The Impact of the Financial Crisis on Fiscal Space for Education Expenditure in Africa

Transcription

1 2010/ED/EFA/MRT/PI/43 Background paper prepared for the Education for All Global Monitoring Report 2010 Reaching the marginalized The Impact of the Financial Crisis on Fiscal Space for Education Expenditure in Africa Matthew Martin & Katerina Kyrili 2009 This paper was commissioned by the Education for All Global Monitoring Report as background information to assist in drafting the 2010 report. It has not been edited by the team. The views and opinions expressed in this paper are those of the author(s) and should not be attributed to the EFA Global Monitoring Report or to UNESCO. The papers can be cited with the following reference: Paper commissioned for the EFA Global Monitoring Report 2010, Reaching the marginalized For further information, please contact

2 THE IMPACT OF THE FINANCIAL CRISIS ON FISCAL SPACE FOR EDUCATION EXPENDITURE IN AFRICA Background Paper for Education for All report 10 July 2009 Matthew Martin Katerina Kyrili Development Finance International 1

3 1. INTRODUCTION This paper has been commissioned by UNESCO to examine the impact of the global financial crisis on fiscal space for (particularly primary) education expenditure in lowincome Sub-Saharan African countries. The achievement of the Millennium Development Goals is the focus of international development policy through to Among these, primary Education For All is a key goal for governments and donors, on which much progress has been made since However, for the EFA goal to be achieved, progress in developing countries (especially in Sub-Saharan Africa) needs to be accelerated in the coming years. The aim of this report is to assess the degree to which this progress - and prospects for future progress - have been undermined by the current global financial crisis. 1 The macro-economic impact of the crisis on Sub-Saharan Africa is by now clear to all (Committee of 10, 2009; IMF REO, 2009). The direct impact has been a reduction in private financial inflows (foreign direct investment, bank loans, portfolio flows and international bonds): many low-income African countries including Nigeria, Ghana, Kenya, Mozambique, Tanzania, Uganda and Zambia have seen a sharp reduction in access to market financing. The indirect impact has been via lower global growth, reducing demand for African exports and commodity prices, remittances from Africans overseas, and prospects for aid flows. All of these have in turn sharply reduced African per capita growth (from an average of 3% in 2008 to -0.6% in 2009) and government budget revenues (from 24.7% of GDP to 20.8%) (IMF REO, 2009). This is particularly disappointing given that in recent years, African LICs growth and progress to the MDGs had been accelerating sharply. The Managing Director of the IMF, Dominique Strauss-Kahn, has recently referred to Africa as an innocent victim of the financial tsunami (IMF, 2009). However, the detailed impact of the crisis on SSA budgets and therefore on education expenditure has not so far been examined. In particular: while the G20 and the IMF have called for country governments to use fiscal stimulus to fight the crisis if they have the fiscal space to undertake stimulus, it is not clear how many African LICs have such space. while the international community and education campaigners have urged African governments not to cut education (or other MDG) spending during the crisis - and indeed to see education spending as part of the solution because it can enhance skills, reduce unemployment, and protect vulnerable households - it is not clear whether budget revenue falls or reduced aid might be resulting in spending cuts. This paper aims to investigate the degree of fiscal space open to African LICs to help them avoid cutting education expenditure, and the consequences of the crisis for education expenditure. It is structured as follows: Chapter 2 of this study defines Fiscal Space; Chapter 3 presents a proposed Fiscal Space Index; Chapter 4 presents country coverage and data sources; Chapter 5 the results of assessment of fiscal space for African LICs; Chapter 6 educational expenditure trends in African LICs budgets and Chapter 7 conclusions. 1 This study does not cover the impact of the food and fuel price crisis, for which an examination of 2008 budgets and outturns would be more appropriate. 2

4 2. DEFINING FISCAL SPACE Fiscal space is a relatively longstanding concept, which has more recently been widely discussed in the international community. A search of the literature reveals three main definitions of the concept: the IMF defines it as the room in a government s budget that allows it to provide resources for a desired purpose without jeopardizing the sustainability of its financial position or the stability of the economy (IMF, 2005). Recent IMF reports have underlined that only countries which have reached relatively stable fiscal and macroeconomic positions can be regarded as having sufficient fiscal space to increase spending to confront the crisis. This definition appears to give primacy to financial and economic stability rather than MDG spending, as well as to a relatively static definition of fiscal space based on a set of current indicators or short-term forecasts, and therefore has been criticised by civil society. international civil society organisations define fiscal space as the amount of room a country has to increase spending (on social, infrastructure and productive sectors) to accelerate progress towards the MDGs (ActionAid, 2007). They argue for governments to focus on mobilising development financing to give them greater space for expenditure, especially if they are way short of spending levels needed to attain the MDGs, and even if the IMF s conditions for macroeconomic stability are not entirely met. They argue that long-term economic stability depends on growth and utilising capital and labour to their highest potential. This definition appears to give primacy to MDG spending rather than economic stability, and therefore has been criticised by the IMF. the UNDP definition of fiscal space is a more neutral one which takes account of both the MDG and sustainability perspectives: Fiscal space is the financing that is available to government as a result of concrete policy actions for enhancing resource mobilization, and the reforms necessary to secure the enabling governance, institutional and economic environment for these policy actions to be effective, for a specified set of development objectives. It is therefore seen as vital to look at the dynamic long-term impact of fiscal space on growth and development in order to assess the sustainability of spending (Rathin et al, 2007). The common aspects of all definitions are: the ability of a government to generate resources to finance expenditure, by mobilising domestic revenue, external grants or external or domestic borrowing. that government should use any space to finance projects essential for their development, and where necessary cut less essential expenditures and improve the efficiency of expenditures. The differences revolve around the relative weight given to financial and economic stability as opposed to MDG goals, and how to define essential projects. However, there is increasing consensus that essential projects should be defined not just by economic rate of return, but also by whether they produce short- or long-term social and economic benefits. For example, investment in (especially primary) education is associated with an increase in the productivity, health and general welfare of pupils and their communities, through higher returns to human capital, and is therefore generally regarded as critical (Barro, 1991; Benhabib and Spiegel, 1994; Pscacharapoulos and Patrinos, 2004). 3

5 3. A FISCAL SPACE INDEX How should a country s fiscal space be assessed? The Bretton Woods Institutions and UNDP have since 2006 used a fiscal space diamond. (Development Committee, 2006; Rathin et al, 2007) This analyses and expresses visually the potential increase in four types of fiscal resources ( pillars) as percentages of GDP: i) external borrowing; ii) domestic resource mobilisation; iii) deficit financing; and iv) reprioritising expenditure or increasing its efficiency. It thereby identifies the potential of a government for resource mobilisation without endangering macroeconomic stability. However, this diamond is based on the assumption that analysis needs to take place country-by-country to assess potential for mobilising each type of resources, and therefore does not set any benchmarks for what would constitute reasonable maximum levels of resource mobilisation. As expressed by the UNDP, it is essential when using this instrument to define precisely the policy assumptions underlying the diamond, the time frame within which the different measures take effect and whether the policy actions that could be taken to tap into a source of fiscal space are endogenous or exogenous to domestic policy making (Rathin et al, 2007). This study takes a slightly different perspective, and goes a step further. It aims to assess overall how much potential fiscal space is available to for low income African countries, by defining maximum desirable levels for objective and comparable indicators relating to three of the pillars (borrowing, revenue mobilisation, and budget deficits). 2 By combining indicators for the three pillars, it then creates a fiscal space index, which assesses the space which governments have to mobilise resources, in order to fund additional expenditures to attain the MDGs The Three Fiscal Space Indicators (and Additional Check Indicators) The first step is to define thresholds for the three indicators comprising the fiscal space index, beyond which countries are likely to have very low or no fiscal space to mobilise additional resources. This has been achieved as follows: 1) Sustainable Borrowing Levels The first way that a country could finance its expenditure is by borrowing externally or domestically. Low-income African countries have relied extensively on debt financing for their development since the 1960s. However, excessive borrowing would risk creating unsustainable debt ie a debt overhang or debt service burden which could undermine growth, as well as (in the case of domestic borrowing) increasing inflation. To assess the maximum level of sustainable borrowing, the index uses internationally accepted indicators for debt distress, which comprise: 2 This study does not define thresholds or indicators for the fourth pillar of expenditure rationalisation, as these have not been defined clearly in any literature, and because we assume that virtually all African LICs need to spend much more on education to reach the MDGs. Further work could include examining indicators used to measure the quality and pro-poor focus of public expenditure to judge relative space for expenditure rationalisation. 4

6 for external debt, (column 4 in Table A) the country s classification according to the BWIs Low Income Countries Debt Sustainability Framework (LIC-DSF) which is in turn based on a series of indicators comparing the present value of debt, and debt service, to the country s GDP, exports and budget revenue. Any country with high risk of debt distress, or which is already in debt distress, is assumed not to have scope for external borrowing. for domestic debt, (column 3 in Table A) the IMF definition of nominal domestic debt stock as unsustainable if it exceeds 15% of GDP (Commonwealth Secretariat, 2009; IMF, October 2008). Any country whose level of domestic debt exceeds 15% of GDP is assumed not to have scope for domestic borrowing. The space for overall government borrowing is then assessed by creating a composite Debt Distress indicator (presented in column 10 of the index table A), whereby a country with either external or domestic debt distress is not assumed to have further scope for government borrowing (value of 1) and those without distress have fiscal space to borrow additionally (value of 0). 3 2) Sustainable Domestic Revenue Levels The second way for a government to generate resources to fund expenditures, is to increase domestic revenue mobilisation. Low-income African countries have made major strides in increasing their revenue levels in recent years (see UNECA 2007; Commission for Africa, 2005), by increasing (especially indirect) taxes, widening the tax base to cover for example land taxes, and improving the efficiency of tax collection through establishing independent revenue authorities and other measures (see also AERC, 2004). However, it is also generally accepted that excessive rises in tax levels, especially for the limited number of individuals and enterprises with sufficient income to be able to pay tax in low-income countries, can be inimical to long-term growth prospects by undermining incentives for earnings and profitability. There is a considerable debate as to what constitutes a sustainable level of tax revenue for low-income African countries, and countries which are endowed with large-scale easily taxable mineral resources may be able to increase taxes to higher levels. However, the bulk of African low-income countries are not so fortunate, and for this reason we use the indicator of acceptable levels of revenue effort accepted as the economic convergence criterion for the CFA Franc Zone. This is that revenue (excluding grants) as percentage of GDP should reach 17%.(UEMOA, 2008). The data for this indicator are presented in column 6 of Table A. 3) Sustainable Aid Levels The third way to finance additional expenditures in low-income African countries would be to mobilise more grants. While aid flows to Africa have increased substantially in recent years, they are still falling considerably short of earlier promises such as those of 2005, including in the field of education (see ONE, 2009; UNECA 2009). If donors live up to their earlier promises, there should be considerable space to increase grant flows for education in order to combat the crisis. 3 This has been done because it is assumed that in a situation of either external or domestic debt distress, additional borrowing is not desirable as the focus should be on reducing the debt burden. However, an alternative method could be to assess the indicators separately and assess a country s space to borrow either externally or domestically. 5

7 However, various studies indicate that excessive aid dependence can be inimical to development, and therefore it is important to set a threshold for aid beyond which countries might seen as excessively aid dependent. The most recent studies have suggested that only aid which actually reaches a recipient country should be included (such as the OECD definition of Country Programmable Aid or CPA), and that this level should be set around 25% of GNI (Foster and Keith, 2003). We therefore assume that countries whose CPA exceeds 25% of GNI have no space to increase their aid dependence. The data are presented in Column 7 and 14 of Table A. 4) Additional Checks Even if a country has scope to mobilise more financing, the IMF and others might argue that it should not if this would provoke macroeconomic instability, through either an excessive fiscal deficit (which could be inflationary) or a high preceding level of inflation. To judge whether either of these factors might constrain the overall fiscal space available to governments, we have assessed two additional indicators: Fiscal balance: the first check indicator used is the fiscal balance (including grants) as percentage of the GDP. According to earlier studies by the IMF and Bevan/Adam, which were examining how to define mature stabilisers, a country s fiscal deficit including grants should be below 3% of GDP to demonstrate that they have a fiscal deficit which is sustainable (FONDAD, 2005). Countries with a larger deficit cannot increase it further to mobilise loan resources. Data for this indicator are in Table A, column 8. Inflation: the second check indicator used is the level of inflation (CPI). According to the IMF, the level of inflation which is optimal for growth and assists in defining mature stabilisers is under 10% (for further references to IMF literature see FONDAD, 2005) 4. Countries with an inflation rate above 10% cannot conduct fiscal stimulus without risking provoking an inflationary spiral. Data for this indicator are column 9 of Table A Combining the Indicators Into An Index To assess overall national fiscal space, we then aggregate these indicators into a fiscal space index. The index consists of 3 indicators, tracking the three main instruments the government can use to generate resources: debt, domestic revenue and grants. For each country, the ability of the government to use each specific instrument is examined, depending on whether the indicator is below its threshold. Countries which are not constrained in relation to any source of resources are described as having high fiscal space; those constrained on one indicator have moderate space; those constrained on two have low fiscal space; and those constrained on all three have no fiscal space. To facilitate the aggregation, the indicator is given the value 1 if its 4 A more recent IMF publication has defined the mature stabiliser level for inflation as below 12.5% and that for the budget deficit as 5.5% of GDP (IMF, March 2009). However, these are not used here, as Fund and African LIC sources indicate that the levels cited in the text above are more likely to be used by the Fund to identify countries which should have fiscal space. 6

8 level is beyond the threshold, and 0 if it is below. Therefore countries with high fiscal space have a score of 0, and those with no space have 3. 5 A second assessment is conducted using the check indicators of fiscal balance and inflation. Here we assume a country could not borrow additionally if its fiscal deficit exceeds 3% including grants but it could still use additional grants or revenue (which would leave the deficit unchanged). It also could not use additional grants as these might push up inflation; however, it could increase revenue as these would make spending inflation-neutral. So we add 1 to a country s score if its debt space is eliminated by the deficit, and 1 to its score if its grants space is eliminated by inflation risk. In this assessment, the scores are analysed the same way as in the first. 5 Further work could aim to track all different space indicators as annual budget flow variables (debt service, revenue, grants) as a percentage of GDP and then to provide an idea of aggregate fiscal space and its composition as a % of GDP for each country. This has not been possible at this stage due to the lack of adequate debt service data. 7

9 4. COUNTRY COVERAGE AND DATA SOURCES In Table A of the Appendix, data for low-income Sub-Saharan African countries are presented, as follows: the countries covered are 37 countries classified by the World Bank as IDA-only (plus one blend country which has an IDA-only income level Zimbabwe). This therefore includes Cameroon, Djibouti and the Republic of Congo even though they are officially lower middle income countries according to the latest World Bank atlas. the source of the debt distress classification data are IMF DSF-DSA documents (found in Commonwealth Secretariat, 2009) analyses conducted during (IMF, Documents reflect 2007 and 2008 data depending on availability, and there were no reliable data for Eritrea or Zimbabwe available from these analyses though other assessments have indicated that both have unsustainable debt levels. 6 data for budget revenue, fiscal balance and consumer price index are taken from the April 2009 IMF Regional Economic Outlook report for Africa. For the fiscal balance and revenue the 2008 data have been used. For consumer prices, the 2007 values have been used, as 2008 data were considered to be upwards biased due to sharp exogenous food and petrol price increases. Due to lack of data in the REO report, IMF country reports were used for Djibouti, Sudan and Mauritania (IMF, May 2007; IMF June 2008; IMF, July 2008). data for aid dependence are based on the total 2007 CPA data of 41 major donors, 7 directly provided by the OECD, with 2007 GNI values taken from the World Bank WDI data.(wdi, 2008). 6 See IMF staff reports on Eritrea and Zimbabwe. 7 These donors are: Austria, Belgium, Denmark, France, Germany, Italy, Netherlands, Norway, Portugal, Sweden, Switzerland, United Kingdom, Finland, Ireland, Luxembourg, Greece, Spain, Canada, United, States, Japan, Korea, Australia, GEF, Montreal Protocol, New Zealand, IDA, IDB Special.Fund, African Development Fund, Asian Development Func, EC, OPEC Fund, BADEA, IMF, UNDP, UNICEF, UNAIDS, UNFPA, Islamic Development Bank, IFAD, GAVI, Global Fund. We are most grateful to the OECD for supplying these data. We have chosen these data, rather than grants reflected in IMF programme budget tables, because a large proportion of aid remains off-budget. 8

10 5. RESULTS 5.1. Results at Indicator Level The results of the analysis are shown in Appendix Table A and in diagram 1 below. For the 37 low-income African countries, the results by indicator are as follows: 17 are at critical levels of external debt and 12 of domestic debt. Overall 21 countries have critical debt levels (6 of which on both external and domestic debt). These countries should not borrow further, to avoid exacerbating the risk of unsustainable debt have sufficient levels of domestic revenue effort (17% of GDP) and should not be expected to increase tax revenue to provide themselves with fiscal space. Only 4 have over-dependent levels of aid (>25% of GNI), and therefore almost all countries 33) could absorb more aid (especially grants) without being excessively aid dependent. Once space for using each type of budget financing is corrected to allow for impossible risks from high budget deficits and/or inflation, as shown in diagram 2 below, countries have considerably less space as follows: 12 have levels of fiscal deficit (more than 3% of GDP) which would indicate that they should not increase the deficit further. 10 have inflation levels (>10%) which indicate an inflationary risk from stimulus. A further 3 countries (making a total of 24 or two-thirds of all countries) have no space to borrow because their deficits are already at more than 3% of GDP; A further 8 countries (making a total of 10 countries) have no fiscal space to use even grants because their inflation rates exceed 10%. Nevertheless, 25 countries (more than two-thirds) still have space to use more grants. 8 CAR and Liberia s levels of debt distress will fall in 2009 as a result of HIPC completion points. 9

11 DIAGRAM 1: FINANCING SOURCES WITH SPACE Eritrea Ghana Debt Distress Benin Malawi Cameroon Niger Chad Nigeria Senegal Lesotho Madagascar Rwanda Mali Tanzania Mozambique Uganda Congo D.R. Gambia Congo Mauritania Togo Djibouti Zambia Kenya Côte d'ivoire Ethiopia B.Faso Sudan Guinea CAR Sierra Leone Comoros Zimbabwe Grants Guinea-Bissau São Tomé & Príncipe No Space Revenue Burundi, Liberia 10

12 DIAGRAM 2: FINANCING SOURCES WITH SPACE (corrected for high fiscal deficits or inflation) Congo Côte d'ivoire Benin Niger Cameroon Djibouti Mauritania Senegal Madagascar Lesotho Chad Nigeria Gambia Eritrea Sudan Togo Mali Rwanda Malawi Debt Distress Tanzania CAR Kenya Uganda Grants Comoros B.Faso Mozambique Ethiopia Guinea Guinea-Bissau Sierra Leone Zimbabwe São Tomé & Príncipe No Space Burundi, DRC, Ghana, Liberia, Zambia Revenue 11

13 5.2. Overall Fiscal Space Results and Implications The results of our research indicate that: without adjusting for risks of deficit and inflation levels, two countries (Burundi and Liberia) have no fiscal space, fourteen have low space (12 to use grants and 2 to raise revenue), 15 moderate space and 6 high space. once adjustments are made for deficit and inflation risks, 5 countries (Burundi, DRC, Ghana, Liberia and Zambia) have no space, 17 have low space (11 to use grants and 6 to raise revenue), 11 moderate space and 4 high space. Box 1: Fiscal Space in Low-Income African Countries No space Low space Moderate space High space Not adjusted Adjusted Burundi Burundi Liberia Liberia Congo, D.R. Zambia Ghana Congo, D.R. Mauritania Congo Malawi Congo São Tomé & Príncipe Côte d'ivoire Mauritania Côte d'ivoire Sudan Djibouti São Tomé & Príncipe Djibouti Togo Eritrea Senegal Eritrea Zambia Ethiopia Sierra Leone Gambia Guinea Sudan Ghana Gambia Togo Guinea-Bissau Guinea-Bissau Zimbabwe Kenya Kenya Benin Lesotho Benin Mozambique Burkina Faso Malawi Burkina Faso Niger Cameroon Niger Cameroon Nigeria Central Af.R. Nigeria Central Af.R. Chad Senegal Chad Comoros Sierra Leone Comoros Ethiopia Zimbabwe Lesotho Guinea Madagascar Madagascar Tanzania Mali Mali Uganda Rwanda Mozambique Tanzania Rwanda Uganda 12

14 What are the implications of these findings for country authorities and the international community, in terms of which types of resources to use to combat the crisis? between 25 and 33 countries still have space to use more grants (depending on the degree to which these are used in ways likely to provoke inflation) (Adam and Bevan,2004) This implies that the primary means for low-income African countries them to avoid cutting education spending should be an increase in grant flows. around 15 countries could raise more revenue according to regional norms. However, in the latest IMF programme projections for the next three years, most of these countries are forecast to be hit by revenue falls in 2009 as a % of GDP, and only 5 are forecast to increase revenues as a % of GDP thereafter. Therefore the space which the IMF assesses countries to have, after realistic analyses of their economies, is very limited, especially over the short-to-medium term and in the midst of lower growth levels countries could borrow more with compromising their overall public debt sustainability. This is only around a third of the total countries and shows the high level of indebtedness (external or domestic) of most low-income sub-saharan countries. Around countries might borrow more externally according to the LIC-DSF, and might borrow more domestically, but when the two types of debt are looked at together, debt-related space is fairly limited. However, there are currently discussions following the G20 Summit about making the DSF more flexible, and this might provide more space for countries to borrow more without compromising their sustainability. only 4-6 countries could use all three types of financing. This shows both the relative narrowness of the options faced by low-income African countries and their relatively high existing levels of debt and taxation. Overall, then, the primary financing response of the international community should be (and to a considerable degree is being) to provide more grants, with the secondary response being to reassess the flexibility of the LIC-DSF in order to expand scope for countercyclical (especially concessional) borrowing. However, the analysis also shows that African governments should be using (and the Bretton Woods Institutions should be allowing them to use) the fiscal space they have available to them, especially to absorb grants and borrow, in order to combat the crisis and avoid cutting education expenditure. Further work could be conducted to assess and monitor on a continuing basis the degree to which fiscal space is being used, and whether this is occurring in the most desirable way (ie using the most appropriate instrument in each case). A preliminary look at IMF programmes for 33 low-income African countries indicates that the deficit including grants as a % of GDP is being allowed to increase in around 60% of the countries (though in many cases only for 2009), and staying broadly the same or being reduced in around 20% of countries. It also indicates that around 60% of programmes are forecasting increased expenditure as a % of GDP (again in many cases just for 2009), while 10% are keeping it the same and 30% are reducing it. This indicates some considerable degree of flexibility by the BWIs, but not enough flexibility, given that this paper assesses that around 90% of lowincome SSA countries could use fiscal space. 13

15 6. EDUCATION EXPENDITURE One of the primary goals of the research underlying this paper was to identify whether the current financial crisis is forcing African countries to cut education expenditure, because they either do not have fiscal space, or are not using it, to mobilize financing to protect this and other spending. In order to identify whether low-income SSA countries are planning to cut their educational spending as a result of the crisis, we attempted to conduct a detailed examination of the 2009 budgets of all 37 low income sub Saharan Africa countries Data Issues However, one major constraint encountered was the lack of publicly available data on 2009 projected education expenditure, for two main reasons: in small part, this reflected the fact that countries have different fiscal years. In particular, East African countries (Ethiopia, Kenya, Rwanda, Tanzania and Uganda and in future years Burundi) publish and approve their annual budgets in Parliament during May-July. As a result, any future assessment of education spending levels should take account of this calendar and be done in the second half of June. As of end-june 2009 Ethiopia and Malawi had not published their 2009/10 budgets. however, the main reason for the lack of data was that very few (especially Francophone) countries publish their budgets on the web. Extensive searches of official finance and planning ministry websites in Africa revealed the results shown in Appendix Table B. The Table presents the availability of educational expenditure data for each country, the source and the relevant link. The outcome of the research for educational expenditure availability is presented in the last column and coded as follows: o A: (15 countries) full or partial 2009 education expenditure information was found. o B: (8 countries, including Ethiopia and Malawi see above) there is no availability of the 2009 budget or data on the official website of the ministry. o C: (4 countries) it was not possible to enter the ministry s official website due to error. o D: (10 countries) it was not possible to find the official Ministry website. in addition, international and regional data are virtually nonexistent. Most IMF and World Bank documents do not break down expenditures by sectoral classifications (except when these are essential to monitor sectoral spending targets, which exist in only a very small number of programmes). Only one of the countries covered by this study (Burkina Faso) had a clear breakdown of recent and projected budget education expenditure specified in its IMF programme documents. UNESCO s own data for the EFA Monitoring Report are historical up to African regional organizations responsible for monitoring human development trends (such as the UNECA, UEMOA etc) also do not keep track of the most recent or current data. It is rather shocking, in view of the strong emphasis given to monitoring progress to the education MDGs, that there does not exist a more current database for analyzing education spending. Data for the previous and current budget year (and in an increasing number of countries for 3-5 year spending plans) do exist in virtually all low-income African countries (with a few exceptions where primary education expenditure is not adequately distinguished from overall expenditure), but are simply not assembled by any regional or international organization. One conclusion resulting from this attempt to collect data is that there is potentially a very vital role for UNESCO to play in monitoring education expenditures on a current year basis, to see whether expenditures are rising or falling in future. This could be achieved by establishing a proactive network of the education ministry planning or budget directors, or the national PRSP coordinators, of most African LICs, who could supply data on 14

16 recent and forecast overall and primary education expenditure more rapidly than current channels. DFI could assist with the initial contacts in this process Are Countries Cutting Education Expenditure? From the 37 low-income Sub Saharan countries investigated, detailed information on education expenditure for the 2009 budget was available in 15. However, further investigation revealed that Congo and Mali data covered only part of their education budgets namely current in the case of Congo and investment in the case of Mali and Nigerian data from different sources were highly inconsistent. As a result, education expenditure data from 12 countries are analysed here (two more have just published budgets and will be included in the final report) However, it is important to note that these 12 countries include most of the countries for which budgets would have been expected to reflect the impact of the global financial crisis. Most African LICs (26 of the 35 for which information is available) presented their last budgets to parliament in December 2008, for the calendar year 2009, and their underlying analysis was finalized in around October, before there was any significant impact of the crisis on their economies. The only way to track the impact of the crisis would be to assess any mid-year restrictions on expenditures, which could be done only via questionnaires or case studies. Four countries (in Southern Africa) present budgets in March for an April-March fiscal year; and five (in East Africa) present budgets in June or July. It is only these 9 governments which could have been expected to reflect the crisis impact in their budgets, and information has been found for 6 of these countries (all except Ethiopia and Malawi whose budgets have just been published and will be included in the final report, and Zimbabwe). The available educational expenditure for the 12 countries is presented in Table 1 (see also Table C of the Appendix). Whenever disaggregated information on expenditure for different levels of education was available, it has been included. In Table 3, the expenditures on education as percentage of the GDP and percentage of total expenditure are presented in order to assess trends. Information on actual amounts in local currency is available in Table C of the Appendix: they are not presented here because GDP is projected to rise in all countries studied, and therefore there is no country where spending is falling in nominal terms through rising as a % of GDP or of the total budget. Increases in indicators are shaded green; stable levels in yellow; and falls in red. From the sample of 12 countries, the following results emerge: 5 countries (Burkina Faso, Liberia, Mozambique, Sierra Leone and Zambia) plan to increase their expenditure on education, in relation to both GDP and the total budget. It is important to note that four of these countries (except Zambia) produced their budgets in December, before the impact of the crisis was felt. Liberia, Sierra Leone and Zambia are increasing education expenditure more sharply as a percentage of the total budget than compared to GDP, reflecting reallocation of funds to education within a growing budget. On the other hand, it is rising faster as a percentage of GDP in Mozambique, reflecting a decision to increase non-social spending even faster. In more recent reports, these countries have started raising concerns about the level of their estimated revenues and consequently their ability to cover their expenditure. One example 15

17 is Mozambique. In a recent IMF report 9 it is mentioned that revenues in 2009 are expected to drop by 1.3% relative to the budget approved earlier by the Parliament. Zambia is an interesting case because it had been planning to spend considerably more on education, as a result of an expected windfall tax on mining companies, but this tax had to be withdrawn when the crisis hit minerals prices, so spending would have been higher if the crisis had not occurred. two countries (Kenya and Uganda) are planning to maintain education spending at current levels in relation to both GDP and total budget spending. Lesotho is increasing overall education expenditure as a percentage of GDP and maintaining it as a percentage of total expenditure. This reflects a rise in overall budget expenditure as a percentage of GDP. Rwanda is increasing education spending as a percentage of GDP but showing a slight fall as a percentage of the budget, because its budget is focusing on agriculture and infrastructure as part of its second-generation PRSP. Similarly, Tanzania is maintaining it as a percentage of GDP but cutting it as a percentage of total expenditure. This is because Tanzania is dramatically increasing expenditure on agriculture, infrastructure and water as part of its Mkukuta poverty reduction strategy. only two countries (Benin and Ghana) are planning to cut education expenditure as a % of GDP and as a % of total expenditure. In Benin this reflects movement of funding within the budget away from education to non-social sectors. In Ghana it reflects drastic cuts in all spending as a result of a budget crisis left to the incoming government by its predecessor (rather than any significant impact of the global financial crisis). Even in the countries which seem to be cutting education expenditure, there is considerable evidence of protecting basic education spending from cuts in Benin and Ghana. Benin will be keeping pre-primary and primary spending at almost the same percentage of GDP, and Ghana will be increasing basic education expenditure considerably as a percentage of the budget. However, interestingly in Lesotho and Rwanda, pre-primary and primary spending are losing out to other education spending, reflecting in Lesotho the government s commitment to increasing skills and vocational education to combat unemployment, and in Rwanda the commitment to increase spending on secondary and vocational education under Rwanda The reluctance to cut basic education expenditure in countries is thoroughly welcome. This is partly because education is an MDG and a key policy objective in all national development strategies, as well as a wish to offset the potential increase in inequality that the financial crisis will cause, by offering a safety net for the social expenditures of the poorest families. It also reflects the continuing concern that the international community has shown for protecting social expenditures during the crisis (and the fact that large amounts of educational investment spending (ie projects) are funded by donors and therefore cannot be cut. 9 IMF Mission Calls for Fiscal Stimulus in Mozambique, May 13, 2009, No. 09/165, available at: 16

18 Table 1: Education Spending for 2008 and 2009 % of GDP % of total budget Burkina Faso edu_exp edu_exp Benin pre-primary&primary secondary&vocational edu_exp Ghana basic education expenditure Kenya edu_exp Lesotho edu_exp pre-primary&primary Liberia edu_exp Mozambique edu_exp Rwanda edu_exp pre primary&primary Sierra Leone edu_exp Tanzania edu_exp Uganda edu_exp Zambia edu_exp Education and Health Expenditure Compared Finally, to test the relative priority being given to education within social sector spending, we have conducted a comparative analysis of education and health spending for four East African countries which have most recently released budgets. The results are shown in Table 2 below. Table 2: Comparison of Health and Education Spending Education Health % GDP (09-08) % Budget (09-08) % GDP (09-08) % Budget (09-08) Kenya Rwanda +0.5 (-0.3 basic) -0.2 (-2 basic) Tanzania Uganda The table shows that, with the exception of Rwanda, education spending is holding up better than health in the other three countries though there was also a sharp increase (form 0.3 to 0.7% of GDP) in public health and sanitation expenditure in Kenya, it is understood this is more related to water and sanitation than to medical services. While this is relatively good news in terms of priority being given to education, the linkages between health and education are so strong that reductions in health spending could undermine education outcomes Two additional types of analysis which it is not possible to do within the scope of this paper are to analyse whether the countries cutting or increasing expenditure are doing so above/below current average spending levels in Africa, or reflecting any clear correlation with the likelihood that they will attain the education MDGs. 17

19 7. OVERALL CONCLUSIONS AND RECOMMENDATIONS 1) The Fiscal Space Index This study finds that a multi-country fiscal space index could be a very useful tool to judge whether LICs have space to respond to the global financial crisis (and future crises) through fiscal stimulus. In non-crisis years it could also help to judge whether education spending could be increased to meet the MDGs, without negative macroeconomic consequences. However, in order for the index to have maximum utility, the following steps would be needed: further discussion of indicators and thresholds with a wider group of stakeholders (especially low-income country policymakers) to confirm optimal definitions and levels. further technical improvements to the index, in particular expansion to include potential for expenditure rationalisation, as well as using budget flow variables as % of GDP (especially debt service rather than stock) in order to be able to quantify fiscal space more clearly. expanding the index to cover at a minimum all low-income countries. comparing the space indicated by the index with estimates of the annual addition al spending needed to reach the education (and other) MDGs, and especially the EFA country strategy financing gaps, to show clearly what funding sources are available to countries. The study finds that most fiscal space is on the grant (and secondarily revenue) side, rather than in relation to debt, leading to four strong advocacy points: fiscal stimulus and increases in MDG spending will be largely dependent on grants in the short-term especially given the undesirability of increasing tax burdens during a recession; many countries can still probably make efforts to increase tax revenue, notably by taxing foreign investors and large corporations, assisted by greater international tax transparency there is a strong case for reviewing the Low Income Countries Debt Sustainability Framework (as is already being done for the BWI Annual Meetings in October) to assess whether more space could be created on the loan element of resource mobilization. only 4-6 countries have space to use all three financing sources, showing the narrowness of options for most African LICs, and underlining the need to live up to grant promises. Following the recent indication by the IMF that it is providing fiscal space to LICs with programmes, a preliminary examination indicates this may be true in 60% of countries. However, as this compares with 90% of African LICs which, according to this study, should be given fiscal space, it would also useful to compare the fiscal space identified in the index with that being provided by the IMF, and look at how this is being done and what financing sources are being used, to assess whether the IMF is fully using countries fiscal space. 2) Impact of the Crisis on Education Expenditures The study examined 2009/10 budgeted education spending data for 12 African LICs and compared them with 2008/09. These countries include all those for which we would expect to see an impact of the crisis, which published budgets in March or June-July. Other countries published their budgets in December before the crisis hit African LICs. The process of this survey also revealed the paucity of web-available budget data and the lack of any international system for monitoring education expenditure trends on a very current basis, 19

20 which is lamentable given that the international community has agreed that increasing education spending is crucial to attaining the MDGs. As a result, the study strongly recommends that UNESCO establish (as part of the UN system s commitment in G-20 and New York to monitor the social impact of the crisis) a system for tracking current expenditure trends, through a network of LIC government officials, and disseminate its results widely in order to mobilise greater campaigning and advocacy force behind the education MDGs. The results of the survey show that: 5 countries plan to increase expenditure (4 of which published budgets before the crisis); 2 countries (Kenya and Uganda) are maintaining it; 2 countries (Rwanda and Tanzania) are cutting it either as a % of GDP or as a % of total expenditure; and 2 countries (Benin and Ghana) are cutting on both indicators. However, virtually all countries are protecting basic education expenditure from cuts, with the exception of Lesotho and Rwanda, which appear to be wanting to spend more on secondary and vocational education, partly because they have dramatically increased primary education spending in recent years, and partly because they see a need to train their workforce to a higher level to combat the crisis and maintain high long-term growth levels. A comparison with health expenditure for the countries with June budgets indicates that education is being protected more than health. This is relatively good news for education, but the interrelations among the MDGs mean health cuts could undermine education progress. Further analysis would be desirable to assess whether countries are cutting or increasing expenditure compared to medium-term trends; compared to average spending levels in SSA or global or regional commitments to spend a particular proportion of the budget on education; and compared to the likelihood that they will attain the education MDGs. 20

21 REFERENCES Action Aid (2007). The IMF s Policy Support Instrument: Expanding Fiscal Space or Continuing Belt-Tightening?, Policy Brief, Johannesburg, October Adam C. S.and David L. Bevan (2004). Aid, Public Expenditure and Dutch Disease, March African Economic Research Consortium (AERC) (2004). Financing Pro-Poor Growth in Africa, AERC Senior Policy Seminar VI, Kampala 2-4 March Barro, Robert (1991). Economic Growth in a Cross-Section of Countries, Quarterly Journal of Economics, 106(2), Benhabib, Jess and Mark M. Spiegel (1994). The Role of Human Capital in Economic Development: Evidence from Aggregate Cross-Country Data. Journal of Monetary Economics, October, 34(2), Commission for Africa (2005). Our Common Interest, Report of the Commission for Africa, March 2005 Committee of 10 (2009), Impact of the Crisis on African Economies Sustaining Growth and Poverty Reduction: African Perspectives and Recommendations to the G20, a report from the Committee of African Finance Ministers and Central Bank Governors established to monitor the crisis, March 17. Commonwealth Secretariat (2009), Fiscal Sustainability of Debt, Paper prepared by Development Finance International (DFI), CEMLA and Commonwealth Secretariat, April Development Committee (2006), Fiscal Policy for Growth and Development: An Interim Report, April 6, Forum on Debt and Development (FONDAD), Helping the Poor? The IMF and Low- Income Countries, The Hague 2005, pp Foster, M. and A. Keith, (2003). The Case for Increase Aid, Final report to the Department for International Development, 1, December International Monetary Fund (2009). Regional Economic Outlook: Sub-Saharan Africa, IMF, Washington DC, April International Monetary Fund (2009). The Global Financial Crisis: The Challenge for Africa, an Address in Celebration for Africa Week 2009 by Dominique Strauss-Kahn, IMF, Washington DV, May

22 International Monetary Fund, (2009). The Dept Sustainability Framework for Low-Income Countries, IMF, Washington DC, February 2009 International Monetary Fund, (2008), Staff Guidance Note on the Application of the Joint Fund- Bank Debt Sustainability Framework for Low-Income Countries, Washington DC, October, available at International Monetary Fund, (2008). Sudan: First Review of Performance Under the Staff-Monitored Program, IMF Country Report No. 08/174, June International Monetary Fund, (2007). Djibouti: 2007 Article IV Consultation Staff Report; and Public Information Notice on the Executive Board Discussion, Country Report No. 07/178, May International Monetary Fund, (2007). Islamic Republic of Mauritania: 2008 Article IV Consultation and Third Review Under the Three-Year Arrangement Under the Poverty Reduction and Growth Facility Staff Report; Staff Supplement; Public Information Notice and Press Release on the Executive Board Discussion; and Statement by the Executive Director for the Islamic Republic of Mauritania, IMF Country Report No. 08/231, July International Monetary Fund (2005). Back to Basics- Fiscal Space: What It Is and How to Get It IMF quarterly magazine, 42 (2), Washington DC, June ONE, (2009). The data Report 2009: Monitoring the G8 Promise to Africa Psacharapoulos, George and Harry A. Partinos (2004), Returns to Education: A Further Update, Education Economics, Vol.12, No.2, August. Rathin, R., H. Antoine and E. Letouze (2007). Fiscal Space for What? Analytical Issues from A Human Development Perspective, UNDP. Paper prepared for the G-20 Workshop on Fiscal Policy, Istanbul, June 30- July 2, 2007 Union Economique et Monetaire Ouest Africaine (UEMOA), (2008). Rapport Semestriel d Execution de la Surveillance Multilaterale, December United National Economic Commission for Africa (UNECA) and Organisation for Economic Co-operation and Development (OECD) (2009). The Mutual Review of Development Effectiveness in Africa: Promise and Performance, Joint UNECA and OECD Report, United National Economic Commission for Africa (UNECA) and Organisation for Economic Co-operation and Development (OECD) (2007). Development Finance in Africa: From Monterey to Doha, Document prepared for the 11 th Meeting of the Africa Partnership Forum in Addis, Ababa November