from catch to ex- port) and more efficient supply chain management, which is often associated with large players in the industry (see Adeleke et al., 2021). Growth, trade, and transformation 128 Spotlight C. Ugandan Coffee: High Quality at a Discount Chapter 2. The role of international trade Source: Coffee and Climate Toolbox (2022) Coffee is Uganda’s main cash crop and Uganda is the largest producer of coffee in Africa by volume. Uganda has one of the highest concentrations of coffee grow- ers in Africa (and the world) with around half a million farmers. They produce 6 per- cent of the world’s Robusta supply at lower altitudes in the center of the country and 1 percent of the world’s Arabica supply in the mountainous borders (Morjaria and Sprott, 2018). Uganda’s Robusta production is about four times more than its Arabica production.8 Coffee is intercropped with other crops such as bananas and beans, and coffee farmers have begun farming vanilla as well (in 2019 vanilla exports were worth USD 30 million). Although large-scale coffee producers are becom- ing increasingly common, the sector is dominated by about 500,000 small-scale farmers, who cultivate about 0.33 hectares per farm on average. These producers use inputs at a low scale and gather low yields, about 369 kilograms per hectare. As prices recovered from the sharp collapse in late 1990, Ugandan producers also increased their production. Production and export receipts have been in- 129 creasing steadily since the 2000s, reaching USD 450 million in 2017, a more than quadruple increase from about USD 100 million in 2001. This trend has been driven by increasing prices and volumes, leading to the highest export receipts in two de- cades. Overall, coffee contributed close to 10 percent of Uganda’s export earnings in recent years. The Ugandan coffee sector has set itself the goal of achieving 20 million bags (60 kilograms each) of coffee production (under the government Cof- fee Roadmap). The expected production for the marketing year 2020–21 is expect- ed to reach 6 million bags (USDA, 2021), up from 5.5 million as new coffee plants are maturing. With insufficient regulations and occasional poor farming practices, Uganda has sold its coffee at prices well below the global average. In 2018, the Robusta variety was sold at an average of USD 0.848 per pound globally, whereas Uganda’s exports earned about 17 percent less9 even though they often include about 20–25 percent arabica, which yields much higher prices internationally.10 Similarly, despite some successes like supplying the US ice-cream brand Ben and Jerry’s and gaining Fairtrade certification by the Rwenzori Farmers’ Cooperative Union, the sector’s growth has been hampered by poor farming practices such as early harvesting and theft issues. Efforts have since been made to regulate marketing and farming practices (Shriver, 2020). With better global branding and business practices, Uganda can increase the value of its coffee exports. While growth of Ugandan coffee exports has been im- pressive, lack of Ugandan brand awareness hinders ambitions for higher growth. This has been identified as a major constraint limiting sales in foreign markets de- spite Ugandan coffee being certified as among the world’s best. This is especially true with respect to arabica coffee, which is traded globally at much higher prices than Robusta; Morjaria et al. (2018) report that, despite its quality, Ugandan Arabica sells at 30 percent less than coffee of similar quality from other destinations. The government’s strategy to raise prices by 15 percent through better branding is well placed and well within reach. Growth, trade, and transformation 130 References Artuc, Erhan, Isambert Leunga, and Harun Onder. 2022. “Uganda’s structural trans- formation.” Mimeo. Artuc, Erhan, Guido Porto, and Bob Rijkers. 2021a. “Trade Poverty and Inequality.” Mimeo. Washington, DC: World Bank. ––––––. 2021b. “Household Impacts of Tariffs: Data and Results from Agricultural Trade Protection.” The World Bank Economic Review 35(3): 563–85. Washington, DC: World Bank. Atkin, David, and Dave Donaldson. 2021. “The Role of Trade in Economic Develop- ment.” NBER Working Paper 29314. Cambridge, MA: National Bureau of Economic Research. Bernard, Andrew, J. Bradford Jensen, Stephen J. Redding, and Peter K. Schott. 2011. “The Empirics of Firm Heterogeneity and International Trade.” NBER Working paper 17627. Cambridge, MA: National Bureau of Economic Research. Coffee and Climate Toolbox. 2022. at https://toolbox.coffeeandclimate.org/; ac- cessed on 5/23/2022. EU-EAC MARKUP. 2020. Policy Brief on SPS and TBT Obstacles on Coffee Trade in the East African Community. Kee, H.L. and Tang, H., 2016. Domestic value added in exports: Theory and firm evi- dence from China. American Economic Review, 106(6), pp.1402-36. Kongsamut, Piyabha, Sergio Rebelo, and Danyang Xie. 2001. “Beyond Balanced Growth.” The Review of Economic Studies 68(4): 869–82. Lagakos, David, and Michael Waugh. 2013. “Selection, Agriculture, and Cross-Coun- Chapter 2. The role of international trade try Productivity Differences.” American Economic Review 103(2). Pittsburgh, PA: American Economic Association. Matsuyama, Kiminori. 2009. “Structural Change in an Interdependent World: A Global View of Manufacturing Decline.” Journal of the European Economic Associ- ation 7(2–3): 478–86. Mugenyi, Robert. 2020. Coffee Sector Trade Obstacle Paper Uganda, report pre- pared for EU-ITC MARKUP Project. Ngai, Rachel L., and Christopher A. Pissarides. 2007. “Structural Change in a Multi- sector Model of Growth.” American Economic Review 97(1): 429–43. Rauschendorfer, Jakob, Anna Stojanov, and Jaime de Melo. 2021. Uganda’s Partici- pation in International and Regional Trade: Patterns, Prospects, and Policy. London: International Growth Centre, and Washington, DC: World Bank. Rauschendorfer, Jakob, and Anna Twum. 2020. “Unmaking of a Customs Union: Re- gional (Dis)integration in the East African Community.” World Trade Review: 1–12. Cambridge, UK: Cambridge University Press. 131 Sotelo, Sebastian. 2020. “Domestic Trade Frictions and Agriculture.” Journal of Po- litical Economy 128(7):2690–2738. Chicago: University of Chicago Press. The Market Project. 2020. Improved Dairy and Livestock Management Grows Farm- er Income and Food Security, at: https://marketproject.org/2020/07/09/improved- dairy-and-livestock-management-grows-farmer-income-and-food-security; ac- cess date 05/23/2022. Growth, trade, and transformation 132 Chapter 3. Future challenges, opportunities, and policies CHAPTER 3. FUTURE CHALLENGES, OPPORTUNITIES, AND POLICIES U ganda’s future economic success will depend on its ability to foster more rapid transformation. The analysis has so far provided a detailed as- sessment of how the Ugandan economy evolved up to now. The country made significant progress especially in the 2000s, increasing its per capita income 3-fold and life expectancy by half. However, this economic dynamism has failed to foster rapid structural transformation and eventually stalled in the 2010s, leaving about two thirds of the country’s labor force in agriculture, who produce less than a quarter of Uganda’s GDP. The COVID-19 pandemic has magnified this trend. Overall, the stalling economic performance has revealed the country’s structural fault lines. Uganda needs to accelerate growth and structural transformation to (i) generate enough jobs for its rapidly growing labor force, which will grow by 2.5-fold in the next three decades; and (ii) to halt the depletion of forests and bushlands, which are being converted to small-scale farms. Chapter 3. Future challenges, opportunities, and policies Fostering more rapid transformation will depend on policies. Uganda’s slow pace of transformation was driven by a weak pull from non-agricultural sectors, fric- tions like job search (lacking digital facilitation) and mobility (low rural connectiv- ity) costs, credit constraints, and the inability to liquidate assets to afford upfront investments for transition, e.g., acquiring human capital. The low provision of public services caused or aggravated many of these factors. Such frictions have become more binding in recent years. These problems are magnified by (i) a trade policy that is influenced by protectionist reflexes (high food tariffs and restrictive import regulations), (ii) frequent problems in regional integration frameworks, and (iii) high trading costs driven by logistics and infrastructure constraints. This chapter analyzes the future dynamics of growth and transformation in Uganda, with a special emphasis on potential opportunities in key areas. The analysis first employs the structural transformation model used in previous chap- ters to simulate economic trends in the absence of a major change (business as usual scenario). Next, the effects of policies aimed at boosting productivity growth and reducing frictions in the economy are simulated. The discussion then turns to five areas where potential opportunities and challenges may affect the future trends in the country. These are: (i) agricultural productivity, (ii) regional integra- tion and trade (especially the AfCFTA), (iii) tourism, (iv) digital transformation, (v) hydrocarbons, and (v) climate change. Policy recommendations for each area are also provided. 135 The Future of Transformation in Uganda With much uncertainty surrounding all economies, the medium-term analysis here employs a scenario approach. Forecasting economic performance over long periods is a challenging task in normal times. During a pandemic with extreme un- certainties, it is daunting. Thus, instead of forecasting future trends, we focus on a ‘what-if’ analysis, where the marginal effects of different assumptions are quanti- fied. We employ the structural transformation model developed in earlier chapters and consider three scenarios regarding the exogenous variables of the model–– international prices, population growth rate, the expansion of arable land, mobility frictions, and sector-specific TFP growth rates––over the next 10 years: • Baseline: All exogenous variables grow at the same rate as they did in the last decade. • Higher manufacturing TFP: The annual TFP growth rate of manufacturing in the next decade exceeds that in the previous decade by 2 percentage points • Lower frictions: The trend growth in mobility frictions is set at zero as op- posed to the trend in the baseline. Without a breakthrough, Uganda’s transformation is projected to remain mod- est in the medium-term. Figure 11 shows the results of the simulation exercise. In the baseline, where all exogenous factors have the same trend as in the last de- cade, the labor share of agriculture decreases by only 1.6 percentage points. This outflow, combined with no productivity growth in the sector, leads to 0.6 shrinkage in the sector’s value added (panel b), and the sector’s share in total value-added decreases by 10.9 percentage points, largely because other sectors keep growing as their momentum from the last decade is preserved. A higher productivity growth in manufacturing accelerates growth and trans- formation significantly. With 2 percentage points higher TFP growth in manufac- turing, Ugandan GDP grows by an additional 20 percent in a decade compared to the baseline projections. Part of this is driven by the labor outflow from agriculture: Growth, trade, and transformation the sector’s labor share decreases by 10.3 percentage points instead of a modest 1.6 percentage points in the baseline. While this leads to significant gains in man- ufacturing value added (additional value-added equivalent to 22 percent of the base-year GDP compared to the baseline projections), the impact on agricultural value added is modest (only lower than the baseline projections by an equivalent of 2.9 percentage points of the base-year GDP). As a result, the value-added share of agriculture (in proportion to current GDP) decreases by an additional 3 percentage points relative to the baseline projections. 136 Figure 42. Medium-term (10-year) projections: baseline vs. policy out- comes In the baseline, the labor share of agriculture decreases by 1.6 pp in 10 years. With 2 pp higher manufacturing TFP growth annually, it decreases by 10.3 pp, and with lower frictions by 15 pp. Labor share projections 70 64.7 63.1 60 54.4 49.7 50 Percent 40 29.1 30 26.9 21.1 21.5 22.0 18.8 20 13.8 14.9 10 0 Agriculture Manufacturing Services In the baseline, the value added of agriculture decreases by 0.6 pp in 10 years. With 2 pp higher manufacturing TFP growth annually, it decreases by 3.5 pp, and with lower frictions by 3.0 pp. Value added projections (as a share of base-year GDP) 120 107.7 105.0 110 102.1 100 95.9 85.7 86.0 Chapter 3. Future challenges, opportunities, and policies 90 80 Percent 70 60 50 42.0 40 36.0 30 21.9 21.3 18.4 18.9 20 10 0 Agriculture Manufacturing Services In the baseline, the value added share (% of current GDP) of agriculture decreases by 10.9 pp in 10 years. With 2 pp higher manufacturing TFP growth annually, it decreases by 13.9 pp, and with lower frictions by 13.2 pp. Value added projections (as a share of current GDP) 70 60 50 46.6 47.1 45.4 44.4 44.6 44.2 42.0 Percent 40 36.0 30 21.9 20 11.0 10 8.0 8.7 0 Agriculture Manufacturing Services Base year Baseline Higher manufacturing TFP Lower frictions Source: Artuc, Leunga, and Onder (2022). The trend growth rates are fixed at historical (10 years) averages for all exogenous vari- ables (international prices, TFP growth, and mobility frictions). 137 With lower frictions, labor reallocation accelerates, and the economy grows faster despite no additional gains in sectoral TFPs. Mobility frictions prevent workers from moving between low-productivity and high-productivity activities. Thus, when they are smaller, the economy registers efficiency gains even in the absence of additional TFP growth in any sector. In simulations, with lower frictions, Ugandan GDP grows by about 12 percentage points more in the next decade rela- tive to the baseline. As TFPs are identical between the baseline and the low friction scenarios, all this additional growth comes from better allocation of labor. The low- er frictions scenario leads to additional labor outflow from agriculture, about 13.4 percentage points more than that in the baseline projections. This outflow reduces agriculture’s value added by 2.4 percentage points relative to the baseline (panel b) but increases that of manufacturing and services by 16.4 percentage points and 9.9 percentage points, respectively. As a result, agriculture’s share in total value-added decreases by 2.3 percentage points by the end of the next decade in the lower frictions scenario, relative to the baseline. Policies aimed at boosting productivity and reducing frictions will be essential to put the country on a more rapid growth and transformation path. The simula- tion analysis here shows the need for higher productivity growth and lower frictions in promoting growth and structural transformation in the next decade in Uganda. The previous two chapters on growth and trade have already identified a number of key constraints, including high food tariffs, costly trade logistics, limited rural ac- cess to infrastructure and information, and certain market imperfections like credit constraints and low liquidity of assets like land. Addressing these constraints can help significantly both to increase sector specific productivities and reduce fric- tions against sectoral and spatial transitions of Ugandans. Besides removing the prevailing constraints, the authorities can also exploit more targeted opportunities to unleash a more dynamic economic path in Uganda. In addition to resolving the current bottlenecks, the authorities should take a more proactive stance for identifying and exploiting opportunities that can transform the economy structurally. In the remaining part of this chapter, we pro- vide an overview of the key candidates, with an emphasis on the opportunities Growth, trade, and transformation they provide, potential risks that can derail attempts to take advantage of them, and suggested policy actions to counterbalance such risks. The key candidates are: (i) the prospects for further regional integration and trade (especially the Af- CFTA); (ii) the development of hydrocarbon resources in Uganda, which can provide important resources that could solve the country’s prevailing bottlenecks; (iii) an expansion of tourism to tap into the sector’s vast potential in Uganda; (iv) the dig- ital transformation agenda, which can help reduce transaction costs and frictions dramatically; and (v) the climate change agenda, which poses significant risks but also provides an opportunity to revalue the global public goods provided by the country’s natural assets. 138 Agricultural Productivity Uganda’s agriculture sector plays a critical role in the economy and is funda- mental to the country’s economic transformation and transition to a middle-in- come country. It accounts for about two-thirds of employment, provides most of the income for rural and poor households, is an important source of export earn- ings and supply of raw materials for industry, and accounts for about one-quarter of GDP. Furthermore, the broader agriculture value chain is critical for providing off- farm jobs and stimulating demand for local service providers and manufacturers. Both domestic and regional demand for agriculture commodities is on a rapid rise, and an increasing number of urban dwellers will demand more processed food and protein-rich diets. This provides massive opportunities for Uganda’s agriculture sector and wider agri-food system. Agricultural value chain is labor intensive and it can contribute to inclusive economic growth. Uganda’s agri-food system has the potential to enhance em- ployment opportunities, both skilled and unskilled, for the country’s predominantly young population, the majority of whom live in rural areas. Diverse agribusiness- es, particularly along the dairy, maize, and coffee value chains, have developed in recent years, linking farmers to inputs, markets and finance, and improving rural livelihoods. Yet to fully harness the sector’s unique opportunities, Ugandan author- ities needs to spur this nascent agribusiness dynamism and quicken the shift from Chapter 3. Future challenges, opportunities, and policies low-value smallholder farming towards a higher value-added and commercialized agri-food sector. Key challenges faced by Ugandan agriculture Despite agriculture’s important role in Uganda’s economy, output is far below potential, and the country will need to overcome a range of challenges to agri- culture productivity. National agricultural output grew at only 2 percent per an- num in the five years prior to the COVID-19 crisis, compared to agricultural output growth of 3 to 5 percent in other EAC members, and 3.3 percent per annum growth in Uganda’s population over the same period. Total factor productivity growth has largely been absent from Ugandan agriculture and has generally been underper- forming compared to peers in the region. Farmers’ limited market participation, low uptake of improved agricultural inputs, and limited adoption of technology are some of the main causes of stagnating yields. This also reflects insufficient pub- lic priority to maintaining innovation – for instance in the area of improved and resilient seeds – and effective extension services in the sector. In fact, Ugandan agriculture has one of the lowest adoption levels of improved seeds, inputs, and mechanization. Political interference accentuates inefficiencies in the farm input market. Through Operation Wealth Creation (OWC), which started in 2014, the extension services system has steadily moved away from its core function of knowledge 139 transfer and has increasingly taken the role of distributing free or highly subsidized agricultural inputs. These are distributed by the military, are often of low quality and marred by inadequate timing. The cost of the OWC distortion is substantial, as inputs are sometimes procured at 20-50 percent above market prices. At the same time, with free seeds being available to many smaller scale farmers, private seed companies have seen their market share and their ability to expand adversely affected — reducing their incentives to invest and, thus, crowding out the private sector from seed distribution. In addition to OWC, perceived and real politization of rural organizations has limited the growth of farmer groups and cooperatives that could serve an important role in organizing access to markets for inputs and outputs among smallholder farmers. Over the last decade, the institutional base for agriculture has continued to de- teriorate and public expenditures have been insufficient and poorly allocated. Institutional weaknesses within the responsible ministries and agencies have been a key bottleneck for agricultural policy design and implementation, regulation and administrative coordination. Furthermore, despite the emphasis on the agricultural sector in many policy plans, agriculture was estimated to account for, on average, just 3.6 percent of total public expenditures between 2013/14 and 2017/18, which is lower than the annual average share of agriculture in total spending in Africa of about 4 percent and far from the 10 percent of total expenditures that the Com- prehensive Africa Agriculture Development Programme (CAADP) compact recom- mends. Furthermore, a significant share of this expenditure went to funding OWC (i.e. free or subsidized distribution of inputs), while very little was invested in digital solutions and data driven analytical tools to help with improved planning in the ag- ricultural sector, including effective soil observation and management techniques to inform the selection of crops. Similarly, very little was invested in irrigation, ac- cess roads, wholesale and livestock markets, veterinary services, sanitary and phy- tosanitary laboratories and equipment, and research and extension services (as shown in Figure 2, the share of funding to extension and advisory services has been steadily declining), which are critical for increasing productivity and building resil- ience to climate change risks. Growth, trade, and transformation Growing rural density has caused increasing land fragmentation and shrinking farm sizes, which constrain productivity and incentives to commercialize. From 2006 to 2016, the share of small household farms, with less than 2 hectares of land, rose from 75 to 83 percent — this compares unfavorably to comparators like Tan- zania and Ghana where 44 and 55 percent of farms in 2012/13, respectively, were operated on land bigger than 5 hectares. As a result, the average net land operated in Uganda fell from 1.7 to 1.2 hectares per household, thus reversing the trend toward larger farm holdings, which are more likely to commercialize due to economies of scale and ease of adopting modern technologies. Land tenure issues are also a critical bottleneck to organizing more productive farms. Overlapping systems result in unclear and unofficial land rights for many smallholders, inhibiting their ability to rent, sell or use land as collateral. This is even more constraining for women farmers as Uganda’s land tenure system is rooted in patriarchy, with customary law usually affording women fewer land rights. 140 Figure 43. Allocation of public expenditure in support of the agriculture sector 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2013/14 2014/15 2015/16 2016/17 2017/18* Input subsidies Capital subsidies Research Extension and advisory services Inspection and quality control Feeder roads Other infrastructure Storage Chapter 3. Future challenges, opportunities, and policies Irrigation Other subsidies Processing and marketing Training Source: Agriculture Sector Public Expenditure Review (2019). Lack of land tenure security for refugees (as well as the estimated 30 percent of the population who still do not have national IDs) also provides a barrier to invest- ment and longer-term self-reliance. Farming is exposed to increasing climate variability and weather shocks: Ugan- da is the most vulnerable to climate change among regional peers on the ND- GAIN index.1 More than 95 percent of cropland is rainfed and based on subsistence farming, making it especially vulnerable to weather variability and climate hazards. In recent years, seasonal rainfall has become more variable and less predictable and, in combination with higher temperatures, is likely to reduce crop productivity. Extreme events such as droughts, floods, or landslides are also projected to be- come more frequent and intense and are exacerbated by unsustainable land use practices and the expansion of agricultural land into other ecosystems such as forests. Soil degradation and erosion, caused by unsustainable land management, has further reduced agricultural productivity and increased vulnerability. In recent years, this lack of resilience has resulted in huge losses in livestock and crops. For example, due to drought and pests such as the armyworm in 2016, output plum- 1 Notre Dame Global Adaptation Initiative (ND-GAIN). Regional peers include: Ethiopia, Kenya, Madagascar, Malawi, Mozambique, Rwanda, Tanzania, Zambia, and Zimbabwe. 141 meted, which resulted in widespread food insecurity; future predictions look even more grim.2 The high probability of weather shocks also reduces the incentives of farmers to invest in higher productivity techniques and inputs, leading to even fur- ther foregone potential growth. Beyond the farm-gate, poor logistics and considerable post-harvest losses impact the quality of crops, access to markets, and incentives to adopt more productive approaches. The agricultural supply chains linking farmers to output markets are underdeveloped and fragmented. Smallholders typically depend on lower quality post-harvest handling and storage practices (leading to significant sanitary and phytosanitary issues), whereas aggregators are constrained by lack of working capital to take on the post-harvest processes on a larger, more centralized level. This results in an unreliable supply for agro-processing companies, reducing their efficiency and dampening their incentives to invest. Because of this, most crop sales end up in informal markets, which generates low and variable farmer income, despite relatively high prices in other countries in the region. Furthermore, the marked seasonality of agriculture leads to high volatility of output and pric- es, creating (largely unexploited) opportunities for intertemporal arbitrage. Lack of connectivity in the region and affordable mobile devices also limit the ability of farmers to stay abreast of daily price changes and coordinate a response amongst each other. Consequently, the lack of formalization, finance, production support services – such as warehousing, quality-controlled post-harvest management, cold chain infrastructure and logistics services – and effective market stabilization mechanisms prevent farmers from smoothing income and consumption. High transportation costs and weak competition in agricultural intermediation squeeze the profit margins of farmers engaged in domestic and international trade. The poor quality of rural roads leaves some rural areas economically isolated from larger markets and reduces the profitability for smaller farmers. For commod- ities grown primarily for export and with limited long-term storage potential, like coffee, remoteness can also make farmers even more vulnerable to intermediaries. Weak competition in market intermediation, likely due to high costs of entry into long-distance subnational trade, and high transport costs, means that intermedi- Growth, trade, and transformation ary traders and aggregators capture a high share of the market surplus. Informa- tion asymmetries, the small scale of production, and weak cooperative marketing further reduce the bargaining power of farmers. In this way, while Uganda’s exports typically receive internationally competitive prices, this is not always reflected in the prices received by farmers. This also then disincentivizes them to use and in- vest in better inputs (Figure 44) and technologies, and to commercialize more ac- tively. 2 For example, by 2050, the production of Arabica and Robusta coffee could fall by 50 percent, and areas suitable for growing tea and beans could be severely affected, resulting in a massive loss of market opportunities. 142 Figure 44. Use of improved inputs in Uganda 25 24 24 Percentage of households engaged in agriculture 20 19 17 18 17 17 16 16 15 15 15 15 14 13 13 10 6 5 5 5 5 4 0 Organic Fertilizer Inorganic fertilizer Pesticide Improved seeds 2005/06 2009/10 2010/11 2011/12 2013/14 Chapter 3. Future challenges, opportunities, and policies Source: World Bank (2021). Key policy solutions Adopting climate-smart agriculture (CSA) and sustainable land management (SLM) practices can increase the productivity and the resilience of smallholder farmers. Crop rotation, integrated soil fertility management and intercropping sta- ples with nitrogen-fixing plants or trees are some of the most promising CSA/SLM practices, which, in addition, provide a vast array of ecosystem services. Improving water management and storage infrastructure also holds significant potential to increase productivity and guard against future climatic risks. However, while the adoption of CSA/SLM appropriate technologies is critical to enhancing the produc- tivity and resilience of smallholder farmers, such gains are only marginal if appro- priate management practices to land use systems are not broadly adopted. In this regard, agricultural productivity and resilience are, to a large extent, determined by actions taken to prevent and reverse overall land degradation, potential disasters, and climate shocks. With increasing demand for agricultural products, opportunities come along the agricultural value chain, especially for emerging or more commercially ori- ented farmers. Demand for meat, fish, milk, and fruits is growing more rapidly than income, and this growth is widespread in both rural and urban areas. Regionally, Africa’s demand for food is projected to more than double by 2050 and the value 143 of the African food market is predicted to rise threefold to USD 1 trillion by 2030. For example, the livestock sub-sector in Uganda has already benefited from this, hav- ing experienced very positive growth in recent years, and holds significant potential to expand further and meet the growing demand. Research, using household data in five countries of East and Southern Africa with income levels similar to Uganda, estimates that demand for processed foods in urban areas will increase by a factor of 8 over the next 3 decades. Furthermore, the full agricultural value chain holds tremendous economic potential and can be a major source of jobs and income, and can also support the growth of manufacturing. It has been estimated that agro/ food processing and food services together can generate about 200,000 new for- mal jobs in Uganda over the next decade. Vertical integration can provide small emerging commercial farmers3 with op- portunities for commercializing and engaging in value addition for domestic and export markets. To incentivize smallholders to invest more in inputs and more productive methods, and for the private sector to increase investments in value addition, credible and stable opportunities for price certainty, market access, and profitable commercialization have to be present. Vertically integrated agri-busi- ness operations in Uganda, such as contract farming and outgrower schemes, are increasingly being used to achieve this for commodities with more organized value chains, such as tea, sugar, coffee, dairy, barley, and sorghum. Expanding this to oth- er crops could also address finance and technology constraints, enabling higher productivity and value addition. Core to this is investing in capacity building of farmer organizations so that they can participate in such opportunities. Capitalizing on promising export trends can create significant opportunities for growth in agricultural commodities and value addition. Regionally, Uganda already handles significant flows of goods in transit between ports in Kenya and Tanzania and other landlocked neighbors. While Uganda’s domestic market com- prises 45 million people, the population of the broader East African region is almost 10 times this, thus implying a much larger potential market. The extension of new railway systems in Kenya and Tanzania towards the Ugandan border, and the revival of goods transport services across Lake Victoria, hold great potential for the private Growth, trade, and transformation sector in Uganda to offer value adding logistics services and to leverage the coun- try’s location as a logistics hub. Demand for coffee is expected to increase up to 3 percent per year and Uganda has the opportunity to increase its supply by further investing in its two coffee varieties and increasing its brand awareness, as well as upgrading along the coffee value chain. Uganda is already the third biggest fish exporter on the African continent, after South Africa and Tanzania, and the sector creates more than 1.5 million jobs. Uganda is also a growing net exporter of livestock products and live animals, dominated by dairy products and eggs, but with meat and meat products also expanding. 3 It is important to note that different segments of smallholder farmers should be targeted by different solutions to enhance their productivity and resilience. For example, subsistence-oriented small- holders may need to focus more on food security objectives and may not have the capacity to engage in vertically integrated opportunities. However, smallholders who have begun diversifying their crops and selling in markets may be able to take advantage of these productivity-enhancing opportunities. 144 Digital technologies can help improve productivity by increasing access to information for all stakeholders and facilitating greater coordination. The re- cent increase in internet connectivity and substantial rise in mobile phone own- ership (almost 70 percent of the population) is laying the foundation for Uganda’s digital transformation and enabling the rapid take-up of digital services. Digitally enabled services can: (i) provide timely access to information on pricing of agri- cultural products (inputs and outputs) and extension services delivered via mobile devices, (ii) support digital payments, and (iii) help identify and connect differ- ent players across the value-chain such as farmers, suppliers, buyers, and logis- tics providers. There are already many existing applications and digital technolo- gy-enabled tools for data collection purposes, including the Jaguza Livestock App, Ensubiko for rural financial institutions, e-Voucher system for agro-input dealers, and MUIIS for weather and agricultural information. Also, as farmers access infor- mation through platforms like M-Farmer and Yo! Uganda, the generated data can be leveraged to support data-based decisions on purchasing inputs and access to financial services like micro-loans. Expanding the uptake of these digital enabled technologies also has significant potential to overcome the challenges of distance and multiplicity of small independent farms in Uganda. In turn, better access to real time digital data by the government can inform timely decision making, including around prioritization of crops, distribution of seedings, and provision of subsidies to the most vulnerable farmer households temporarily affected by weather or pan- demic related events. Chapter 3. Future challenges, opportunities, and policies Regional Integration and Trade In the medium-term, two factors can influence Uganda’s regional trade sub- stantially: relations with EAC members and the AfCFTA. Uganda’s recent ten- sions with EAC neighbors have once again shown that trade cooperation is subject to the rules of bicycle theory, a term coined by economist Bhagwati (1988), which emphasizes that the building of trade cooperation among countries is a perpetual task: if cooperation is not progressing, it is regressing. Going forward, EAC members, including Uganda, should consider a careful enhancement of cooperation. However, progress in the implementation of the AfCFTA may complicate the future dynamics within the EAC as discussed below. Implementation of the AfCFTA will be a major breakthrough. On March 21, 2018, at the 10th Extraordinary Summit of the African Union, almost all countries on the African continent signed the AfCFTA agreement, thereby creating the largest free- trade area in the world. The agreement officially entered into force on May 30, 2019, after ratification by 22 countries, including Uganda. In July 2019, the Heads of State adopted the Niamey Declaration, launching the operational phase of the Af- CFTA with the aim of resolving the long-standing economic fragmentation of Africa, where trade barriers remain high, especially for sensitive sectors. The agreement will also cover policy areas such as trade facilitation and services, as well as regula- 145 tory measures such as sanitary standards and technical barriers to trade. It aims to complement existing subregional economic communities and trade agreements in Africa by offering a continent-wide regulatory framework and by regulating policy areas such as investment and intellectual property rights protection. What does the AfCFTA mean for Uganda? To assess the economic impact of the AfCFTA for Uganda, the World Bank (2020) has developed a detailed methodology. The methodology relies on a global computable general equilibrium model linked with a microsimulation (World Bank, 2020). Most prominently, the simulations estimate the static gains from trade in a scenario where reductions in tariffs and non-tariff barriers (NTBs), as well as the implementation of trade facilitation measures that reduce transport costs across the continent, take place over the medium-term. Specifically, these rely on the following assumptions: • Tariffs on within-Africa trade are progressively reduced starting from 2020. Accordingly, 90 percent of those are eliminated over a 5-year period (10- year period for the least developed countries––LDCs). Starting in 2025, tar- iffs on an additional 7 percent of tariff lines are eliminated over a 5-year period (8-year period for LDCs). A maximum of 3 percent of tariff lines that account for no more than 10 percent of intra-Africa imports can be excluded from liberalization by the end of 2030 (2033 for LDCs). • NTBs on both goods and services are reduced on a most-favored-nation (MFN) basis. It is assumed that 50 percent of NTBs are actionable within the context of the AfCFTA—with a cap of 50 percentage points. These are imple- mented as AVEs. It is also assumed that reduction of NTBs benefits African exporters to non-AfCFTA markets with an additional reduction of NTBs by 20 percent. • Trade facilitation measures as envisaged by a trade facilitation agreement. Growth, trade, and transformation Estimates of the size of these trade barriers were obtained from de Melo and Sorgho (2019) and are assumed to be halved (capped at 10 percentage points) for the scenario analysis. Simulations show that some EAC members are among the countries that should benefit most from the AfCFTA. Figure 45 shows the simulations results, where gains from the AfCFTA are estimated as percentage point deviations from those in the no-AfCFTA scenario. Estimates show that Kenya and Tanzania are amongst the largest beneficiaries, with more than 10 percentage point income gains com- pared to no-AfCFTA by 2035. In comparison, although Uganda’s potential gains stand lower at about 3.3 percent, they are still significant. For all EAC members, the majority of these income gains are driven by trade facilitation measures and reductions in NTBs. It is, however, important to note that the likelihood of such gains is conditional on these countries’ abilities to remove trade barriers as outlined by the assumptions described above. 146 Figure 45. The estimated impact of the AfCFTA by country and sector in Uganda . Uganda’s benefits from the AfCFTA are moderate and driven mostly by trade facilitation measures and NTBs. a. Income gains from the AfCFTA in 2035 (% change with respect to baseline scenario in 2035) In Uganda, agriculture and services benefit from the AfCFTA by 5.1 and 1.4 percent, respec- tively, while there are mixed outcomes for specific manufacturing sectors, and large losses for natural resources. b. The impact of AfCFTA on output in Uganda Chapter 3. Future challenges, opportunities, and policies (Percentage change relative to baseline in 2035) Processed foods 7,9 Agriculture 5,1 Water transport services 5,0 Road and rail transport services 3,5 Construction 3,2 Air transports services 2,3 Other services 1,9 Minerals n.e.s. 1,3 Communication services 1,0 Trade services -0,1 Hospitality services -0,3 Insurance, real estate services -0,5 Energy intensive manufacturing -1,3 Other financial services -1,6 Wood and paper products -2,6 Petroleum, coal products -3,8 Other business services -6,2 Textiles and wearing apparel -6,9 Chemical, rubber, plastic products -10,9 Fossil fuels -11,7 Manufactures, n.e.s. -14,6 Total Agriculture 5,1 Total Natural Resources -10,4 Total Manufacturing -0,8 Total Services 1,4 Total 1,3 -20 -15 -10 -5 0 5 10 Percent Source: World Bank (2020). 147 In Uganda, gains accrue mainly through agriculture. The second panel in Figure 45 shows the simulation results by sector in Uganda. Food processing and agri- culture are the biggest winners from the AfCFTA-linked reforms with 7.9 and 5.1 percent gains (about USD 1 billion and USD 1.6 billion), respectively. In comparison, natural resources and manufacturing register losses of 10.4 and a modest 0.8 per- cent, respectively. While some services like insurance and financial services are set to register small losses, major gains are expected in other areas like transportation services. As a result, the overall services sector is projected to gain about 1.4 per- cent from the AfCFTA-linked reforms. These gains can be larger with further reforms. Alternative simulations include (i) AfCFTA broad, where the impact of foreign direct investment (FDI) from a preferen- tial trade agreement among all countries on the continent, representing a shallow but broad integration, are included; and (ii) AfCFTA deep, where additional coop- eration, notably in investment policy, competition policy, and intellectual property rights are included. For AfCFTA deep, Uganda’s income gains can increase up to 5.7 percent. Notwithstanding the prospects for gains, the implementation of the AfCFTA may raise complex challenges for Uganda as well. Most importantly, synchro- nizing the EAC (especially the CET) with the AfCFTA, and maintaining cooperation in this process, will require considerable efforts. The AfCFTA has only been ratified by three of the six EAC members. Without further ratification by remaining mem- bers, the integrity of the CET may be under further pressure. The other particularly problematic issue for Uganda is the potential impact of liberalization on protected markets such as the dairy sector. Uganda’s booming dairy sector has largely relied on the high CET in the EAC and may lose competitiveness in an integrated African market to other producers like South Africa and Egypt. It is essential for the Ugan- dan government to take policy actions to boost the competitiveness of domestic businesses through better provision of public services, lower economic frictions and trade costs, and an overall business-friendly policy environment. Growth, trade, and transformation Trade policies for the future Maintaining regional cooperation is vital, and the authorities should focus on re- inforcing regional integration. Regional integration policies should aim to remove trade barriers, not erect new ones for creating protected markets in the region. In this regard, the Ugandan authorities should seek to reset cooperation with EAC members, and thereby underscore the potential for Uganda’s diversified economy to benefit the region and enhance shared interests (i.e., transport and trade facil- itation links, and the management of lake and water resources). The recourse to stay of applications for the EAC CET should be reduced and further efforts made to resolve trade tensions with neighbors (through mutually beneficial removal of trade barriers). In addition, an agreement with EAC members on future schedules of liberalization and the identification of Uganda’s interests should be pursued. 148 Uganda should prepare its producers for the implementation of the AfCFTA, including the development of a transparent roadmap of policy actions. The authorities, in close cooperation with businesses in key sectors (e.g., dairy) can develop a transition plan. The main objective of this plan would be to boost the international competitiveness of key sectors gradually and soften the impact of sudden exposure to stronger competition in export markets. In the dairy sector, for example, authorities could review the sector strategy and perform a market study for the setting up of MCCs in the Central region; looking at the quality of local infrastructure (e.g., community access roads still need improving) and the govern- ment services needed (e.g., phytosanitary, veterinary services) in the target areas. In fisheries, the new fisheries and aquaculture law needs to be adopted to further develop the aquaculture industry, and stricter stock conservation policies in coop- eration with riparian neighbors are also needed. The authorities should scale up trade facilitation efforts and remove obstacles to both imports and exports. Uganda should soften import substitution policies that can hinder exporters. The analysis in this report showed that the ability in Uganda to import strongly correlates with the ability to export. With difficulties in importing, this reduces both the prospects for small-scale producers and the competitiveness of Ugandan producers. Another obstacle for more dynamic ex- port performance is the poor infrastructure conditions in the country. Given the importance of the agricultural sector, improving rural access to producing regions Chapter 3. Future challenges, opportunities, and policies should yield important benefits. Investments in infrastructure and systems that reduce external trade costs, such as investments along international trade cor- ridors, can also be beneficial. Uganda remains a price taker in international freight and trade markets. This is largely due to its geographic situation as a dependent, landlocked country. Greater private participation and further regional cooperation opportunities can be exploited to invest in infrastructure for diversifying sea-ac- cess corridors. International branding, quality standards, and conservation efforts are essen- tial for supporting the country’s main foreign exchange earners, tourism, and coffee. Uganda’s potential in coffee and tourism has not fully materialized yet. More efficient quality controls and concerted efforts in building Uganda as a quali- ty brand in coffee and tourism markets can help realize the country’s true potential. For coffee, authorities should review controls on exports to seek more efficient and less costly procedures, authorize less than LCL options, and seek to improve traceability. In tourism, a better conservation policy is required to ensure that for- ests and green spaces are protected, and better forestry practices are developed. These measures are particularly crucial now as the country gears up to produce its first oil, which is discussed next. 149 Tourism Tourism has been an important source of economic activity in Uganda. In 2019, Uganda’s travel and tourism exports reached USD 1.33 billion, which made it the leading foreign exchange generating sector in Uganda (about 32 percent of all ex- port revenues). The number of international visitors increased steadily in the 2010s (at about 5.8 percent per year) to reach 1.4 million in 2017 (Figure 46). In addition to direct revenue generation, the demand for these services also cascades down to other domestic activities through intersectoral linkages. The demand for Uganda’s tourism services has largely been regional. According to the World Bank (2020b), the majority of international visitors (80.1 percent) in 2017 came from elsewhere in Africa, with the largest number of visitors coming from neighbors––Rwanda (441,994), Kenya (334,788), and Tanzania (89,253). As a region, Europe sent the second-largest number of visitors (7.4 percent), followed by the Americas (5.5 percent) and Asia (5.1 percent). Outside of Africa, the five most common countries of origin of visitors were the US (61,775), India (35,676), the UK (33,564), China (16,842), and Canada (13,109). With several diverse attractions, Uganda’s tourism potential is much larger than the current level. These include the East African savanna, the tropical rainforest, Rwenzori’s or Mountains of the Moon, Lake Victoria, Murchison Falls, and the source of the Nile. The country’s national parks and wildlife reserves house more than half of the mountain gorilla population in the world along with several hundred bird spe- cies, some of which are endemic to Uganda. With such natural capital, Uganda is well placed to tap into a structurally growing global and regional leisure market in the long-term. The main pillars of this approach can comprise: • Cultural heritage tourism attractions such as historic palaces; burial sites, ruins, archaeological sites; places associated with industrial, scientific, and agricultural heritage; museums, cultural sites; and intangible cultural attrac- tions like marriage ceremonies, dances, and rituals. Growth, trade, and transformation • Health and wellness tourism attractions including mainstream, traditional, or alternative healthcare and therapy services, spas, health farms, and reha- bilitation services. • Ecotourism attractions focusing on community participation, conservation and management of biodiversity, and indigenous knowledge systems. These comprise scenic natural and cultural areas; protected wildlife habitats; areas for camping, trekking, and climbing; and prominent sites for traditional or indigenous environmental practices. • General leisure tourism attractions including recreational facilities with high visitor density: golf parks/resorts, theme parks and amusement centers, convention, and meeting centers, among others. 150 Figure 46. International visitor arrivals in Uganda International visitor numbers increased steadily in the last decade, but data for more recent years is missing. International Visitor Arrivals 1600 1400 Persons (Thousands) 1200 1000 800 600 400 200 0 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Source: World Bank (2020b) based on UBOS statistics (2015–2019) and UNWTO Barometers (2010–2014). Notes: International visitor arrivals data for 2018 and 2019 have not been collected due to the introduction of e-visas and the phasing out of immigration cards in 2018. Chapter 3. Future challenges, opportunities, and policies • Religious tourism attractions such as the Namugongo Shrine and various sites linked to the Uganda Martyrs, which are already significant tourism at- tractions for domestic and regional tourists and with further development can also receive international attention. The pandemic-driven downturn in tourism can provide an opportunity to im- plement policies for lifting the sector up to its potential. The impact of COVID-19 on tourism has been debilitating. International travel came to a standstill on March 21, 2020, when Uganda officially closed all of its borders. According to a business survey conducted by the Uganda Bureau of Statistics (UBOS), April 2020 earnings for accommodation and food service establishments were 70 percent lower than the previous year (with the largest losses concentrated around national parks); about 30 percent had closed their doors and 77 percent had laid off staff (World Bank 2020b). By May 2020, about 88 percent of tour operators were unable to pay their workers and 38 percent anticipated filing for bankruptcy. Although these effects have been devastating for Ugandans’ livelihoods, they have also provided the authorities with ample incentive to support the sector in a way that builds its potential. This requires solving several structural bottlenecks faced by the sector. 151 Figure 47. Uganda’s Travel and Tourism Competitiveness Index rankings Growth, trade, and transformation Source: World Tourism Organization. Key bottlenecks and how to resolve them Compared to other countries, Uganda has favorable natural resources and competitive prices for travel and tourism. Overall, Uganda does not rank very high in the Travel and Tourism Competitiveness Index (Figure 47). In 2019, it was ranked 112 among 140 countries around the world— just slightly better than the SSA average. However, Uganda stands out in two areas, where its scores exceed even world averages. The first is its abundant natural resource endowment, for which Uganda ranked 38 among all countries, well above world and SSA averages. The second is its competitive prices, for which it ranked 39, again much better than world and SSA averages. In addition to these, there are a few other areas where 152 Uganda ranked above SSA averages, most notably international openness (82), pri- oritization of travel and tourism (93), and human resources and labor markets (98). Weak infrastructure is one of the most important bottlenecks for tourism in Uganda, along with poor hygiene. Insufficient infrastructure has consistently weakened Uganda’s rankings on the Travel and Tourism Competitiveness Index. The country ranked towards the bottom, even lower than the SSA averages, in all cat- egories of the infrastructure subindex: ground and port infrastructure (113), tourist service infrastructure (124), and air transport infrastructure (127); these also reflect limited road access and national park trails, and an insufficient mobile (3G) net- work. Ironically, some large-scale infrastructure projects targeted for oil develop- ment have hindered rather than helped tourism through the loss of iconic natural attractions. Uganda’s hygiene ranking (136) was worse than that in infrastructure. Tourism management is another area that calls for improvement. Ugandan tour- ism can benefit substantially from improved soft infrastructure, including several elements. The first is more strategic product development and diversification. The lack of tailored market intelligence, digital marketing skills and access to markets; undeveloped potential for community-based tourism to support conservation; and lack of affordable funding and dependence on foreign investment limit the sec- tor’s potential dramatically. The second is better human resources development. Problems in this area include inadequate skills training; lack of digital expertise; challenges to meeting customer expectations; lack of a nationally accredited qual- Chapter 3. Future challenges, opportunities, and policies ification for tour guides; and poor employment conditions, including long working hours and sexual exploitation. The third is coordination and facilitation. Problems here are weak destination branding and marketing strategies; limited coordination in planning and regional development; inadequate data collection; very limited sus- tainability accreditation; insufficient enforcement of regulations; and the preva- lence of corruption. For future development of the tourism sector in Uganda, policies should con- sider the following: (i) coherent strategies and flexible action plans for tourism development; (ii) improved national-subnational coordination by empowering local communities and fully sharing benefits from tourism locally; (iii) digital transforma- tion of tourism by modernizing regulatory frameworks, strengthening capacity, and expanding accessible digital infrastructure; (iv) greater focus on the environmental and sociocultural pillars of sustainability; (v) access to tourism data in a timely manner to help individual destinations develop appropriate niches; and (vi) iden- tifying tourism tradeoffs from other economic activities like deforestation and oil infrastructure explicitly to be used in medium-term planning and policies. Digital Transformation For a landlocked country like Uganda, where physical connectivity is costly, digital transformation is essential. The information and communications tech- 153 nology (ICT) sector accounted for less than 2 percent of Uganda’s GDP in 2019. However, according to the Digital Economy for Africa (DE4A) report (World Bank, 2020c), the sector has the potential to play a catalytic role in enhancing econ- omy-wide productivity growth and trade. Globally, the Internet has encouraged inclusion, efficiency, and innovation by lowering the cost of transactions, expanding markets and services to underserved communities, and making regional and global supply chains more efficient. Uganda can take full advantage of this unique op- portunity by addressing remaining gaps that hinder use of digital technologies to transform the country. In Uganda, digital transformation can foster structural transformation and trade. Some of the key obstacles to trade and structural transformation in Uganda can be overcome with better digitalization. Most prominently, these include high job search costs and low access to finance. Enhancing digital services in labor mar- ket transactions (e.g., online labor supply, demand, and matching mechanisms) and accessibility of such services can help reduce information asymmetry and uncer- tainty in spatial and sectoral transitions. Similarly, improvements in digital finan- cial services (DFS) can help Uganda’s poor and underbanked consumers, such as women and rural dwellers, access the formal financial system, thereby reducing the credit constraints that slow their spatial and sectoral transitions. Finally, e-govern- ment services for trade facilitation and value chain digitalization can bring down transaction costs. From smallholders seeking to improve farm management for productivity gains to SMEs’ compliance with quality standards and large firms en- tering export value chains through improved logistics management and use of big data, digital technologies offer value addition for enterprises across supply chains. Uganda has made significant progress in digital transformation, but many chal- lenges remain. Higher public and private investments in ICT infrastructure, includ- ing in fiber infrastructure, and improvements in the policy and regulatory environ- ment, has led to a major increase in ICT penetration in Uganda. In 2018, mobile infrastructure reached 83 percent of the Ugandan population and 44 percent of its territory. According to the DE4A report (World Bank, 2020c), the use of digital technologies such as mobile phones, mobile Internet, and social media has risen Growth, trade, and transformation significantly in recent years. The percentage growth of digital payments in Uganda (2014–2017) is the highest in the region at 12 percent, followed by Kenya with 10 percent. Despite this progress, however, Uganda has one of the lowest (14 percent) internet penetration rates in the world, and internet costs are still high (2.3 fold of that in neighboring Kenya), which precludes wider take-up and usage of services. Communications infrastructure is limited to key urban centers, while rural areas, particularly the northern region of the country, and the country’s 1 million or more refugees and host communities (RHCs) have limited or no connectivity. For those who do have access, cost of services and digital literacy constrains usage. 154 Table 11. Digital Evolution Index components Kenya Nigeria Rwanda Tanzania Uganda Access Infrastructure 31.94 24.25 32.80 20.29 20.50 conditions Supply Transaction Infrastructure 27.85 24.12 27.49 22.99 24.46 Fulfillment Infrastructure 53.73 45.41 53.64 50.36 42.63 State of the Human Condition 34.12 17.05 16.01 9.57 17.91 conditions Demand Device and Broadband Uptake 27.97 27.74 20.08 19.97 20.17 Digital Inclusion 60.04 21.56 38.92 53.23 51.16 Digital Payment Uptake 64.89 24.64 40.05 47.46 50.47 Institutional Effectiveness and Institutional en- 17.49 5.48 46.28 21.13 15.19 Trust vironment Institutions and the Business 62.37 49.88 74.78 64.27 60.62 Environment Chapter 3. Future challenges, opportunities, and policies Institutions and the Digital 66.81 54.47 75.05 57.54 60.27 Ecosystem Innovation and Inputs 41.29 25.91 38.36 26.57 31.34 change Processes 56.87 38.63 49.71 37.51 38.55 Outputs 13.35 9.66 11.08 10.14 9.25 Source: Chakravorti et al. (2020). Key bottlenecks The challenges faced by Uganda’s digital economy are many, but supply side issues stand out as a key constraint compared to other countries. According to the Digital Intelligence Index, Uganda ranks 81 out of 90 economies on key metrics including infrastructure access, digital inclusion, institutional readiness, and level of innovation in the digital economy (Table 11). Although the index is focused more on overall digital conditions and not specifically on the environment for firms, it offers key insight into the readiness of Uganda to scale up digitalization in domestic and cross-border economic transactions. Compared to its regional peers, Uganda also ranks particularly low on access and fulfillment infrastructure (supply conditions) and innovation, although it exhibits average demand conditions (digital access and inclusion). 155 The expansion of better physical infrastructure to support a digital transfor- mation in Uganda also requires progress in soft infrastructure and solutions for problems. These include the following: (i) stagnating mobile telephony prices; (ii) unsatisfactory quality of mobile service; (iii) high prices for digital devices and in- ternational access; (iv) lack of policies and regulatory instruments; (v) low availabil- ity of skilled professionals in cybersecurity; (vi) the need to strengthen protection of critical national information infrastructures; and (viii) the need to strengthen in- teroperability and information-sharing between national and sectoral Computer Emergency Response Teams (CERTs). With low digital literacy, usage beyond ba- sic services is limited. Authorities have introduced important legislation like digital signatures for financial transactions, but key gaps remain, e.g., in supporting digital transactions in the land sector. Digital taxation is one of the key constraints to broad market development of the digital economy and to digitalization of export value chains in particular. In April 2021, the Government introduced a 12 percent internet access tax, which replaced the Over-the-Top (OTT) social media tax imposed in 2018 (which failed to meet projections, according to the Uganda Revenue Authority). There are major design flaws to this internet access tax. First, it is a tax on an enabling service rather than on a value-added service (like e-commerce transactions on a platform); thus it introduces a friction against economic transactions. Second, Uganda already has one of the highest internet costs in the region and the additional tax will further inhibit access for a broad section of the population, especially in rural areas (where smallholders/agriculture producers are located). Third, the costs of digitalization could rise for businesses, which in turn can disincentivize adoption of digital tools, and raise costs throughout the value chain and potentially make Ugandan exports more expensive. The tax on mobile money withdrawals is regressive and restraining. Mobile money withdrawals were initially taxed at a rate of 1 percent of their value, which was later reduced to 0.5 percent. This tax poses a challenge to market develop- ment and access to DFS and associated digital products and services, especially among the most price-sensitive consumers. It also creates an unlevel playing field Growth, trade, and transformation since traditional financial services are not subject to such a tax. Increased taxation on access and services reduces the ability of start-ups to leverage digital services and develop a consumer base. With relatively low adoption and market develop- ment, a review of the taxation policy may be useful to balance the need to incen- tivize higher usage of digital services and to meet domestic resource mobilization priorities. An improved institutional environment is needed to boost e-commerce in Uganda. Domestic and cross-border e-commerce opportunities are determined by data protection and privacy, logistics and warehousing, electronic transactions, including financial transactions, and consumer protection. According to the United Nations Conference on Trade and Development (UNCTAD) e-Trade Readiness As- sessment, Uganda is making progress in developing the environment for e-com- merce in the country. While investment in the expansion of internet services is 156 continuing, rural access remains limited. The 2019 Data Protection and Privacy Law has been passed, but its implementation is not yet completed. The specific digital skills needed for e-commerce, including content development, digital marketing, and logistics management, are in short supply. Access to finance for e-commerce firms is low, with most financial institutions reluctant to lend to new entrants and small firms in the sector. Uganda is undertaking an ambitious plan to expand digital access. The NDP III FY20/21–FY24/25 includes digital transformation and innovation as key pillars of economic growth for Uganda. There is a recognition of ICT as a critical enabler of trade and economic development. The NDP III also highlights challenges including limited network coverage, poor quality services, high cost of devices and services, limited digital skills, and limited innovation capacity. Innovation and technology de- velopment and transfer is another key area in the NDP III; the priority is to invest in R&D domestically and to develop Uganda’s innovation capacity for increased industrialization. This reform momentum provides a great opportunity to enhance digital inclusion for firms and communities, by improving the policy and regulatory environment to encourage investments in rural and underserved communities; re- leasing newly available spectrum; making more efficient use of existing infrastruc- ture through infrastructure sharing; and reducing the cost of internet and digital devices. The Uganda Digital Acceleration Program (UDAP) is aimed at scaling up the national fiber infrastructure, mobile broadband for government offices, and Wi-Fi Chapter 3. Future challenges, opportunities, and policies hotspots. Key policy solutions The medium- to long-term development of the digital economy in Uganda re- quires a combination of regulatory and market development policies. There is a need to invest in soft infrastructure––policies, institutional capacity, digital literacy programs––to capitalize on Uganda’s advantages. The AfCFTA provides opportuni- ties for Uganda to access regional markets and for Ugandan exporters to broaden their customer base. Capitalizing on this opportunity requires continued invest- ments in expanding digital infrastructure, including the network and access to the network. Digital access, particularly for underserved populations (women, youth, rural residents) depends on the availability of internet and digital technologies, af- fordability of devices and services, and the skills to use these. Authorities should aim to enhance digital literacy programs, especially for small businesses and low-skilled workers, for market development. Uganda can increase its domestic value added and export market share through enhanced dig- italization. This is particularly true for its main export items, tourism, and coffee. In the coffee sector especially, the low level of digital skills among smallholders raises expenses for producers seeking to digitalize value chains. The UCDA undertakes extension services but its capacity to add smallholder training to the portfolio of services on offer is limited; however, low-tech digital training, e.g.,via Unstructured Supplemental Service Data (USSD) may offer a viable path for national programs 157 to train cultivators to better use digital technologies and access information in a timely manner. Effective strategies to lower costs of digital devices and services are needed, including removal or reduction of taxes on enabling services. Compared to its regional peers, Uganda lags on digital connectivity, access, and availability of locally developed digital solutions, and access to digital financial services. This may partly stem from Uganda’s taxes on enabling services. For example, the recent internet access tax is likely to inhibit the growth of the digital economy and slow down the emerging digitalization in key domestic and export value chains. Similarly, the taxes on mobile finance have a disproportionate effect on communities with no alterna- tive banking access. Revenue mobilization is an important priority for Uganda, but it needs to be balanced with prospects for growth from a digitally driven reduction in transaction costs. Policies should proactively promote domestic technology solutions and their integration into strategic value chains. One of the challenges for digitalization stems from the lack of locally available, affordable, and customizable solutions. Digital firms in Uganda are not well connected to traditional sectors like agribusi- ness and manufacturing. These linkages can be advanced through a program to digitalize priority sectors. For example, a technology adoption program (which in- cludes training, business advisory, and financial incentives) for the agriculture and manufacturing sectors can boost uptake and yield significant productivity benefits for the real economy, as well as grow the market for digital solutions’ providers. High-quality support services are needed to establish linkages between local dig- ital firms and industry. There are a small number of service providers in Kampala who are engaged in this matchmaking. Alternatively, public infrastructure could be established and run by the private sector as in the case of the National Information Technology Authority-Uganda (NITA-U)’s subsidized Internet to selected digital en- trepreneurship hubs. This can be scaled up in locations with agglomerated small- holders and manufacturing suppliers. Existing legislation should be fully implemented, and additional reforms un- Growth, trade, and transformation dertaken to deepen digitalization. Legislative initiatives like the Data Protection Act are only as effective as their implementation. While strides have been made in bringing some legislation in line with international good practices, implementation and intergovernmental coordination remain stumbling blocks. Moreover, some key sectors remain outside of the recent efforts to digitalize citizen services. One of the main examples is the land sector, where the low level of digitalization of land administration and business processes distributed among different stakehold- ers constrains the processing of vital land information and issuance of land titles. Adopting legislation and corresponding regulations to digitalize land administra- tion and management processes fully, such as use of digital signatures and e-con- veyancing, are essential for the land sector’s ability to address these constraints. These gaps limit the full use of land as a productive asset in Uganda and limit its contribution to economic growth. 158 Policies should support the AfCFTA negotiations on digital trade. There is a strong demand in the private sector, among exporters and potential exporters in Uganda, to realize the potential from the opening up of the regional market. This re- quires swift movement on the negotiations, especially on mutual recognition of na- tional laws and harmonization across the region. Digital taxation is one area where governments are struggling to balance market development with capitalizing on the profits generated in their respective national jurisdictions. As one of the countries with operational laws on data protection and e-transactions, Uganda has the op- portunity to take a leadership role in the AfCFTA negotiations on digital trade. Such leadership will also provide opportunities to align Uganda’s domestic policies with regional dynamics. Hydrocarbons Starting in 2006, Uganda discovered sizeable oil reserves in the Lake Albert re- gion. This region (about 23,000 square kilometers) runs along Uganda’s western border with the DRC and is also about 1,200 kilometers from the nearest coast (Figure 48). The exploration area is spread over two projects: (i) the Tilenga project, covering the districts of Nwoya and Buliisa, consists of 10 fields with more than 400 wells (to be brought into production at different times); and (ii) the Kingfisher Chapter 3. Future challenges, opportunities, and policies project, in the Kikuube District, with 31 wells. According to the World Bank (2020), the stock of oil in the Lake Albert region is estimated at approximately 6.0 billion barrels, with a recoverable resource of 1.38 billion barrels. Forty percent of the pro- spective resource has been explored, and 10 percent has been licensed. Several factors have slowed field development and delayed production. Fifteen years after the first discovery, oil production has not yet begun. The delays have largely been driven by three factors (World Bank, 2021). The first concerned several tax disputes between the Ugandan government and the international oil compa- nies (IOCs) regarding the taxation of portfolio sales. The second concerned nu- merous disagreements between the government and IOCs regarding downstream arrangements: diverging views about a prospective refinery (whereas the govern- ment wanted to refine all oil domestically, the IOCs wanted to export most, if not all, in crude form); and a standoff on the pipeline route for the crude export pipeline (Tullow and the government favoring a route through Kenya, and Total favoring a route through Tanzania). Finally, the third factor concerned the commercial viability of the project. With the relatively lower quality of the Albertine oil, high pipeline transportation costs, along with high operating costs stemming from geographic attributes (remoteness and environmental sensitivity) and infrastructure condi- tions, the Albertine development had a relatively high breakeven oil price. Thus, the collapse of oil prices in 2014 threatened the viability of the project for several years. 159 Figure 48. Major oil fields and refinery and pipeline positions Ugandan fields with proven reserves lie in the Albert-Edward Rift Basin in northwest Uganda, near the border with the DRC. This area comprises part of the East African Rift System. a. Fields, operators, and the proposed refinery location The pipeline that will export the Ugandan oil stretches from Hoima, where a refinery is scheduled to be built, to the Tanga Port in Tanzania (pink line). b. Pipeline route Growth, trade, and transformation Source: Wood Mackenzie (2021). 160 Figure 49. Oil production and fiscal projections Latest projections put the beginning of oil production in 2025. At its peak, production reaches about 230,000 b/d. a. Oil Production 300 250 Thousand BBL/D 200 150 100 50 0 2040 2050 2030 2044 2020 2046 2048 2049 2045 2034 2036 2043 2038 2035 2042 2024 2026 2039 2029 2025 2028 2033 2047 2023 2032 2022 2037 2027 2041 2031 2021 Kingfisher Block 2 Block 1 Downstream Capacity Fiscal revenues from oil production are projected to reach their peak (USD 3.3 billion in 2021 dollars) by 2030, with ‘profit oil’ (yellow bars) contributing the most. b. Government Upstream Revenue 4000 3500 3000 Chapter 3. Future challenges, opportunities, and policies Million USD 2500 2000 1500 1000 500 0 2040 2050 2030 2020 2044 2045 2046 2048 2049 2034 2035 2036 2038 2043 2024 2025 2026 2028 2029 2033 2039 2042 2023 2032 2047 2022 2037 2027 2041 2031 2021 Income Taxes UNOC Participation Profit Oil Royalties With frontloaded investments in the refinery and the pipeline, net cash flows will remain negative until 2025. c. Government Revenue Total 3500 3000 2500 Million USD 2000 1500 1000 500 0 -500 2040 2050 2030 2044 2020 2046 2048 2049 2045 2034 2036 2043 2038 2035 2039 2042 2024 2026 2029 2033 2025 2028 2032 2047 2023 2022 2037 2027 2041 2031 2021 Refinery Pipeline Upstream Source: World Bank staff calculations using data from the Petroleum Authority of Uganda (PAU) and Wood Mackenzie. 161 Progress has now been made on downstream plans for the refinery and pipeline. Uganda’s oil will be exported through a 1,443-kilometer heated pipeline with trans- port capacity of over 200,000 barrels per day (b/d), stretching from Hoima in the Lake Albert Basin to the Port of Tanga in Tanzania. The pipeline will be owned and operated by a third-party joint venture, with Ugandan upstream partners (Tullow Oil, Total, and CNOOC) holding the largest share. In Hoima, a 60,000 b/d refinery is scheduled to be built by a Saipem-led consortium. The capacity exceeds the current domestic demand, which is 35,000 b/d, but is expected to grow by about 7 percent annually. Additional pipelines connecting the refinery to Kampala (200 kilometers) and from there to the Mombasa-Eldoret pipeline in Kenya (320 kilome- ters), and possibility also to Rwanda, are expected to be built. According to latest estimates, oil production will begin in 2025. The Final In- vestment Decision (FID) was scheduled in 2021, but had not taken place by the time this report was prepared. Assuming no further delays, the first production of oil is expected in 2025 with all three fields becoming operational. Production is projected to reach its peak at 230,000 b/d quickly, and then, in the absence of fur- ther discoveries, gradually decline after 2030. The refinery is expected to become operational in 2026, with a peak load of 51,000 b/d. This level will be sustained for about two decades before lowering workload as oil production phases out. By that time, the oil pipeline is expected to cease transporting oil, as any oil produced will be refined for domestic consumption. Authorities expect that with more explora- tion efforts and technological improvements a higher recovery and production rate may be feasible in the future. However, these expectations for potential increases in production capacity are counterbalanced by climate change uncertainties sur- rounding fossil fuels in the future. Why do hydrocarbons matter? In the medium-term, fiscal revenues from oil production can boost Uganda’s fiscal space significantly. Assuming an oil price of USD 70 per barrel throughout Growth, trade, and transformation in real terms, the analysis in Figure 49 suggests that fiscal revenues from oil pro- duction can reach about USD 3.3 billion per year during peak production (2030). This would likely correspond to about 4.9 percent of GDP and about a quarter of fiscal revenues from other sources at that time11. Today, that revenue would cover a large share of the deficit (net borrowing) in the fiscal budget. Therefore, although Uganda’s hydrocarbon reserves are by no means large, the fiscal revenues they may generate can help accelerate the country’s economic and social developments. If used well, the additional revenues from hydrocarbons can help close Ugan- da’s human development and infrastructure gaps. Better provision of infrastruc- ture services, like transportation and electricity, will be a crucial determinant of productivity growth and structural transformation. The authorities have already begun undertaking investments in critical infrastructure to support oil production and its commercialization, starting with roads. The Hoima-Kaiso-Tonya Road, which connects the oil wells, was completed in 2014, and the Nyamasoga Oil Treatment 162 Plant was completed in 2015. The government is in the process of securing more resources to construct and rehabilitate roads in the oil region, including the Ma- sindi-Biiso, Kabaale-Kiziranfumbi, and Hohwa-Nyairongo-Kyarusesa-Butoole roads. In addition, an international airport to serve the oil region, and an industrial park, are in the works. These investments will undoubtedly contribute to the domestic economy, initially in their construction as some of them could be tapped by local suppliers, according to the country’s local content policy, and by local communi- ties, and later on when they can be used by other economic players and industries. In the short-term, however, the pursuit of oil production will add to the coun- try’s fiscal burden. Investment in the exploration phase until now has already cost an estimated USD 3.5 billion, largely in the private sector. Before pumping the oil, additional investment is required for upstream activities, including the development of the fields and associated infrastructure development, some of which will have public-sector financing. This will increase both overall costs and the fiscal burden. The National Oil Company estimates that construction of the refinery could cost around USD 4 billion, excluding working capital costs that are yet to be estimated. In addition, the total cost for the construction of the pipeline is estimated at about USD 3.65 billion. Overall, the project capex is estimated at USD 13.1 billion, more than a third of Uganda’s current GDP. The annual gross project capex is projected to reach a peak of about USD 3.3 billion by 2023, which will generate a USD 350 million gross financing need for the Ugandan government (Figure 50). Chapter 3. Future challenges, opportunities, and policies What are the key risks and how to mitigate them? The most prominent risk is a potential delay in production, or even a complete shelving of the project, which can add to the country’s fiscal burden. Uganda’s fiscal outlook has deteriorated in recent years: the public debt-to-GDP ratio has increased from 27 percent to 48 percent since 2015 (about two thirds of which is external). As a result, a Debt Sustainability Analysis (DSA) conducted in June 2021 downgraded Uganda’s rating to a moderate risk of external and overall public debt distress, with limited space to absorb shocks, especially export shocks. In the medium-term, the government’s capex obligations will add to this deteriorating outlook before oil revenues can provide some relief. A potential delay in produc- tion, dissolution of prospects, or a sharp decrease in oil prices can all distort the fiscal outlook dramatically. The World Bank’s SCD Update in 2021 highlighted these risks on the basis of high cost and marginal economics of the oil fields, which in- clude high pipeline transportation costs, poor crude oil quality, and the remoteness and environmental sensitivity of the Lake Albert region resulting in high operating costs. Even if oil production remains economically viable in the long-term, with a narrow fiscal space and frontloaded capex requirements at the outset, the govern- ment can face financial bottlenecks. The authorities are thus advised to consider conservative projections that allow for such downside risks while preparing medi- um-term fiscal frameworks. 163 Figure 50. Capex and government financing needs projections Gross capex for oil production is projected to peak in 2023 (at USD 3.3 billion), comprising mainly upstream and pipeline investments. Refinery costs will run until 2028. a. Gross Capex 3500 3000 2500 Million USD 2000 1500 1000 500 0 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 Government obligations in capex will translate into a gross financing need peaking at about USD 345 million by 2023. b. Government Financing Requirements 400 350 300 250 Million USD 200 150 100 50 0 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 Refinery Pipeline Upstream Growth, trade, and transformation Source: World Bank staff calculations using data from the PAU and Wood Mackenzie. Once the oil starts flowing, one of the most important fiscal challenges be- comes the volatility of oil revenues. Oil prices are highly volatile. Most notably, in April 2020, the price of oil futures contracts decreased to below zero as petroleum demand fell and crude oil inventories increased, leading to a storage glut. At the time of this analysis, they were trading above USD 82. Such extreme volatility and uncertainty can be very harmful if transmitted to the non-oil economy through sudden adjustments in public expenditure. Therefore, authorities are advised to pursue a proactive stabilization policy: saving a portion of revenues when prices are high, and using the savings to offset gaps when prices are low, thereby effectively injecting a more stable and predictable flow of finances into the domestic economy. 164 Box 4. Hydrocarbons: curse or blessing? Oil was named the “devil’s excrement” by Venezuela’s former Development Minister and OPEC Founder, Perez Alfonso. Numerous studies have suggest- ed that countries with windfall revenues from oil and similar natural resources have not done better than others in economic performance, or have even done worse, a phenomenon known as a resource curse. Frankel (2012) suggests five key factors behind these findings: price volatility, permanent crowding out of manufacturing, autocratic/oligarchic institutions, anarchic institutions (con- flict), and Dutch disease. The underlying processes involve one or more of the following: • Resource windfalls boost demand for goods and services; while the pric- es of tradable sectors are fixed, those of non-tradables increase in the short-term; thus labor is allocated towards non-tradables. • If tradable sectors have productivity advantages (e.g., learning by doing), this may lead to permanent losses in competitiveness (with or without an appreciation in exchange rates). • The windfall also leads to an increase in rent-seeking (voracity effect) as incentives are tilted towards predatory activities rather than productive use of resources. Chapter 3. Future challenges, opportunities, and policies The scale and composition of these effects in a country will often depend on institutional factors and policy responses. For example, barriers to exports (and the imports of intermediary goods) can further aggravate Dutch disease dynamics. Similarly, Behzadan et al. (2017) show that a higher inequality in ben- efiting from resource revenues can worsen Dutch disease effects. Similarly, Cordella and Onder (2020) show that while sharing only a small share of re- source revenues subnationally can increase rent-seeking and conflict, sharing a large share can reduce them. Some research suggests that a resource curse can take hold even before oil flows (see Mihalyi & Scurfield (2020), Cust & Mihalyi (2017), for example). The Ugandan growth performance has not been systematically worse than World Economic Outlook (WEO) projections (in vintage series) since 2006 (Figure 51). However, the recent hike in public debt, on the expectation of future oil revenues, can justify such concerns, especially if the production is delayed further and the fiscal burden increases. 165 Box 4. (continue) Figure 51. Is there a resource curse in Uganda already? GDP growth in Uganda (blue line) does not exhibit a systematic gap vis-à-vis vintage WEO forecasts (dotted-gray lines), reducing concerns for a possible resource curse in the real economy. Real GDP Growth Projections 11 Second oil discovery 9 Growth rate, percent 7 5 3 First oil discovery 1 -1 Covid-19 pandemic -3 2000 2005 2010 2015 2020 2025 Actual WEO Forecasts 2021 Forecast Source: WDI and WEO vintage forecasts by IMF. Source: WDI and WEO vintage forecasts by IMF. The World Bank (2020) analyzed alternative frameworks for such stabilization poli- cies, including Sovereign Wealth Funds that can serve as an instrument for depos- iting and managing savings, and fiscal rules that regulate the limits of discretion in saving/spending decisions. The next important fiscal challenge is the efficient spending of oil revenues, considering the economy’s absorptive capacity constraints. In countries with weak institutional resilience, windfall resources can boost rent-seeking and lead to Growth, trade, and transformation a more than proportionate increase in spending (or transfers) of public funds for private interests. This voracity effect (Tornell and Lane, 1999) can also take place through a rapid increase in public investments, which can overwhelm the country’s public investment management system and lead to the selection of bad invest- ments with poor economic rationale, thus reducing potential growth. These risks require a multi-pronged approach: • Build public investment management (PIM) capacity to assess the economic viability of investment opportunities, and implement a stringent project se- lection procedure. Some reforms and institutional strengthening have taken place over the last 5 years (World Bank SCD Update, 2021), yet critical steps remain to be completed. • In light of PIM assessment, and the cash flow projections in the medium- and long-term, impose a speed limit on public investment, selecting only the best 166 projects, and ‘park’ the money in other investments (e.g., a portfolio managed by a Sovereign Wealth Fund) until more such projects become available. • Manage spending expectations with a comprehensive public-relations pro- gram before the oil starts flowing. Communicate with full transparency to preempt public perception of being excluded from sharing the benefits. In the long-term, oil revenue inflows can also pose structural challenges to the Ugandan economy. While the fiscal revenues from oil can help build much needed physical and human capital in Uganda, the economic gains from this process are neither guaranteed nor automatically generated. Most importantly, adverse eco- nomic impacts that may be generated by windfall revenues can limit the com- petitiveness of Uganda’s tradable sectors (Dutch disease) and lower the country’s growth potential (a resource curse), as explained in Box 3.1. Fundamentally, this is a problem about Uganda’s transformation: given the country’s relatively small mar- ket size, tradable sectors like manufacturing, agriculture, and tradable services like tourism will likely play a crucial role in future growth. However, the windfall revenues can reallocate productive resources to non-tradable sectors, thereby reducing growth potential (if the tradable sectors have a greater productivity growth po- tential). In Uganda, oil’s toll on tradable sectors can also manifest through its environ- mental impact. Uganda’s oil reserves are concentrated in an area with one of the Chapter 3. Future challenges, opportunities, and policies greatest levels of biodiversity on the planet, including the rare mountain gorilla. There are several national parks, forest reserves, and wildlife sanctuaries in the vi- cinity of oil exploration zones. Some of these areas are fully, or in part, inside the oil blocks, and others that are not in the zone are still likely to be affected as a result of developments related to oil. In addition to the likely ethical concerns re- garding biodiversity, the environmental impact of oil sector-driven activities can also directly reduce the country’s tourism potential. Therefore, such costs of en- vironmental degradation should be factored into economic planning, and relevant mitigation policies should be adapted. The enactment of a right of nature law in 2019 recognizes that considering natural capital is a step in the right direction, but it is the enforcement and legal interpretations of that law (which has proved to be challenging in other countries) that will shape the outcomes. Will oil be a blessing or a curse for Uganda? The answer to this question will largely depend on policies going forward. Oil revenues provide an excellent op- portunity to extricate Uganda from a low equilibrium by enhancing infrastructure, building human capital, and promoting private economic activity. However, cap- italizing on this opportunity is not easy. First, the auhorities will need to get the fiscal management right to achieve stabilization and efficiency objectives. Second, policies will need to target conditions that can reinforce Dutch disease dynam- ics in the economy. These include trade barriers (i.e., difficulties in exporting final products or importing intermediary inputs), non-inclusive prospects for economic participation, and institutional weaknesses that can facilitate rent-seeking, among others. Third, policies should respond effectively to concerns about graft, environ- mental impact, lagging regions, and disadvantaged groups to preempt grievances 167 and social unrest. In the final analysis, there is only one way to make the most out of oil, and that is to get these policies right. In comparison, there are many ways to fail, as one can sadly see in several countries around the world. Climate Change Climate change poses broad-based challenges to all economies and Uganda is no exception. Higher temperatures and variability in climate-driven natural con- ditions affect agriculture systems, natural landscapes, and the environment; and threaten not just rural livelihoods but the economy as a whole. The increased fre- quency and intensity of droughts, floods, and locust invasions in recent years in Uganda has shown how these events may manifest in the coming decades. Uganda’s climate has already changed and will change further. Observations to date show an uptick in extreme precipitation events, particularly during the shorter rainy season. The region can also expect to see reduced precipitation between rainy seasons, decreasing annual rainfall. The onset of rainy seasons can shift by 15–30 days (earlier or later), while the length of the rainy season can change by as much as 20–40 days from year to year. In all areas, the projected increase in rainfall includes December, January, and February, typically the months of the dry sea- son. Rainfall is predicted to increase significantly and consistently for the western shores of Lake Victoria and the central Western region, the Mount Elgon region, and the region extending from Mount Rwenzori to the southern parts of Lake Kyoga. The greatest change in the intensity and frequency of extreme rainfall events is likely to take place between the current and mid-century period. In Uganda, temperatures have become warmer, and the number of hot days has increased. The average annual temperature has increased by 1.3 degrees Cel- sius since 1960, and the average number of hot days and nights per year has also increased. Reports from the Famine Early Warning Systems Network indicate that there has been an increase in seasonal mean temperature in many areas of Uganda Growth, trade, and transformation over the last 50 years. With increasing global average temperatures, this trend in Uganda is likely to continue and intensify. Key challenges One of the most prominent effects of climate change in Uganda is likely to be in agriculture. There is a predicted climate-induced yield loss for certain specific climate-sensitive crops: the hardest hit will be coffee, tea, cassava, potato, and sweet potato. Rising temperatures are expected to increase conditions for crop diseases, pest infestations such as blast and bacterial leaf blight in rice, fungal and viral diseases in banana and beans, and coffee rust in coffee trees. The loss of food crops is estimated at about USD 1.5 billion per year by 2050. For example, the ar- abica coffee-growing area will significantly shrink by 2050, with yield losses in the 168 order of 50–75 percent. There is a similar shrinking of suitable tea areas. In addi- tion, reduced water availability and watershed recharge is likely to stress fisheries, resulting in disrupted livelihoods and significant economic losses. Several of these crops and commodities are already important export products and have great po- tential for further increasing Uganda’s exports in the future. Besides their direct impact on agriculture, the rising temperatures and climate variability also have important indirect effects, including a greater incidence of disease and pests. Climate extremes can be conducive to pathogens, diseas- es, and pests, which in turn affect agricultural production, productivity, and entire economies. For example, in 2020, after a series of cyclones, the most serious des- ert locust outbreak in 25 years hit Uganda. Large swarms can eat the same amount of food as 80 million people in a single day, devastating crops and elevating the risk of food insecurity. Changing weather patterns help the migratory locust to thrive, as flood conditions leave sandy soil in arid and semi-arid areas moist; and in the last three years the increase in the frequency of cyclones in the Indian Ocean has played a role in this upsurge by supporting their breeding. During the 1997/98 El Niño, a cholera outbreak affected 39 districts in Uganda. Finally, climate-driven ecological and economic pressure can aggravate migra- tion or conflict. The World Bank (2018) investigated the scale of climate-induced migration driven by slow-onset climate factors (water stress, drops in crop produc- tivity and sea level). This study identified the Lake Victoria region as a climate mi- Chapter 3. Future challenges, opportunities, and policies gration hotspot, with between 7 percent (optimistic scenario) and 11 percent (pes- simistic scenario) of the population becoming internal climate migrants by 2050 (Figure 52). These trends are accompanied by high levels of climate in-migration to the capital, around Mount Elgon, and in the southeast and along the border with Rwanda. High levels of out-migration take place in the North-west and Cen- tral-west regions. Overall, these areas correspond to zones of positive and negative changes in water availability and livelihood viability, and the magnitude of internal climate migration will be on a par with or exceed other forms of development-re- lated migration. Key policies Investing in adaptation will play a central role in coping with the challenges posed by climate change and extreme weather events. In Uganda, a broad- based approach to adaptation is essential in the country’s climate change coping strategy. This will include both targeted interventions like technological adaptation solutions and across-the-board policies like policy reforms that aim to correctly value the country’s natural assets. In agriculture, a climate-smart strategy for building resilience and enabling ad- aptation is required. A focus on climate-smart adaptations (CSAs) for agriculture, integrated watershed management, and climate-resilient landscapes (drylands and forests) is key to boosting productivity in a changing climate. In coffee, for 169 instance, drought and flood prone varieties can reduce the yield pressure driv- en by changing weather conditions. A more widespread use of digital technology in the design and delivery of integrated weather and market advisories can help farmers make informed decisions. To identify, design, and implement such CSAs, a multi-stakeholder platform can be established. This platform can identify and promote cost-effective CSA interventions by monitoring and managing new trends in pests and disease, promoting no-burn agricultural practices, and helping adopt better varieties (e.g., semi-stabled cattle systems) to increase the country’s agri- cultural resilience. Climate resilience and food security are important components for guiding all future considerations around infrastructure investments. In addition to enhanc- ing trade and regional connectivity and limiting frictions to rural-urban migration, the design, appraisal, and selection of new infrastructure should prioritize climate resilience for food security as put forward in the Next Generation Africa Climate Business Plan. For example, a key question is whether new water and sewage sys- tems can withstand the gradient of variation in rainfall. Hallegatte et al. (2019) find that in low- and middle-income countries, for every dollar invested in more resilient infrastructure, four dollars are gained in benefit. This can include solar-powered irrigation pumps, soil conservation and tillage practices, and off-grid, solar-pow- ered agricultural appliances suitable for smallholder farmers; as well as adaptation support, e.g., precision farming systems, such as precipitation detectors combined with irrigation optimization, for real-time crop management advice. Microgrid sys- tems with small-scale public/private investments should be encouraged. Ensuring the resilience of infrastructure against storm surges, coastal and lake erosion, and high winds is also critical for cities. An important part of building resilience includes an adaptive social protection system that can help reduce the persistent effects of transitory shocks and variability. Adaptive social protection enhances the socioeconomic resilience of vulnerable households by providing the triple wins of social protection, climate change adaptation, and disaster risk reduction. Uganda has already scaled up safe- ty net programs to respond to shocks, focused on early warning information, in- Growth, trade, and transformation cluding the use of seasonal assessment and the creation of new triggers based on data from satellites. Further social safety nets, including targeted agricultur- al insurance schemes that use improved climate data, can allow more vulnerable households to avoid costly coping strategies and help build community resilience and address gender-specific vulnerabilities. 170 Figure 52. Predicted climate-driven migration patterns Chapter 3. Future challenges, opportunities, and policies Source: World Bank (2018). Source: World Bank (2018). For better implementation of these policies, Ugandan authorities should have better access to data and information on disaster risk and natural resource management. A better understanding of disaster risk and the role of natural re- sources in an economy is necessary to design effective policies for climate adap- tation. For Disaster Risk Management (DRM) progress in community-based early warning systems, hydrometeorological service delivery systems and climate infor- mation are needed. As an example, the Northern Uganda Social Action Fund Project already provides well-known DRM data; however, demand for these services will in- crease as damages affecting both people and assets unfold. Similarly, ecosystems and associated hydrological systems provide a first line of defense against current and future climate impacts. Hydroinformatics and surface flow, and weather fore- casting, can systematically improve the understanding of hydrology, which can help flood and drought management and support resilient water systems. Remote sens- ing can also be used for monitoring trends. 171 Synopsis This chapter analyzed future options for growth and transformation in Uganda. The analysis first presented the likely trends under a business as usual scenario (with the economy continuing its trends of the last decade) by using the struc- tural transformation model. Next, it considered the effects of policy interventions that lead to a 2 percentage points increase in manufacturing TFP growth and lower mobility frictions. This was followed by a discussion on potential factors that can provide opportunities or challenges, and policy recommendations for growth and transformation. These included (i) regional integration and trade (especially the AfCFTA); (ii) hydrocarbons; (iii) tourism; (iv) digital transformation; and (v) climate change. To accelerate growth and transformation, Uganda needs a breakthrough. With- out a breakthrough from recent trends (business as usual scenario), Uganda’s transformation is projected to remain modest in the medium-term (agriculture’s labor share decreases by only 1.6 percentage points). However, policies aimed at boosting productivity and reducing frictions can put the country on a more rapid growth and transformation path. With 2 percentage points higher TFP growth in manufacturing, Ugandan GDP can gain an additional 20 percent in a decade com- pared to the baseline projections. Similarly, with lower frictions, labor reallocation accelerates, and the economy grows by 12 percentage points more despite no ad- ditional gains in sectoral TFPs. To facilitate such a breakthrough, authorities can adopt a two-pronged ap- proach that removes prevailing constraints and exploits more targeted oppor- tunities. The first includes removing key constraints, including high food tariffs, costly trade logistics, limited rural access to infrastructure and information, and market imperfections like credit constraints and low liquidity of assets like land. The second involves taking a more proactive stance for identifying and exploiting opportunities that can transform the economy structurally, as follows: i. Boosting agricultural productivity. Uganda’s low and stagnant productivity Growth, trade, and transformation in agriculture is further threatened by population trends and climate change. To boost productivity and resilience, authorities should (i) encourage vertical integration, (ii) reduce transaction costs through better logistics and (iii) fa- cilitate the adoption of better inputs, digital technologies, and climate-smart agriculture practices including sustainable land management techniques. ii. Enhancing regional integration and trade. The AfCFTA provides potential opportunities in the form of reform-driven income and efficiency gains (e.g., a 3.3 percent boost to Uganda’s GDP by 2035). However, there are major risks, e.g., coordination challenges within the EAC and the loss of protec- tion-driven competitiveness. Success will depend on enhancing regional co- operation and preparing businesses for open competition. iii. Tapping into the untapped potential of tourism. The sector is already the leading foreign exchange generator for the country, but its potential is much larger––it can diversify products and attract visitors globally. But risks 172 abound: a lingering pandemic, environmental hazards from oil development, failure to strengthen infrastructure and improve hygiene standards, and inad- equate tourism management. Authorities can take advantage of the current downturn to develop a comprehensive strategy for the sector. iv. Transforming with digital transformation. In Uganda, digital transformation can help reduce many of the frictions (e.g., costly connectivity and access to finance and information) that stall the country’s structural transforma- tion. Challenges, however, are comprehensive and some are policy driven, e.g., taxes on digital systems and mobile money withdrawals. Reforms should focus on a combination of regulatory and market development policies. v. Utilizing hydrocarbons for development. If successful, oil revenues can gen- erate significant fiscal revenues (about 4.9 percent of GDP by 2030), which are much needed for financing development. However, substantial risks re- main before this can happen, i.e., fiscal exposure driven by delays or even cancellation, environmental costs, and the possibility of a resource curse. There are many ways to fail, only one way to succeed, and that is by getting the policies right at the outset. vi. Climate change adaptation. Uganda’s climate change has already begun and will get worse. The country faces direct effects in agriculture (e.g., a 75 percent yield loss in arabica coffee) and a wide spectrum of indirect effects such as disease and migration. Uganda’s trade potential has been thwarted Chapter 3. Future challenges, opportunities, and policies in the recent past by its inability to monitor and control pests and disease. Adaptation should focus on (a) a climate-smart agriculture strategy, (b) cli- mate-smart infrastructure, (c) an adaptive social protection system, and (d) better data and information on disaster risk and natural resource manage- ment. References AGRA. 2017. Africa Agriculture Status Report: The Business of Smallholder Agricul- ture in Sub-Saharan Africa (Issue 5), Nairobi, Kenya: Alliance for a Green Revolution in Africa. Behzadan, N., Chisik, R., Onder, H. and Battaile, B., 2017. Does inequality drive the Dutch disease? Theory and evidence. Journal of International Economics, 106, pp.104-118. Bhagwati, Jagdish. N. 1988. Protectionism (Vol. 1). Cambridge, MA: MIT Press. CIAT, BFS/USAID. 2017. Climate-Smart Agriculture in Uganda. Country Profiles for Africa Series. International Center for Tropical Agriculture (CIAT); Bureau for Food Security/United States Agency for International Development (BFS/ USAID), Wash- ington, D.C. Cordella, T. and Onder, H., 2020. Sharing oil rents and political violence. European Journal of Political Economy, 63, p.101882. 173 Cust, J.F. and Mihalyi, D., 2017. Evidence for a presource curse? Oil discoveries, el- evated expectations, and growth disappointments. World Bank Policy Research Working Paper, (8140). Delgado, C. 2018. Uganda: A Reform Agenda for More and Better Jobs through Ag- riculture. A background paper for the World Bank Uganda Jobs Strategy. Jayne, T.S. and Kray, H. 2018. Unmistakable Signs of Agri-food Systems Transforma- tion in Africa. Agricultural Working Group (AWG) seminar, Dar es Salaam, Tanzania (9 April 2018). Frankel, J.A., 2012. The natural resource curse: A survey of diagnoses and some pre- scriptions. Commodity price volatility and inclusive growth in low-income coun- tries, pp.7-34. Mihalyi, D. and Scurfield, T., 2020. How Did Africa’s Prospective Petroleum Producers Fall Victim to the Presource Curse?. Muratori L. 2016. Price Gap along the Ugandan Coffee Value Chain. FAO Working Papers series. ISSN 2385-2755. Onder, Harun. 2012. “What Does Trade Have to Do With Climate Change?” VoxEU 12. London: Centre for Economic Policy Research. ReliefWeb. 2015. Economic Assessment of the Impacts of Climate Change in Uganda  https://reliefweb.int/report/uganda/economic-assessment-impacts-cli- mate-change-uganda-final-study-report-november-2015  Sheahan, M. and Barrett, C. B. 2014. Understanding the agricultural input landscape in sub-Saharan Africa: Recent plot, household, and community-level evidence. World Bank Policy Research Working Paper 7014, August 2014. Tornell, A. and Lane, P.R., 1999. The voracity effect. American economic review, 89(1), pp.22-46. Tschirley, D., Reardon, T., Dolislager, M. and Snyder, J. 2015. The Rise of a Middle Class in East and Southern Africa: Implications for Food System Transformation. Journal Growth, trade, and transformation of International Development. (27) 5, July. 628–646. USDA International Agricultural Productivity database (https://www.ers.usda.gov/ data-products/international-agricultural-productivity/). World Bank. 2018. Closing the potential-performance divide in Ugandan agriculture. Washington, D.C.: World Bank Group. World Bank. 2019. Republic of Uganda, Agriculture Sector Public Expenditure Re- view. Washington, D.C.: World Bank Group. World Bank. 2020. The African Continental Free Trade Area: Economic and Distri- butional Effects. Washington, DC: World Bank. World Bank. 2021. Uganda Systematic Country Diagnostic Update. Washington, D.C.: World Bank Group. 174 Appendices Appendix A. A Simple Calculus of Migration Decisions Consider a simple dynamic environment with two periods, where the second peri- od has a variable length l > 0, reflecting differences in the planning horizon (i.e., the age of the worker). Each worker is endowed with an income in the first period. If a worker stays in the home region, she is endowed with the same endowment in the second period as well. Alternatively, a worker may choose to migrate at the end of the first period. Once attempted, the migrant may find a job with a probability , which generates an income in the second period. In the absence of access to credit, and abstracting from transfer of resources across periods, we can now define the lifetime utilities as follows: where is the period utility function with and and denotes costs associated with migration. Thus, gains from migration are defined as: Even this simple framework can help us to derive some interesting observations regarding the factors that promote or prevent migration. • A higher wage gap between source and destination locations increases mi- gration. For a given , • Higher probability of finding a job at the destination increases migration: • A higher migration cost reduces migration: • Younger workers are more likely to migrate. Note that, for migra- tion even to be considered, , and with that: , i.e., the longer the remaining lifetime, the higher the gains from migration. Finally, the impact of initial income on migration is ambiguous even in such a simple Appendices framework. 177 When is sufficiently small and satisfies the Inada conditions, the following properties are observed for a given : Intuitively, a small improvement in the initial income has two effects on the com- parison of expected lifetime utilities across options. First, it reduces the opportuni- ty cost of the migration, i.e., reducing consumption by is easier when the first-pe- riod income is greater. Second, it makes migration less attractive. That is because the gap between second-period utilities in migration and no-migration scenarios is lower with a larger . When the initial income is high, i.e., , the second effect dominates, and migration becomes less attractive. In contrast, starting from an income that is too close to , i.e., , a small increase in income relaxes the opportunity cost of migration dramatically, which dominates the second effect. Figure A.1 shows how these factors affect outcomes over the range of . The top panel shows the payoffs associated with not migrating ( ), which are identi- cal regardless of mobility costs, and migrating with and without mobility costs ( and , respectively). Inclusion of a mobility cost makes the migration payoff steeper than the non-migration payoff for low income levels. As a result, as shown in the second panel, whereas the gain from migrating decreases mono- tonically in initial income in the case with no mobility costs ( ), it becomes non-monotonous in the case with mobility costs ( ). Thus, for low income levels, a small increase in initial income can make migration more desirable. In fact, for some ranges of and , a double crossing of the non-migration and migration payoff streams is also possible. In that case, only workers with an intermediate range of initial income would migrate. Growth, trade, and transformation 178 Figure A.1. Initial income and the gain from migration The top panel compares lifetime welfare in the case of not migrating ( ), and for migra- tion with and without mobility costs ( and , respectively). The bottom panel compares gains from migration without mobility costs ( ) and with mobility costs ( ). Only workers with initial incomes between and will choose to migrate. Appendices Source: World Bank based on Beaman, Onder, and Onder (2022). 179 Appendix B. The Structural Transformation Model Model setup We consider a three-sector small open economy model with the production and consumption of three final goods: agriculture ( ), industry/manufacturing ( ), and services ( ), indexed with . Agricultural production uses labor and land, while in- dustry and services production use only labor. Agriculture and industry goods are tradable, and services are not. Preferences in consumption are non-homothetic with a subsistence requirement, with trade balanced in each period. We incorpo- rate friction as sector-specific distortions face by producer firms. Our framework is like those of Teignier (2018), but we do not use capital. Production and technology There is a competitive market in each sector and each has its own aggregate pro- duction function. We posit the following production function for agriculture: The production function in industry and services is given by where denotes production in sector , the proportion of labor employed in Growth, trade, and transformation and is the per-capita land stock need to produce agricultural good, is exog- enous productivity in sector . Per capita land stock falls over time because the stock of land is fixed, and population growth is positive. Firm’s Problem Producers in the agriculture sector choose labor and land stock inputs to maximize profit: and the firms in industry and services solve the following problem: 180 where is price of sector , is wage rate and the rent of one unit of land stock and is sectoral wedge. This wedge can refer to reallocation expenses (transport and housing costs), formal training to acquire the skills used in another sector or an opportunity cost (the time spent looking for a job in a different sector). Perfect competition implies that each production factor is paid according to its marginal product so that: In the absence of differential, the intersectoral labor wedge (i.e., ) labor costs are equal across sectors. Then, the introduction of the sectoral distor- tions implies that wages or labor costs may be different across sectors. We define the intersectoral wedges as relative wage between the two sectors by: Preferences The economy is populated by a representative household which derives utility from consuming the agricultural good ( ), industrial good ( ), and services good ( ). Preferences are summarized by a Stone–Geary utility function, which incorporates the impact of income growth on the secular decline in agriculture’s share and in- creases the services’ share of economic activity. The period utility function of each household is of the following form. where is non-homothetic term, is interpreted as the subsistence con- Appendices sumption in agriculture while is interpreted as a constant level of produc- tion of service goods at home. When income is low, fewer resources are allocated to the production of services, and when income increases, resources are reallo- cated to services. The share parameters are positive and sum up to one, and 181 is a parameter of elasticity between sectoral goods. With these specifications, the income elasticity of substitution is less than one in agriculture and more than one in services and the industry’s income elasticity of demand is close to one (See Kongsamut etc al. (2001), Duarte and Restuccia (2010), Herrendorf etc al. (2013). Households inelastically supply one unit of time so that the aggregate labor supply is inelastic and equal to unity. They obtain income from renting land and labor to firms and from profits generated by production. This income is devoted to con- suming, investing, or paying the cost of moving to another sector. Therefore, the budget constraint of the household is where and are the profit of industry and services goods producers. The representative household chooses the sectoral composition of units of labor and land that they supply every period to maximize the utility function subject to the budget constraint above. By using a standard procedure, we obtain the fol- lowing necessary conditions for optimality: Competitive equilibrium Given technology and preference parameters and set of intersectoral wedges , a competitive equilibrium in this small open economy can be summarized Growth, trade, and transformation as a collection of prices allocation and profits for firms in industry and services, such that: (i) given prices, intersec- toral wedges and profits, , solve the utility maximization problem of the representative household; (ii) given prices and intersectoral wedges , solve the profit maximization problem of firms in each sector; (iii) all markets clear, i.e., Labor market Good market 182 Trade is balanced where and denote the agricultural and nonagricultural net exports respec- tively. Data This subsection briefly summarizes all data used as inputs to calibrate and simulate the models. Data used in this paper come from five sources. Employment, real and nominal value-added use in this paper come from Economic Transformation Database, covering the period 1990-2018 for Uganda. We aggregate these 10 sectors’ data into three sectors: agriculture, industry, and services, using the International Standard Industrial Classification of All Economic Activities (ISIC). Agriculture corresponds to Agriculture, Industry includes mining, manufactur- ing, utilities, and construction. Services correspond to the rest and include Trade, Transport, Business services, financial services, Real estate, Government services, and Other services. For land stock, we use data from the Food and Agriculture Organization of the Unit- ed Nations which provides for several countries the total farmland includes areas devoted to crop cultivation, animal breeding, and logging.1 Trade data come from UN COMTRADE Database in which provides data on trade of all commodities of Standard International Trade Classification (SITC) Revision 2. We aggregate these data in our two broad sectors as follows. Agriculture corresponds to “ 0: Food and live animals chiefly for food”, “ 1: Beverages and tobacco”, “2: Crude materials, inedible, except fuels”, “ 4: Animal and vegetable oils, fats, and waxes” Minus “27: Crude fertilizer and crude minerals”, “28: Metalliferous ores and metal scrap”. The industry sector corresponds to non-agriculture commodities minus “9: 1 Food and Agriculture Organization of the United Nations provides for each country the total of Appendices areas under temporary and permanent crops, under temporary meadows and pastures, land with temporary fallows, Land under permanent meadows and pastures, Land under protective cover. This category includes tilled and fallow land and naturally grown permanent meadows and pas- tures used for grazing, animal feeding, or agricultural purpose. Scattered land under farm buildings, yards and their annexes, and permanently uncultivated land, such as uncultivated patches, banks, footpaths, ditches, headlands, and shoulders are traditionally included. 183 Commodities and transactions not classified elsewhere in the SITC”2. Exogenous prices of tradable sector goods come directly from Penn World Table. Agriculture price corresponds to the price level of food and beverages exports and Industry price correspond to the price level of industrial supplies exports. Calibration Model parametrization consists of three steps: the preference parameters, the pro- duction function parameters, and exogenous variables. We do not have detailed data on consumption expenditure in Uganda. We employ time-series data on South Africa aggregate consumption expenditure, sectoral consumption expenditure shares, and sectoral prices,3 which is the only country in Africa whose data are produced by the OECD, to estimate preference parameters by minimizing the sum of squared deviations between the actual sectoral expendi- ture shares and the model-implied sectoral expenditure share given the observed sectoral prices and aggregate consumption expenditure. Since the data for South Africa is available for only 12 years, we set the elasticity of substitution to 0.754 and estimate only the sectoral weights and the non-homo- theticity terms Specifically, we minimize: Growth, trade, and transformation where is the total expenditure at period t. As in Herrendorf and al. (2013), we transform the constrained parameters into un- constrained parameters as 2 Precisely Industry includes “27: Crude fertilizer and crude minerals”, “28: Metalliferous ores and metal scrap”, “3: Mineral fuels, lubricants, and related materials”, “5: Chemicals and related products”, “6: Manufactured goods classified chiefly by materials} “, “7: Machinery and transport equipment”, “8: Miscellaneous manufactured articles”. 3 The prices of sectoral final goods are obtained by dividing nominal consumption expenditure by reel consumption expenditure. 4 The value estimated Uy et al. (2013) using long time-series data on Korean. 184 Table B-1. Value of parameters Description Value Source Elasticity of substitution 0.75 Agriculture consumption weight 0.25 Industry consumption weight 0.31 Services consumption weight 0.44 Subsistence agriculture consumption 38.57 Home production services 21.60 Labor-income share 0.5 Agriculture TFP growth rate 3.11% Industry TFP growth rate 1.86% Services TFP growth rate 7.20% Land stock FAO data Price of agriculture and industry goods Penn World Table and we estimate the parameters and . The Table below shows the results of estimations. Noting that all coefficients estimated are significant at the 1 percent level. We now turn to the calibration of production parameters . Following Teignier (2018), we set labor-income share in agriculture , which implies that the land intensity is 0.50. These values are in the range of values used in the literature for other countries. 5 As presented above, we compute labor cost gap directly in data as follows. Our model contains exogenous variables: per capita stock of land , sectoral pro- ductivities , and as well as the agriculture and industry prices and Appendices 5 Teignier (2018) sets in agriculture sector capital-income share at 0.10, labor-income share at 0.50, and land intensity at 0.40 for both South Korea and Great Britain. Valentinyi and Herrendorf (2008) find that the labor income share in agriculture for the United States is 0.46 and the land income share is 0.18, while Hayashi and Prescott (2008) use 0.545 and 0.1932 for the case of Japan in the period 1885–1940, and Stokey (2001) uses 0.387 and 0.45 for the case of Great Britain between 1780 and 1850. 185 respectively. We construct land stock per capita by dividing total land stock provided by the Food and Agriculture Organization of the United Nations and total workers from Economic Transformation Database. For productivities, we calibrate initials values to match initial sectoral employment share. Then, we compute the growth rate of TFP in sector , by dividing the real value-added sector of sec- tor by the number of workers in that sector, and then we calculated the average annual growth rate of the obtained variable. Finally, we take and directly in the Penn World Table. References Beaman, L., Onder, H. and Onder, S., 2022. When do refugees return home? Evidence from Syrian displacement in Mashreq.  Journal of Development Economics,  155, p.102802. Duarte, M. and Restuccia, D., 2010. The role of the structural transformation in ag- gregate productivity. The Quarterly Journal of Economics, 125(1), pp.129-173. Hayashi, F. and Prescott, E.C., 2008. The depressing effect of agricultural institutions on the prewar Japanese economy. Journal of political Economy, 116(4), pp.573-632. Kongsamut, P., Rebelo, S. and Xie, D., 2001. Beyond balanced growth. The Review of Economic Studies, 68(4), pp.869-882. Herrendorf, B., Rogerson, R. and Valentinyi, A., 2013. Two perspectives on preferenc- es and structural transformation. American Economic Review, 103(7), pp.2752-89. Stokey, N.L., 2001, December. A quantitative model of the British industrial revolu- tion, 1780–1850. In Carnegie-Rochester conference series on public policy (Vol. 55, No. 1, pp. 55-109). North-Holland. Teignier, M., 2018. The role of trade in structural transformation. Journal of Develop- ment Economics, 130, pp.45-65. Growth, trade, and transformation Uy, T., Yi, K.M. and Zhang, J., 2013. Structural change in an open economy. Journal of Monetary Economics, 60(6), pp.667-682. Valentinyi, A. and Herrendorf, B., 2008. Measuring factor income shares at the sec- toral level. Review of Economic Dynamics, 11(4), pp.820-835. 186 Endnotes 1 Measured by the headcount ratio at USD 1.90 a day, using 2011 PPP series. 2 Dennis et al. (2016) shows that a large share of the observed within-industry growth in labor productivity represented allocative efficiency gains from the correction of intra-industry interfirm misallocation of labor. 3 For the purposes of this study, a migrant household is defined as one whose head of household lived in a different district in Uganda for more than 6 months in the 5 years preceding the interview. This is admittedly an imperfect measure since it captures only those who had a migration experience since 2012 and focuses on only household heads’ migration experience. For brevity, migrant heads are referred to as migrants throughout the text. 4 Product is defined at the 6-digit level of the Harmonized System. 5 This reasoning draws on recent trade models with heterogeneity across firms. See Bernard, Jensen, Redding, and Schott (2011) on the empirical implications of such trade models in NBER Working Paper No. 17627, “The Empirics of Firm Heterogeneity and International Trade.” 6 See: https://www.eac.int/documents/category/eac-common-external-tariff. 7 See for example: https://allafrica.com/stories/202001190208.html. 8 For example, according to the UCDA, in February 2021 Uganda exported coffee worth USD 50.55 million, of which about USD 40.9 million was Robusta and USD 9.6 million wasArabica. 9 The average price was calculated using Uganda customs transactions firm-lev- el data. 10 According to the UCDA monthly reports, Arabica makes up 10–30 percent of Uganda’s coffee exports, depending on the month of export. The average ex- port price for Arabica is about 70–80 percent higher than the one for Robusta. Many firms export both Robusta and Arabica, but some focus only on Robusta. 11 To find the GDP share of fiscal revenues from oil in the future, IMF-WEO Oc- tober 2021 projections for Ugandan GDP and fiscal revenue to GDP ratios are extrapolated beyond 2026 using an identical trend growth for the periods 2021–-2026 and 2026–-2030. In addition, a 2 percent annual USD inflation is assumed throughout. Actual fiscal revenues from hydrocarbons are calculated using a detailed fiscal model that captures contractual details of oil production, refining, and transportation. Overview 187 Uganda has made impressive progress since the 1990s. With policy reforms unlocking productivity growth, and with favorable external conditions in the 2000s, the country’s per capita income increased 3-fold and life expectancy by 50 percent in the last three decades. However, this growth performance failed to foster rapid structural transformation and eventually stalled in 2010s, leaving about two thirds of the country’s labor force in agriculture, where they produce less than a quarter of the national income. Without a rapid transformation, and with a high population growth rate, the country also suffered from one of the fastest deforestation rates in the world. Going forward, Uganda faces two major challenges. First, generating enough jobs for its labor force, which will grow 2.5-fold in the next three decades. Second, halting the depletion of woodlands and bushlands, which have been converted to small-scale farms under economic pressure. To support Uganda’s search for higher economic growth and more rapid transformation, this report provides a detailed analysis of the drivers of growth and structural transformation in Uganda in recent years, the role of international trade in shaping the country’s economic dynamics, and future risks and opportunities faced by policy makers in Uganda.