South Africa. UNICEF/Bart de Ruigh

South Africa UNICEF/Bart de Ruigh MACRO BUDGET SOUTH AFRICA 2017/2018 1

0.5 % Real average annual rate of decline of provincial government financing over the MTEF UNICEF/Bart de Ruigh Preface This budget brief is one of four that explore the extent to which the national budget and social services sector budgets address the needs of children under 18 years in SOUTH AFRICA. The briefs analyse the size and composition of budget allocations for fiscal year 2017/18, as well as offer insights into the efficiency, effectiveness, equity and adequacy of past spending. Their main objectives are to synthesise complex budget information so that it is easily understood by stakeholders and to present key messages to inform financial decision-making processes. Key Messages and Recommendations Aggregate spending trends and priorities: Lower tax revenues as a result of poor economic growth have reduced the speed at which the South African government is able to finance commitments outlined in the National Development Plan (NDP). Furthermore, departmental baseline spending is being reduced over the Medium Term Expenditure Framework (MTEF). The government is encouraged to: 1. Protect programmes and services that benefit children without compromising service delivery in the neediest communities; 2. Build capacity-enhanced planning, managing, monitoring, and execution in departments that serve children to spend resources more effectively; 3. Avoid using spending performance as the only criterion for reducing departments baseline spending. Decentralisation and financing of subnational governments: Provincial government financing is under pressure and declining at a real average annual rate of 0.5 per cent over the MTEF. Programmes that service children appear to absorb most of the cuts. The government is encouraged to: 1. Prioritise basic education, primary healthcare and social welfare services that support poor and vulnerable children and their families; 2. Encourage provincial governments to build linkages with tertiary educational institutions to maximise poor children s gains from the improved funding of universities; 3. Build capacity in provincial governments to spend all the available resources that have been set aside for the various infrastructure grants; 4. Avoid aggressive cuts to frontline service delivery staff in health, education, social welfare and public works programmes that serve poor and vulnerable communities. UNICEF/Hearfield 2

Sources of financing for the national budget: Public debt is set to rise from 51 per cent of the Gross Domestic Product (GDP) in 2016/17 to 52.4 per cent of GDP in 2019/20. In addition, debt service costs are projected to increase from 3.4 per cent in 2017/18 to 3.6 per cent at the end of the MTEF. Given this scenario, the government is encouraged to: 1. Implement the much-discussed sugar tax and earmark these revenues for programmes and services that benefit children; 2. Continue to pursue cost-cutting measures that do not affect services for children and vulnerable households; 3. Reinforce the commitment to adhere to agreed-to spending ceilings so that fiscal space is created for programmes and services that benefit children. Developments in Public Finance Management (PFM): Implementing a revenue-sharing formula in a country where provinces do not have the same fiscal capacity, or where there is differential capacity within provinces to use resources effectively, makes it difficult to frame allocation criteria that satisfy all relevant stakeholders. Nonetheless, the government is encouraged to: 1. Refine the indicators in the education component of the provincial revenue-sharing formula to reflect the relative poverty of learners; 2. Adopt the principle that funding should follow the learner; 3. Continue the conservative approach for revising the formula to avoid disrupting services for children at the provincial level. UNICEF/Hearfield 3

Section 1. Macro and Socioeconomic Context UNICEF/Bart de Ruigh The performance of the South African economy must be viewed against the backdrop of persistent socio-economic challenges that show no clear signs of being resolved. Table 1 indicates that the real GDP growth is forecast to be 1.0 per cent for the last financial year, while the unemployment rate remains high, at almost 27 per cent of the working-age population. Furthermore, the price of food is increasing at a faster rate than general price increases. Table 1: Macro-economic trends in South Africa: 2015/16 to 2017/18 between the ages of 0 and 17 years and 51 per cent of children are considered to be growing up in poor households. 1 South Africa s Human Development Index (HDI) falls in the medium range and continues to be negatively impacted by stagnant HIV prevalence rates and the poor quality of the basic education system. Table 2: Social development trends in South Africa, 2011 to 2016 2 Key social development indicators Key economic indicators Total population, 2016 55.2 million GDP, 2016/17 ZAR4.4 trillion Total child population, 2016 (0 17 years) 18.6 million GDP per capita, 2016 (ZAR) ZAR79,925 Children as a % of population, 2016 33.6% Real GDP growth, 2016/17 1.0% Unemployment rate, 2016 26.5% Headline inflation, June 2017 5.1% Food inflation, June 2017 6.9% Budget deficit, 2017/18 3.1% Taxation as a % of GDP, 2016/17 26.7% Debt service costs (interests) as a % of GDP, 2017/18 Donor funding as a % of revenue, 2015/16 3.4% 0.8% Poverty rate, 2015 (Lower Bound Poverty Line) Child poverty rate, 2015 (Lower Bound Poverty Line) Human Development Index, 2015 40.0% 51.0% Expected years of schooling, 2015 (male) 12.5 Expected years of schooling, 2015 (female) 13.6 HIV prevalence rates (overall) 12.6% HIV prevalence rates, 15 49 years (women) 21.3% HIV prevalence rates, 15 24 years (youth) 4.8% 0.67 (medium HDI ranking) Source: Budget Review and General Household Survey (GHS) 2016 (own calculations) Life expectancy: male, 2015 55.5 years The pessimism that marks the general economy is mirrored in some of the socio-economic indicators displayed in Table 2. More than a third of the country s population are children Life expectancy: female, 2015 59.5 years Sources: Statistics South Africa 2017 and GHS 2016 (official reports); UNDP 2016 4

The performance of the South African economy can best be described as sluggish as evidenced by the low real GDP growth rate over the last six years. To put economic growth into perspective, the country s National Development Plan (2011) estimated that the economy needed to grow at an average rate of 5 per cent per annum until 2030, to address structural challenges such as unemployment and inequality. Over the entire period represented in the graph below, the economy was unable to breach the 3 per cent growth mark. Given the challenges, it is admirable that the headline inflation rate was kept within the targeted 3 6% band. Tax revenue as a percentage of GDP consistently increases from about 25% of GDP to a forecast 27.2% of GDP at the end of the present MTEF cycle. Figure 1: Performance of the South African economy, 2013/14 to 2019/20 (%) 2013/14 Outcome 2014/15 Outcome 2015/16 Outcome 2016/17 Revised estimate 2017/18 MTEF 2018/19 MTEF 2019/20 MTEF 2.3 1.8 0.6 1.0 1.3 2.1 2.3 5.8 5.6 5.2 6.4 6.3 5.7 5.6 24.8 25.5 26.2 26.0 26.7 27.0 27.2 Tax revenue as a % of GDP Headline Inflation (%) Real GDP growth (%) Percentage 0 5 10 15 20 25 30 Source: Budget Review 2017 The severe drought curtailed the contribution of agriculture to the country s GDP in 2016, but the government has predicted better returns for this sector because of rains returning to some part of the country (Budget Review, 2017). Table 3 shows that the largest contributor to economic growth and job creation was the services sector, as can be seen in the manner in which the growth rates across these sectors mirror the overall GDP growth rate during the past three years. Table 3: Real sector growth trends compared to GDP growth, 2011 to 2016 2011 2012 2013 2014 2015 2016 Agriculture, forestry and fishing 2.0 1.8 3.6 6.9-5.9-7.0 Mining and quarrying -0.7-2.9 4.0-1.4 3.2-4.1 Manufacturing 3.0 2.1 0.8 0.1-0.3 0.7 Electricity and water 1.5-0.4-0.6-1.3-1.0-2.9 Construction 0.4 2.6 4.6 3.6 2.0 1.4 Wholesale and retail trade 4.1 4.0 1.9 1.4 1.4 1.1 Transport and communication 3.5 2.4 2.8 3.1 1.4-0.1 Finance, real estate and business services 4.3 3.0 2.5 2.4 2.8 2.1 Personal services 2.5 2.1 2.2 1.7 1.1 1.1 General government 4.7 3.0 2.9 2.7 0.7 1.7 GDP 3.3 2.2 2.3 1.6 1.3 0.4 Source: Budget Review 2017 5

The main instruments that affect planning in government are the National Development Plan (NDP), which sets out long-term planning goals for various facets of social and economic life in South Africa and the Medium- Term Strategic Framework (MTSF), 3 which serves to domesticate the electoral mandate of the governing party. The most recent MTSF (2014 2019) is the first of its kind to directly engage with the NDP and attempts to ensure greater alignment between the goals of the NDP and the priorities that guide budget planning and implementation. Some of the key MTSF goals that have a direct bearing on the government s budget are: Rapid economic growth and job creation; Rural development, land and agrarian reform and food security; Ensuring access to adequate human settlements and quality basic services; Improving the quality of and expanding access to education and training; Ensuring quality health care and social security for all citizens; Fighting corruption and crime. The Annual Reports of departments include reporting on the performance of most of the adopted indicators. UNICEF/Hearfield Takeaways: The below-par economic growth has reduced the speed at which the South African government is able to finance commitments outlined in the NDP and the MTSF. Given that the low growth scenario is projected to be unchanged over the present MTEF, serious questions are raised about the ability of the government to protect spending on programmes and services that benefit children. A shrinking resources base will require a combination of cost-savings and the strategic prioritisation of spending commitments. Government departments that serve children will be under pressure to spend their existing resources effectively to avoid any changes to their baseline spending plans. 6

Section 2. Aggregate Spending Trends and Priorities UNICEF/Bart de Ruigh Size of Spending Consolidated government expenditure and proposed allocations constitute between 27 28 per cent of GDP over recent and projected years (Figure 2). For instance, government spending as a share of the economy totalled 29 per cent in 2015/16, which was more than a percentage point higher than prior years. However, with planned reductions over departments baseline spending plans, allocations are predicted to constitute less than 28 per cent of GDP at the end of the MTEF in 2019/20. Figure 2: Consolidated government expenditure and allocations as a % of the Gross Domestic Product (GDP), 2013/14 to 2019/20 2013/14 Outcome 27.4 2014/15 Outcome 27.6 2015/16 Outcome 2016/17 Revised estimate 2017/18 MTEF 27.9 28.0 28.9 2018/19 MTEF 2019/20 MTEF Percentage of GDP 26.5 27.0 27.5 28.0 28.5 29.0 27.7 27.8 Source: Budget Review 2017 Spending Changes The gap between nominal and real government spending (and allocations) remains largely unchanged over the same period, suggesting a relatively predictable inflationary environment that should strengthen budget credibility. Consolidated government spending is projected to grow by less than 2 per cent over the present MTEF and has maintained a similar trajectory over the past six years (Figure 3). Given the government s commitment to curtail departmental spending, it is highly likely that the growth in allocations at the end of the MTEF will be much smaller than the projected 2.5 per cent. 7

Figure 3: Annual growth in consolidated government expenditure and allocation, 2013/14 to 2019/20 (2016/17=100) Annual percentage change 12 10 8 6 4 2 0 7.1 1.4 5.4 10.8-1.5 4.7 7.2 7.5 0.8 1.7 8.3 2.5 Nominal Real -2 2014/15 Outcome 2015/16 Outcome 2016/17 Revised estimate 2017/18 MTEF 2018/19 MTEF 2019/20 MTEF Source: Budget Review 2017 Spending and Allocation Priorities in the Consolidated Government Budget Non-interest expenditure (exclusive of debt servicing costs) is projected to grow from R1.2 trillion in 2016/17 to R1.5 trillion at the end of the present MTEF, at a real average annual rate of 1.7 per cent. Figure 4 shows that spending and proposed allocations to the Social Development function takes precedence over the present MTEF as measured in terms of its real average annual growth rate (1.8%), which is slightly above that of consolidated government expenditure (1.7%). The proposed allocations to basic education and health are projected to remain stagnant over the MTEF. Smaller functions, such as the Housing and Amenities function, however, are projected to have real average annual increases in line with the overall government average. It should be abundantly clear that South African departments operate in a restrained spending environment. Given the limited economic growth mentioned earlier, some of these proposed allocations might actually be revised downward. Figure 4: Consolidated government expenditure and allocation by function, 2015/16, 2017/18 and 2019/20 (ZAR billion) South Africa Housing and community amenities Defence Public order and safety Health Social Development Education 45 48 53 135 151 174 122 137 156 158 182 207 152 179 206 202 230 257 1,200 1,323 1,540 2015/16 Outcome 2019/20 MTEF 2017/18 MTEF 0 200 400 600 800 1,000 1,200 1,400 1,600 1,800 Source: Budget Review 2017 and Provincial Estimates of Revenue and Expenditure 2017 The resolve of the South African government to adhere to spending ceilings is demonstrated by consistently delivering spending outcomes in line with those set in budget documents (Figure 5). For example, in 2015/16, estimates that were presented in February 2015 were virtually the same when new estimates were presented in October 2015 via the MTBPS 2015 (0.3% difference) and a full year later in February 2016 via the 2016 Budget Review when the 2015 ceiling only differed by 0.4 per cent. 8

Figure 5: Adherence to expenditure ceilings, 2014/15 and 2015/16 2015 MTBPS 2016 Budget Review 99.5 99.5 99.6 99.7 2014/15 2015/16 2016 MTBPS 99.4 2017 Budget Review 99.4 99.0 99.2 99.4 99.6 99.8 100.0 Source: Budget Review 2017 (own calculations) Recurrent and Capital Spending and Allocations in the Consolidated Government Budget Spending and allocations on compensation remain around 37 per cent of the total budget, which is the largest expense. The OECD (2017) 4 estimates that wages constitute about 14 per cent of South Africa s GDP compared to other middle-income countries where the average share of wages is between 4 and 12 per cent. Transfers to households come in at around 18 per cent of total government spending, and the bulk of these transfers is on social grants. Given the centrality of spending on tertiary institutions, transfers to universities alone consumed 2.4 per cent of total government resources in 2017/18, while in 2019/20, that share is projected to grow to 2.5 per cent. Spending pressures are felt in the acquisition of capital assets, whose shares of total government expenditure dropped from 4.6 per cent in 2013/14 to 4.1 per cent in 2017/18 and fell to only 3.7 per cent at the end of the present MTEF. Figure 6 represents the spending intentions of government as it tries to protect the real value of social grants, as well as extending coverage and responding to the financing pressures that emanated in the university sector. Figure 6: Expenditure and allocation by type in the consolidated government budget, 2013/14 to 2019/20 (%) Percentage of consolidated expenditure/allocation 40 35 30 25 20 15 10 5 0 37.3 17.5 2.3 2.4 4.6 2013/14 Outcome 37.3 17.5 2.3 2014/15 Outcome 4.6 2.4 36.3 17.3 2.2 4.8 2.2 2015/16 Outcome 37.5 18.2 2.3 4.1 2.3 2016/17 Revised estimate 37.6 18.2 2.4 2017/18 MTEF 4.1 2.3 3.7 18.1 2.5 2018/19 MTEF 3.9 2.3 37.0 17.9 2.5 2.2 3.7 2019/20 MTEF Compensation Transfer to households Transfers to universities Transfers to NPOs Payment for capital assets Source: Budget Review 2017 Takeaways: A critical feature of the present MTEF is the downward revisions to departmental baselines. The information presented thus far suggests that further spending cuts are now inevitable. Given the government s commitment to maintain the real value of key social grants, the Social Development function performs better than Education in attracting additional resources. In addition to elevated social development spending, transfers to universities and spending on wages of government employees absorb whatever fiscal space is available. Overall, most departments are operating in a constrained spending environment and unless the economy starts growing again, sharp intra-departmental and inter-sectoral trade-offs will become the norm over the next few years. There is evidence that the spending proposals for the 2017 MTEF are virtually cast in stone and that the MTEF represents the government s actual spending stance and commitments. 9

Section 3. Decentralisation and Subnational Spending UNICEF/Hearfield Decentralisation Context and Subnational Funding Guidelines The Constitution of the Republic of South Africa grants equal status to the three spheres of government in South Africa and promotes an overall principle of interdependent governance (Constitution of Republic of South Africa, 1997). Furthermore, the Constitution (Chapter 6) spells out the legislative authority of provinces via the provincial legislatures and Schedule 4 provides for functional areas of concurrent national and provincial legislative competence. Table 4: Distributing the equitable shares by provinces, 2017 In terms of the financing of provincial governments, Section 214 (a-c) of the Constitution 5 makes provision for three different sources, namely a vertical division of revenue (among the three spheres of government), a horizontal division of revenue (using a formula to divide resources among provinces), and through special allocations (conditional grants that have strict conditions for their use). The vertical division of revenue is decided politically, while the horizontal process relies on a funding formula based on relative need (demographic data). The provincial equitable shares (PES) formula contains six components with different weights. Education (48%) Health (27%) Basic share (16%) Poverty (3%) Economic activity (1.0%) Institutional (5%) Weighted average Eastern Cape 15.1% 13.5% 12.6% 16.3% 7.6% 11.1% 14.0% Free State 5.3% 5.3% 5.1% 5.2% 5.0% 11.1% 5.6% Gauteng 18.0% 21.8% 24.1% 17.3% 34.3% 11.1% 19.8% KwaZulu-Natal 22.3% 21.7% 19.8% 22.2% 16.1% 11.1% 21.1% Limpopo 13.0% 10.3% 10.4% 13.6% 7.1% 11.1% 11.7% Mpumalanga 8.4% 7.3% 7.7% 9.1% 7.5% 11.1% 8.1% Northern Cape 2.3% 2.1% 2.1% 2.2% 2.1% 11.1% 2.7% North West 6.5% 6.7% 6.8% 8.0% 6.5% 11.1% 6.9% Western Cape 9.1% 11.3% 11.3% 6.1% 13.6% 11.1% 10.1% Total 100% 100% 100% 100% 100% 100% 100% Source: Budget Review 2017 Note: Green represents the highest value, while red represents the lowest value for each component and the total weighted share by province. 10

Box 1 Let s imagine how this formula works in practice. If R100 billion was available to be spent on provinces, then the Eastern Cape provincial government should receive 14% of R100 billion (or R14 billion). Almost half of this money (48%) would have been decided using the education component. That means of the R48 billion decided through the education component, the Eastern Cape government would have received 15.1% of R48 billion (or R7.2 billion). For the health component, the Eastern Cape is entitled to 13.5% of the R27 billion (or R3.6 billion). The formula assumes that the cost of running government is the same for all provinces and the Eastern Cape will receive 11.1% of R11.1 billion (or R1.2 billion). The same calculation is made for all the components; these are tallied and the final shares will be paid over in quarterly instalments during the financial year. Provincial Spending and Allocation Trends, 2015/16 to 2019/20 Provincial spending grew from R487 billion in 2015/16 to an expected R618 billion in 2019/20, which amounts to a real average annual growth rate of -0.5 per cent (Figure 7). This is generally not good news for children, as some of the key services for children are funded and delivered at the provincial level. The provincial annual growth curves are bunched together, with the exception of the present financial year, where the extent of funding decline is much larger for the Gauteng, Limpopo and Northern Cape provincial governments. In 2017/18, consolidated provincial spending is planned to decline by almost 2 per cent, while the next financial year (2018/19) merely moderates the extent of the decline in financing provincial governments (-0.2%). Figure 7: Real annual growth rates in subnational expenditure and allocations, 2015/16 to 2019/20 8 Eastern Cape Real annual change (percentage) 6 4 2 0-2 -4-6 2016/17 2017/18 2018/19 2019/20 Free State Gauteng KwaZulu-NaTal Limpopo Mapumalanga Northern Cape North West Western Cape South Africa Source: Provincial Revenue and Estimates of Expenditure 2017 Spending Disparities by Provincial Government The Limpopo and KwaZulu-Natal provinces have a smaller share of total provincial spending than would be predicted by their weighted equitable shares, suggesting lower capacity to leverage own funding than other provinces. For example, KwaZulu-Natal receives 21.1 per cent of national transfers through the equitable share formula, but its share of total provincial allocations (equitable shares plus own revenue) is only 20.9 per cent, while the corresponding numbers for Limpopo are 11.7 per cent (weighted shares) and 11.1 per cent (equitable shares plus own revenue) (Figure 8). The Gauteng and Mpumalanga provinces have more or less the same weighted and total provincial allocation shares, while total allocations for the remaining provinces are higher than their corresponding shares of the transfers made available by the national government to provinces. 11

Figure 8: Comparing actual allocations by provincial governments to their weighted shares of national transfers in 2017/18 (%) Western Cape North West Northern Cape Mpumalanga Limpopo KwaZulu-Natal Gauteng Free State Eastern Cape 2.7 2.9 6.9 7.1 8.1 8.1 5.6 6.0 10.1 10.8 11.7 11.1 14.0 13.5 21.1 20.9 19.8 19.7 PES, 2017/18 2017/18 MTEF Percentage 0 5 10 15 20 25 Source: Provincial Revenue and Estimates of Expenditure 2017 and Budget Review 2017 Adjusting total provincial spending for the child population, the Northern Cape (smallest population overall) has the largest spending per child over these two years, while Limpopo and the Eastern Cape have the lowest per-child spending in 2015/16 and 2016/17 respectively. In 2015/16, the national per-child spending was R26,000 and three provinces, namely the Eastern Cape, Kwazulu-Natal and Limpopo spent below the national average. In 2016/17, the national per-child spending was R28,000 and the same three provinces, with the addition of Mpumalanga, spent below the national average. Figure 9: Per-child spending by provincial governments in 2015/16 and 2016/17 (ZAR) Per-child spending by province (ZAR) 40,000 35,000 30,000 25,000 20,000 15,000 10,000 5,000 0 24,651 26,283 Eastern Cape 33,231 36,205 Free State 26,205 Gauteng 29,383 25,684 27,049 KwaZulu-Natal 24,091 27,062 Limpopo 25,225 Mpumalanga 27,483 2015/16 Outcome 2016/17 Revised estimate 35,295 Northern Cape 38,663 26,534 29,018 North West 27,186 29,296 Western Cape 26,195 28,495 South Africa Source: Provincial Revenue and GHS 2015 and 2016 (own calculations) Note: Children between the ages of 0 and 17 years. Takeaways: Provincial government expenditure is under severe strain and is only projected to recover at the end of the present MTEF. Provincial spending and allocations grew from R487 billion in 2015/16 to R618 billion in 2019/20 at a real average annual rate of -0.5 per cent. The largest adjustment in allocations is made in the present financial year (2017/18) where provincial budgets lose close to 2 per cent in real value, while the next financial year merely moderates the real losses in allocations. Two of the traditionally poor provinces, KwaZulu-Natal and Limpopo, have a lower share of total provincial allocations in 2017/18, which is at odds with their higher overall share of national transfers. When total provincial spending is adjusted for the size of the child population, the Northern Cape has the largest per-child spending in 2015/16 and 2016/17, while the Eastern Cape, KwaZulu-Natal, and Limpopo spent below the national averages in both years. 12

Section 4. Financing the National Budget Domestic Revenue To off-set the negative impact of lower economic growth on the country s revenue position, tax revenue (mostly personal income tax) is set to rise from 26.2 per cent in 2015/16 to 27.2 per cent of GDP at the end of the present MTEF (Figure 10). Consolidated domestic revenue is projected to achieve a surplus over consolidated non-interest expenditure in 2017/18 and 2019/20, while the country s gross borrowing UNICEF/Pawelczyk requirement increases from 15.8 per cent in 2017/18 to 17.7 per cent at the end of the present MTEF. Furthermore, debt service costs, as a percentage of the GDP, are projected to increase from 3.4 per cent in 2017/18 to 3.6 per cent in 2019/20. The government argues that a combination of consistent reductions of departmental baseline spending and higher taxes will narrow the consolidated budget deficit from 3.4 per cent of GDP in 2016/17 to 2.6 per cent in 2019/20. Figure 10: Government expenditure financing, 2015/16 to 2019/20 Government expenditure financing, 2015/16 to 2019/20 (percentage) 100 90 80 70 60 50 40 30 20 10 0 27.8 28.4 29.9 29.4 29.8 Domestic revenue as a % of GDP 26.2 26.0 26.7 27.0 27.2 Tax revenue as a % of GDP 3.2 3.3 3.4 3.5 3.6 Debt service costs as a % of GDP 99.5 100.4 101.5 103.1 103.8 Domestic revenue as a % of total expenditure 16.3 18.8 15.8 15.0 17.7 Gross borrowing requirement as a % of total expenditure 2015/16 Outcome 2016/17 Revised estimate 2017/18 MTEF 2018/19 MTEF 2019/20 MTEF Source: Budget Review 2017 Note: Percentages are indicated for 2017/18 and 2019/20 only. Borrowing Total government debt is set to increase from 49.4 per cent in 2015/16 to 52.4 per cent in 2019/20 and the OECD latest review notes that the country is fast approaching the upper limit of 40 55% debt-to-gdp ratio considered prudent in developing country contexts (Figure 11). Around 10 per cent of the government s debt portfolio consists of international loans, while long-term domestic debt makes up 78 per cent of total government debt. 13

Figure 11: Public debt as a percentage of GDP, 2015/16 to 2019/20 54 53 52.3 52.9 52.4 Percentage of GDP 52 51 50 49 49.4 50.7 48 47 2015/16 2016/17 2017/18 2018/19 2019/20 Source: Budget Review 2017 The costs of servicing (the country s rising) debt increased from R129 billion (or 3.2% of GDP) in 2015/16 to more than R197 billion (or 3.6% of GDP) at the end of the present MTEF. As can be seen from Figure 12, the largest debt servicing costs are on domestic loans, while a much smaller part of the costs involve servicing interests on international loans. UNICEF/Schermbrucker Figure 12: Rising costs of servicing public debt, 2015/16 to 2019/20 (ZAR billion) 2019/20 MTEF 18 180 197 2018/19 MTEF 16 164 181 2017/18 MTEF 14 148 162 2016/17 Revised estimate 11 135 146 2015/16 Outcome 10 118 129 ZAR billion 0 20 40 60 80 100 120 140 160 180 200 Total Foreign loans Domestic loans Source: Budget Review 2017 14

Additional Financing Options A UNICEF-commissioned study into fiscal space in South Africa 6 has highlighted several issues. Some of the salient findings include: (i) Levying taxes on the consumption of luxury goods and introducing the much-discussed sugar tax could lead to a sizeable amount of new revenue that could be used to finance programmes and services that benefit children; (ii) Building on the ongoing efforts to achieve cost-savings, such as reducing unnecessary travel, curbing the use of external consultants and generally decreasing spending on non-priority areas that have a poor spending record. At the same time, it is unrealistic to expect all additional savings to be channelled to priority expenditures for children because of the government s other stated priorities. These include: funding higher education; contributing to the New BRICS Development Bank; increasing the contingency reserve fund for small business development; strengthening government s planning, monitoring and evaluation capacity; and decreasing the country s debt levels. At a minimum, the fiscal space study notes that priority sectors (for children) need to show that they are managing their funds well and maximizing the effectiveness of their current budgets, which is the starting point for receiving more resources. UNICEF/Pirozzi Takeaways: In order to fill the void left by lower revenues, the government is set to increase taxes (especially personal income tax) to close the revenue gap. Tax revenue (mostly personal income tax) is set to rise from 26.2 per cent in 2015/16 to 27.2 per cent of GDP at the end of the present MTEF. The fact that domestic revenues are still able to finance non-interest expenditure is encouraging and this in itself provides a powerful incentive for government to rein in wastefulness, focus on priority spending for children and improve the quality of spending. Public debt is set to rise from 51 per cent of GDP in 2016/17 to 52.4 per cent in 2019/20, which is indicative of the financial pressures the government is under to deliver vital services. With an increase in public debt, the cost of servicing government debt is expected to increase from 3.2 per cent in 2015/16 to 3.6 per cent in 2019/20. 15

Section 5. New Developments in PFM UNICEF/Hearfield PFM Challenges in South Africa The main challenge is that during a time when targeted spending on the poor should be accelerated, the economy is underperforming, thus depriving the budget of muchneeded resources to tackle the country s formidable social and economic problems. This is best summed up when examining the reductions to the baseline budgets of all three spheres of government. Table 5 shows that provincial governments are required to surrender almost R1.8 billion in 2017/18, of which the bulk of the reductions are taken from the various conditional grants. A much smaller amount (between R500 and R550 million) will be deducted from the unconditional block grant, which to some extent preserves the small fiscal space that provinces have at their disposal. One can surmise that variable spending rates on the conditional grants might have prompted a more conservative implementation of conditional grant funding. Table 5: Baseline reductions by sphere of government, 2017/18 to 2019/20 (ZAR million) 2017/18 2018/19 2019/20 National government 3,910 2,297 2,770 Compensation of employees 437 471 497 Goods and services 649 667 787 Transfers to public entities 2,850 1,240 1,539 Other national spending items 411 390 444 Provincial government 1,757 1,882 1,955 Provincial equitable share 500 529 558 Provincial conditional grants 1,257 1,353 1,397 Local government 791 813 837 The Challenge of Providing Resources to Provincial Governments on an Equitable Footing Because more than 80 per cent of provincial governments revenue are allocated through the equitable share formula, it is imperative to understand whether this formula can be improved to better serve the needs of the most marginalised people in South Africa. The introduction of the Social Security Agency (SASSA) removed the need to retain the social development component, while infrastructure backlogs are now funded through a conditional grant. Weights for the remaining components were adjusted to reflect expenditure patterns at the provincial level. Table 6: Comparing the original and existing PES formula (weights are in brackets) Old PES formula (introduced in 1996) PES formula in use (introduced since 2004) Education (41%) Education (48%) Health (19%) Health (27%) Social development (18%) Component was removed Economic activity (7%) Economic activity (1%) Backlog (3%) Component was removed Basic share (7%) Basic share (16%) Institutional (5%) Institutional (5%) Poverty (3%) Local government conditional grants 791 813 837 Source: Budget Review 2017 16

UNICEF/Hearfield Implementing a revenue-sharing allocation formula in a country where provinces do not have the same fiscal capacity or where there is differential capacity within provinces to use resources in the best possible way, makes it difficult to frame allocation criteria that satisfy all relevant stakeholders. Rao and Khumalo (2004) 8 argued that any allocation formula needs to consider cost factors that are beyond the immediate control of provincial governments. The Financial and Fiscal Commission (FFC) (2000) took this process one step further when it called for a costed norms approach, which requires the government to define an output standard for a service (65% pass rate in school-leaving examinations) and then estimate the resources needed to achieve that goal. 9 The intention was to apply this approach to all services that are constitutionally required to be delivered by provincial governments. The FFC (2014 and 2017) 10 made further submissions on the allocation formula and argued that if the formula were to be revised in the long term, a conditional block grant for education and health should be provided to ensure that provinces meet their constitutional obligations with regard to these key services. Equal Education (2017), 11 one of the leading civil society advocacy groups, makes a strong case for the incorporation of an explicit rural and poverty element in the education formula of the provincial equitable share. Using Statistics South Africa s demarcation of geographic types, Equal Education computes population weighted schoolage distributions that account for rurality and consequently allocates substantially more resources to rural provinces (Eastern Cape, KwaZulu-Natal and Limpopo) and takes away resources from the more urban provinces (Gauteng and the Western Cape). It is understandable that the National Treasury has taken a cautious approach to reviewing parts of the allocation formula because of the disruptive effect this may have on provincial revenue. This approach is likely to be continued and we should not expect large swings in provincial revenue either way. However, what is clear is that the allocation formula needs to change so that it provides an improved understanding of what it means to deliver services to poor and vulnerable communities. The education formula needs to find a balance between recognising that rural areas have implicit cost factors that increase financing costs, while recognising that urbanisation trends are irreversible and urban poverty is as real as rural poverty. Takeaways: Reductions in the baseline funding for provinces that were motivated by poor spending at the subnational level should be complemented with capacity-building interventions aimed at improving the effectiveness of spending. Furthermore, reductions should not be based only on poor spending performance, but constitutional requirements to deliver basic services to poor and vulnerable individuals should be factored into such decisions. Mooted changes to the education component in the allocation formula should consider poverty and other factors of need, but changes need to be negotiated with all relevant stakeholders. 17

Key Events in the Budget Calendar National budget process July 2016: National agencies submit their first draft budgets to the National Treasury September 2016: MTEC process concludes: Recommendations tabled to MINCOMBUD October 2016: Tabling of Medium Term Budget Policy Statement October 2016: Preliminary allocation letters issued to departments November 2016: Submission of draft 2017 national budget chapter and database by departments/entities November 2016: Cabinet approved final allocations distributed to departments February 2017: Budget tabled in Parliament. Provincial budget process July 2016: Technical Committee on Finance Lekgotla August 2016: Provincial treasuries submit first draft 2017 Budgets to National Treasury: Estimates of Provincial Revenue and Expenditure and database September 2016: Budget Council and Budget Forum meeting (inter-governmental meeting) October 2016: Tabling of Medium Term Budget Policy Statement October 2016: Preliminary allocation letters issued to provinces equitable share and conditional grant allocations December 2016: Final conditional grant frameworks and allocations submitted to National Treasury by national departments End Jan/Early Final allocation letters issued to provinces Feb 2017: End Feb/Early Provincial 2017 Budgets tabled at provincial legislatures. March 2017: 18

Endnotes 1 The poverty measure used to determine that 51 per cent of children are poor is the lower bound poverty line (which is ZAR647 per person per month). The number would be much higher if the upper bound poverty line (ZAR992 per person per month) is used. 2 Data for the textbox were drawn from the United Nations Development Programme (UNDP), Human Development Report 2016: Human Development for Everyone. New York, UNDP, 2016 <http://hdr.undp.org/sites/default/files/2016_human_development_report.pdf> [accessed 15 August 2017]. Poverty data were drawn from Statistics South Africa, Poverty trends in South Africa: an examination of absolute poverty between 2006 and 2015. Pretoria, Government Printers, 2017. Population data were drawn from the official reports of the General Household Survey 2016. HIV prevalence rates were drawn from Statistics South Africa, Mid-year population estimates 2017. Pretoria, Government Printers, 2017. 3 The MTSF report was downloaded from: <http://www.gov. za/sites/www.gov.za/files/mtsf_2014-2019.pdf> [accessed 18 July 2017]; The Presidency, The National Development Plan 2030: Our future make it work. Pretoria, Government Printers, 2011. 4 OECD, OECD Economic Surveys South Africa. Paris, OCED Publishing, 2017. 5 Legally, these constitutional provisions find expression in the annual Division of Revenue Bill/Act that accompanies the main budget documentation at the start of the new financial cycle. The Division of Revenue Bill/Act is particularly helpful in listing in detail the conditions under which (conditional) grant funding should be used, including reporting requirements and the duration or life-span of the grant. 6 UNICEF, National political economy analysis and fiscal space profiles of countries in Eastern and Southern Africa: fiscal space analysis-south Africa. Rotterdam, UNICEF, 2017. 7 Rao, MG and Khumalo, B. Sharing the Cake: A Review of the Provincial Equitable Share Formula in South Africa: <06.2%20 Rao%20G_Khumalo%20B_Sharing%20the%20cake%20 (32pp)%20(1).pdf> [accessed 09 August 2017]. 8 The Financial and Fiscal Commission (FFC). A costed norms approach for the Division of Revenue. Midrand, FFC. 9 The Financial and Fiscal Commission (FFC), Submission for the 2018/19 Division of Revenue. Midrand, FFC and Submission for the 2010/11 Division of Revenue. 10 Equal Education (EE). Adjusting the equitable share formula to improve opportunities for equal education across rural and urban areas. Cape Town, Equal Education. UNICEF/Schermbrucker 19

United Nations Children s Fund Equity House 659 Pienaar Street Brooklyn Pretoria 0181 www.unicef.org/southafrica