91887 THE WORLD BANK GROUP Country Development Diagnostics Post-2015 January 2015 Country Development Diagnostics Post-2015 Susanna Gable Hans Lofgren Israel Osorio-Rodarte Development Prospects Group World Bank January 2015 We thank Mahmoud Mohieldin, Marilou Uy, and Jos Verbeek for overall guidance in this project, and Hans Tim- mer, Elena Ianchovichina, and Punam Chuhan for their valuable suggestions as peer reviewers. We are also grateful for comments from Lily Chu, Anton Dobronogov, Eric Feyen, Marcelo Giugale, Gloria Grandolini, Raj Nallari, Alberto Portugal, Sajjad Shah, Marco Scuriatti, Chris Thomas, and Debrework Zewdie. The findings, interpreta- tions, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Contents Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Step 1:  Benchmarking SDG Progress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Step 2:  SDG Business-as-Usual Projections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Step 3:  Benchmarking Determinants and Identifying Spending Priorities . . . . . . . . . . . . . . . 11 Current Performance of Determinants. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Identifying Spending Priorities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Step 4:  Identifying Fiscal Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 List of Figures Uganda – Primary School Net Enrollment/GNI per capita; Figure 1:  Primary School Completion/GNI per capita. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Figure 2: Uganda – Secondary School Gross Enrollment/GNI per capita; Secondary School Completion/GNI per capita. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Figure 3: Uganda – Historical Data and Projections for Real GDP per capita. . . . . . . . . . . . . . . . . . . . . . . 7 Figure 4: Uganda – Expenditure per Primary Student/GNI per capita; Expenditure per Secondary Student/GNI per capita . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Figure 5: Uganda – Primary Pupil-Teacher Ratio/GNI per capita; Secondary Pupil-Teacher Ratio, secondary/GNI per capita. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Figure 6: Uganda – Tax Revenues 1990–2011 (% of GDP); Tax Revenues (% of GDP)/GNI per capita . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Figure 7: Uganda – ODA (% of GNI)/GNI per capita; ODA (per capita)/GNI per capita. . . . . . . . . . . . . . 20 iv Country Development Diagnostics Post-2015 List of Tables Table 1: Uganda – Historical and Projected Growth from Various Sources . . . . . . . . . . . . . . . . . . . . . . . . 8 Table 2: Uganda – SDG Projections for 2030. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Table 3: Uganda – Policy-Relevant SDG Determinants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Table 4:  Government Fiscal Space – Recent Indicators and Future Directions of Change . . . . . . . . . . . . 17 List of Boxes Box 1:  Using GNI per capita for SDG Benchmarking. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Box 2:  Projecting GDP and GNI. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Box 3: SDG Business-as-Usual Projections for 2030. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Box 4:  Measures of Government Effectiveness. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Executive Summary W ith the 2015 deadline for the current Millen- relative to other countries, given GNI per capita, nium Development Goals (MDGs) drawing a variable that is highly correlated with most de- near, the global community is shaping a new velopment indicators, including SDGs and their set of international development goals for the longer determinants. Accordingly, in this analysis, GNI term. The process has involved consultations led by the per capita is treated as a summary indicator of UN Open Working Group guided by the 2013 report, the capacity of a country to achieve outcomes, for “A New Global Partnership” of the UN High-level both SDGs and their determinants. Panel. The work so far indicates that the post-2015 de- Step Two projects the country’s business-as-usual velopment agenda will encompass goals for social, eco- (BAU) GNI per capita and values for SDGs by nomic, and environmental sustainability with broader 2030. coverage than the current MDGs.1 This paper refers Step Three turns to the determinants of SDG out- to these post-2015 development goals as Sustainable comes—many of these are related to policies, in- Development Goals, or SDGs. cluding those that affect the efficiency and levels This paper presents the Post-2015 Country De- of public spending—pointing to ways of achiev- velopment Diagnostics, a framework developed by the ing outcomes that are more ambitious than those World Bank Group to assess the implications of im- of the BAU projections. Policies may influence plementing the post-2015 global development agenda an SDG directly—health services may promote at the country level. The framework has been applied better outcomes for health SDGs—or indirectly, to a pilot case study on Uganda, and some of the re- such as when measures that promote growth in sults of this study are highlighted here for illustrative household incomes per capita or increased access purposes. The World Bank Group has also developed to sanitation have an indirect positive influence on a multi-country database that provides a starting point health SDGs. In this step, therefore, we bench- for similar diagnostics in other countries. Subject to mark Uganda’s current levels of SDG determi- data availability, the framework may be used to analyze likely progress in SDGs and their determinants and to discuss policy and financing options to accelerate 1  According to the HLP, the overall goals are to: end poverty; their progress. This work has been shared with the In- empower girls and women and achieve gender equality; provide tergovernmental Committee of Experts on Sustainable quality education and lifelong learning; ensure healthy lives; Development Financing. ensure food security and good nutrition; achieve universal ac- cess to water and sanitation; secure sustainable energy; create The purpose of this paper is to demonstrate the jobs, sustainable livelihoods and equitable growth; manage nat- application of this framework, drawing on the pilot ural resource assets sustainably; ensure good governance and study of Uganda. The framework consists of four steps: effective institutions; ensure stable and peaceful societies; and create a global enabling environment and catalyze long-term Step One benchmarks the current level of prog- finance. The Open Working Group is currently discussing a set ress for each SDG for the country being analyzed of 17 development goals. vi Country Development Diagnostics Post-2015 nants in relation to its GNI per capita and discuss GDP is higher than expected (9.5 compared to an ex- potential changes in policies and spending in pri- pected 5.9 percent of GDP), above the expected level ority areas. for private but below for public spending. At the same Step Four discusses ways to expand fiscal space for time, however, the level of dollar spending per capita priority SDG spending, including additional do- is well below the recommended minimum for achiev- mestic or foreign financing (including taxes and ing even current health MDGs, and even if projected foreign aid) and efficiency gains (achieved by re- growth rates are maintained, Uganda is not expected allocating spending from areas of lower priority to achieve this minimum spending level before 2020. and/or reducing spending in areas with technical On the other hand, the ability of the health sector to efficiency gains without any service reduction). absorb additional spending while maintaining efficien- This analysis is applied to the specific case of cy in the short to medium term is severely constrained Uganda: how and to what extent may it be able to by a lack of qualified manpower while waste is sub- create room for increased public spending in pri- stantial. Accordingly, policymakers need to assess alter- ority areas? Would such adjustments be advisable? native ways of making progress: What can be done to What trade-offs may be involved? increase absorptive capacity in the public health sector? Could partnering with the private sector enhance ab- Empirically, the results for Uganda indicate that sorptive capacity? BAU performance would fall substantially short of the In addition to investment in education and ambitious goals of the evolving global SDG agenda. health, infrastructure development—in water, sanita- However, the country could make stronger progress by tion, roads, electricity, communications and internet 2030 in key areas, including poverty reduction, edu- technology—is a major SDG-related spending area cation, health, and infrastructure development. This for a low-income country like Uganda. Despite infra- would depend on policy changes that raise per capita structure spending over the last several years averaging income growth, generate greater fiscal space for need- over 10 percent of GDP, or US$1 billion per year, the ed expenditures, and enhance the efficiency of public country continues to lag in electricity provision, while spending. Improved creditworthiness would further shortcomings remain severe in sanitation, water and increase Uganda’s capacity to borrow from interna- roads (especially secondary roads). Given high costs tional financial markets. and public financing limitations, could part of the In primary education, net enrollment is higher needs gap be met via mobilization of private invest- and completion lower than expected given Uganda’s ments, leveraged by the allocation of additional fiscal GNI per capita, findings that may be explained by space to infrastructure? low spending per student and/or low efficiency. At The fiscal space analysis suggests that Uganda the secondary level, expenditures per student are as will be able to increase fiscal space for priority spend- expected. Given the fact that the completion rate is ing during the period up to 2030. This assessment is as expected while the enrollment rate is below expec- highly dependent on expected but uncertain oil reve- tations, this suggests that the system—considering its nues. Among other fiscal space sources, foreign aid (as level of spending—performs relatively well in terms of percent of GDP) is expected to decline. Increases in bringing enrolled students to completion. However, as spending on human development and infrastructure Uganda in the future meets the challenge of increasing of this magnitude (or more) could easily be advocated the number of entrants that proceed from primary, the considering the size of unmet needs. However, the gov- demands for public spending on secondary education ernment faces the challenge of increasing spending at will increase. the same time as it maintains and preferably improves In health, Uganda’s key indicators for under-five government efficiency, translating additional spending and maternal mortality rates are as expected. Total into services that significantly contribute to more rapid (public and private) health spending as a percent of progress on the SDG agenda. Executive Summary vii The Post-2015 Country Development Diagnostics moderate cost, given that the multi-country database framework, used in this paper, and the accompanying is readily accessible and can be used for cross-country database offer analysts in developing countries and the analysis and benchmarking. However, it is important broader international community useful starting points to note that, in order to permit more specific policy for assessing SDG targets and related policy and financ- conclusions, the cross-country diagnostics that the ing priorities in virtually any low- or middle-income framework offers should be linked to more detailed country. Such diagnostics can be conducted at a fairly country-specific studies at country and sector levels. Introduction2 W ith the 2015 deadline for the current Millen- key parts of the global SDG agenda in their countries. nium Development Goals (MDGs) drawing The Post-2015 Country Development Diagnostics near, the global community is shaping a new framework is designed for application in countries set of international development goals for the longer with a wide variety of characteristics, including dif- term. The process involved consultations led by the ferences in initial conditions and access to financing, UN Open Working Group guided by the 2013 report, and provides a starting point for more detailed analy- “A New Global Partnership” of the UN High-level sis. It benchmarks a country’s achievements, provides Panel (HLP). The work so far indicates that the post- projections up to 2030, and helps policy makers ask 2015 development agenda will encompass social, eco- questions about SDG targets and policy options. It nomic, and environmental sustainability goals with covers the following SDG areas: (i) poverty reduction broader coverage than the current MDGs.3 This paper and shared prosperity, (ii) infrastructure (water, sanita- refers to these post-2015 development goals as Sustain- tion, electricity, roads, and information and communi- able Development Goals, or SDGs. cations technology, or ICT), access to (iii) education, In setting the post-2015 SDGs, the global commu- (iv) health, and (v) climate change. Several indicators nity will need to take cognizance of various challenges are used to measure progress of goals in each of these to implementation and financing at the country level. areas, limited by what is available in cross-country data This will necessitate integrated discussion of the devel- opment goals and the associated financing framework. Financing in particular will have to be structured in a 2  This paper was prepared as part of collaborative work on the way that taps into and leverages a variety of financing post-2015 global agenda, involving the Development Pros- pects Group and the Office of the World Bank Group Corpo- sources beyond aid, and the policy framework will have rate Secretary and President’s Special Envoy, led by Mahmoud to ensure private sector efficiency and improved public Mohieldin. sector productivity.4 The ability to leverage diverse fi- 3   According to the HLP the overall goals are to: end poverty; nancing will differ from country to country, typically empower girls and women and achieve gender equality; provide with less ability for low-income and/or conflict-affected quality education and lifelong learning; ensure healthy lives; countries.5 Given the vastly different capabilities, histo- ensure food security and good nutrition; achieve universal ac- ries, starting points and circumstances of the countries cess to water and sanitation; secure sustainable energy; create concerned, the HLP has suggested that each govern- jobs, sustainable livelihoods and equitable growth; manage nat- ment be allowed to choose the appropriate level of am- ural resource assets sustainably; ensure good governance and effective institutions; ensure stable and peaceful societies; and bition for each target, since every country cannot be create a global enabling environment and catalyze long-term expected to reach the same absolute target. finance. The Open Working Group is currently discussing a set Against this background, the World Bank Group of 17 development goals. has developed a framework, with Uganda as the pi- 4   World Bank Group, “Financing for Development Post- lot study, to provide an initial understanding of the 2015”, October 2013. challenges policymakers will face in implementing 5   Ibid. 2 Country Development Diagnostics Post-2015 sets. Given that the aim of the current paper is to con- availability, the database covers key aspects of the post- cisely present the analytical framework and selected 2015 agenda that can be meaningfully analyzed in a results for Uganda, it is more selective in terms of both framework of the type developed here. An SDG analy- SDGs and the indicators used.6 sis for any given country is expected to make selective More concretely, the framework benchmarks use of the data. The database will become part of the country performance in SDGs, policies, and other public domain, making it possible for analysts to draw determinants (factors that influence SDGs). It makes on it in analyses of the SDG agenda for any low- or projections for SDGs to the year 2030, analyzes spend- middle-income country. ing adjustments in priority areas, and discusses sources The purpose of this paper is to illustrate our frame- of fiscal space. Cross-country regressions of SDGs and work, drawing on the more detailed case study applica- their determinants on GNI per capita play a central tion to Uganda. The analysis is made up of four steps. role in the analysis. The advantages and disadvantages In each step, we explain the methodology and present of (typically more elaborate) cross-country regressions an excerpt from the more comprehensive Uganda pa- have been discussed extensively.7 Our use of this tool per, with a focus on education. Throughout the paper, is very simple and transparent, drawing on the obser- we emphasize how the framework can be used as a tool vation that many development indicators, including to identify priority policy areas and fiscal alternatives SDGs and their determinants, are highly correlated to progress on the post-2015 agenda at the country with GNI per capita. For such indicators, we view GNI level. The paper is structured as follows: per capita as a summary indicator of the basic capacity of a country to bring about outcomes, both for SDGs Step One benchmarks Uganda’s current SDG out- and their determinants. This does not translate into an comes against those of other countries, given the assumption of GNI being a direct determinant of out- levels of GNI per capita. comes—it is merely a benchmark and starting point Step Two projects BAU levels for the SDGs in year for discussion about how a country performs relative 2030, drawing on GNI per capita projections. to others at its income level. It is noteworthy also that Step Three tries to assess how to achieve more certain indicators, such as the income share of the bot- ambitious targets than those suggested by the tom 40 percent (the key measure of shared prosperity) BAU projections. To this end, it benchmarks the are largely unrelated to GNI per capita. This points to the fact that purposeful measures are crucial to change for many development outcomes: in this case, growth 6  We will strive to contrast SDGs according to their closeness of linkage to GNI per capita, and in terms of whether Uganda does not, in any regular fashion, directly or indirectly, is over- or under-performing. stimulate processes that bring forth shared prosperity. 7   Among the potential advantages is the ability to control for The questions that the framework helps to address various alternative determinants, and—when robust results are include: For any country, what would be a set of feasi- found—to generalize results beyond the country-specific con- ble development targets for 2030 if the country were text. However, as noted by many (for example, ADB 2006), to develop with business-as-usual (BAU) assumptions? cross-country regressions are often unable, for various interre- What policy areas should the country’s government lated reasons, to successfully address the role of different de- consider in order to accelerate progress? How could it terminants, severely limiting the usefulness of these results to create the fiscal space needed to achieve more ambi- policymakers. More specifically, the regressions tend to suffer from a lack of robustness to different specifications; difficulty in tious development outcomes? assessing the direction of causality between different indicators Underpinning the analysis is a database that cov- (causality may often go in both directions); high correlations ers all low- and middle-income countries, designed to and complex interactions between determinants; variable re- include available indicators relevant to the post-2015 lationships (across time and space); and imperfect indicators agenda, including SDGs, their determinants, and in- (for example, spending on human development is an imperfect dicators related to financing options. Subject to data indicator of real services in human development). Introduction 3 current levels of the determinants of the various (through additional financing and government SDGs for Uganda and compares them to those efficiency gains), again by looking at Uganda’s of other countries in order to assess spending pri- current situation compared to what is expected orities. Determinants for which Uganda is sig- for a typical country at its GNI per capita. These nificantly lagging behind other countries with a findings for fiscal space are then compared with similar level of GNI per capita are singled out for the assessment of spending priorities identified in special consideration. Step Three. Step Four addresses challenges related to expand- The report concludes with a summary of findings ing fiscal space. In this context, the analysis con- for Uganda and a discussion of how this frame- siders Uganda’s options for creating fiscal space work may be applied to a variety of countries. Step Benchmarking SDG Progress 1 I n this step, cross-country regressions are used to as- primary completion rates are significantly lower than sess the performance of the case study country in expected. Figure 2 shows similar information for sec- terms of SDGs, relative to its level of GNI per capi- ondary education in Uganda: gross enrollment rates ta. (Box 1 provides the rationale.) are significantly lower than expected but completion Here we will exemplify the SDG benchmarking rates are as expected.9 approach analysis of primary and secondary education in Uganda.8 Figure 1 shows two scatter plots with each observation representing a country’s position relative to its GNI per capita and the SDG, the latter repre- 8  In addition, the analysis may also review the evolution of the sented by primary school enrollment on the left and SDG in recent decades as part of the assessment of initial coun- primary completion on the right. The fitted, straight try SDG performance. In addition to benchmarking country performance against what is expected, it may also be relevant line represents expected school enrollment or com- to benchmark against top performance within countries that pletion levels for countries at different levels of GNI in other important respects remain similar to the case-study per capita. Countries outside the shaded area are sig- country. nificantly over- or under-performing relative to their 9   Uganda’s secondary completion rate is highly uncertain. GNI per capita. Hence, for Uganda, net enrollment Drawing on population, enrollment, and repetition data in Ed- in primary is significantly higher than expected, while Stats, a rate of 9.4 percent was calculated for 2011. Box 1: Using GNI per capita for SDG Benchmarking GNI per capita plays a central role in the analysis. Its level is highly correlated with SDG indicators for several reasons, perhaps most impor- tantly due to the fact that GNI per capita is highly correlated with determinants of SDGs, including (i) per capita household incomes, parts of which is spent on items that contribute to SDGs (for example, on health, education, and electricity); and (ii) tax revenue, which contributes to the fiscal space for government spending in areas that, directly or indirectly, contribute to SDGs (most importantly, government services and infrastructure). Causality may also go in the opposite direction: the levels for different SDGs (for example, those related to health and education) may influence GNI per capita. Cross-country, constant-elasticity regressions are first used to benchmark current SDG outcomes—i.e., to assess whether a country is over- or under-performing for an SDG relative to its GNI per capita.a Hence, for individual countries, deviations from predicted SDG values may be viewed as an indication of how well a country does relative to its capacity to achieve outcomes and provide inputs (determinants). Instead of GDP per capita (a production measure), GNI per capita, an income measure, is used since it conceptually is more closely related to a country’s capacity to achieve SDGs. a   These simplified regressions are useful for current purposes (benchmarking and projections). However, they do not claim to sort out interactions between dif- ferent indicators, a difficult task given high degrees of correlation, lagged effects, complex time- and space-specific relationships, and data limitations. Tests of al- ternative functions indicated that the simplicity of the constant-elasticity function dominated any gains in fit for some SDG indicators with alternative functions. 6 Country Development Diagnostics Post-2015 Figure 1:  Uganda – Primary School Net Enrollment/GNI per capita (left); Primary School Completion/GNI per capita (right) 5.0 5.0 Determinant, logs Ln Primary completion rate, Ln School enrollment, primary (% net) total (% of relevant age group) 4.5 UGA 4.5 Determinant, logs 4.0 4.0 UGA 3.5 3.5 150 400 1100 2980 8100 150 400 1100 2980 8100 Ln GNI per capita (constant 2005 US$) Ln GNI per capita (constant 2005 US$) Measure of Income Per Capita, Log−scale Measure of Income Per Capita, Log−scale Ln(DET) = 3.924*** + .073*** Ln(INC) ; R2: .198 Ln(DET) = 3.315*** + .153*** Ln(INC) ; R2: .421 Sources: WDI, EdStats. Figure 2:  Uganda – Secondary School Gross Enrollment/GNI per capita (left); Secondary School Completion/GNI per capita (right) 5 6 Ln School enrollment, secondary (% gross) Ln DHS: Secondary completion rate 4 Determinant, logs Determinant, logs 4 2 UGA 3 0 UGA –2 2 150 400 1100 2980 8100 150 400 1100 2980 8100 Ln GNI per capita (constant 2005 US$) Ln GNI per capita (constant 2005 US$) Measure of Income Per Capita, Log−scale Measure of Income Per Capita, Log−scale Ln(DET) = 2*** + .297*** Ln(INC) ; R2: .55 Ln(DET) = −.348 + .48 Ln(INC) ; R2: .072 Sources: WDI, EdStats. Step SDG Business-as-Usual Projections 2 I f the relationship between GNI per capita and an Figure 3:  Uganda – Historical Data and SDG is considered tight enough, then the GNI Projections for Real GDP per capita data for the country in question is used, not only to (2011=100) benchmark the initial SDG outcome but also to proj- ect business-as-usual SDG outcomes for 2030. For 250 this, we need projections of GNI per capita. Box 2 discusses alternative sources for GDP and 200 GNI projections, which are available for most coun- 150 tries. Figure 3 uses three of these sources to show Ugan- da’s projected (indexed) levels of GDP per capita up to 100 2030 (and, for comparison, the historical development since 1990), while Table 1 presents growth rates. We 50 opted for the CEPII projection, which for Uganda has 0 a growth rate for GNI per capita of 4.0 percent per 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030 year (at constant 2005 US dollars), translating to an increase from US$378 in 2011 to US$817 in 2030 IASA OECD CEPII Historic (both at constant 2005 prices), a level similar to the current levels of countries such as Vietnam, India, and Sources: WDI, IIASA, OECD, and CEPII. Box 2: Projecting GDP and GNI Aggregate growth projections covering most countries are produced by various international organizations, including the World Bank, IMF, CEPII, OECD, and IIASA, but also by most governments and other sources, such as Hausmann et al. (2011). From the projections, it is difficult to determine which source is most reliable. Moreover, given the fact that available sources only project GDP while this paper uses GNI data, we have to assume, for most countries quite reasonably, that projected GNI growth will not deviate substantially from projected GDP growth (both expressed in constant 2005 US dollars).a In any country case study, it is good practice to compare different projections and, if necessary, refine what is available. a   As indicated by the names of the terms, GDP is primarily a measure of production while GNI is an income measure, more specifically GNI = GDP plus net receipts from abroad of primary income (compensation of employees and property income). For most countries, the two measures are highly correlated; among low- and middle-income countries, they tend to diverge most strongly in countries where (net) FDI over time has represented a substantial share of total private investment, often in natural resource sectors, generating substantial profit remittances to the foreign investors. If additional information is available on how future GNI and GDP growth may differ for a country, then such information should be reflected in the GNI projections. 8 Country Development Diagnostics Post-2015 Table 1:  Uganda – Historical and Projected Growth from Various Sources Average annual Time Indicator Source growth (%) period (real values) Comment WDI 3.3 1990–2012 GDP per capita Data used in Figure 3 for period up to 2012 WDI 3.2 1990–2011 GNI per capita GDP per capita growth for 1990-2011 was 3.5 percent CEPII 4.0 2013–2030 GDP per capita OECD 3.8 2013–2030 GDP per capita IIASA 2.5 2013–2030 GDP per capita IMF (2013) 3.7 2013–2030 GDP per capita Including oil revenues, adjusted for population growth Hausmann et al. (2014) 3.3 2009–2020 GDP per capita Based on the Economic Complexity Index Republic of Uganda 2014, 5.6 2014–2040 GDP per capita Calculation based on data for GDP growth and popula- pp. 27, 30, 53 tion in Uganda’s Vision 2040 GNI per capita of US$817. As explained under Step 1, Box 3:  SDG Business-as-Usual Projections for Uganda is currently over-performing in its primary 2030 school net enrollment rate (indicated by green text in Table 2); however, the cross-country relationship is not If the fit between GNI per capita and an SDG indicator is tight enough to make a relevant BAU projection for reasonably tight (which tends to be the case), the results of 2030. For the primary school completion rate, Uganda a cross-country regression permits us to compute projected is under-performing (indicated by red text). The pro- business-as-usual 2030 values. A tight or moderately tight jected BAU value in 2030 is 66.1 percent, an increase relationship refers to a significant GNI per capita variable and a good enough explanatory power of the regression (“tight” due mainly to GNI per capita growth but influenced R2 > 0.3, “moderately tight” 0.1 < R2 < 0.3). also by the convergence effect. Substantial progress is recorded for other indicators, but without realizing global ambitions: for example, the extreme poverty rate declines very strongly. Senegal.10 Considering the range of alternative projec- tions, an annual per capita growth rate of 4 percent seems realistic, if perhaps erring on the moderately op- timistic side. The levels of selected SDGs are projected to 2030. These BAU projections reflect what can be expected 10   We chose the projections of CEPII due to a combination given a country’s initial conditions, projected growth of factors, including a transparent model structure, clear docu- in GNI per capita, typical rates of progress according mentation, and comprehensive country coverage. 11   Given that (i) SDGs have extreme values (such as 100 per- to cross-country patterns, and gradual convergence to close gaps between observed and expected values.11 cent for improved water access) and (ii) the current SDG level never is exactly as expected relative to GNI per capita, it is nec- For any SDG, projections are presented only if the fit essary to incorporate convergence toward the expected value between GNI per capita and the SDG is considered into the projections. It is here assumed that such convergence sufficiently tight (Box 3). is gradual. For example, for a country that over-performs in Table 2 presents recent values and BAU projec- water access, as GNI per capita increases the extent of over-per- tions to 2030 for Uganda for a set of SDG indicators, formance gradually declines, so that when the expected value is including those shown in Figures 1 and 2, using a 2030 100, over-performance has reached zero. SDG Business-as-Usual Projections 9 Table 2: Uganda – SDG Projections for 2030 SDG Recent value BAU projection for 2030 Poverty rate at $1.25 a day (PPP) (% of population 38.0 11.5 Malnutrition (weight for age: % of children under 5) 14.1 8.8 Income share, bottom 40% (% of total income) 15.5 — GINI index 44.3 — Access to improved sanitation (% of population) 33.9 44.8 Access to improved water (% of population) 74.8 80.7 Access to electricity (% of population) 14.6 31.0 Road density (km road per 100 sq. km of land area) 32.2 35.8 Internet use (% of population) 14.7 — Mobile cellular subscriptions (% of population) 45.0 — Net enrollment, preprimary (%) 13.6 20.4 Net enrollment, primary (%) 90.9 — Primary completion rate (%) 53.1 66.1 Gross enrollment, secondary (%) 27.6 41.6 Secondary completion rate (%) 9.4 — Maternal mortality (modeled estimate, per 100,000 live births) 310.0 146.3 Under 5 mortality (per 1,000 live births) 68.9 42.7 Prevalence of HIV total (% of population ages 15-49) 7.2 — Malaria reported 7.3 1.3 Prevalence of tuberculosis 175 109 CO2 emissions per capita 0.11 0.39 Note: Green = Currently significantly over-performing; Red = Currently significantly under-performing; Black = Performing as expected; No projection = Too loose relationship with GNI per capita. Whether a specific deviation (positive or negative) reflects a stronger or weaker performance varies across indicators. For example, a positive deviation reflects weaker performance for poverty but stronger performance for water access. The terms over-performance and under-performance are used normatively; for example, with regards to the maternal mortality rate, a lower-than-expected rate is reflected as over-performance. Benchmarking Determinants and Step Identifying Spending Priorities 3 Current Performance of Determinants covers—SDGs related to access to infrastructure—the basic approach is simpler: deviations are viewed mainly In Step 3, we regress SDG determinants against GNI as indicating insufficient levels of efficient investments. per capita (in Step 1, we did this for SDG indicators; Shared prosperity is not addressed in a separate section cf. Box 1). The identification of determinants is guided but rather highlighted throughout. Wherever data al- by previous country and cross-country research, lim- lows, the results of the sample of the bottom 40 per- ited to indicators that are available in cross-country cent is presented, and indicators such as those related databases. We emphasize those determinants that may to education and health, access to finance, and second- be influenced by policy in the short to medium terms. ary road infrastructure are given special attention. It is The purpose is to assess the feasibility of policy changes important to note that some determinants influence that accelerate SDG progress and make more ambi- several SDGs, and that SDGs may be determinants of tious targets possible. Policies may influence SDGs in other SDGs.12 Of course, the fact that cross-country two ways, by: (i) raising the level of GNI per capita, analysis has shown that a certain determinant matters which in turn, through various channels, affects SDGs, for an outcome does not necessarily mean that it is and (ii) improving country SDG outcomes relative to important in a specific country setting; conversely, a what is expected given its GNI per capita. lack of evidence on the cross-country level does not To illustrate, if a country underperforms in both an necessarily mean a determinant is unimportant for a SDG and its more important determinants, then policy specific country. In order to arrive at more definitive actions may be both feasible and rewarding. Examples conclusions for a given country, it is necessary to assess include government spending in various areas and the and enrich the findings of our analysis, drawing on ad- related provision of inputs crucial to SDG progress. ditional country information. Such policies may have an influence directly (by having To demonstrate this step, we look at expenditures a direct bearing on specific services—e.g., health ser- per student at the primary and secondary school levels, vices targeted to reduce maternal mortality) and/or in- highlighting data for Uganda (Figure 4): at the prima- directly (by contributing to capacity-creating economic ry school level, spending is significantly lower than ex- growth). The discussion of major policy changes has pected while, at the secondary school level, it is within direct implications for costs and financing needs. the expected range. These findings may help to explain The determinants—in our cross-country database the enrollment-completion puzzle presented in Step 1: represented by over 200 indicators—may be classified Uganda’s lower than expected primary completion rate according to which of the following four areas they impact: economic growth, education, health, and For example, access to electricity is an SDG in its own right 12   climate change. In the fifth area that our approach and is likely also to influence both education and health SDGs. 12 Country Development Diagnostics Post-2015 Figure 4:  Uganda – Expenditure per Primary Student/GNI per capita (left); Expenditure per Secondary Student/GNI per capita (right) 4.0 5 per student, primary (% of GDP per capita) per student, secondary (% of GDP per capita) Determinant, logs Ln Expenditure Determinant, logs Ln Expenditure 3.5 4 2.5 3 UGA 3.0 UGA 2 2.0 1.5 1 150 400 1100 2980 8100 150 400 1100 2980 8100 Ln GNI per capita (constant 2005 US$) Ln GNI per capita (constant 2005 US$) Measure of Income Per Capita, Log−scale Measure of Income Per Capita, Log−scale Ln(DET) = 1.581*** + .138*** Ln(INC) ; R2: .09 Ln(DET) = 3.476*** −.082 Ln(INC) ; R2: .027 Sources: EdStats, World Bank. Figure 5:  Uganda – Primary Pupil-Teacher Ratio/GNI per capita (left); Secondary Pupil- Teacher Ratio, secondary/GNI per capita (right) 4.5 4.5 Ln Pupil−teacher ratio, secondary Ln Pupil−teacher ratio, primary 4.0 4.0 UGA Determinant, logs Determinant, logs 3.5 3.5 3.0 3.0 UGA 2.5 2.5 2.0 2.0 150 400 1100 2980 8100 150 400 1100 2980 8100 Ln GNI per capita (constant 2005 US$) Ln GNI per capita (constant 2005 US$) Measure of Income Per Capita, Log−scale Measure of Income Per Capita, Log−scale Ln(DET) = 5.561*** −.315*** Ln(INC) ; R2: .508 Ln(DET) = 4.751*** −.247*** Ln(INC) ; R2: .408 Sources: EdStats, World Bank. Benchmarking Determinants and Identifying Spending Priorities 13 Table 3: Uganda – Policy-Relevant SDG Determinants SDG Recent value Government consumption (% of GDP) 11.3 Public investment (% of GDP) 6.7 Logistic Performance Index 2.8 Ease of doing business rank 132.0 Public expenditure per student, primary (% of GDP per capita) 7.6 Public expenditure per student, secondary (% of GDP per capita) 20.7 Public expenditure per student, tertiary (% of GDP per capita) 45.6 Public expenditure, primary (% of GDP) 1.8 Public expenditure, secondary (% of GDP) 0.8 Public expenditure, tertiary (% of GDP) 0.4 Pupil-teacher ratio, primary 47.8 Pupil-teacher ratio, secondary 18.5 Public health expenditures (% of GDP) 2.5 Contraceptive use (% of population) 30.0 Physicians (per 1,000 people) 0.12 Skilled staff at birth (% of births) 57.4 Adolescent fertility rate (per 1,000 girls 15-19) 131.0 Fertility rate (births per woman, 15+ years of age) 6.1 Note: Green = Currently significantly over-performing; Red = Currently significantly under-performing; Black = Performing as expected; No projection = Too loose relationship with GNI per capita. The terms over-performance and under-performance are used normatively; for example, with regards to the maternal mortality rate, a lower-than-expected rate is referred to as over-performance. may be due to lower-than-expected expenditure per per capita (highly correlated with GNI per capita) and student and, as a related matter, a higher-than-expect- some of the other SDGs, including those related to in- ed pupil-teacher ratio. As for secondary schools, the frastructure—for example, access to safe water affect- expenditures per student are as expected but the pu- ing health indicators—may also matter. For those in pil-teacher ratio is lower than expected. The fact that red text, performance is significantly weaker than ex- the completion rate is as expected while the enrollment pected relative to Uganda’s GNI per capita, suggesting rate is below expectations (both rates are computed rel- that improvements in policies and outcomes in these ative to the total population in relevant age groups) areas may be most feasible. suggests that the system performs relatively well for its spending level in bringing enrolled students to com- Identifying Spending Priorities pletion. A more detailed investigation is needed to assess the room available for efficiency improvements. A cross-country perspective can shed useful light on Table 3 presents findings for a longer list of de- spending decisions, which are especially difficult when terminants, chosen from those that are directly poli- made in a situation such as Uganda’s, where large un- cy-relevant, not only for education but also for other met needs coexist with a constrained capacity to scale SDGs, giving a flavor of the type of determinants that up spending with retained efficiency. may be analyzed in a more detailed study. In addition At the aggregate level, Uganda’s spending-to-GDP to the determinants in the table, household incomes ratio is low relative to its GNI per capita for aggregate 14 Country Development Diagnostics Post-2015 public consumption (at 11.3 percent of GDP in 2011, In addition to education, health and infrastructure falling short by 2 percentage points) and, to a lesser ex- are two major SDG-related spending priorities for a tent, for aggregate public investment, suggesting that low-income country like Uganda. In health, key indi- some expansion would not put excessive pressures on cators such as under-five and maternal mortality rates, financing or institutional capacity. are at expected levels while total health spending is The above analysis focused mainly on primary and higher than expected (9.5 percent compared to an ex- secondary education. At the primary level, Uganda’s pected 5.9 percent of GDP). At a more disaggregated government spent around 7.6 percent of GDP per cap- level, public spending is roughly as expected (2.5 per- ita per student in 2011 (Table 3), which is less than cent of GDP) and private spending higher (7.0 percent the expected 11.0 percent. However, while spending of GDP compared to an expected level of 3.0 percent) per student as percent of GDP is less than expected, (Gable at al. 2014). In the short to medium runs, the its spending on primary education as percent of GDP ability of the public health sector to absorb addition- is as expected. The reason for this seeming contradic- al spending while maintaining efficiency is severely tion is that enrollment is relatively high, largely due constrained by a lack of qualified manpower, while to high rates of repetition and enrollment of students waste is substantial, estimated at 13 percent of spend- who are older than the expected age for their grade. If ing for 2005/2006 (Okwero at al. 2010, pp. 47, pp. repetition rates can be reduced and completion rates 65-68). Meanwhile, the level of spending on current increased—something that may require more spend- health MDGs is well below the recommended mini- ing per student—the GDP share for primary spend- mum—US$54 per capita at 2005 prices (Task Force ing required to offer services similar to those of other on Innovative International Financing for Health Sys- countries will eventually decline as students graduate tems 2009, p. 11; WHO 2010, pp. 36–37); if pro- from the primary level. All things considered, an ini- jected growth rates are achieved, Uganda’s total health tial jump in the GDP spending share to 2.5 percent of spending would not reach this level until about 2020. GDP (compared to the current 1.8 percent of GDP) In other words, further financing for increased health would raise spending to the expected level. Howev- services will be a high priority, especially if the govern- er, even though such increased spending would raise ment managed to overcome the manpower and oth- per-student resources to what is typical for countries er constraints to increased absorptive capacity in the at Uganda’s GNI per capita, it still remains far below health sector. what may be needed to offer a quality primary educa- tion.13 For secondary education, the enrollment rate and spending as percent of GDP are both lower than expected while completion rates (measured relative to 13   In 2011, at PPP in constant 2010 US dollars, average pub- the population in the relevant age cohorts) and spend- lic spending per primary student in low-income, middle-in- ing per student as percent of GDP are as expected. As come, and high-income countries was US$94, US$554, and Uganda in the future meets the challenge of increasing US$6,353, respectively (UNESCO 2014a, p. 383; UNESCO the number of entrants that proceed from primary, the 2014b, Table 11). 14   For Uganda and many other low-income countries, the ed- demands for public spending on secondary education will increase. As a result of expansion at lower levels, ucation quality gap and challenge is particularly strong at the primary level. This is because enrollment is higher at this level the demand for tertiary education will also increase, and spending per student tends to grow faster than GDP per albeit with a lag. In 2011, public spending on tertiary capita (raising the value for spending per student as percent of education was 0.4 percent of GDP, less than expected. GDP per capita), reflecting initial over-enrollment relative to Like primary education, keeping spending per student resources. At higher levels of education it is easier to manage as percent of GDP at expected levels may not be suffi- the challenge: enrollment is smaller while growth in spending cient to offer a quality education.14 per student tends to be slower than growth in GDP per capita. Benchmarking Determinants and Identifying Spending Priorities 15 Regarding infrastructural development, invest- water and other services. According to Ranganathan ments, and spending on operations and maintenance and Foster (2012, p. 42), a program for accelerat- (in such sectors as water, sanitation, roads, electricity, ed (but still not unreasonable) progress may require and information and communications technology, annual spending of an additional US$400 million or ICT) are crucial for Uganda’s SDG agenda. But, per year (in 2011 US dollars) through 2015, corre- despite having spent heavily on infrastructure during sponding to around 2.4 percent of GDP. Given the 2001–2009—at slightly above 10 percent of GDP, importance of infrastructure access within the SDG or US$1 billion per year—Uganda still lags behind agenda, and its key role in raising growth and con- comparator countries in electricity supply, is severely tributing to a wide range of development goals, it challenged in achieving universal access to sanitation would be crucial to continue to improve services in and considerably lacking in provision of running this area up to 2030. Step Identifying Fiscal Space 4 T he level and efficiency of public spending are of course, may be difficult) may release substantial re- typically among the determinants of the de- sources for additional high-priority spending without velopment of SDGs and their determinants. It additional financing. If efficiency initially is high, then is important to keep in mind that any given level of this source of fiscal space is less important. However, spending may take place within a wide range of policy if so, the government is likely in a better position to frameworks, among other things, with varying roles use additional financing to scale up services and invest- for public and private service delivery. In order to raise ments in priority areas while maintaining acceptable spending in priority areas, additional fiscal space is efficiency.15 needed. Also, the means by which resources are mo- Drawing on the summary in Table 4, among the bilized makes a difference to outcomes—for example, potential sources of fiscal space for priority spending, the effects of additional aid are different from the ef- we find the following: fects of additional taxes. Here we primarily address fiscal space from a budgetary perspective since, by definition, budget re- 15  The challenges of raising government efficiency in service sources are most directly controlled by policymakers. delivery in general, and for services benefitting poor people in However, as will be noted, financing from NGOs and particular, is addressed in the seminal World Development Re- private investors may play an important complemen- port of 2004, “Making Services Work for Poor People” (World tary role. Our framework is comprehensive, analyzing Bank 2003). According to the report, the key to improved ser- vice delivery is institutional changes that strengthen relation- the scope for creating additional fiscal space from tax- ships of accountability between policymakers, providers, and es, fossil fuel subsidy cuts, Official Development Assis- citizens. A large body of research stimulated by this report sug- tance (ODA—i.e., grants and concessional loans), and gests that such institutional changes are possible but not easily other borrowing (domestic or foreign). It is also im- implemented, largely because politicians in many settings may portant to bring government efficiency into the anal- be able to resist accountability to citizens (Devarajan 2014; see ysis: if it is low initially, then improvements (which, also ODI 2014). Table 4:  Government Fiscal Space – Recent Indicators and Future Directions of Change Income and Efficiency Indicators Recent value Impact on future fiscal space Comment Taxes (% of GDP) 13.0 + Likely increase (mainly due to revenues from oil sector) Fuel subsidies (% of GDP) 1.3 + Potential (and desirable) decrease. ODA (% of GNI) 10.1 – Likely decrease. External Debt Stocks (% of GNI) 22.5 + Potential room to increase borrowing. Government efficiency + Potential (and desirable) increase. 18 Country Development Diagnostics Post-2015 Non-oil taxes. Tax revenues are the main source of is as low as 4.2 percent of GDP or, in an average government financing in Uganda. Figure 6 shows year during 2016–2030, around 6.1 percent of how they have evolved since 1990, and bench- GDP, i.e., a loss of 3.4 percentage points. To limit marks their current GDP share against those of this loss, it may be possible to tap into global ini- other countries.16 As shown, Uganda’s tax reve- tiatives, such as the Global Fund to Fight AIDS, nue, at 13 percent of GDP in 2011, is as expect- Tuberculosis and Malaria. ed. The relationship with GNI per capita is not Borrowing. Uganda’s external debt stocks have tight enough to project future changes on the decreased substantially, not least following the basis of projected income growth. If non-oil tax HIPC initiative, and the current 22.5 percent policy were to change, then it would be important of GNI is lower than expected. Again, the rela- to consider the detailed design and likely effects tionship to GNI per capita is not tight enough to on the SDG agenda of such changes, comparing make projections based on cross-country results. the benefits from additional spending to the costs However, a recent IMF-World Bank Debt Sus- related to a reduction of the resources controlled tainability Analysis (DSA) considers as sustainable by households and enterprises.17 an increase in Uganda’s external public or public- Oil taxes. While considerable uncertainty is relat- ly-guaranteed debt from 16 percent of GDP in ed to the oil sector—currently, 2018 is the expect- 2012 to 22 percent in 2033; this permits addi- ed starting year for production—it is likely that tional annual borrowing of roughly 0.3 percent the sector will generate a substantial increase in tax of GDP. In the DSA, it was assumed that other revenues. According to one set of projections, the debt stocks—public domestic and external private tax revenues from oil will reach 8 percent of GDP non-guaranteed—would not change from their by 2023, after which they will decline gradually current GDP shares of 13 percent and 10 percent, until 2045, when production ends and reserves are respectively (IMF 2013). depleted; for the period 2016–2030, oil revenues Government efficiency. A number of government may amount to an average of roughly 4.9 percent efficiency measures are available (Box 4). Ac- of GDP per year (IMF 2013, p. 57). cording to both the health and education indi- Fossil fuel subsidies. Currently Uganda’s subsidy ces, Uganda’s performance is below the expected level is at around 1.3 percent of GDP. Subsidy re- levels; among these two indices, GNI per capita duction is thus a potential source of fiscal space is strongly correlated with the education index and would contribute positively to the climate but largely uncorrelated with the health index. change agenda. It is difficult to assess the likeli- Uganda is performing as expected in terms of the hood of reforms in this area. more general Public Investment Management Official Development Assistance (ODA). Ugan- Index and better than expected according to the da’s net ODA is at around 10.1 percent of GNI World Bank Governance Indicators. Given that (9.4 percent of GDP), also roughly at the expect- the different indices measure different aspects of ed level (11.1 percent of GNI). The cross-country government performance, such mixed findings relationship between GNI per capita and ODA (as may not be inconsistent. Among other coun- percent of GNI, or GDP) suggests that Uganda’s ODA will decline relative to both GNI and GDP 16   Figure 6 suggests, interestingly, that ODA per capita is un- (Figure 7, left panel) while remaining constant in related to GNI per capita—i.e., there is no significant tendency per capita terms. The likely advent of large oil rev- to give higher aid per capita to the countries where needs are enues may lead to further cuts as donors turn to highest. countries with more severe fiscal constraints. The 17   IMF (2013) suggests that, by 2018, an increase of 1.5 per- projected 2030 level of ODA for Uganda—taking centage points of GDP for non-oil would be feasible; Uganda only the increased GNI per capita into account— would still remain within its expected range. Identifying Fiscal Space 19 Figure 6:  Uganda – Tax Revenues 1990–2011 (% of GDP) (left); Tax Revenues (% of GDP)/GNI per capita (right). 14 4 12 UGA Ln Tax revenue (% of GDP) 10 2 Determinant, logs 8 0 6 4 −2 2 0 −4 150 400 1100 2980 8100 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Ln GNI per capita (constant 2005 US$) Measure of Income Per Capita, Log−scale Tax revenue (% of GDP) Ln(DET) = 1.794*** + .12 Ln(INC) ; R2: .027 Sources: WDI, World Bank. try-specific sources, scattered survey evidence especially for infrastructure investments, leveraged by also points to inefficiencies. For example, on any additional government spending in this area. To pro- given day, roughly 15–20 percent of the teachers vide context, according to recent figures, total govern- (including head teachers with supervisory respon- ment spending amounts to around 20 percent of GDP sibilities) are absent, with illness accounting for (IMF 2013, p. 28); it would be a severe challenge to an almost-negligible share of absences (UNESCO raise spending by 4–5 percent of GDP while maintain- 2014a, pp. 31 and 267–268). Similarly, an anal- ing acceptable efficiency. If it were achieved, then gains ysis of local governments suggests, if all districts in the SDG area could be considerable. For the sake of could be brought up to the health and education efficiency, if spending is to be increased, it may be wise outcome-to-spending ratios of the best perform- to do so gradually and seek guidance from frequent ing districts, then about one-third of their bud- impact assessments. gets could be saved (World Bank 2013b, p. xiii). In sum, even though they are unpredictable, effi- ciency gains have the potential to add consider- able fiscal space. 18  Using figures from the preceding discussion, a high esti- mate of the fiscal space increase may be as follows (all percent On balance, this information suggests the fiscal of GDP for an average year 2016-2030): 4.9 (oil taxes) + 1.5 space for SDG priority spending could increase by as (non-oil taxes) + 1.3 (fuel subsidy cuts) – 3.4 (ODA) + 0.3 much as 4–5 percent of GDP.18 However, the extent of (foreign borrowing) = 4.6. In addition, the government may the increase is highly uncertain, not least due to uncer- be able to raise efficiency. However, as noted, the changes for tainty regarding the future of the oil sector. In addition individual items are uncertain, difficult to bring about, and/ to the sources included in the table, it may be pos- or subject to drawbacks (especially if increased spending is not sible to attract additional external private financing, efficient). 20 Country Development Diagnostics Post-2015 Figure 7:  Uganda – ODA (% of GNI)/GNI per capita (left); ODA (per capita)/GNI per capita (right) 4 10 received per capita (current US$) Ln Net ODA received (% of GNI) UGA Determinant, logs Ln Net ODA 2 Determinant, logs 5 0 UGA −2 0 −4 −6 −5 150 400 1100 2980 8100 150 400 1100 2980 8100 Ln GNI per capita (constant 2005 US$) Ln GNI per capita (constant 2005 US$) Measure of Income Per Capita, Log−scale Measure of Income Per Capita, Log−scale Ln(DET) = 8.348*** −1.04*** Ln(INC) ; R2: .303 Ln(DET) = 4.607*** −.095 Ln(INC) ; R2: .004 Sources: WDI, World Bank. It is important to note that trade-offs are involved, Box 4: Measures of Government Effectiveness to varying degrees, when fiscal space is freed up and spending is increased according to priorities: policy On the basis of relationships between inputs and outputs, makers need to think through scenarios for Uganda Grigoli and Kapsoli (2013) and Grigoli (2014) constructed in- with and without major policy changes, and the im- dices for government efficiency in health and education spend- plications for the SDG agenda. The trade-offs may be ing; Dabla-Norris et al. (2011) developed a Public Investment least severe for success in raising government efficiency Management Index (PIMI) that reflects actual practices in four areas (appraisal, selection, implementation, and evaluation). and ODA. For alternatives with different tax and sub- In addition, the World Bank Governance Indicators provide sidy policies, the net short- and long-run impacts on cross-country data on rule of law, government effectiveness, different population groups should be considered. Ad- control of corruption, political stability and absence of vio- ditional borrowing increases the risk of unsustainable lence, quality of regulations, and voice and accountability. future debt levels. Conclusions I n this paper, we present the Post-2015 Country accelerate progress, policymakers and country leaders Development Diagnostics framework for analyzing will have to prioritize government effectiveness and ef- the implications for the SDG agenda at the level ficiency and ensure that development spending is raised of individual low- and middle-income countries. The and allocated to areas critical to the SDG agenda. framework that we present is divided into a sequence The Post-2015 Country Development Diagnos- of distinct steps; each step is illustrated here with se- tics framework and the accompanying database is in- lected findings from a more detailed country diagnos- tended to give analysts in developing countries and the tic of Uganda (Gable et al. 2014). The fact that, in broader international community useful pointers for spite of accelerating progress, most countries will not assessing policy priorities, targets, and financing op- achieve most of the MDG targets by the 2015 deadline tions for virtually any low- or middle-income coun- indicates that this is an important undertaking: while try. The marginal cost of additional applications of ambitions should be global, in order to be effectively this diagnostic framework is relatively low since the embraced, strategies and targets in individual countries cross-country database and related regressions and should be locally owned and anchored in individual graphs have already been done and are easy to access country realities and priorities (UN 2013).19 and use. The framework does not say what policymak- The findings for Uganda—illustrating the nature ers should do but it should help them pose important of country-specific insights that the framework may questions and find answers, also drawing on more de- lead to—reveal a mixed picture of how the country is tailed, country-specific studies.20 Together, this infor- performing compared to what is expected at its GNI mation should provide helpful guidance for stronger per capita. The fact that the country underperformed SDG accomplishments. in various indicators may set off alarms and prompt more detailed analysis, with the initial hypothesis that improvements are clearly attainable in those areas. The 19   On the basis of data for 2010, Uganda seemed on track analysis suggests that in some areas certain linkages are to achieve the MDGs for extreme poverty, education gender at work (e.g., between relatively weak primary educa- parity, under-five (and infant) mortality, and water access. On tion outcomes and the allocation of relatively few re- the other hand, Uganda was off track for undernourishment, sources per primary student). With regard to the SDG primary completion, maternal mortality, and sanitation access agenda, the results suggest that substantial yet only (World Bank 2014). 20   Such studies may be sector-focused or economy-wide. An moderate progress should realistically be expected by economy-wide approach is needed to consider the many in- 2030. This is true even for an economy like Uganda’s teractions between policies, financing, growth, and SDG out- that is expected to grow at a relatively rapid pace and comes. MAMS (Maquette for MDG Simulations), initially de- have access to additional foreign exchange resources veloped at the World Bank for analysis of MDG strategies, is an (from oil). In other words, business as usual clearly is example of such an approach. For more on MAMS, visit www. insufficient to achieve the global SDG ambitions. To worldbank.org/mams. References Asian Development Bank (ADB). 2006. “Key Indica- Working Paper No. 30, Effective States and Inclu- tors of Developing Asian and Pacific Countries”. sive Development Research Centre (ESID), Man- Manila, Philippines. chester, U.K. Devarajan, S. 2014. “What the 2004 WDR Got IMF. 2013. “Uganda: 2013 Article IV Consultation Wrong”. Blog. (http://blogs.worldbank.org/ and Sixth Review Under the Policy Support In- futuredevelopment/what-2004-wdr-got-wrong) strument, Request for a Three-Year Policy Support Fouré, J., A. Bénassy-Quéré, and L. Fontagné. 2010. Instrument and Cancellation of Current Policy “The Great Shift: Macroeconomic Projections for Support Instrument—Staff Report”; Public Infor- the World Economy at the 2050 Horizon”, CEPII mation Notice and Press Release on the Executive Working Papers, No. 2012-03. Board Discussion; and Statement by the Execu- Gable, S., H. Lofgren, and I. Osorio-Rodarte. 2014a. tive Director for Uganda. Country Report No. “Post-2015 Country Development Diagnostics: 13/215. July. Washington, D.C. Uganda”. Forthcoming. World Bank, Washington, Overseas Development Institute (ODI). 2014. “The DC. Impact of 2004 World Development Report”. Grigoli, F. 2014. “A Hybrid Approach to Estimating (http://www.odi.org/events/wdr-conference-mak- the Efficiency of Public Spending on Education ing-services-work-for-poor-people) in Emerging and Developing Economies”. IMF Okwero, P., A. Tandon, S. Sparkes, J. McLaughlin, Working Paper 14/19. International Monetary J. G. Hoogeveen. 2010. “Fiscal Space for Health Fund, Washington, D.C. in Uganda”. Africa Human Development Series. Grigoli, F. and J. Kapsoli. 2013. “Waste Not, Want Working Paper No. 186. World Bank, Washing- Not: The Efficiency of Health Expenditure in ton, D.C. Emerging and Developing Economies”. IMF Ranganathan, R. and V. Foster. 2012. “Uganda’s Infra- Working Paper 13/187. International Monetary structure: A Continental Perspective”. World Bank Fund, Washington, D.C. Policy Research Working Paper 5963, World Bank Hausmann, R., C. A. Hidalgo, S. Bustos, M. Cos- Africa Region and Sustainable Development De- cia, S. Chung, J. Jimenez, A. Simoes, and M. A. partment, February 2012. Yıldırım. 2011. “The Atlas of Economic Com- Republic of Uganda. 2014. “Uganda Vision 2040”. plexity: Mapping Paths to Prosperity”. Center for Kampala. International Development, Harvard University; Task Force on Innovative International Financing for John F. Kennedy School of Government, Harvard Health Systems. 2009. “More Money for Health, University; and Macro Connections, MIT Media and More Health for the Money”. World Health Lab, Cambridge, Ma. Organization, United Nations, New York. Hausmann, R., J. Matovu, R. Osire, and K. Wyett. UN. 2013. “A New Global Partnership: Eradicate 2014. “How Should Uganda Grow?” ESID Poverty and Transform Economies through 24 Country Development Diagnostics Post-2015 Sustainable Development”. Report of the High-level ———. 2013b. “Republic of Uganda: Service Delivery Panel of Eminent Persons on the Post-2015 Develop- with More Districts in Uganda Fiscal Challenges ment Agenda, United Nations, New York. and Opportunities for Reforms”. Public Expendi- UNESCO. 2014a. Education for All Global Monitoring ture Review. Report No. ACS4421. June. Wash- Report 2013/2014. Paris. ington, D.C. ———. 2014b. Statistical Tables for UNESCO (2014a). ———. 2014. “MDG Dashboard” (accessed July 3) (http://www.unesco.org/new/fileadmin/MULTI- (http://data.worldbank.org/mdgs/trends-and-pro- MEDIA/HQ/ED/GMR/pdf/gmr2013/2013_ jections-of-each-mdg-indicator-for-each-country). Report_Statistical_tables-School_year_end- World Health Organization. 2010. “Constraints to ing_2011.zip) Scaling Up the Health Millennium Development World Bank. 2003. “Making Services Work for Poor Goals: Costing and Financial Gap Analysis”. Back- People”. World Development Report 2004. Wash- ground Document for the Taskforce on Innovative ington, D.C. International Financing for Health Systems, Inter- ———. 2013a. “Financing for Development Post- national Health Partnership. World Health Orga- 2015”. World Bank, October. Washington, D.C. nization and World Bank. THE WORLD BANK 1818 H Street, NW Washington, DC, 20433, USA www.worldbank.org