Uganda: Fiscal Space Analysis

Transcription

1 Uganda: Fiscal Space Analysis

2

3 Uganda: Fiscal Space Analysis MARCH 2018 UNICEF UGANDA

4

5 Contents Preface... iii Abbreviations... iv Executive summary... v 01 Introduction and methodology Social categories for children and vulnerable groups Social identity and analysis Data limitations Organization of the FSA Uganda s current macroeconomic and fiscal situation Structure and characteristics of the economy Recent economic developments and medium-term prospects Fiscal policy Revenue Expenditures Fiscal deficit and debt sustainability Looking ahead Uganda s social and fiscal space Social and fiscal space in recent years Fiscal space in recent years Projections post FY2015/ Current and future challenges and implications for fiscal space in the social sectors The base scenario Base scenario assumptions Base scenario projection results Base scenario and fiscal space mapping Alternative scenarios Options to enhance fiscal space Alternative scenarios and projections compared with the base scenario Alternative Scenario 1: Increasing tax and non-tax revenue Alternative Scenarios 2 and 3: Increasing s in social sectors Alternative Scenarios 4 and 5: Increasing social through the World Bank s Intergovernmental Fiscal Transfers Program Alternative Scenarios 6 and 7: Higher (or lower) GDP growth Other possibilities for enhancing fiscal space Further increase in domestic revenue (through oil revenue) Reducing other Reducing external debt service through agreements with creditors Increasing external debt disbursements Increasing net internal borrowing flows Conclusion...35 Conclusions...36 Appendix 1: Bibliography...37 Appendix 2: Fiscal space projections...38 i

6 ii

7 Preface This Fiscal Space Analysis is part of a series of country studies carried out by Ecorys for UNICEF during 2016 and 2017 in various sub-saharan African economies. As such, it aims to provide a better understanding of the role of political economy factors in processes and decisions around the creation and use of fiscal space for investments in children in Uganda. It is published jointly with its sister publication Uganda: Political Economy Analysis. The report was written by a team of Ecorys staff and consultants headed by Ecorys Project Director Ivo Gijsberts, and including consultants Jonathan Wolsey, Andrea Dijkstra, Paul Beckerman, Gabriele Pinto and Gerald Twijukye. It is based on a literature review and a seven-day fact-finding mission to Kampala, Uganda carried out by Jonathan Wolsey, with support from Gerald Twijukye, a Ugandan-based consultant. The writers of this report wish to thank UNICEF Uganda Country Office staff in Kampala, in particular Diego Angemi, Arthur Muteesasira, Nathalie Meyer, and the UNICEF Eastern and Southern Africa Regional Office staff for their support. They also express gratitude to the various Government of Uganda officials, officials of international agencies based in Kampala and representatives of non-governmental entities for generously taking time to meet with the consultants. The team would also like to thank Frances Ellery who provided editorial inputs, and Rachel Kanyana who designed this report and its associated advocacy materials. Photo Uganda. The content of this report does not reflect the official position of UNICEF. Responsibility for the information and views expressed in the report lies entirely with the author(s). iii

8 Abbreviations BFP BoU EAC FY CPI GDP IMF LG MTEF MoFPED NDP PAF PEAP PSI UNICEF VAT Budget Framework Paper Bank of Uganda East African Community Fiscal year Consumer price index Gross domestic product International Monetary Fund Local government Medium-term framework Ministry of Finance, Planning and Economic Development National Development Plan Poverty Action Fund Poverty Eradication Action Plan Policy Support Instrument United Nations Children s Fund Value-added tax iv

9 Executive summary Fiscal space for social s has remained tight in Uganda, in particular given the prioritizing of infrastructure spending. Due to the scaling up in infrastructure investment, which has been ongoing in recent years, the balance in spending between social sectors on the one hand and economic and productive sectors on the other has shifted significantly towards the latter. This shift has thus far not generated higher growth of gross domestic product (GDP), which slowed to 3.8 per cent in financial year (FY) 2016/17. The combination of low growth and inefficient tax administration has in turn limited the revenue base. Facing a tight revenue envelope, and with rising development spending, the government has had to tighten recurrent and increase borrowing to finance infrastructure. Taken together, these actions have led to a progressive increase in the debt/gdp ratio, placing social spending under pressure, and putting social indicators, such as primary completion rates, at risk. Among social sectors (education, health, social development, and water and environment), s have been broadly stable in recent years, both as percentages of GDP and in per-child real terms. While fiscal space for social s (excluding externally financed s due to data limitations) has benefited from a recent modest increase in domestic revenues, external debt disbursements and net internal financial flows, it was negatively affected by the increase in s in other sectors. Over the fiscal years FY2012/13 to FY2015/16, Uganda s social s have averaged about 3.4 per cent of GDP. They have also stagnated in real terms. In 2016 dollars, they amounted to US$45 per child in FY2012/13 and US$47 in FY2015/16 (excluding externally financed s). The overall level of social sector as a percentage of GDP can be considered low, including by regional standards. The education sector has accounted for about 60 per cent of social. On the other hand, the resources spent in the water and environment and social development sectors have been especially low (respectively 0.1 per cent and 0.2 per cent of GDP). The significant differences in recurrent/development s ratios across social sectors suggest that some sectors (education and health) are more constrained than others by the weaknesses in domestic revenue mobilization. In a base scenario with economic growth in line with recent trends, higher spending on social sectors based on projected needs would lead to a fiscal space financing gap. Under this scenario, overall social would average 3.2 per cent of GDP v

10 over the FY2016/17 FY2025/26 projection period, translating into an average of US$62.00 per child at 2016 prices and exchange rate. The gap would presumably have to be covered with internal financing, and the required internal financing flow would average 1.2 per cent of GDP over the projection years. The government s total external and internal debt would rise from an estimated 33.7 per cent of GDP at the end of FY2015/16 to 38.8 per cent of GDP at the end of FY2025/26. The most promising option for covering the fiscal gap is the upcoming World Bank Intergovernmental Fiscal Transfers Program. Assuming the Government respects its commitment in terms of counterpart funding, the World Bank loan would lead to an average increase in social s (in the health and education sectors) of 0.23 per cent of GDP or US$4.96 per child over the FY2016/17 FY2025/26 projection period compared with the status-quo scenario. This would represent an important contribution to reducing the fiscal gap but, on the other hand, the Government may prefer to avoid heavy reliance on donors for social s, especially in a context where it wishes to decrease aid dependency. Another option to enlarge fiscal space would be to enhance tax revenue through improvements in tax administration. While no changes in tax policy are envisaged in the near future, the Uganda Revenue Authority has been taking steps to improve tax compliance, especially for value-added tax (VAT). Assuming that the efficiency of domestic VAT collection increases gradually from 12 per cent in FY2016/17 to 16 per cent in FY2025/26, while import VAT collection efficiency increases gradually from 62 per cent in FY2016/17 to 83 per cent in FY2025/26, the average tax revenue-to-gdp ratio would improve 0.5 percentage points over the projection period compared to the base scenario. If all the additional tax revenue were channelled toward social sectors, average total social would increase by 0.19 percentage points or by US$4.01 per child (assuming debt ratios remain the same) compared to the base scenario. If recent efforts to increase infrastructure investment bring about increased growth, fiscal space for social s could also increase through increased tax revenue. Assuming GDP growth starts to pick up from its current modest levels (and some of the downside risks to growth, such as the crisis in South Sudan, are avoided), average social s would grow in per child real terms even though they decrease slightly as percentages of GDP over the projection period. Assuming the real GDP growth rate gradually increases from 5 per cent in FY2016/17 to 8 per cent in FY2025/26, average social per child would decrease by 0.02 percentage points of GDP but increase in real terms by US$1.66 per child compared to the base scenario. Each of the options analysed would lead to an average increase of roughly 0.1 per cent of GDP or US$3 per child compared to the base scenario. Improvements indicated by the scenarios selected reflect Uganda s relatively weak economic conditions as well as the lack of comprehensive measures to increase revenues. Prospects for increasing fiscal space more significantly for social s remain limited for Uganda. Other options for enhancing fiscal space such as reducing infrastructure or taking on more debt seem less likely to succeed, or even to prove feasible. In theory, once large investment projects are ready (i.e. oil, roads and dams, etc.), the ratio of infrastructure to GDP would gradually decrease (and oil could start flowing). Notwithstanding the challenging environment facing social and the limited scope for increasing fiscal space in the short term, while the ongoing prioritization of infrastructure spending is expected to continue, there are some signs that Government may become more flexible with regards to the overall balance between infrastructure and social spending. vi

11 01 Introduction and methodology This chapter analyses the Ugandan government s recent and future financial capacity to carry out on which children and vulnerable groups depend for their human development and general welfare. This financial capacity is understood to be the fiscal space underlying such. The fiscal space analysis has been carried out using a fiscal-projection exercise in Excel. 1.1 Social categories for children and vulnerable groups This report uses the term social to refer to categories regarded as beneficial to children and vulnerable groups. In the interest of this exercise, social for children and vulnerable Ugandans constitutes the following institutional categories: 1. Education 3. Social development 2. Health 4. Water and environment. Notably, this fiscal space analysis focuses on public. Donor contributions that do not pass through the government budget are not taken into account. This can be relevant, especially for the health and social protection sectors, for which off-budget s are estimated to be as high as on-budget ones in Uganda. The present focus, however, is the flows in the social sectors that flow through Uganda s fiscal accounts. 1 1 While it would be possible to carry out the kind of analysis this chapter describes using an enhanced set of accounts going beyond the official budget accounts, it may prove challenging to identify and incorporate all relevant programmes and funding sources. 1

12 1.1.1 Social identity and analysis To analyse fiscal space for social s, the methodology first sets out the identity that governs the relationship of social spending with its underlying fiscal space. This identity states that total s (comprising current, non-interest, interest and capital ) less the sum of total revenue and external grants is (identically) equal to the overall deficit, which is in turn equal to the net flow of external and internal financing to the government. If total is broken down in the three categories of (1) social and (2) other non-interest and (3) interest, this identity can be rearranged for any year as shown in the box. Fiscal identity Social = Tax and non-tax revenue + External grants - Other - External debt service - Internal interest + External debt disbursements + Net internal financing flows The below-the-line accounts taken together constitute fiscal space for the social flow. For a retrospective analysis that is, for analysis of fiscal performance in historical years this structure can be applied directly to show how the below-the-line flows (the retrospective fiscal space) combined to finance social flows. Chapter 3 describes the historical quantitative analysis for Uganda, for the fiscal years FY2012/13to 2015/16 (the fiscal years in Uganda run from July to June). For the projection analysis, the accounting identity is applied in a different way. For each projection year, the social- flow is projected on the basis of programming assumptions, encompassing the various determinants of recurrent and non-recurrent in the education, health, social development and water and environment categories. Similarly, the below-the-line accounts, except for the net internal financing flows, are projected based on programming assumptions. The total net internal financing flow for each year is then calculated residually, to ensure that the accounting identity is satisfied. For any projection year, this net internal financing flow is the fiscal space gap, that is, the difference between the projected social- flow and fiscal space. If this gap is too large, then the programming assumptions, taken together, would be considered unfeasible. The criteria for too large include the limits on the government s capacity to borrow in domestic financial markets and the implied increase in the government s debt-gdp ratio. Policy makers would presumably want to avoid the net internal borrowing flow as a percentage of GDP exceeding nominal GDP growth in coming years, to prevent the internal-debt stock from rising as a percentage of GDP. The projection exercise is formulated by applying various assumptions, together constituting a scenario, to the historical data base. The relatively simplified, illustrative projection exercise applies scenarios to historical data (as discussed in the Appendix). Each scenario comprises programming assumptions for the years FY2016/17 to FY2025/26, covering: world economic conditions basic Uganda macroeconomic variables merchandise exports and imports tax and non-tax revenue external grants to the government government in social and other categories external and internal debt. For each scenario, some of the assumption lines are set as simple numbers (growth rates, percentages of GDP, etc.). Many of the assumptions, however, are constructed from other assumptions. For example, the growth rates of real GDP and of the price level are numbers that the analyst chooses for any given scenario. It is straightforward to combine these assumptions into an assumed growth rate for nominal GDP. 2

13 1.1.2 Data limitations The core analysis is based on budgetary data covering budget execution for FY2012/13 to FY2015/16. The main data source is the Ministry of Finance, Planning and Economic Development (MoFPED). Additional data sources include the World Bank/International Finance Corporation, the International Monetary Fund (International Financial Statistics) and the Uganda Bureau of Statistics (UBOS). The social sectors are identified using the budget execution sector breakdown as reported by the MoFPED. Data on for education and health reported by the MoFPED in FY2015/16 revealed inconsistencies in the wage economic classification, caused by changes in the local government grant system in the middle of FY2016/17, which led to changes in the reporting of education and health s. To cope with this issue, which would otherwise limit the consistency of the model, we estimated those s based on budgeted amounts (since previous years execution rates were close to 100 per cent in those categories). Moreover, complete budget execution data with an external funding breakdown were not available, making it difficult to disentangle directly funded by external financing (we made use of budgeted amounts, and not actual s, for the externally financed development budget). With respect to population data, we used the projections and the actual number given by the UBOS 2014 revision, 2 while for the previous years (2012 and 2013) we assumed a constant increase equal to the 2014/15 growth Organization of the FSA The remainder of this report is organized as follows: Chapter 2 summarizes Uganda s present macroeconomic context in particular, its fiscal context. Chapter 3 analyses the recent evolution of the flows in the categories of social and outlines some specific challenges in the various areas relevant for social. Chapters 4 and 5 discuss various options available to policy makers to enhance fiscal space, using an illustrative projection exercise for the social flows and fiscal space that would fund them for the years FY2016/17 to FY2020/21. The exercise is a sensitivity analysis, consisting of a base scenario projection based on a broad range of macroeconomic and fiscal-policy assumptions, with various alternative scenarios based on changes to the base scenario assumptions. Chapter 6 summarises the main findings from the analysis. The Appendix provides further details of the sensitivity analysis. 2 Uganda Bureau of Statistics, Population Projections

14 02 Uganda s current macroeconomic and fiscal situation This chapter summarizes Uganda s macroeconomic situation its fiscal conditions. Understanding the broad economic context, and especially the evolution of the government budget, is important in order to first determine the current and likely trends in social s (Chapter 3) and then to assess the scenarios for increasing fiscal space for social s (Chapter 4). This chapter will focus on two major characteristics of that framework that have a direct (and negative) impact on social spending. The first is the high priority the government gives to infrastructure spending, especially for transport and energy. As will be further outlined in the following section, there has been a major shift in the balance between social and productive spending over the last 20 years. While the balance may have been geared too much towards social spending in the 1990s and early 2000s, declining social indicators, in particular primary education school completion rates, suggest it now tilts too much towards infrastructure spending. The second major issue is the continued weak domestic revenue mobilization, notwithstanding some recent progress. In combination, these two circumstances imply that the overall macroeconomic context is not favourable to an increase in the allocation of resources to social sectors. The scaling up of infrastructure investment has - so far - not led to increased GDP growth rates. 2.1 Structure and characteristics of the economy Despite high growth rates over two decades, Uganda has remained primarily an agricultural economy. While the sectoral composition of GDP has changed, with services now constituting about half of GDP and industry and agriculture each account for about 25 per cent of the total value of GDP, agriculture employs the majority of the labour force: an estimated three quarters of the population still depends on low-paying jobs in 4

15 the sector. While it contributes to about 80 per cent of the exports of the country, with coffee being the most important export commodity, it is estimated that the majority (two-thirds) of the agricultural workforce is subsistence farmers. Low productivity growth, especially in agriculture, has been a major hindrance to the growth of the economy over recent years. While real growth was strong up to 2010 (averaging 6 7 per cent between 1995 and 2010), growth has slowed since then, averaging 4 per cent, and lagging behind neighbouring economies. Rates of labour productivity growth have been particularly low (approximatively 2 per cent per annum), reflecting underlying structural features of the economy, including the lack of medium- and large-scale enterprises and of agricultural modernization. Significantly, the productivity in agriculture has declined, in contrast with neighbouring countries. This is explained both by the slow growth of physical capital accumulation and weak human-capital accumulation. While Uganda s previous period of high growth benefited from strong physical capital accumulation, driven by high levels of private investment, since FY2012/13, Uganda s gross capital formation has increased by only 1.7 per cent per annum, again far lower than the increases in neighbouring countries, including Rwanda (8.3 per cent) and Tanzania (6.2 per cent), and despite high public investment. Challenges to human capital accumulation are associated with the poor quality of education, with Uganda s primary completion rate declining from 60 per cent in to about 55 per cent in , 3 in a context of a rapid (youth) population increase. To address growth bottlenecks, the Government has embarked on a large infrastructure programme. It has continued and even scaled up its public investment programme, with development s increasing from 6.5 per cent of GDP in FY2012/13 to 8.6 per cent in 2015/16. This is in line with the priorities of the National Development Plan II (NDP II 2015/16 to 2019/20), which aims to enhance infrastructure develop- ment to provide affordable power and lower transportation costs for value addition and enhanced market access. The delivery of many flagship projects with economy-wide impacts (NDP II Core Projects) is intended to address growth bottlenecks and support Uganda s goal of graduating to an upper middle-income country by A major challenge to that strategy and a major weakness in the economy though is that, so far, an increase in public investment has not been accompanied by increases in private investment, including in commercial farming, manufacturing and agro-processing. Nonetheless, the Government views this strategy as having a likely trickle-down effect on social services through higher growth and domestic revenue generation, in a context where the demographic challenges are putting heavy pressures on such services. The current strategy to prioritize infrastructure spending, which is also driven by the prospect of oil production in the coming years, represents a major shift from previous years, during which the priority was more on social spending, in line with the priorities of the Poverty Eradication Action Plans (PEAPs) implemented in the 2000s. Figure 1 highlights the historical trends in the balance between sectoral, in particular between economic and productive and social sectors. While the balance may have been geared too much towards social sectors in the 1990s and early 2000s, recent declines in social indicators have raised questions on whether it is now tilted too much towards infrastructure. Most notably, primary completion rates have declined recently (from 59 per cent in 2014 to 53 per cent in 2015) 4 and are now lower than those of regional peers. At the same time, these poor outcomes in social sectors cannot be simply attributed to the squeeze on public spending in those sectors but also reflect inefficient service delivery, such as as the poor quality of teaching, teacher absenteeism and mismanagement of funds. In its last country report, the IMF has raised some concerns about this lack of attention to social spending. 5 3 Source: Ministry of Education and Sports. 4 World Bank Development Indicators. 5 Country report 17/2016, July

16 Figure 1: Selected government composition, 6 base year 2003/04= / / / / / / / / / / / /15 Economic and productive sectors Social sectors Public administration Security and justice Source: Boost, World Bank 2.2 Recent economic developments and mediumterm prospects Amid declining growth trends, the recent year (FY2016/17) has been marked by a temporary slowdown. According to the IMF, growth slowed down to 3.9 per cent in FY2016/17, just above population growth. Much of the slowdown can be attributed to the 2016 drought, which has led to a significant decline in agricultural output. Tightened bank lending, affecting private sector credit growth, has been another factor behind the slowdown, along with the civil conflict in South Sudan, which is Uganda s largest trading partner. The supply side shock to agriculture has also pushed up prices, with food price inflation reaching 23.1 per cent in May (headline inflation remained however relatively stable at 5.1 per cent, and is expected to stabilize at around 5 per cent over the medium term). Since some of the causes of the current slowdown are temporary, the economy is expected to rebound in the short to medium term. As the current decline is largely due to temporary factors, growth is widely expected to rebound in the short term, with the IMF estimating growth in the next three years between 5.5 per cent to 6.5 per cent, similar to projections from the World Bank and the Bank of Uganda (BoU). Such growth rates could be considered low for a country in 6 Using the BOOST dataset, government s have been divided into four main groups that compromise several sectors: Economic and productive sectors (Agriculture, Land, Housing and Urban Development, Energy and Mineral Development, Works and Transport, Information and Communication Technology, Tourism, Trade and Industry); social sectors (Education, Health, Social Development); public administration (Public Sector Management, Accountability, Legislature, Public Administration); security and justice (Security and Justice, Law and Order). 6

17 which the population grows by 3 per cent per year and is estimated to reach 130 million in An important risk to growth is the possibility that public investment will fail to yield the expected growth and welfare dividends, 8 and that absorption capacity limitations will slow the investment. The Government is taking steps to improve public investment management with the support of the IMF and the World Bank. The apparent failure of the current investment boom to generate higher GDP growth rates is raising questions about the appropriateness of the current growth strategy. The scaling up of public investment is not translating into higher growth rates, partly because implementation of key projects is constantly delayed. In addition, there are indications that strong achievements on key social indicators face reversal (e.g. primary education completion rates). Development partners, as well as the IMF, have called for growth to be more inclusive, which would require a stronger focus on social spending and on strengthening the social safety net. There are signs that, to a minimum extent at least, the Government is becoming more open to acting on these calls. Table 1: Key macroeconomic indicators (FY2012/13 to FY2015/16) FY12/13 FY13/14 FY14/15 FY15/16 Gross domestic product (US$)* Growth rate (%) Per-capita: Gross domestic product (US$)** Growth rate (%) Non-government consumption (US$)** Growth rate (%) Per cent of GDP: Gross fixed capital formation*** Central government fiscal surplus*** Merchandise-trade surplus*** Growth rate (%): Consumer prices (December) Exchange rate (December) Growth rate (%): Population Population under fifteen Source: International Financial Statistics, World Development Indicators. * US$ million at 2015 prices and exchange rate. ** US$ at 2015 prices and exchange rate. *** Data for 15/16 are estimations based on the assumption of the previous year growth rate UN World Population Prospects. 8 Over the past decade, it is estimated by the World Bank that for every dollar invested in Uganda s capital infrastructure, only seven-tenths of a dollar has been generated. This is far below countries that have successfully undergone structural transformation. 7

18 2.3 Fiscal policy Uganda has a history of relatively stable fiscal management, thanks in part to a strong Ministry of Finance that has enforced fiscal discipline across government. Despite residual bouts of fiscal slippages (especially in election years), the country has had fiscal deficits (after grants) of between 3 per cent and 5 per cent of GDP in recent years. The deficits have been rising as percentages of GDP. The Charter for Fiscal Responsibility, signed in December 2016, emphasizes: sufficiency in revenue mobilization; sustainable levels of fiscal balance and public debt over the medium term; management of revenues from natural resources; management of fiscal risks in a prudent manner; and consistency of the medium-term framework with the National Development Plan (NDP). 9 Within that framework, the Government has been implementing a large infrastructure programme. Uganda has implemented a series of successive programmes with the IMF. In the latest review of the recently completed Policy Support Instrument (PSI), the Fund considers the current fiscal trajectory consistent with the Charter for Fiscal Responsibility. Table 2: Fiscal operations (as % of GDP) FY2012/13 FY2013/14 FY2014/15 FY2015/16 (est) FY2016/17 (proj) FY2017/18 (proj) Total revenue and grants Revenue Grants Total and net lending Recurrent Wages and salaries Other recurrent Interest payments Development Local External Overall balance External financing Domestic financing Source: IMF * All figures based on the 2009/10 rebased GDP provided by the authorities. 9 The Charter requires keeping government debt below 50 per cent of GDP in net present value terms and achieving an overall fiscal deficit (excluding large infrastructure projects) of no more than 3 per cent of GDP by FY2020/21 consistent with the East Africa Community (EAC) convergence criteria for monetary union. 8

19 2.3.1 Revenue Fiscal policy has been constrained by persistently weak domestic revenue mobilization, including by regional standards. One of the characteristics of Uganda s fiscal landscape has been weak revenue mobilization, which has oscillated around 12 per cent of GDP for many years, with some improvement in FY2014/15 and FY2015/16 (reaching 13.8 per cent of GDP). Despite recent progress, domestic resource mobilization (both taxes and non-taxes) has been low by regional standards, as Figure 2 shows. There are several reasons for this modest performance. One relates to the structure of the economy, which has remained very informal, limiting the scope for widening the tax base. Another relates to weak tax compliance, through the granting of tax exemptions. There does not appear to be a strong focus on pursuing high value and high wealth taxpayers, with the IMF s 2015 Tax Administration Diagnostic Assessment Tool review concluding that initiatives to detect businesses and individuals that fail to register are weak and fall short of international good practice. The annual value of revenue foregone from tax exemptions is estimated at 1.2 per cent of GDP for FY2014/15 by the Uganda Revenue Authority. In 2011, the IMF estimated that Uganda s tax effort, understood as the proportion Uganda collects out of its potential, was 67 per cent. 10 There is evidently scope to increase tax revenue, in particular through better compliance. 11 Figure 2: Uganda s fiscal stance compared with regional peers (FY2014/15) as % of GDP Uganda Tanzania Kenya Rwanda Ethiopia Revenue Grants Recurrent wage Other recurrent Devevelopment local Development external Source: IMF staff reports 10 IMF (International Monetary Fund) Revenue Mobilization in Developing Countries. Fiscal Affairs Department. 11 It is also noticeable that, according to IMF estimations (2014), the VAT compliance gap in Uganda is particularly significant and increasing. Eliminating noncompliance would lead to an increase in VAT revenues by 6 per cent of GDP. 9

20 On the other hand, aid flows have remained significant, but there has been a shift from grants to (non-concessional) loans. Aid (grants and loans) constituted more than half of government revenue up to and including the late 2000s. There was a strong focus on and priority given to funding the Poverty Eradication Action Plans (PEAPs), which targeted pro-poor development programmes in many social sectors and in rural infrastructure and agriculture. Budget support funds, which were dominant, were used to fund Poverty Action Fund (PAF) s, which contributed positively to an increase in pro poor s. In the last decade, there has been a shift in aid flows to more productive sectors, notably energy and transport. About half of the current scaling up of infrastructure is financed by development partners, mostly through non-concessional loans. To finance the infrastructure programme, the Government has also resorted increasingly to domestic (debt) financing, although more recently it has decided to limit the growth of its internal debt Expenditures Low domestic revenue mobilization has constrained recurrent spending. The share of allocations for recurrent has been constantly low (including by regional standards, as outlined in Figure 2). It declined from an average value of 64 per cent of total from FY2003/04 to FY 2007/08 to 55 per cent from FY2012/13 to FY2015/16. While the Government s fiscal policy includes no explicit fiscal rule requiring a positive recurrent balance, the Government has, in line with standard fiscal practices, ensured that domestic taxation covers recurrent s. Partly because domestic revenue has grown slowly, recurrent s have stagnated, especially since domestic revenues have also financed (increasing) development s. Service delivery has been affected by this tightening of the recurrent envelope, espe- cially in social sectors. Whereas schools have been built, the average pupil-teacher ratio has remained at about 55, because of the inadequacy of domestic resources to increase the wage bill. Similarly, while health facilities have been constructed, staffing and medicines have been lacking for the same reason. Meanwhile, the share of the development budget has remained high, partly to support the ongoing infrastructure investment plan. As a percentage of GDP, the value of development s has increased by about 50 per cent to around 7 per cent in the last five years, (FY2010/ /16) compared to the preceding five years. Since the launch of the first NDP, the Government has been increasingly emphasizing allocations to the development budget, with a focus particularly on the energy and mineral development and works and transport sectors (which have seen their allocation totalling more than 5 per cent of GDP over recent years, or about one-third of total s). This shift in emphasis reflects the Government s efforts to build the nation s physical capital base as a means of facilitating accelerated economic growth and transformation, in line with the NDP objectives. There are growing concerns though that the balance in spending may have been geared too much towards productive sectors. As discussed above, a significant part of the development budget is externally financed (increasingly through non-concessional loans). 13 While the recurrent budget execution has often exceeded 100 per cent, the development budget execution has been weaker, raising concerns about the effectiveness of the current spending strategy. While allocations to the development budget have increased, the level of execution of the development budget has been mixed, especially on large externally financed projects, implementation of which is generally facing severe delays. While this is broadly in line with other countries experiences, this has raised question 12 In the budget speech FY2017/18, the Government has indicated its intention to reduce domestic borrowing from 2 per cent to 1 per cent of GDP. 13 According to FY2017/18 estimates, 57 per cent of the development budget in FY2017/18 is externally financed. 10

21 about both the returns on the large infrastructure boom as well as about the government s absorption capacity, taking account of recurrent finance constraints. Under the recurrent budget, execution has generally been strong, with some ministries (defence, state house) regularly overspending benefiting from reallocations from the (local) development budget and/or supplementary budgets. This took place in FY2015/16, as the development budget underspent (84 per cent), which led to a reallocation to the recurrent budget, which overspent (115 per cent) Fiscal deficit and debt sustainability The combination of weak revenue performance and the scaling up of the development budget has widened fiscal deficits. The size of both Uganda s overall and primary deficits has increased over recent years. The fiscal deficit is estimated to have reached 5.3 per cent of GDP in the FY2015/16 budget 14 but was expected to decrease slightly in FY2016/17 and FY2017/18. In light of the continued increases in infrastructure spending, Uganda s fiscal policies have been among the more expansionary in the region (as Figure 2 also shows). Its fiscal deficit is second only to that of Tanzania s as a percentage of GDP. Even so, in its latest PSI Review, the IMF has indicated that it remains comfortable with the current fiscal path, with the government s medium-term fiscal framework remaining consistent with the Charter of Fiscal Responsibility. The weak execution of the development budget has limited the size of the deficit over recent years. The growing fiscal deficits have come with a corresponding increase in the debt stock, with the value of public debt increasing from 24.9 per cent of GDP in FY2009/10 to an estimated 37.3 per cent in FY2015/16. With the continued increase in infrastructure investment, government debt is expected to rise and reach 41 per cent of GDP in FY2019/20. In the latest PSI review, the IMF continues to rate Uganda as having a low risk of debt distress, but notes that the risk has clearly increased. There are now large committed but undisbursed external project-related debts amounting to about 20 per cent of GDP, which are likely to be disbursed in the medium term. In addition, Government plans to contract substantial new debts for the railway and other big projects. To safeguard debt sustainability, improved domestic revenue mobilization and sound project implementation will be necessary to realize the envisaged growth dividend. Interest payments amounted to 2 per cent of GDP in FY2015/16, up from 1.4 per cent in FY2012/13, and are expected to rise to 2.7 per cent in FY2017/ Looking ahead The Government s medium-term fiscal framework projects a continued, albeit modest, increase in domestic revenue mobilization. In its medium-term fiscal strategy, the Government is committed to increasing domestic revenue mobilization by 0.5 per cent of GDP per year. Nevertheless, there are no notable measures in the FY2017/18 budget aiming at increasing domestic revenue. The Ministry of Finance has explicitly recognised that, with the current economic downturn, the time may not be ripe for significant tax reforms. The only significant measure concerns tax compliance, with the Uganda Revenue Authority being allocated an additional 90 billion Ugandan shillings to buttress compliance (according to the Parliament Budget Office, this would increase tax revenue by 452 billion Ugandan shillings or 0.5 per cent of GDP in FY2017/18). On the other hand, prospects for a reduction in large tax exemptions appear limited, with large new ones having been granted recently. Prospects for revenue increase may be brighter over the medium term (two to three years), since the Government has recently launched (with the support of the World Bank and IMF) a comprehensive review of both tax administration and policy and is planning to develop a Medium Term Revenue Strategy. Looking further forward, 14 GoU 6.1 per cent. 11

22 oil production is expected to start in about five years, which could boost domestic revenue in the medium term (by between 0.5 and 4 per cent of GDP according to the IMF). However, with no final investment decision yet made on the Uganda/ Tanzania pipeline, no tax regime (for the pipeline) prepared and no policy on the use of petrol revenues (for now, withdrawals from the Petroleum Fund are limited to infrastructure spending), major uncertainties remain about the level of revenue this could generate. As development s are expected to continue to be prioritized in the next few years, there is a growing recognition that the balance between productive and social sectors may have tilted too much towards the former. The FY2017/18 budget accommodates infrastructure investment needs, which implies a very tight current envelope (10.5 per cent of GDP). Looking further forward, the medium-term framework (MTEF) projects a continued high level of development spending in FY2018/19, with a progressive decrease starting in FY2019/20. As the increased infrastructure has yet to deliver a clear growth dividend, and its implementation has been slow, development partners and the IMF have expressed concern about the lack of balance between spending for infrastructure and social sectors. There are also concerns that the current stance, and the tight current envelope, is leading to a deterioration of social indicators. 15 In its most recent PSI reviews, the IMF has expressed concern about the lack of inclusiveness of the current growth and fiscal path, noting that more high-quality social spending was required to make growth sustainable. There are indications that the government is becoming more receptive to these voices. There was for instance a major shift (away from infrastructure spending) between the allocation in the FY2017/18 Budget Framework Paper and the final estimates, following protests about planned reductions in social spending. The tight fiscal envelope is also pushing the Government to place more emphasis on improving efficiency. As the overall fiscal envelope has remained tight, and weak capital budget execution has been noted, there has been a renewed emphasis on budget efficiency measures and improving allocative efficiency. The Government has introduced programmebased budgeting from FY2017/18 onwards. In parallel, a comprehensive consolidation of local government (LG) grants is under way, with a plan to rapidly devolve more resources through LG, particularly in education. Concerning the development budget, a comprehensive programme to strengthen public investment management has been launched. Overall, the combination of low growth, resulting largely from slowing productivity growth despite significant infrastructure investment, and inefficient tax administration has limited the revenue base. Facing a tight revenue envelope, and with rising development spending, the Government has had to tighten recurrent and increase borrowing to finance infrastructure. Taken together, this is leading to a rapid increase in the debt/gdp ratio. This situation has put severe pressure on social spending. 15 The concern about the tight envelope for social sectors has been exacerbated by the rapid population growth and by the influx of refugees from South Sudan to the Northern part of the country, which put significant pressure on social services. 12

23 03 Uganda s social and fiscal space This chapter analyses the recent evolution of, as well as challenges associated with, flows in social sectors (education, health, social development and water and environment). As the government s budget execution datasets do not include data on externally funded projects, the analysis below uses data for government-financed s only. For the externally financed development budget, budget estimates data are used. For that reason, the analysis below refers to data excluding externally financed s Social and fiscal space in recent years Over the last five years, social s have been broadly stable as a percentage of GDP, with a marked decrease however between FY2014/15 and FY2015/16. Table 3 shows the recent evolution of social s. From FY2012/13 to FY2015/16, these s remained on average at around 3.4 per cent of GDP (excluding externally financed s) and 4.5 per cent (including externally financed s). They have also stagnated per child in real terms, reaching US$45 in FY2012/13 and US$47 in FY2015/16 (excluding externally financed s). The relative stability in social levels can be observed across categories (recurrent and development). Within the recurrent budget, a slight increase in wages has been compensated by a decrease in non-wage s. The overall level of social-sector as a percentage of GDP can be considered quite low, including by regional standards. Compared with its most similar neighbour (Tanzania), for which a similar analysis was recently conducted, the total level of social s in FY2015/16 (4.2 per cent versus 5.2 per cent) is particularly low. Using a slightly different definition, the IMF has calculated that Uganda s spending for social categories 13

24 (health, education, pensions and social assistance) has been about half of the East African Community average. 16 The education sector has accounted for about 60 per cent of social s. The allocation of social s across sectors is relatively similar to the one in other East African Community (EAC) countries, according to the IMF. Nevertheless, the fact that 90 per cent of social s are spent on education and health underscores the limited amount of resources allocated to water and environment on one hand and social development on the other (respectively 0.1 per cent and 0.2 per cent of GDP excluding externally financed s). Notably, while 90 per cent of the education sector has been spent on recurrent s (a large proportion of which on salaries), the ratio is very different to water and environment (and to some extent social development), for which development s predominate. These differences mainly reflect the nature of the sectors. While for example investment in water provi- sion fails if schemes do not have a sound and sustainable means of covering maintenance and other recurrent costs, those costs in the budget are relatively low when compared with those of education and health, which are salary intensive. The implication of this is that the ongoing weaknesses in mobilizing domestic revenues (which fund recurrent s, particularly wages) represent a major financial constraint, especially for the expansion of the education and health sectors. Social s also remained broadly stable in per-child US$s at 2016 prices over the last five years. Total social grew in per-child real terms from US$44.6 in FY2012/13 to US$46.60 in FY2015/16 (excluding external finance), as total social decreased as a percentage of GDP from 3.4 per cent of GDP in FY2012/13 to 3.3 per cent in FY2015/16. It is important to remember that this measure includes all flows defined as social, which as explained above include some items that do not directly benefit children but exclude some that do. Figure 3: Composition of social s by categories (as % of GDP) Percentage (%) / / / /16 Total development foreign s (estimates data) Total development local Total non-wage recurrent Total wage Source: MoFPED 16 Uganda, Selected Issues, IMF Country Report No. 17/207, July Data is for 2014 or latest year available. 14

25 Table 3: Evolution of social s in Uganda (as % of GDP) FY2012/13 FY2013/14 FY2014/15 FY2015/16 Total social : Total social (including external financing) Total Wage Total non-wage recurrent Total Development local Total external financing (estimates data) Total education Total education (including external funding) Wages Non-wage recurrent Development local External Financing (estimates data) Total health Total health (including external funding) Wages Non-wage recurrent Development local External Financing (estimates data) Total social development Total social development (including external funding) Wages Non-wage recurrent Development local External Financing (estimates data) Total water and environment Total water and environment (including external funding) Wages Non-wage recurrent Development local External Financing (estimates data) Source: MoFPED Fiscal space in recent years Fiscal space analysis in Table 4 shows how the flows of social for children were effectively funded. As explained in Section above, tax and non-tax revenue, external grants, external-debt disbursements and the net internal debt flow (including internal interest) contribute positively to fiscal space that funds social s for children and vulnerable groups, while other and external debt service contribute negatively to fiscal space. The retrospective analysis of fiscal space highlights the following: Tax and non-tax revenue has risen from 11.3 per cent of GDP in FY2012/13 to 13.5 per cent of GDP in FY2015/16 External grants have decreased from 1.45 per cent of GDP in FY2012/13 to 1.3 per cent of GDP in FY2015/16 Expenditure in other sectors rose from 11.2 per cent of GDP in FY2012/13 to 14.7 per cent of GDP in FY2015/16 15

26 External debt disbursements amounted to 2.5 per cent of GDP in FY2012/13, and then rose to reach 3.3 per cent of GDP in FY2015/16 The net internal financial flow increased from minus 0.2 per cent of GDP in FY2012/13 to 1.3 per cent of GDP in FY2015/16. External debt service has risen gradually from 0.5 per cent of GDP in FY2012/13 to 1.4 per cent in FY2015/16 Table 4: Fiscal space (as % of GDP) FY2012/13 FY2013/14 FY2014/15 FY2015/16 Total fiscal space Tax and non-tax revenue (excl. external grants) (+) External grants (+) Total other non-interest (-) External debt disbursements (+) External debt service (-) Net internal financial flows (incl. internal interest) (+) Overall, fiscal space has been broadly stable over the four years analysed, in accordance with the analysis of the trends in social s above. The main constraint to its increase has been the increase in in other sectors. While fiscal space for social s has benefited from the increase in domestic revenues, external debt disbursements and net internal financial flows, it was negatively affected by the increase in s in other sectors. This analysis is consistent with the analysis of fiscal trends in Chapter 2, which emphasised the significant priority given to infrastructure in recent years, in line with the NDP II priorities. Over the medium term, however, this policy may enhance revenue growth, assuming the investment in infrastructure increases the growth rate. The continued prioritization on economic and productive sectors represents the major constraint to the increase in fiscal space for social Projections post FY2015/16 Based on the MTEF projections and discussions with key stakeholders, we can describe likely trends for some of the variables determining the available fiscal space for social s post FY2015/16. As discussed below, some of the variables in question suggest that these trends are less than encouraging. These basic projections will contribute to the definition of alternative scenarios for increasing fiscal space, as discussed in Chapter 5. MTEF projections indicate that allocations to the social sectors will rise modestly in Ugandan shilling terms, but with health declining slightly. On the other hand, sectors like works and transport would benefit from continued increase. This suggests that the growth in other non-interest observed during FY2012/13 and FY2015/16 would continue over the coming years, which would reduce fiscal space for social s, although the potential increase in growth rates resulting from the investment in infrastructure should enhance revenue growth. As discussed in Chapter 2, there are indications that the Government may become a bit more flexible regarding its policy of heavily prioritizing infrastructure spending. Once large investment projects are ready (i.e. oil, roads and dams, etc.), the ratio of infrastructure to GDP is also likely to decrease. According to projections from international financial institutions and the BoU, GDP growth is likely to be around 5.5 per cent to 6.5 per cent in the coming years, subject to a number of risks (such as the conflict in South Sudan, Uganda s main trading partner, and question marks over returns from key infrastructure projects) and thus relatively high uncertainty. This will affect domestic revenue, together with revenue policy and administration measures, which are now unclear as the Government begins a comprehensive review of tax administration and policy. The Government forecasts a continued increase in domestic 16

27 revenue mobilization of 0.5 per cent per year, but in the absence of solid revenue enhancing measures, it recognizes that this target will be a challenge. Meanwhile, oil production is expected to start in about five years, which could boost domestic revenue in the medium term (between 0.5 and 4 per cent of GDP according to the IMF). The uncertainty regarding the level of revenue oil production could generate is significant, however, and relates partly to the fact that no final investment decision has yet been made for the Uganda-Tanzania pipeline and no tax regime (for the pipeline) is yet in place. Meanwhile, there is no policy on the use of petrol revenues. Projections for grants and loans in the three sectors are also not encouraging, especially for water and environment, where the main donors have indicated that they will soon withdraw. On the other hand, the World Bank s Intergovernmental Fiscal Transfers Program consists of a US$200 million loan for both the education and health sectors over the next four years, in the form of budget support for both development and non-wage recurrent LG grants. Meanwhile, the current heavy reliance on domestic and external non-concessional loans to finance the ongoing infrastructure programme implies a planned increase in external and internal debt service Current and future challenges and implications for fiscal space in the social sectors This section provides more detail on trends across social sectors and the challenges they face in terms of increasing fiscal space. Education As Figure 2 shows, education s have remained stable as a share of GDP in recent years. Over the years, as the government has overseen a continued increase in enrolment rates, the majority of the sector s budget has been allocated to wages (54 per cent in FY2015/16). In terms of programme allocation, primary education has dominated sector s over the years with a widening gap between it and other programmatic lines (53 per cent of budget allocation in FY2015/16). Ensuring universal participation in the primary education system is indeed a key sector priority, as outlined in the NDP, with the primary enrolment rate standing at 91 per cent in FY2013/14. Execution rates have been high for the recurrent budget but have varied between 60 per cent and 86 per cent for the local development budget over the last four years. Implementation of the present sector policy objectives focused on reaching universal primary participation and on the quality of education is likely to face significant financial constraints given the current fiscal path. As the government is currently in the process of reviewing its education sector plan, the NDP II remains the sector s main policy anchor. Its priorities for the sector are to achieve equitable access, ensure delivery of quality education and enhance efficiency and effectiveness. Discussions with sector stakeholders suggest that the tight recurrent envelope in particular is a major constraint on achieving these objectives, especially those linked to improving the quality and efficiency of service delivery. Key services such as inspections, which have been critically lacking, are directly affected by the lack of resources on the recurrent budget. In that respect, the MTEF projections are not encouraging as they show only modest increases for both wage and non-wage recurrent s, after years of relative stagnation of those s (as per cent of GDP). The sector s challenges in delivering key services is also epitomized by the decrease in LG funding to (pre-primary, primary and secondary) education over recent years. The Government plans to reverse that trend in the coming years. It will receive World Bank support to achieve this through the Uganda Intergovernmental Fiscal Transfers Program, a Program for Results loan, which will contribute to both development and recurrent non-wage LG grants from FY2018/19 onwards. The Government has committed to annual increases in its own contributions to those LG grants corresponding to the annual allocations from the World Bank. (However, its record in providing counter-part funding to donor projects has been mixed in the past). 17

28 Figure 4: Composition of education s by categories (as % of GDP) Percentage (%) / / / /16 External finance (estimates data) Non-wage recurrent Development local Wage Source: MoFPED. Source: MoFPED Health Health s funded through the budget have remained relatively stable as a share of GDP in recent years (moving from 1 per cent of GDP to 0.9 per cent in FY2015/16 excluding external finance and from 1.3 per cent to 1.4 per cent including external finance). A slight decrease can be observed for non-wage recurrent s and development s, while the wage bill has remained broadly stable. There are indications that the funding gaps have been particularly severe in primary health care, with reports of primary health facilities not functioning due to lack of staff or medicines. Like education, LG grants, which primarily support health care, have declined significantly in recent years. The World Bank s Uganda Intergovernmental Fiscal Transfers Program will also support LG grants in the health sector, but the Ministry of Health has not indicated the same commitment as the Ministry of Education and Sports to increase LG funding in coming years. Fiscal constraints are particularly acute for the wage bill. While the wage bill represents about 25 per cent of the sector budget, the number of health workers per 1,000 population in Uganda is still far below the WHO threshold of 2.3 doctors, nurses and midwives per 1,000 population. In FY2015/16, the number of doctors, nurses and midwives per 1,000 population was 0.74/1,000. Although there has been some progress in recruitment of health personnel in recent years, 17 the shortage of health personnel, due partly to recurrent financing constraints, represents a major limitation to the expansion of health services. This point was already emphasized in the fiscal space analysis of 2008, 18 and may be more severe 17 Health sector staffing improved slightly to 71 per cent (42,530/60,384) in 2015/16 from 69 per cent in 2014/15. This was a net increase of 747 health workers during the year. The objective of the Health Sector Development Plan is to fill the current staffing norms in the public sector to at least 80 per cent of the current staffing norms by 2019/20, by which time the structure of the whole health workforce should have been reviewed. 18 Fiscal Space for Health in Uganda, World Bank,

29 now given the evolution of the recurrent budget and, in particular, the primary health care budget, in recent years. Efforts to increase health salaries in recent years aim to partly address that constraint, but appear insufficient. Health donors now focus less on infrastructure and more on social delivery, as building more facilities without adequate staffing will not improve service delivery. The Health Financing Strategy (2015/16 to 2024/25) calls for government to increase its funding to prevent out-of-pocket s from exploding, 19 and in parallel for continued donor support to the sector. This varied between 25 per cent and 30 per cent of the sector budget in the last few years, excluding off-budget projects, which are significant. This suggests that unless the Government increases its support to primary health care, challenges in absorption of donor funding will remain severe, and the Health Sector Strategic Plan s main objectives, including universal health coverage, are unlikely to be met. Figure 5: Composition of health s by categories (as % of GDP) Percentage (%) / / / /16 External finance (estimates data) Non-wage recurrent Development local Wage Source: MoFPED. Source: MoFPED Social development Social development is the priority sector that has seen the most positive budgetary evolution in recent years, albeit from a low base. Its main interventions focus on the mainstreaming of gender in the policies, plans and programmes of different sectors and on the establishment and expansion of social protection programmes for vulnerable people. The sector is led by the Ministry of Gender, Labour and Social Development and its largest allocation is for social protection for vulnerable groups. However, it still represents only 0.1 per cent of GDP and less than 3 per cent of total priority s. Most of the recent increase comes from the local development budget. Although the recurrent budget has also increased, there is evidence that it is grossly insufficient. For example, since the signing of the Children (Amendment) Act in 2016 and the establishment of a National Children Authority, there has been no budget to make it operational. The Government has also been unable to meet its commitment in terms of counterpart funding for development programmes, such as the Social Assistance Grant for Empowerment. 19 Public financing of health is estimated at 17 per cent of total health s in the latest National Health Accounts (FY2013/14), with the level out-of-pocket estimated at 40 per cent, compared with the WHO maximum of 15 per cent. 19

30 Fiscal space for the integrated social protection schemes under consideration appears limited. Recently the Government has shown a firm commitment to social protection, and a stronger policy framework has been developed with the launch of the National Protection Policy in The crux of this framework is the broadening of social protection and support systems beyond the small schemes financed mostly by development partners, such as the Senior Citizen Grant, to include vulnerable groups across the whole lifecycle. A recent investment case commissioned by UNICEF 20 analysed the costs and fiscal space for more integrated social protection schemes. A grant covering the period between the first antenatal care visit and the second year of a child s life, for example, would cost between 0.24 per cent and 0.36 per cent of GDP, at full scale. Extended to the eighth year of life, the cost would increase to 0.95 per cent to 1.43 per cent of GDP. Based on those figures and the cost estimates of other schemes on the one hand and the current sector budget on the other, the financial gap in covering these schemes is significant. Figure 6: Composition of social development s by categories (as % of GDP) Percentage (%) / / / /16 Development local Non-wage recurrent Wage Source: MoFPED *Data, including estimates, for externally funded development budget, are not available. Water and environment Expenditure in the water and environment sector has remained broadly stable in recent years. Access to safe water has been assessed as a key determinant for limiting child deaths but only 0.3 per cent of GDP is allocated to the supply of save water. By far the greatest is on the urban water supply, followed by rural water supply and sanitation and natural resources management. As with the social development sector, development constitutes the largest share of water and environment (about 90 per cent). This is largely financed by donors, partly through pooled funding. The main changes in over FY2012/13 to FY2015/16 relate to the reduction in local development, and an increase in non-wage recurrent s. 20 UNICEF/Republic of Uganda/Economic Policy Research Institute, Social protection investment case, Uganda, November

31 Recurrent finance constraints appear to reduce the development budget absorption capacity. There is no sector strategy, but the NDP II highlights a few policy priorities, in particular the promotion of value addition to natural resources and promotion of sound management of hazardous chemicals and electronic waste, including the establishment of modern waste-management infrastructure. Nationwide tree planting and the restoration of degraded natural forests and community forests is also mentioned, as is the provision of safe and clean water and improved sanitation facilities. While by its nature this sector requires strong infrastructure investment, a sustainable financing strategy will require increased support for recurrent s, especially for maintenance, which is recognized by many sector stakeholders as a major issue. The recurrent budget (at 0.01 per cent of GDP) is plainly inadequate given the scope of the challenges the sector faces. Figure 7: Composition of water and environment s by categories (as % of GDP) Percentage (%) / / / /16 External finance (estimates data) Development local Non-wage recurrent Wage Source: MoFPED 21

32 04 The base scenario 4.1 Base scenario assumptions The projection analysis is carried out first with a base scenario (Scenario 0), comprising a set of programming assumptions for the financial years 2016/ /26 intended to be neutral and non-controversial. This scenario includes several key assumptions for the growth rates of such variables as GDP, the exchange rate and population, on which many of the other assumptions depend. 20 Table 29 in the Appendix lists the base scenario assumptions and provides brief explanations for them. Key assumptions in the base scenario are presented in Table 5. Table 5: Key assumptions for the base scenario (Scenario 0) Growth rates FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/26 Real GDP Consumer price index Population growth

33 Table 30 in the Appendix lists and briefly describes the numerical values for all the base scenario assumptions. The real GDP growth rate is assumed to rise gradually from 4.8 per cent in FY2016/17 to 6.5 per cent in FY2025/26. The December December growth rate of the consumer price index (CPI) is assumed to decline gradually from 6 per cent in FY2016/17 to 5 per cent in FY2025/26.The overall population growth rate would remain at 3.04 per cent over the projection period. 4.2 Base scenario projection results Table 6 shows some of the key projection results for the years 2016/ /26, based on the base scenario assumptions. Table 6: Key projection results for the base scenario Social % of total FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/ % of GDP Per child in US$ at 2016 exchange rate and prices Net internal financing gap (fiscal gap) % of total % of GDP Fiscal deficit (surplus/deficit) % of GDP Under the projections resulting from the base scenario assumptions, overall social would average 3.2 per cent of GDP over the FY2016/17 FY2025/26 projection period, while total social would average US$62.00 per child at 2016 prices and exchange rate. Under the base scenario assumptions regarding tax and non-tax revenue, external grants, other, and external- and internal-debt stocks and flows, the projected social flows would produce a fiscal space financing gap that would presumably have to be covered by internal financing. For the specific assumptions of the base scenario, the required internal financing flow would average 0.68 per cent of GDP over the projection years. The government s total external and internal debt would rise from an estimated 33.7 per cent of GDP at the end of FY2015/16 to 38.8 per cent of GDP at the end of FY2025/26. Table 7: Results for the other elements of the fiscal account Results Scenario 0 Average tax and non-tax revenue/gdp, FY2016/17 FY2025/ Average social /GDP, FY2016/17 FY2025/ Average social per child (US$ at 2015 prices and exchange rate), FY2016/17 FY2025/ Net internal debt flow/gdp, FY2016/17 FY2025/ Total government debt/gdp, FY2025/ Table 33 in the Appendix shows the full projection results, based on the base scenario assumptions. 23

34 4.3 Base scenario and fiscal space mapping The funding flows for government comprise (1) tax and non-tax revenue, (2) external grants, and (3) the net (external and internal) financing of the fiscal deficit. Over any time interval, total funding flows equal total flows. Overall comprises the three broad categories of (1) social, (2) other non-interest, and (3) interest. With just a few exceptions, no category of can be said to be directly linked with any specific funding source, and no funding source can be said to be linked with any specific. The exceptions are that certain grants and loan disbursements are provided to fund specific project s, and these are in specific sectors. Apart from these exceptions, all funding effectively flows into a single general fund, and this general fund is used for any budgeted. Figure 8 is a fiscal-mapping chart, with FY2016/ /26 projections according to the base scenario. The projections are shown as percentages of GDP. In the stacked-bar presentation, funding sources are above and flows below the horizontal axis. The sum of everything shown above the horizontal axis effectively funds everything shown below. By definition, the sum of the net external and internal financing flows is the same as the overall fiscal deficit. For the base scenario, the flows would imply a negative net internal government debt flow. Under the base scenario, the government deficit would diminish from 5.2 per cent of GDP in FY2015/16 to 1.10 per cent in FY2025/26, while the net internal debt flow would decrease from 2.2 per cent of GDP in FY2015/16 to1.5 per cent in FY2025/26. Figure 8: Mapping of fiscal space profile base scenario FUNDING (+) Percentage (%) of GDP SPENDING (-) FY12-13 FY13-14 FY14-15 FY15-16 FY16-17 FY17-18 FY18-19 FY19-20 FY20-21 FY21-22 FY22-23 FY23-24 FY24-25 FY Net internal financial flows (incl. internal interest) (+) External-debt disbursements (+) External grants (+) Tax and non-tax revenue (excl. external grants) (+) Total non-priority noninterest (-) External debt service (-) Total priority (-) Source: Estimates and calculations from the projection workbook UgFs.xlsm 24

35 05 Alternative scenarios The projection exercise is set up to carry out a sensitivity analysis. The analysis consists of calculating projections based on alternative scenarios. The assumptions of each alternative scenario are the same as those of the base scenario, except that one or two assumption lines are changed. Comparing each alternative scenario with the base scenario indicates the order-of-magnitude consequences for the fiscal accounts of the changed assumptions, given the exercise s other assumptions. In this chapter, Section 5.1 describes possible approaches to enhance fiscal space. Section 5.2 provides some illustrative alternative scenarios and their projection results, and describes how their results compare with those of the base scenario. 5.1 Options to enhance fiscal space In principle, policy makers have the following general options for enhancing fiscal space for priority : (1) increasing tax and non-tax revenue; (2) reducing spending in social sectors (possibly by increasing efficiency); (3) reducing other ; (4) reducing external debt service through agreements with creditors; (5) increasing external debt disbursements; and (6) increasing net internal borrowing flows. Apart from government policy choices, changes in the macroeconomic context can affect fiscal space. For example, increased GDP growth would increase fiscal space by increasing tax revenue. The following sections discuss these options in the light of the circumstances relatively likely to prevail in Uganda. The policy options are analysed by means of sensitivity analysis, under which projections under alternative scenarios are compared with the base scenario. The alternative scenarios are intended to be illustrative, and should not be understood as specific policy recommendations. Reflecting Uganda s current political and economic context, seven scenarios have been selected. The first three scenarios consider increases in tax and non-tax revenue from improved tax administration, combined with an increase in social. The fourth and fifth scenarios consider an increase in social funded by an increase in external aid (World Bank). Eventually, the sixth and seventh consider the impact of increasing and decreasing GDP growth rates in Uganda. 25

36 5.2 Alternative scenarios and projections compared with the base scenario Alternative Scenario 1: Increasing tax and non-tax revenue Uganda has a relatively low ratio of tax and non-tax revenue to GDP. The IMF has argued that Uganda has the potential to increase tax collection, and the Government is now committed to increase domestic revenues by 0.5 per cent of GDP each year over the medium term. While no specific tax policy changes are envisaged in the short term, the Government has committed itself to improving tax administration by making the Uganda Revenue Authority more efficient. 22 Scenario 1 considers an improvement in the efficiency of VAT collection. Domestic VAT collection efficiency increases gradually from 12 per cent in FY2016/17 to 16 per cent in FY2025/26, while import VAT collection efficiency increases gradually from 60 per cent in FY2016/17 to 83 per cent in FY2025/26. With these assumptions, the average FY2016/17 FY2025/26 net internal debt flow would be 0 per cent of GDP (compared with 0.7 per cent of GDP in Scenario 0). The total (external and internal) government debt stock would amount in FY 2025/26 to 33.3 per cent of GDP (compared with 38.8 per cent in Scenario 0). Table 8: Key assumptions in Scenario 1 Assumptions FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/26 Domestic VAT collection efficiency Import VAT collection efficiency Table 9: Key projection results for Scenario 1 FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/26 Social % of total % of GDP Per child in US$ at 2016 exchange rate and prices Net internal financing gap (fiscal gap) % of total % of GDP Fiscal deficit (surplus/deficit) % of GDP Table 10: Enhanced VAT administration (20% increase) Results Scenario 0 Scenario 1 Variation Average tax and non-tax revenue/gdp, FY2016/17 FY2025/ Average social /GDP, FY2016/17 FY2025/ = Average social per child (US$ at 2015 prices and exchange rate), FY2016/17 FY2025/ = Net internal debt flow/gdp, FY2016/17 FY2025/ Total government debt/gdp, FY2025/ That scenario does not factor in a possible increase in domestic revenue from oil production, which is expected to start in about five years and which could boost domestic revenue in the medium-term (by between 0.5 and 4.0% of GDP according to the IMF). 26

37 5.2.2 Alternative Scenarios 2 and 3: Increasing s in social sectors As noted in Chapter 3, there is a significant scope for increasing social across all sectors: shortage of personnel, weaknesses in service delivery, inadequate maintenance, and constrained implementation of social protection programmes are all the consequence of insufficient social-sector. Against that background, Scenario 2 considers the consequence of an increase in domestically-financed social. More precisely, Scenario 2 considers an increase in the elasticities of the numbers of staff with respect to the child population in the education and health sectors, from 1 (as in the base scenario) to In addition, the elasticity of total real in social protection and water and environment with respect to the child population would increase from 1 to 1.3. With these assumptions, the average FY2016/17 FY2025/26 net internal debt flow would be 0.9 per cent of GDP (compared with 0.7 per cent of GDP in Scenario 0). The total (external and internal) government debt stock would amount in FY2025/26 to 40.7 per cent of GDP (compared with 38.8 per cent in Scenario 0). In effect, a fairly substantial improvement in social could be purchased for a relatively modest increase in the government s net internal financing flow. Table 11: Key assumptions in Scenario 2 Growth rates FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/26 Elasticity of education staff size with respect to child population Elasticity of health staff size with respect to total population Elasticity of central government recurrent social development with respect to total population Elasticity of Water and Environment staff size with respect to child population % of GDP Total Social Expenditure External Debt Disbursement % change compared to the base scenario Total Social Expenditure External Debt Disbursement The magnitude of the increase in education staff has been chosen randomly as we do not have information on teacher hiring trends. The purpose is to consider a substantial increase in education. 27

38 Table 12: Key projection results for Scenario 2 FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/26 Social % of total % of GDP Per child in US$ at 2016 exchange rate and prices Net internal financing gap (fiscal gap) %of total % of GDP Fiscal deficit (surplus/deficit) % of GDP Table 13: Increased social Results Scenario 0 Scenario 2 Variation Average tax and non-tax revenue/gdp, FY2016/17 FY2025/ = Average social /GDP, FY2016/17 FY2025/ Average social per child (US$ at 2015 prices and exchange rate), FY2016/17 FY2025/ Net internal debt flow/gdp, FY2016/17 FY2025/ Total government debt/gdp, FY2025/

39 Scenario 3 combines the assumptions of Scenario 2 and Scenario 1, so that the increase in social would effectively be funded through higher tax revenue resulting from enhanced VAT administration. When combining those assumptions, the average FY2016/ /26 net internal debt flow would be 0.3 per cent of GDP (compared with 0.7 per cent of GDP in Scenario 0). The total (external and internal) government debt stock would amount in FY2025/26 to 35.2 per cent of GDP (compared with 38.8 per cent in Scenario 0). Table 14: Key assumptions in Scenario 3 Growth rates FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/26 Elasticity of education staff size with respect to child population Elasticity of health staff size with respect to total population Elasticity of central government recurrent social development with respect to total population Elasticity of Water and Environment staff size with respect to child population % of GDP Total Social Expenditure External Debt Disbursement % change compared to the base scenario Total Social Expenditure External Debt Disbursement Tax collection efficiency Domestic VAT collection efficiency Import VAT collection efficiency Table 15: Key projection results for Scenario 3 FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/26 Social % of total % of GDP Per child in US$ at 2016 exchange rate and prices Net internal financing gap (fiscal gap) % of total % of GDP Fiscal deficit (surplus/deficit) % of GDP

40 Table 16: Increased social and enhanced VAT administration (20% increase) Results Scenario 0 Scenario 1 Variation Average tax and non-tax revenue/gdp, FY2016/17 FY2025/ Average social /GDP, FY2016/17 FY2025/ Average social per child (US$ at 2015 prices and exchange rate), FY2016/17 FY2025/26 Net internal debt flow/gdp, FY2016/17 FY2025/ Total government debt/gdp, FY2025/ Alternative Scenarios 4 and 5: Increasing social through the World Bank s Intergovernmental Fiscal Transfers Program As noted in Chapter 3, to reverse the negative trend in the funding of health and education, especially at LG level, the Government plans to receive a credit from the World Bank starting from FY2018/19 and to apply the proceeds in a Program for Results, providing direct support for LG grants (the Intergovernmental Fiscal Transfers Program). As counterpart funding, the Government has committed itself to contribute an amount similar to the World Bank s participation. The following two scenarios consider this approach. Scenario 4 considers the case in which the government fulfils this commitment, meaning it will increase domestically funded by the same amount; Scenario 5 considers a situation in which the government increases its own by only 50 per cent of the World Bank disbursement. Scenario 4 assumes an increase in social on health and education and an increase in loan disbursements equal to half the total increase. The increase is assumed each year, up to FY2020/21. With these assumptions, the average FY2016/17 FY2025/26 net internal debt flow would be 0.9 per cent of GDP (compared with 0.7 per cent of GDP in Scenario 0). The total (external and internal) government debt stock would amount in FY2025/26 to 41.7 per cent of GDP (compared with 38.8 per cent in Scenario 0). Table 17: Key assumptions in Scenario 4 Assumptions FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/26 Elasticity of health staff size with respect to total population Elasticity of central government recurrent social development with respect to total population % of GDP Total Social Expenditure External Debt Disbursement Compared to the base scenario Total Social Expenditure External Debt Disbursement

41 Table 18: Key projection results for Scenario 4 FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/26 Social % of total % of GDP Per child in US$ at 2016 exchange rate and prices Net internal financing gap (fiscal gap) % of total % of GDP Fiscal deficit (surplus/deficit) % of GDP Table 19: World Bank Loan and increase in social (100% counterpart) Results Scenario 0 Scenario 1 Variation Average tax and non-tax revenue/gdp, FY2016/17 FY2025/ = Average social /GDP, FY2016/17 FY2025/ Average social per child (US$ at 2015 prices and exchange rate), FY2016/17 FY2025/26 Net internal debt flow/gdp, FY2016/17 FY2025/ Total government debt/gdp, FY25/ Scenario 5 combines an increase in external loans for recurrent and development starting from FY2018/19, with an equal increase in recurrent social. The World Bank loan will cover approximately two-thirds of the increase. With these assumptions, the average FY2016/17 FY2025/26 net internal debt flow would be 0.9 per cent of GDP (compared with 0.7 per cent of GDP in Scenario 0). The total (external and internal) government debt stock in FY2025/26 would amount to 41.1 per cent of GDP (compared with 38.8 per cent in Scenario 0). Table 20: Key assumptions in Scenario 5 Growth rates FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/26 % of GDP Total Social Expenditure External Debt Disbursement Compared to the base scenario Total Social Expenditure External Debt Disbursement Elasticity of health staff size with respect to total population Elasticity of central government recurrent social development with respect to total population

42 Table 21: Key projection results for Scenario 5 FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/26 Social % of total % of GDP Per child in US$ at 2016 exchange rate and prices Net internal financing gap (fiscal gap) % of total % of GDP Fiscal deficit (surplus/deficit) % of GDP Table 22: World Bank loan and increase in social (50% counterpart) Results Scenario 0 Scenario 5 Variation Average tax and non-tax revenue/gdp, FY2016/17 FY2025/ = Average social /GDP, FY2016/17 FY2025/ Average social per child (US$ at 2015 prices and exchange rate), FY2016/17 FY2025/26 Net internal debt flow/gdp, FY2016/17 FY2025/ Total government debt/gdp, FY2025/ Alternative Scenarios 6 and 7: Higher (or lower) GDP growth The projection exercise could be used in a straightforward way to evaluate the consequences of a real GDP growth rate that is higher or lower than what the base scenario assumes. Although both the Bank of Uganda and international financial institutions project relatively high growth in coming years (around 5.5 to 6.6 per cent of GDP), there are several risks (including the conflict in South Sudan and questions over returns from key infrastructure projects). On the other hand, if all goes well Uganda could also achieve higher growth rates than those assumed for the base scenario. Scenario 6 reverts to the Scenario 0 assumptions, but with a single difference: the growth rate of real GDP gradually increases from 5 per cent in FY2016/17 to 8 per cent in FY2025/26. With these assumptions, the average FY2016/17 FY2025/26 net internal debt flow would be 0.4 per cent of GDP (compared with 0.7 per cent of GDP in Scenario 0). The total (external and internal) government debt stock in FY2025/26 would amount to 34.1 per cent of GDP (compared with 38.8 per cent in Scenario 0). Table 23: Key assumptions in Scenario 6 Growth rates FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/26 Real GDP

43 Table 24: Key projection results for Scenario 6 FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/26 Social % of total % of GDP Per child in US$ at 2016 exchange rate and prices Net internal financing gap (fiscal gap) % of total % of GDP Fiscal deficit (surplus/deficit) Per cent of GDP Table 25: Higher real GDP growth Results Scenario 0 Scenario 6 Variation Average tax and non-tax revenue/gdp, FY2016/17 FY2025/ Average social /GDP, FY2016/17 FY2025/ Average social per child (US$ at 2015 prices and exchange rate), FY2016/17 FY2025/26 Net internal debt flow/gdp, FY2016/17 FY2025/ Total government debt/gdp, FY2025/ For Scenario 7, real GDP growth would decrease gradually from 4.3 per cent in FY2016/17 to 3 per cent in FY2020/21. With these assumptions, the average FY2016/17 FY2025/26 net internal debt flow would be 1.4 per cent of GDP (compared with 0.7 per cent of GDP in Scenario 0). The total (external and internal) government debt stock in FY2025/26 would amount to 51.8 per cent of GDP (compared with 38.8 per cent in Scenario 0). Table 26: Key assumptions in Scenario 7 Growth rates FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/26 Real GDP Table 27: Key projection results for Scenario 7 FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/26 Social %of total % of GDP Per child in US$ at 2016 exchange rate and prices Net internal financing gap (fiscal gap) % of total % of GDP Fiscal deficit (surplus/deficit) % of GDP

44 Table 28: Lower real GDP growth Results Scenario 0 Scenario 5 Variation Average tax and non-tax revenue/gdp, FY2016/17 FY2025/ Average social /GDP, FY2016/17 FY2025/ Average social per child (US$ at 2015 prices and exchange rate), FY2016/17 FY2025/26 Net internal debt flow/gdp, FY2016/17 FY2025/ Total government debt/gdp, FY2025/ Other possibilities for enhancing fiscal space There are, of course, other possibilities for fiscal space enhancement besides those analysed for the sensitivity analysis, but these are less likely to prove feasible. because it is more likely to be regarded as technical rather than political and it can be sustained through multiannual efforts, often with foreign technical support. By contrast, a reduction in other is more likely to be politically controversial and difficult to carry out in sustained multiannual efforts Further increase in domestic revenue (through oil revenue) As discussed above, oil production is expected to start in about five years, which could boost domestic revenue in the medium term (by between 0.5 and 4 per cent of GDP according to the IMF). There is, however, significant uncertainty about when production is likely to start and the level of domestic revenue oil production could generate (as a result of which MoFPED refuses to factor oil revenue in its medium-term fiscal framework). This uncertainty relates partly to the fact that no final investment decision has yet been made on the Uganda-Tanzania pipeline and no tax regime (for the pipeline) is yet in place. Meanwhile, there is no policy on the use of petrol revenues. Once there is further clarity and certainty on those aspects, this could be considered as a scenario for increasing fiscal space Reducing external debt service through agreements with creditors Uganda s public debt stood at 34.5 per cent of GDP in FY2015/16 (60 per cent of it external). Although this is below the 50 per cent of GDP ceiling in the Charter for Fiscal Responsibility, the debt level has increased faster than economic growth over recent years (it was at 26 per cent in FY2012/13). The IMF has pointed out that the risk of debt distress, while still low, has increased (debt service to exports ratio was 15 per cent in FY2016/17). The increase in debt levels has resulted largely from increased external non-concessional loans from non-paris club creditors such as China to finance the Government s infrastructure programme. In the short term, at least, there is essentially no possibility of agreements with creditors (multilateral, bilateral) to reduce debt service Reducing other For any economy, and especially for Uganda, limiting other as a general approach for sustainably enhancing fiscal space for social presents a complex set of practical issues. By comparison, improving tax administration is likely to be politically more acceptable Increasing external debt disbursements In general, macroeconomic policy specialists concur that it is inadvisable to use commercial external debt to fund education, health or social protection. The reasoning is that eventual returns to education and health are realized over decades, while debt service 34

45 on commercial external debt is generally due within a decade. Concessional debt, with multi-decade terms and a near-zero interest rate, is more sensible for these sectors. Policy statements by government officials as well as NGOs broadly concur with this. The Government recognizes that there is limited scope for new external borrowing, especially for social s. On the other hand, the Government has just reached agreement on a new (concessional) loan with the World Bank to support LG grants in health and education Increasing net internal borrowing flows In the analytical structure of the projection exercise, net internal borrowing is calculated residually. In effect, it is the consequence of all the programming assumptions taken together. Evaluation of its feasibility therefore amounts to evaluation of the feasibility of all the programming assumptions taken together. In principle, if it is inadvisable to fund social with external debt, it is even more inadvisable to use internal debt, which usually carries a higher interest rate and shorter maturities. Uganda s net internal financing flow averaged about 2.2 per cent of GDP over the FY2011/12 FY2015/16 period. Given the economy s (nominal) GDP growth, this caused Uganda s internal debt to rise over those years from about 9.2 per cent to about 12.8 per cent of GDP. To prevent internal debt from rising further, policy makers have begun setting policy with a view to reducing the internal debt flow in coming years. (The Government has recently committed itself to reduce net internal borrowing to 1 per cent of GDP in FY2017/18). For this reason alone, it is unlikely that policy makers would consider issuing internal debt to increase fiscal space. 5.4 Conclusion Reflecting Uganda s current political and economic contexts, different scenarios and their combinations for increasing fiscal space have been analysed in this chapter (linked to an improvement in tax administration, increase in social, variation in GDP growth rates, and increase in external aid, through the World Bank Intergovernmental Fiscal Transfers Program). Each of the options analysed would lead to a maximum average increase of roughly 0.1 per cent of GDP or US$3 per child compared to the status quo scenario. The alternative scenarios selected reflect the relatively weak economic conditions as well as the lack of comprehensive measures to increase revenues. The prospects for significantly increasing fiscal space for social remain relatively limited in Uganda over the projected period. The scenario linked to the World Bank Intergovernmental Fiscal Transfers Program offers, based on the model, the highest potential in terms of increasing social, assuming the Government respects its commitment in terms of counterpart funding. In some respects, this is a disappointing finding. The Government may not want to rely heavily on a donor for its social, especially in a context where it wishes to decrease aid dependency. 35

46 Conclusions This report has underscored the significant challenges Uganda faces in increasing its fiscal space for social s, at a time when the lack of prioritization of social s appears to endanger the country s progress on social indicators, including on children and other vulnerable groups. The continued prioritization of economic and productive sectors represents the major constraint to the increase in fiscal space for social, but there are no immediate signs of a significant shift in policy in that regard. Key findings include: 1. The lack of significant progress in domestic revenue mobilization is a major constraint on fiscal space for social s. Unless this situation changes, the fiscal space will remain tight, in particular in terms of recurrent. The recently agreed review of tax policy/administration and the forthcoming medium-term revenue strategy offer hope that new revenue enhancing measures will be introduced from FY2018/19 onwards, although the political climate remains unconducive to major changes in revenue policies such as a reduction in tax exemptions. Over the medium term, domestic revenues could be boosted by oil revenues, but there remain major uncertainties regarding the timing and size of the revenues that could be generated by oil production. 2. Growth levels have also been disappointing in recent years, which has contributed negatively to fiscal space for social s. Looking forward, much depends on the capacity of the infrastructure investment programme to start delivering higher returns, which necessitates improvements in the project implementation capacities of the Government; 3. The World Bank s Intergovernmental Fiscal Transfers Program can contribute to an increase in fiscal space for social s, more particularly in health and education. Much of the effect on fiscal space, however, will depend on the capacity of the Government to provide counterpart funding, which is not assured. Overall, donor funding is unlikely to contribute significantly to an increase in fiscal space in social sectors. 4. The ongoing prioritization of other s (and especially infrastructure), which is expected to continue, constitutes another constraint to increasing fiscal space for social s. On the other hand, there are signs that the Government may be becoming more flexible with regards to the overall balance between infrastructure and social spending. 5. In general (and in line with the Government s policy preferences), non-concessional external debt should not be used to fund education and health. The basic reason is that the yields from education and health come only in the long term, beyond the terms typical of non-concessional external debt. For similar reasons, domestic debt should not be used to fund education and health. 36

47 Appendix 1: Bibliography 1. Government of Uganda, Sector Budget Framework Papers FY2017/18 Education, Health, Social Protection, and Water and Environment, February IMF, Revenue Mobilization in Developing Countries. Fiscal Affairs Department, IMF, 7 th Review under the PSI, Country Report No. 17/7, January IMF, 8 th Review under the PSI, Country Report No. 17/206, July IMF, Uganda: Selected Issues, Country Report No. 17/207, July IMF, Tax Administration Diagnostic Assessment Tool, Performance Assessment Report Uganda, October Martin Brownbridge (Bank of Uganda), The State of the Economy, Power Point Presentation, MoFPED, Charter for Fiscal Responsibility, December MoFPED, Budget Framework Paper FY2017/18, March MoFPED, Draft estimates, Volumes 1 and 2, April MoFPED, Budget Speech for FY 2017/18, June Press release, EU envoys ask government to invest more in people, February Parliament of the Republic of Uganda, Report of the Budget Committee on the Annual Budget Estimates FY2017/18, May Uganda Bureau of Statistics, Population Projections Uganda Debt Network, Performance of Uganda s Debt Portfolio and Development Challenges, Key Lessons, Issues Paper, May UNICEF/Republic of Uganda/Economic Policy Research Institute, Social protection investment case, Uganda, November UNICEF, Budget briefs for Education, Health, Social Protection, and Water and Environment, February World Bank, Fiscal Space for Health in Uganda, World Bank, Uganda economic update, 7th edition, November World Bank, Intergovernmental Fiscal Transfers Program, Project Appraisal Document, June

48 Appendix 2: Fiscal space projections This Appendix describes the details of the base scenario (Scenario 0) projection exercise discussed in Section 4.11, and then describes the results of a sensitivity analysis. The base scenario programming assumptions are simplified to make the calculation relatively easy to carry out and understand. The following general explanatory points are noted: 1. The assumptions are programming assumptions. They are not intended, and should not be understood, as forecasts but rather as plausible possibilities for planning purposes. In particular, the growth rates of government are intended as plausible policy settings. 2. In general, the aim of Scenario 0 is to set programming assumptions that are neutral in character. For example, Uganda`s merchandise export volumes are assumed to grow at the same rates as the world trade volume, so Uganda s exports maintain the same share of the world trade volume. The volume of Uganda s merchandise imports is assumed to grow at the same rates as real GDP, so merchandise imports would tend to maintain the same percentage of GDP. For recurrent, the assumption that staff sizes will grow at the same rate as the population would be neutral in a similar sense. So is the assumption that government wage rates would grow at the same rate as per-capita nominal GDP. The elasticities that help determine the government s revenue performance are taken to be unitary for Scenario 0. This is also a neutral assumption. (In general, it is inadvisable to apply econometric point estimates based on historical data for these values, for at least two reasons. The first is that future elasticities of tax revenue with respect to their underlying determinants are likely to differ from historical elasticities. The second is that, say, if the elasticity of a given revenue line with respect to nominal GDP is assumed to exceed (be less than) one, the projected revenue flow would rise (diminish) indefinitely as a percentage of GDP. 3. It is straightforward to set programming assumptions that adjust gradually over the projection period, using ( geometric ) adjustment formulas. This is useful for several different assumption lines. For example, a large proportion of the assumptions are set as growth rates. These can be assumed to rise or diminish gradually from their initial projection values towards their final projection values. Another way to use a gradual adjustment would be for the elasticity of a given revenue line with respect to nominal GDP to take on an initial value somewhat different from one, but then gradually adjust towards a long-term value of one. 38

49 Programming assumptions For Scenario 0, the programming assumptions are listed in Table 29. Table 29: Programming assumptions base scenario (Scenario 0) (A) World economic conditions (1 3): (1) The growth rate of the world trade volume rises gradually from its estimated FY16/17 value of 3% to a FY25/26 value of 5%. (2) The growth rate of the US$ world price level rises gradually from its estimated FY16/17 value of 1.1% to a FY25/26 value of 2%. (3) The London Interbank Offer Rate rises gradually from its FY16/17 value of 0.6% to a FY25/26 value of 1.5%. (B) Basic Uganda macroeconomic variables (4 10): (4) The growth rate of real GDP gradually increases from 4.5% in FY16/17 to 3% in FY25/26 (5) The GDP deflator grows at the same rate as the year-average consumer price index. (6) The December December growth rate of the CPI declines gradually from 5.8% in FY16/17 to 5% in FY25/26. (7) The December December growth rate of the US$ exchange rate grows at a rate (approximately) equal to the differential of the Ugandan and the world US$ inflation rates. (8) The overall population growth rate stays at 3.04% over the projection period. (9) The under-15 population growth rate stays at 2.35% over the projection period. (10) The headcount poverty incidence declines gradually from 32.8% in FY16/17 to 29.4% in FY25/26. Exports and imports of goods and non-factor services (11 17): (11) The export volume grows at the same rate as the world trade volume. (12) Export prices grow at the same rate as the world US$ price level. (13) The import volume grows at the same rate as real GDP. (14) Import prices grow at the same rate as the world US$ price level. (15) Non-factor service exports grow at a rate equal to the combined growth rates of world trade volume and the world US$ price level. (16) Non-factor service imports excluding insurance and freight charges for merchandise imports grow at a rate equal to the combined growth rates of world trade volume and the world US$ price level. (17) Insurance and freight charges decline gradually from 11.96% of the value of merchandise imports in FY16/17 to 11.6% in FY25/26. National accounts (18 20): (18) Consumption by government entities outside the central government remains at 0% of GDP over the projection period. (19) Gross fixed capital formation remains at 17.5% of GDP. (20) The net increase in inventory stocks remains at 0% over the projection period. (C) Tax and non-tax revenue (21 31): (21) The elasticity of personal income tax with respect to nominal GDP declines from 1.5 in FY16-/7 to 1 in FY25/26. (22) The elasticity of company-tax revenue with respect to nominal GDP declines from 1.5 in FY16/17 to 1 in FY25/26. (23) The elasticity of other income-tax revenue with respect to nominal GDP declines from 1.5 in FY16/17 to 1 in FY25/26. (24) The elasticity of customs revenue with respect to merchandise-imports value declines from 1.5 in FY16/17 to 1 in FY25/26. (25) The elasticity of excise revenue with respect to nominal GDP declines from 1.5 in FY16/17 to 1 in FY25/26. (26) The elasticity of export duty revenue with respect to export value declines from 1.5 in FY16/17 to 1 in FY25/26. (27) The internal value-added tax rate remains unchanged at 18%. (28) The internal VAT collection increases gradually from 11.8% in FY16/17 to 13.9% in FY25/26 (29) The import-based VAT rate remains at 18%. (30) The import-based VAT collection increases gradually from 58.4% in FY16/17 to 68.8% in FY25/26 (D) External grants to the government (32 33): (32) Central government external grants for current would remain at 0.4% of GDP. (33) Central government external grants for capital (projects) would remain at 0.9% of GDP. 39

50 (E) Government in the social and other categories (43 55): (E.1) For non-interest recurrent (E.1.a) In the education sector (34) The staff size grows at the same rate as the number of children. (35) Staff salaries grow at a rate equal to the growth rate of per-capita nominal GDP. (36) Expenditure on current goods and services grows at a rate equal to the combined growth rates of the year-average CPI and the sectoral staff size. (37) Expenditure on non-staff recurrent excluding current goods and services grows at a rate equal to the combined growth rates of the year-average CPI and the number of children. (E.1.b) In the health sector (38) The staff size grows at the same rate as the population. (39) Staff salaries grow at a rate equal to the growth rate of per-capita nominal GDP. (40) Expenditure on current goods and services grows at a rate equal to the combined growth rates of the year-average CPI and the sectoral staff size. (41) Expenditure on non-staff recurrent excluding current goods and services grows at a rate equal to the combined growth rates of the year-average CPI and the population growth rate. (E.1.c) In the social protection sector (42) Central government recurrent grows at a rate equal to the combined growth rates of the year-average CPI and the population. (E.1.d) In the water and environment sector (43) The staff size grows at the same rate as the number of children. (44) Staff salaries grow at a rate equal to the growth rate of per-capita nominal GDP. (45) Expenditure on current goods and services grows at a rate equal to the combined growth rates of the year-average CPI and the sectoral staff size. (46) Expenditure on non-staff recurrent excluding current goods and services grows at a rate equal to the combined growth rates of the year-average CPI and the population growth rate. (E.1.f) In the other sectors (47) The staff size grows at the same rate as the population. (48) Staff salaries grow at a rate equal to the growth rate of per-capita nominal GDP. (49) Expenditure on current goods and services grows at a rate equal to the combined growth rates of the year-average CPI and the sectoral staff size. (50) Expenditure on non-staff recurrent excluding current goods and services grows at a rate equal to the combined growth rates of the year-average CPI and the population growth rate. (E.2) For non-recurrent, over the projection years (51) Education non-recurrent central government increases gradually from the FY16/17 value of 0.1% of GDP to a FY25/26 value of 0.2% of GDP. (52) Health non-recurrent central government increases gradually from the FY16/17 value of 0.1% of GDP to a FY25/26 value of 0.1% of GDP. (53) Social protection non-recurrent central government increases gradually from the FY16/17 value of 0% of GDP to a FY25/26 value of 0.1% of GDP. (54) Non-recurrent water and environment increases gradually from the FY16/17 value of 0.2% of GDP to a FY25/26 value of 0.3% of GDP. (55) Other non-recurrent central government decreases gradually from the FY16/17 value of 8.9% of GDP to a FY25/26 value of 10% of GDP. (F) For external and internal debt (56 58): (56) Average interest rates on the previous year s year-end external debt stock increase (decrease) with London Interbank Offer Rate (LIBOR). (57) Average interest rates on the previous year s year-end internal debt stock decline gradually from 14.8% in FY16/17 to 7.9% in FY25/26. (58) External debt repayments in each projection year amount to 2.2% of the preceding year s year-end external debt stock; and (58) External debt disbursements in each projection year amount to 35.8% of total non-recurrent. 40

51 Table 30 shows the assumption values for the base scenario, based on the reasoning as presented in Table 29. Table 30: Programming assumptions for the fiscal space projection exercise (base scenario) FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/26 (A) EXTERNAL STATE-OF-THE-WORLD VARIABLES: Growth rates: World trade volume World US$ price level Average world US$ oil price (US$/ bbl.) Average world US$ copper price (US$/MT) Interest rates: London Interbank Offer Rate (LIBOR) (B) BASIC MACROECONOMIC VARIABLES: Growth rates: GDP (national currency millions) GDP at 2016 prices and exchange rate (US$ million) GDP deflator Consumer prices (year-average) Consumer prices (December) Exchange rate (year average) Exchange rate (December) Population (millions) Population under 15 (millions) Population in poverty Headcount poverty incidence Growth rates (US$ million): Merchandise exports: Unit value Volume Merchandise imports: Unit value Volume Non-oil imports: Unit value Volume Growth rates: Non-factor services receipts Non-factor services payments. excluding merchandise imports insurance and freight Ratios: Ratio. insurance and freight costs/ merchandise imports value Incremental capital output ratio

52 % of GDP: Consumption by governments excl. central government FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/ Gross fixed capital formation Net increase in inventory stocks GENERAL GOVERNMENT FINANCIAL ACCOUNTS: Tax and non-tax revenue (excl. external grants) (+): (C) TAX REVENUE: Central government: Elasticities of... Personal income tax with respect to nominal GDP Company tax revenue with respect to nominal GDP Other income tax revenue with respect to nominal GDP Customs revenue with respect to merchandise imports value Excise revenue with respect to nominal GDP Export duty revenue with respect to export value Internal VAT revenue with respect to nominal GDP VAT revenue from imports with respect to the value of merchandise imports VAT: Internal VAT: Internal VAT rate Internal VAT collection efficiency VAT revenue from imports: External VAT rate External VAT collection efficiency Elasticities of... Central government non-tax revenue with respect to nominal GDP (D)External grants (+): % of GDP: Central government external grants for current Central government external grants for capital (projects) (E) CENTRAL GOVERNMENT EXPENDITURE: Growth rates: Recurrent education :

53 FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/26 Central government recurrent education : Education staff Education remuneration rates Non-staff recurrent education : Recurrent education on goods and services Other non-staff recurrent education Recurrent health : Central government recurrent health : Health staff Health remuneration rates Non-staff recurrent health : Recurrent health on goods and services Other non-staff recurrent health Recurrent social development : Central government recurrent social development : Recurrent water and environment : Central government recurrent water and environment : Water and environment staff Water and environment remuneration rates Non-staff recurrent water and environment : Recurrent water and environment on goods and services Other non-staff recurrent water and environment Other recurrent : Central government other recurrent : Other staff Remuneration rates in other sectors Non-staff recurrent other : Recurrent other on goods and services Other non-staff recurrent other % of GDP: Non-recurrent education :

54 Central government non-recurrent education Non-recurrent health Central government non-recurrent health Non-recurrent social development FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/ Central government non-recurrent social development : Non-recurrent water and environment Central government non-recurrent water and environment Other non-recurrent Central government other non-recurrent (F) EXTERNAL AND INTERNAL DEBT: Average interest rates (applied to preceding year-end debt stock): Average interest rates on external debt Average interest rates on internal debt % of preceding year-end debt stock: External debt repayments (-) % of GDP: External debt disbursements (+): External debt disbursements/total non-recurrent External debt repayments (-) Net internal debt flow (+): Source: Estimates and calculations from the projection workbook UgFS.xlsm Projection results base scenario Table 31 shows the base scenario projection results. The top line of the projection (A) shows the evolution of priority, and the lines below show the evolution of the components of its fiscal space: tax and non-tax revenue (B), external grants (C), non-priority (-) (D), external-debt disbursements (E), external-debt service () (F), and the fiscal gap (G). The net internal-debt flow as a percentage of GDP is the bottom-line result, determined residually for each year from the assumed priority- and fiscal space flows. For each projection year, it indicates the fiscal gap implied by the programming assumptions, including those that determine priority and those that determine the other fiscal accounts. In the base scenario, the average FY2016/17 FY2025/26 net internal debt flow would be 0.7 per cent of GDP. The total (external and internal) government debt stock would conclude in FY2025/26 at 38.8 per cent of GDP. Policy makers would presumably consider the fiscal balance projections satisfactory per se, although one of the main reasons for this outcome is that in education, health and other sectors relevant to children s needs would remain low, and perhaps inadequate. These results would seem to imply there is scope for increasing beneficial to children. 44

55 Table 31: Projection results for the fiscal space projection exercise (base scenario) (A) Total social non-interest FY16/17 FY17/18 FY18-19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/ Total education Total health Total social development Total water and environment Social recurrent Recurrent education Central government recurrent education Expenditure on education staff Non-staff recurrent education Recurrent education on goods and services Other non-staff recurrent education Recurrent health Central government recurrent health Expenditure on health staff Non-staff recurrent health Recurrent health on goods and services Other non-staff recurrent health Recurrent social development Central government recurrent social development Recurrent water and environment Central government recurrent water and environment Expenditure on water and environment staff Non-staff recurrent water and environment Recurrent water and environment on goods and services Other non-staff recurrent water and environment Non-recurrent education : Central government non-recurrent education Non-recurrent health : Central government non-recurrent health Non-recurrent social development :

56 Central government non-recurrent social development Non-recurrent water and environment Central government non-recurrent water and environment LG authorities non-recurrent water and environment (B) Tax and non-tax revenue (excl. external grants) (+): FY16/17 FY17/18 FY18-19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/ a Tax revenue: Central government tax revenue Income tax Personal income tax Company tax Other income tax VAT VAT on internal transactions VAT on imports Customs and excise duties Customs duties Excises Export duties Non-tax revenue (excl. external grants) (+) Central government non-tax revenue (C) External grants (+): External grants for current Central government external grants for current External grants for capital (projects) Central government external grants for capital (projects) (D) Total other non-interest (-): Non-priority recurrent Central government other recurrent Other on staff Non-staff recurrent other Recurrent other on goods and services Other non-staff recurrent other Other non-recurrent : Central government other non-recurrent

57 FY16/17 FY17/18 FY18-19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/26 (E) External debt disbursements (+): External debt disbursements (+) (US$ million) (F) External debt service (-): External interest (-) External interest (-) (US$ million) External debt repayments (-) External debt repayments (-) (US$ million) (G) Net internal financial flows (incl. internal interest) (+): Net internal-debt flow (+): Internal debt disbursements (+) Internal debt repayments (-) Internal interest (-) Discrepancy (+) Source: Estimates and calculations from the projection workbook UgFS.xlsm. Table 32 shows the evolution of social in real terms (US$s per child at 2015 prices and exchange rate) under Scenario 0. Table 32: Projected priority (US$s per child at 2016 prices and exchange rate), FY2016/17 FY2025/26 (base scenario) (A) Total social non-interest FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/ Total education Total health Total social development Total water and environment Social recurrent Recurrent education Central government recurrent education Expenditure on education staff Non-staff recurrent education Recurrent education on goods and services Other non-staff recurrent education Recurrent health Central government recurrent health Expenditure on health staff Non-staff recurrent health

58 Recurrent health on goods and services Other non-staff recurrent health Recurrent social development Central government recurrent social development Recurrent water and environment Central government recurrent water and environment Expenditure on water and environment staff Non-staff recurrent water and environment Recurrent water and environment on goods and services Other non-staff recurrent water and environment FY16/17 FY17/18 FY18/19 FY19/20 FY20/21 FY21/22 FY22/23 FY23/24 FY24/25 FY25/ Non-recurrent education : Central government non-recurrent education Non-recurrent health : Central government non-recurrent health Non-recurrent social development : Central government non-recurrent social development Non-recurrent water and environment Central government non-recurrent water and environment Local government authorities non-recurrent water and environment Source: Estimates and calculations from the projection workbook UgFS.xlsm

59 Projection results alternative scenarios In the base scenario (Scenario 0), the average FY2016/17 FY2025/26 net internal debt flow would be 0.7 per cent of GDP. The total (external and internal) government debt stock would conclude in FY2025/26 at 38.8 per cent of GDP. Scenario 1 considers an increase in (domestic and import) VAT collection efficiency, by changing one line of assumptions from Scenario 0. Under Scenario 1, all assumptions are the same as in Scenario 0, except that domestic VAT collection efficiency rises gradually from 12.6 per cent in FY2016/17 to 16.6 per cent in FY2021/2026, and import VAT collection efficiency rises gradually from 62.1 per cent in FY2016/17 to 82.6 per cent in FY2021/25. With these assumptions, the average FY2016/17 FY2025/26 net internal debt flow would be 0 per cent of GDP (compared with 0.7 per cent of GDP in Scenario 0). The total (external and internal) government debt stock would conclude in FY2025/26 at 33.3 per cent of GDP (compared with 38.8 per cent in Scenario 0). Scenario 2 considers an increase in the elasticities of staff with respect to child population in the education and health sector, from 1 in Scenario 0 to 1.3. Also, the elasticity of total real in social protection and water and environment, with respect to overall population, increases from 1 to 1.3. With these assumptions, the average FY2016/17 FY2025/26 net internal debt flow would be 0.9 per cent of GDP (compared with 0.7 per cent of GDP in Scenario 0). The total (external and internal) government debt stock would conclude in FY2025/26 at 40.7 per cent of GDP (compared with 38.8 per cent in Scenario 0). Scenario 3 combines the assumptions of Scenarios 1 and 2. With higher VAT collection efficiency and the faster growth in priority, the average FY2016/17 FY2025/26 net internal debt flow would be 0.3 per cent of GDP (compared with 0.7 per cent of GDP in Scenario 0). The total (external and internal) government debt stock would conclude in FY2025/26 at 35.2 per cent of GDP (compared with 38.8 per cent in Scenario 0). Scenario 4 combines a yearly increase of external grants of 175 Uganda shillings (approx. US$50 million), starting from FY2018/19. This assumption is combined with an increase in social in health and education equal to twice this amount. The World Bank aid would therefore cover half the increase in social, leaving the remainder to be financed with the government s counterpart funds. Because of the government s counterpart funding, with these assumptions, the average FY2016/17 FY2025/26 net internal debt flow would be 0.9 per cent of GDP (compared with 0.7 per cent of GDP in Scenario 0). The total (external and internal) government debt stock would conclude in FY2025/26 at 41.7 per cent of GDP (compared with 38.8 per cent in Scenario 0). Scenario 5 combines an increase in external grants for recurrent compared with the base scenario, starting from FY2018/19 (similar to Scenario 4). In Scenario 5, however, the counterpart funds would cover only one-third of the increase in the recurrent social, so the World Bank aid would cover two-thirds of the increase. With these assumptions, the average FY2016/17 FY2025/26 net internal debt flow would be 0.9 per cent of GDP (compared with 0.7 per cent of GDP in Scenario 0). The total (external and internal) government debt stock would conclude in FY2025/26 at 41.1 per cent of GDP (compared with 38.8 per cent in Scenario 0). Scenario 6 reverts to the base scenario assumptions, but with a single difference: the growth rate of real GDP gradually rises from 5.3 per cent in FY2016/17 to 8 per cent in FY2025/2026 (compared with 6.5 per cent in Scenario 0). With these assumptions, the average FY2016/17 FY2025/26 net internal debt flow would be 0.4 per cent of GDP (compared with 0.7 per cent of GDP in Scenario 0). The total (external and internal) government debt stock would conclude in FY2025/26 at 34.1 per cent of GDP (compared with 38.8 per cent in Scenario 0). Scenario 7 again reverts to the baseline scenario assumptions, but has the growth rate of real GDP gradually decline from 4.3 per cent in FY2016/17 to 3 per cent in FY2020/21. With these assumptions, the average FY2016/17 FY2025/26 net internal debt flow would be 1.4 per cent of GDP (compared with 0.7 per cent of GDP in Scenario 0). The total (external and internal) government debt stock would conclude in FY2025/26 at 51.8 per cent of GDP (compared with 38.8 per cent in Scenario 0). 49

60 Table 33 presents a summary of the projection results of Scenarios 1 7. Table 34 shows the scenario descriptions. Table 33: Summary scenario results for the projection exercise (% of GDP) Scenario: Average over the projection years: US$ per child at 2015 prices and exchange rate Total social non-interest : Total education Total health Total social development Total water and environment % of GDP General government surplus: General government primary surplus Tax revenue Other revenue External grants Total non-interest (-) Discrepancy General government external and internal interest Total social non-interest : Total education Total health Total social development Total water and environment Total child protection Net financing (gross of interest) Net external financing Net internal financing Final-year general-government debt stock: Final-year general-government external-debt stock Final-year general-government internal-debt stock Source: Estimates and calculations from the projection workbook UgFS.xlsm 50

61 51