85773 H N P D i s c u s s i o n P a p e R REPRIORITIZING GOVERNMENT SPENDING ON HEALTH: Pushing an Elephant Up the Stairs? Ajay Tandon, Lisa Fleisher, Rong Li and Wei Aun Yap January 2014 REPRIORITIZING GOVERNMENT SPENDING ON HEALTH: Pushing an Elephant Up the Stairs? Ajay Tandon, Lisa Fleisher, Rong Li, and Wei Aun Yap January 2014 Health, Nutrition, and Population (HNP) Discussion Paper This series is produced by the Health, Nutrition, and Population (HNP) Family of the World Bank's Human Development Network (HDN). The papers in this series aim to provide a vehicle for publishing preliminary results on HNP topics to encourage discussion and debate. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. Citation and the use of material presented in this series should take into account this provisional character. For information regarding the HNP Discussion Paper Series, please contact Martin Lutalo at mlutalo@worldbank.org or Erika Yanick at eyanick@worldbank.org. © 2014 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW, Washington, DC 20433 All rights reserved. ii Health, Nutrition, and Population (HNP) Discussion Paper Reprioritizing Government Spending on Health: Pushing an Elephant Up the Stairs? Ajay Tandon,a Lisa Fleisher,b Rong Li,c Wei Aun Yapd a East Asia and Pacific Human Development Department, the World Bank Group, Washington, DC. b Consultant, the Health and Economy Program, Health, Nutrition, and Population Unit, the World Bank Group, Washington, DC. c, d Consultants, East Asia and Pacific Human Development Department, the World Bank Group, Washington, DC. This paper was prepared for the Health and Economy Program, Health, Nutrition, and Population Unit, the World Bank Group, Washington, DC; November 2013. This work was funded by the Rockefeller Foundation Trust Fund Abstract: Countries vary widely with respect to the share of government spending on health, a metric that can serve as a proxy for the extent to which health is prioritized by governments. WHO data estimate that, in 2011, health’s share of aggregate government expenditure in the 170 countries for which data were available averaged 12 percent. However, country differences were striking: ranging from a low of 1 percent in Myanmar to a high of 28 percent in Costa Rica. Some of the observed differences in health’s share of government spending across countries are unsurprisingly related to differences in national income. However, significant variations exist in health’s share of government spending even after controlling for national income. This paper provides a global overview of health’s share of government spending and summarizes key theoretical and empirical perspectives on allocation of public resources to health vis-à-vis other sectors from the perspective of reprioritization, one of the modalities for realizing fiscal space for health. Theory and cross-country empirical analyses do not provide clear-cut explanations for the observed variations in government prioritization of health. Standard economic theory arguments that are often used to justify public financing for health are equally applicable to many other sectors including defense, education, and infrastructure. To date, empirical work on prioritization has been sparse: available cross-country econometric analyses suggests that factors such as democratization, lower levels of corruption, ethnolinguistic homogeneity, and more women in public office are correlated with higher shares of public spending on health; however, these findings are not robust and are sensitive to model specification. Evidence from case studies suggests that country-specific political economy considerations are key, and that results-focused reform efforts — in particular efforts to explicitly expand the breadth and depth of health coverage as opposed to efforts focused only on government budgetary targets — are more likely to result in sustained and politically-feasible prioritization of health from a fiscal space perspective. Keywords: Fiscal space, government health expenditure, prioritization, political economy. iii Disclaimer: The findings, interpretations, and conclusions expressed in the paper are entirely those of the authors, and do not represent the views of the World Bank, its Executive Directors, or the countries they represent. Correspondence Details: Ajay Tandon, World Bank, 1818 H St., NW, Washington, DC 20433, USA; telephone: 202-473-6338; e-mail: atandon@worldbank.org; website: www.worldbank.org/hnp. iv Table of Contents ACKNOWLEDGMENTS .......................................................................................................... vii I. INTRODUCTION....................................................................................................................1 II. REPRIORITIZING HEALTH FOR FISCAL SPACE ........................................................2 III. HEALTH SECTOR SHARE OF GOVERNMENT SPENDING .......................................6 IV. REPRIORITIZING HEALTH: THEORETICAL PERSPECTIVES..............................10 NORMATIVE ECONOMIC THEORY..............................................................................................10 POSITIVISTIC THEORIES ON GOVERNMENT EXPENDITURES .......................................................11 V. REPRIORITIZING HEALTH: CROSS-COUNTRY AND CASE STUDY EVIDENCE ............................................................................................................................13 CROSS-COUNTRY EVIDENCE.....................................................................................................13 COUNTRY-SPECIFIC EXPERIENCES WITH REPRIORITIZATION ....................................................14 VI. SUMMARY AND CONCLUSIONS ....................................................................................21 REFERENCES .............................................................................................................................24 v vi ACKNOWLEDGMENTS This report was prepared by Ajay Tandon, Lisa Fleisher, Rong Li, and Wei Aun Yap as part of the Human Development Network Health Economy Program, led by Rafael Cortez (Task Team Leader, Senior Economist, HDNHE). The program work was conducted under the overall supervision of Nicole Klingen (Sector Manager, HDNHE) and Timothy Evans (Sector Director, HDNHE). The authors are grateful for comments and inputs on an earlier version of the paper from Ece Ozcelik, David Evans, William Savedoff, Toomas Palu, George Schieber, and Jack Langenbrunner. The authors are also grateful to the World Bank for publishing this report as an HNP Discussion Paper. vii viii I. INTRODUCTION Countries vary widely with respect to the share of government spending on health, a metric that can serve as a proxy for the extent to which health is prioritized by governments. WHO data estimate that, in 2011, health’s share of aggregate government expenditure in the 170 countries for which data were available averaged 12 percent. However, country differences were striking, ranging from a low of 1 percent in Myanmar to a high of 28 percent in Costa Rica. Some of the observed differences in health’s share of government spending across countries are not surprisingly related to differences in national income: cross-country comparisons show that higher-income countries generally spend a larger share of aggregate government expenditure on health than those at the lower end. Health care costs tend to be higher in richer countries, driven by relative price differences as well as the availability of higher-technology care, among other factors. Richer countries also tend to have more educated and ageing populations with a preference structure that expects higher levels of public financing for social protection programs. Higher costs of and more demand for publicly financed health care — combined with a greater fiscal and institutional ability to do so — are some reasons governments tend to spend a greater share of expenditure on health as countries become richer. However, significant variations exist in health’s share of government spending even after controlling for national income. This paper provides a global overview of health’s share of government spending and summarizes some of the key theoretical and empirical perspectives on why some governments spend more (or less) of public resources on health than others from the perspective of reprioritization, one of the modalities for realizing fiscal space for health. There are a variety of reasons why a focus on reprioritizing health’s share of government spending is important and merited from a fiscal space perspective. Foremost among these are indications that the sector is under resourced and that additional public financing for health is key for many low- and middle-income countries wanting to improve the levels and distribution of population health outcomes [including the achievement of Millennium Development Goals (MDGs)], reduce out-of-pocket (OOP) spending for health, and attain and sustain universal health coverage (UHC) for their citizens. In addition, ageing populations and the rising incidence of noncommunicable diseases (NCDs) are impending challenges that will imply higher levels of health expenditure across developing countries. In many countries, lack of government prioritization for health is a major constraint to increasing public financing for health. Theory and cross-country empirical analyses do not provide clear-cut explanations for the observed variations in government prioritization of health. Standard economic theory arguments often used to justify public financing for health are equally applicable to many other sectors including defense, education, and infrastructure. To date, empirical work on prioritization has been sparse: available cross-country econometric analyses suggests that factors such as democratization, lower levels of corruption, ethnolinguistic homogeneity, and more women in public office are correlated with higher shares of public spending on health; however, these findings are not robust and are sensitive to model specification. Evidence from case studies suggests that country-specific political economy considerations are key, and that results-focused reform efforts — in particular efforts to explicitly expand the breadth and depth of health coverage as opposed to efforts focused only on government budgetary targets — are more likely to result in sustained and politically feasible prioritization of health from a fiscal space perspective. Some caveats must be noted in assessing the share of health in aggregate government expenditure. First, a higher share of government expenditure does not imply a higher level of government health spending: countries with higher health shares may have lower government expenditure per capita (and vice versa). 1 In looking at shares of government expenditure, the focus is really on the issue of prioritization and not on comparing levels of government health expenditure across countries. In addition, efficient allocations of government health expenditure within the overall envelope are also important as is the extent to which public financing for health is pro-poor in its outlays. 2 Leaving aside that health has multisectoral determinants — and that governments may choose to improve health by investing in other sectors such as sanitation, education, infrastructure, or in economic growth more generally — what matters is where and how the government spends on health, not just how much. Therefore, any assessments and analyses of the share of public spending on health must be made within the broader context that prioritization is only one aspect of a bigger picture on how governments can attain population health objectives. Section II of this paper explores the issue of reprioritization within the context of fiscal space. Section III provides some facts and issues regarding health’s share of government expenditures. Section IV presents theories on government allocations to health vis-à-vis other sectors. Section V describes findings from cross-country and country-specific case studies of efforts to reprioritize health. Section VI summarizes the paper and discusses some key implications. II. REPRIORITIZING HEALTH FOR FISCAL SPACE Fiscal space for health refers to the ability of a country to increase public spending for health without jeopardizing the government’s long-term financial sustainability. 3 Assessing fiscal space for health basically entails an evaluation of the different sources of financing that might be available for increasing government health spending, assuming a clear case has been made that such an increase is merited and that the net societal benefits of increasing government health spending are positive. A conducive macrofiscal environment, higher revenues, increased borrowing, seigniorage (inflationary finance), and higher levels of foreign aid are all potential sources of fiscal space (not just for health, but for any sector). Each option brings its own costs and benefits. While increasing revenues may ease fiscal constraints, the way they are raised is crucial: regressive, inefficient, and excessive taxes can do more harm than good to the overall economy. Similarly, borrowing to finance current spending may seem like a good idea in the short run, but could become unsustainable over time (as the recent debt crisis in some European countries demonstrates). Aid may ease budget shortfalls in countries that lack the domestic finances to cover the costs of high disease burdens, but it can bring its own set of negative 1. A companion paper entitled “An Assessment of Sociopolitical Determinants, International Aid and Government Health Expenditure: A Cross-Country Analysis over 1995–2010,” coauthored by Lilin Liang and Andrew Mirelman, examines the relationship between levels of public health expenditure and macroeconomic factors. Specifically, the authors attempt to answer whether larger governments have higher per capita public health spending, whether public debt limits the government’s capacity to expand public health spending and whether different polities affect the level of government health expenditure. 2. Tandon 2007. 3. Heller 2006. 2 externalities and inefficiencies. Seigniorage is rarely, if ever, a serious option to consider. Given clearly defined needs, the issue in any fiscal space assessment is one of identifying and assessing feasible, low-cost means of financing additional spending that minimize potential unintended adverse consequences, assuming multiple options have been identified and are available. From a sector-specific perspective, reprioritization — the focus of this policy note — is an additional option for realizing fiscal space and implies that a government would decide to increase a sector’s share of total government spending, preferably at the expense of spending on relatively less meritorious activities. Prioritization is a key intermediating link between the overall macrofiscal context of a country and how much a government chooses to spend on health (see box 1.1). Box 1.1 The Mathematics of Reprioritization and Fiscal Space Prioritization is a key intermediating link between the overall macrofiscal context of a country and the share of government expenditure on health. This can best be understood using a simplified version of the algebra behind a government’s intertemporal budget constraint: Gt + rtBt-1 = Tt + Bt + At + Ot, Where Gt is government non-interest expenditure in time t; rBt-1 is nondiscretionary debt interest payments; Tt is taxes, fees, and other government revenues, including those from seigniorage (inflationary finance); Bt is total government borrowing (domestic and foreign net of use of deposits); At is grants; and Ot is other sources of funds, such as the sale of assets. The right-hand side represents the government’s aggregate sources of revenue, and the left-hand side represents total spending. Fiscal space for health depends not only on overall government budget constraints, but also on the priority assigned to health. Government health spending, Ht, is a proportion kt of the overall government budget, or, in equation form: Ht = kt Gt. Whether the priority for health (kt) is a constant or variable parameter is a key policy question. The discussion above suggests that, based on the importance accorded the health sector and discounting other sources of financing Ot, public financing for health ought to be closely related to a country’s (a) overall fiscal capacity and effort (both in terms of general taxes as well as earmarked ones, such as under social insurance); (b) its ability to borrow to finance current spending, sustainably (in general and for health); and (c) its ability to attract — as well as its preferences for — grants, including foreign aid (in general and for health). The focus from this perspective would be to analyze increases in G and derive the implications for H. Also, a focus on reprioritization implies finding ways to increase k. 4 Improvements in a country’s macrofiscal context could increase G with only a marginal effect on H, if k remained constant, stressing the importance of reprioritization if the share of health spending needs to be increased. Source: Authors 4. Fiscal space can also be realized through efficiency gains. Assuming Y represents some measure of government health system outputs (e.g., effective coverage of key interventions), then getting the most Y out of H is creating effective fiscal space. Interventions aimed at improving the technical and allocative efficiency of health spending by, for example, using cost- effectiveness criteria to inform resource allocations, reducing leakages in interfiscal transfers, or addressing absenteeism of health workers are examples of policies that could increase efficiency. 3 There are a variety of reasons for increased policy attention in recent years to the issue of fiscal space for health, especially from the perspective of reprioritization. Foremost among these are indications that the sector is underresourced and that additional public financing for health is key for many low- and middle-income countries seeking to improve the levels and distribution of population health outcomes (including MDGs), reduce OOP spending for health, and attain and sustain UHC for their citizens. In addition, ageing populations and the rising incidence of NCDs are looming challenges that will necessitate higher levels of health expenditures across developing countries. Despite recent progress, several low-income countries — especially in sub-Saharan Africa and South Asia — are far from attaining health-related MDGs by 2015. 5 This lack of progress in health outcomes appears even more egregious when one considers the fact that a large proportion of child and maternal mortality are easily preventable via well-known cost-effective interventions. One key constraint to the attainment of health outcomes in low-income countries is the lack of adequate financial resources for health, recent increases in development assistance for health notwithstanding. 6 And the MDGs themselves are explicit in acknowledging additional resource needs: included among the targets is a call for developed countries to commit at least 0.7 percent of their gross national income (GNI) toward overseas development assistance. WHO’s Report of the Commission on Macroeconomics and Health estimates that a minimum of US$34 — or about US$40 in 2007 prices — in per capita health expenditures would be needed in low-income countries to provide a basic package of essential health services. 7 A more recent estimate by the Taskforce on Innovative Health Financing places the number at US$54 per capita. 8 However, very few low-income countries spent even these minimal amounts on health in 2011 (Figure 1.1). Figure 1.1 Total and Government Health Expenditures Per Capita in Low-Income Countries, 2011 Health expenditure in low-income countries, 2011 Eritrea Total US$54 Ethiopia Central African Republic Government Madagascar Congo, Dem. Rep. Niger Myanmar Burundi Gambia, The Guinea Malawi Chad Mozambique Kenya Benin Guinea-Bissau Burkina Faso Tanzania Uganda Comoros Mali Togo Cambodia Tajikistan Liberia Afghanistan Haiti Rwanda Sierra Leone Kyrgyz Republic 0 20 40 60 80 Health expenditure per capita (US$) Source: WHO 5. United Nations 2012. 6. Gottret and Schieber 2006. 7. Sachs 2001. 8. Taskforce on Innovative Health Financing for Health Systems 2009. 4 UHC — the objective of providing everyone access to quality health care when needed, without creating financial hardship as a result — is now an explicit and prominent policy objective in many middle-income countries. For example, countries such as China and Thailand now provide near-universal coverage; others such as Indonesia, Philippines, and Vietnam cover 40 to 60 percent of the population. Coverage rates are lower in some lower-income countries, but even they have made progress in removing financial barriers for certain subgroups such as the poor and for services such as those related to maternal and child health. UHC is also a likely post-MDG international development target. However, while strong policy commitments are evident, the design, organization, and delivery systems for attaining UHC vary considerably, and remain a challenge. In particular, financing UHC programs is a key constraint given the high levels of informality in labor markets, which make it difficult to collect premiums. At present, in many countries the poor are covered by general revenues, and the formal sector is financed by contributions, leaving uncovered a large middle section of the population, consisting mainly of the nonpoor who work in the informal sector. Across many countries, the extent of UHC remains relatively shallow while OOP spending is generally high, even among those with coverage. The fiscal implications of expanding UHC to those still without coverage will largely depend on the extent to which costs are subsidized by governments. Also, governments are likely to face higher costs for supply-side expenditures to improve access to and the quality of care to meet growing demand, as well as to improve services to those already covered. Given the size of the informal sector and supply-side deficiencies, it is estimated that added fiscal resources of 1 to 2 percent of GDP will be needed to attain UHC targets in many low- and middle-income countries. 9 Ageing and NCDs are added challenges to health expenditures as both imply significantly higher health-related utilization rates and resource outlays. The share of the elderly population is increasing rapidly in many developing countries, as are NCDs in the overall disease burden. Recent estimates indicate that NCDs have already overtaken communicable diseases as the dominant share in most middle-income countries. In summary, trends and policy commitments — attaining the unfinished MDG agenda, increasing and improving UHC, ageing, the rise of NCDs — are increasing fiscal pressures for health spending across developing countries. Some argue that reprioritization (combined with donor financing in low-income countries) is necessary to address the fiscal space for health challenge, and that fiscal space contributions from other modalities including general revenue increases, additional borrowing, and inflationary financing are likely to be minimal at best. 10 For example, Haacker (2010) highlights some simple calculations to demonstrate that the fiscal benefits from reprioritization efforts far outstrip those that might be realized from expansionary fiscal or monetary policy efforts, given realistic magnitudes of the expected impact on government spending and of the need to allocate across sectors. 11 Reprioritization is a key challenge for health ministries when dealing with ministries of finance and planning, especially as health is often perceived to be an unproductive and inefficient sector. If reprioritization is to help realize at least some fiscal space for health, what can we glean from a 9. Guerard, et al. 2011. 10. Sachs 2012. 11. Haacker 2010. 5 theoretical and empirical review of the landscape to better inform policy debates related to this issue? Prior to delving into this question, the next section provides some key background information and stylized facts related to health’s share of government spending across countries, regions, and over time. III. HEALTH SECTOR SHARE OF GOVERNMENT SPENDING As noted earlier, WHO data estimate that in 2011, health’s share of government spending in 170 countries where data were available averaged about 12 percent. However, the shares varied greatly across countries (see Figure 1.2). Country differences were striking: ranging from a low of 1 percent in Myanmar to a high of 28 percent in Costa Rica. Costa Rica, Rwanda, Bhutan, Jordan, Bosnia and Herzegovina, and the Solomon Islands had the highest share compared to other countries in their regions. Figure 1.2 Health Share of Government Expenditure, 2011 Health share of government expenditure, 2011 30 Costa Rica Solomon Islands Percent of government expenditure (%) Rwanda 20 Jordan Bosnia and Herzegovina 10 Nepal Myanmar 0 Source: WHO Note: Countries with population less than 250,000 excluded Health’s share of government spending has increased slowly over the past few decades: the average share across countries was 10.2 percent in 1995, 10.7 percent in 2000, and 11.5 percent in 2005; and from 1995–2011, it increased by an average 2.5 percent a year. Not all countries saw an increase in health’s share over time: 29 countries (about 16 percent of the sample) saw a declining share of health in the government expenditure (see table 1.1 for the five with the highest and lowest annual growth rates). Most countries with the largest changes (negative and positive) were in sub-Saharan Africa; in addition, some of the countries seeing the biggest changes have been those that recently faced conflict, civil strife, or natural disaster, implying that the changes in health’s share of government expenditure in these countries were likely not a result of intrinsic changes in the prioritization of health but due more to exogenous factors. 6 Table 1.1 Annual Change in Health's Share of Government Expenditure, 1995–2011 (as percent) Average annual change in Country health share of budget Positive Democratic Republic of Congo 28.2 Afghanistan 27.6 Tuvalu 17.2 Liberia 15.1 Guinea-Bissau 14.9 Negative Chad -9.1 San Marino -5.6 Haiti -4.7 Mongolia -3.9 Mozambique -2.1 Source: WHO. Note: Countries with population less than 250,000 excluded. In 2011, countries in Latin American and the Caribbean (LAC) spent the greatest share of government expenditure on health (almost 14 percent) compared to those in South Asia (SAS), which allocated the lowest share (an average of 7 percent), while other regions allocated 9 to 11 percent. It should be noted that in sub-Saharan Africa (SSA) the average share increased steadily from 1995 to 2011 (Figure 1.3). Figure 1.3 Health Share of Government Expenditure across Regions, 1995–2011 Health share of government expenditure across regions, 1995-2011 16 Percent of government expenditure (%) LAC 12 14 EAP SSA ECA 10 MNA 8 SAS 6 1995 2000 2005 2010 Year Source: WHO Note: Countries with population less than 250,000 excluded In general, health is given a lower priority than education, but a higher priority relative to military spending. On average, from 2006 to 2011, health’s share of government expenditure across countries was 11.4 percent, while education was 15.6 percent, and military spending 8.8 percent (Table 1.2). The SAS and MNA (Middle East and North Africa) regions are notable exceptions, where the share of military spending in government expenditure was greater than health spending. 7 Table 1.2 Health, Education, and Military Share of Government Spending, 2006–11 (as percent) Region Share of government expenditure (%) Health Education Military Latin America & Caribbean 13.7 16.5 7.0 East Asia & Pacific 9.5 17.3 7.8 Sub-Saharan Africa 10.6 17.8 10.5 Europe & Central Asia 10.3 14.4 9.4 Middle East & North Africa 8.6 17.1 11.7 South Asia 7.6 13.4 13.9 Global 11.4 15.6 8.8 Source: WHO. Note: Unweighted country averages; countries with population less than 250,000 excluded. Public spending for health tends to increase as countries become richer: governments not only spend more on health, but also allocate a higher share of their resources to health (and to social sectors, in general). 12 For example, in 2011, in low-income countries, health represented only around 10.5 percent of government spending while in higher-middle, and high-income countries, it was progressively higher (see figure 1.4). The average health share among high-income countries was close to 13 percent. Figure 1.4 Health Share of Government Spending by Income Classification, 2011 Health share of govt budget by income classification, 2011 15 Percent of total govt budget (%) 5 0 10 Lower middle Low income Higher middle High income Source: WHO Note: Countries with population less than 250,000 excluded There are several reasons health’s share of government expenditure increases with income. The reasons for this are grounded, in part, on the macrofiscal environment within which a government operates, as well as the relaxation of budgetary constraints with rising income. As economies grow and the population becomes richer, the nature of the disease burden, demographics, and the preference structure for the demand for public financing for health also tend to evolve. The provision of health services — a relatively labor-intensive process — tends also to be more expensive in richer countries, driving up public (and private) spending on health. Health care costs tend to be higher in richer countries, driven by relative price differences as well as availability of higher-technology care, among other factors. Richer countries also tend to have 12. Also see Fleisher, Leive, and Schieber 2013. 8 more educated and older, ageing populations with a preference structure that generally emphasizes greater levels of social protection. Higher costs and more demand for publicly financed health care — combined with a greater fiscal and institutional ability to address these issues — are some reasons governments spend more as a share of expenditure on health, on average, as countries become richer. The rise in public financing for health is part of a general increase in the relative size of the public versus private sector: this phenomenon is known as Wagner’s Law (based on the work of a 19th century German political economist, Adolf Wagner). He noted two main reasons for government prominence with rising national income, similar to those for the rise in public financing for health mentioned above: (a) richer countries have more complex economies, requiring greater public regulations and administration; and (b) education, culture, and health are luxury goods, and public demand for governments to provide them grows as countries become richer. 13 Another notable characteristic of the global landscape is international demand for benchmarking shares of health in government expenditure or in GDP, partly in response to a perceived lack of prioritization of health in developing countries, noted earlier in the paper. 14 One of the earliest examples of these benchmarks is the oft-quoted WHO recommendation that countries spend at least 5 percent of their GDP on health. As clarified in Savedoff (2007), there actually never was a formal WHO recommendation to this effect. 15 Rather there was a suggestion in some WHO documents beginning in the early 1980s that countries should use this 5 percent number to monitor health financing flows; even then it seems the recommendation was to monitor the magnitude of total health spending, not just public spending on health. More recently, however, WHO’s Western Pacific and South-East Asia Regional Offices (WPRO-SEARO) have made more explicit recommendations to this effect: WHO’s Health Financing Strategy for the Asia- Pacific Region (2010–2015), which covers 48 countries recommends that total health expenditure for health be 4 to 5 percent of GDP, and that the OOP share of total health expenditure not exceed 30 to 40 percent. 16 The number is derived from empirical evidence indicating that it will likely be difficult for countries to attain universal coverage and concomitant reductions in OOP spending and catastrophic payments if public financing for health is less than at least 4 to 5 percent of GDP. WHO’s Eastern Mediterranean Regional Office (EMRO) has similarly proposed that governments allocate 8 percent of their budgets to health.17 Several sub-Saharan governments that are signatories of the Abuja Declaration made a commitment in 2001 to earmark at least 15 percent of government budgets for health. 18 As discussed later in the paper, it is also notable that most international calls for benchmarking health’s share of a government’s or a country’s resources have been largely aspirational and nonbinding. 19 By 2011, less than half of the EMRO countries had reached 8 percent of the 13. Shelton 2007. 14. The proliferation of earmarking attempts has not been limited to the health sector. In 2000, at the World Education Forum, 164 governments pledged to the Dakar Framework to commit at least 7 percent of GDP to education by 2005 and 9 percent by 2010. In 2003, the Assembly of the African Union agreed to the Maputo Declaration, which called for governments to commit at least 10 percent to agricultural and rural development by 2008. 15. Savedoff 2007. 16. WHO 2009. 17. WHO 2010c. 18. African Union 2001. 19. Benchmarking is not always in response to international calls. 9 government budget on health. In the same year, only about six sub-Saharan countries were close to or had exceeded the 15 percent earmark target. Importantly, those countries that achieved the Abuja target or were close to doing so were also countries that received high levels of budget support, blurring the lines between allocations for health from “pure” own-sourced domestic revenues versus from donor funding. 20 More than half of WPRO-SEARO countries spend far less than 5 percent of GDP, and some — including Malaysia and Thailand — have reached or are close to UHC with levels of public spending that are lower than the 5 percent of GDP benchmark. IV. REPRIORITIZING HEALTH: THEORETICAL PERSPECTIVES There are several theoretical perspectives — in economics, in political science, as well as in other related disciplines — that address the role of government in the economy and in society. There are two strands of distinct theoretical perspectives in this regard: a normative approach versus a more positivistic one. The former focuses on how governments should make choices regarding overall expenditures and allocations to health, while the latter stresses the reasons behind government policy choices. This section summarizes some of the theories and their implications in reprioritizing health. NORMATIVE ECONOMIC THEORY Normative mainstream economic theory says that governments should provide key public goods and enhance individual or private capabilities. 21 Specifically, public finance theory views the government as conducting two basic activities: those that offset market failures and those that are redistributive. 22 It assumes that government interventions can — under certain conditions — improve efficiency when market failures lead to suboptimal social welfare outcomes, and improve equity when market allocations lead to outcomes perceived to be unfair. 23 The rationale for government involvement in the health sector stems from some of these same considerations. Jack (1999), for example, summarizes arguments for the role of government interventions based on the “merit good” nature of health, with the perspective that health outcomes and consumption of medical services ought to be related to some concept of need and not to one’s ability to pay, as well as the presence of three broad forms of market failure prominent in the sector: (a) externalities; (b) the public benefit of certain health interventions; and (c) large information asymmetries in the sector. 24 In addition to economic theory arguments, the human rights perspective is often used to justify public financing for health from a normative perspective. This perspective underlies recent policy commitments to UHC in many countries. The constitutions of many countries guarantee access to health or to health care services to all their citizens. This is often taken to imply that 20. Atim et al. 2008. 21 More nuanced justifications for government interventions — for ensuring property rights, enforcing contracts, addressing market failures related to externalities and information asymmetries, and enhancing equity, for instance — have been emphasized as well, but these can be subsumed under broader categories of normative government functions and responsibilities. For more discussion, see World Bank 1997. 22. Gruber 2011. 23. Ibid. 24. Jack 1999. 10 governments have the legal imperative to ensure that the constitutionally mandated right to health is realized for all by ensuring adequate public resources are available. Unfortunately, the same economic theories that justify public financing of health also apply to government interventions in other sectors such as defense, education, food, housing, water, sanitation, and infrastructure. Indeed, theories justify military expenditures on the grounds that national defense is a public good and cannot be provided privately; also, that it is nonrivalrous (that is, use by one does not reduce use by others) and nonexcludable (that is, consumption cannot be restricted to certain subpopulation groups), making it impossible to provide and consume privately. The education sector is another example. While education, unlike national defense, is largely a private good, theoretical arguments underpinning public financing for education are strong and relate to the presence of large human capital and productivity- enhancing externalities that can lead to social benefits not taken into account by individuals when making choices. The normative human rights perspective also does not imply that the health sector is by any means exceptional in this regard: other sectors such as food, housing, and education can and do also make plausible cases for public financing from a human rights angle. Given competing and equally meritorious priorities across many sectors, how much should total government expenditure be in an economy and how should public resources be allocated across sectors? Again, mainstream public finance theory argues for public expenditure levels to increase as long as social benefits from additional expenditures exceed the cost of raising revenues for financing the additional outlays. With regard to allocations across sectors, the implications are that public resources should be utilized up to the point where the marginal benefit of an additional dollar of spending is equalized across sectors. 25 The operational implications of this are not easy to derive given the difficulties of estimating the costs and social benefits of public expenditure. Nevertheless, this implies that — to merit reprioritization — the health sector would need to demonstrate that the social benefits of additional public spending outlays exceed the costs of financing this increase in spending, both in terms of the additional costs of raising revenues and also in terms of foregone public spending in other sectors (something that ministries of health are often not able to articulate persuasively to ministries of finance and planning). POSITIVISTIC THEORIES ON GOVERNMENT EXPENDITURES Unlike the normative perspective, the positivistic approach to understanding government behavior involves theories that describe “…why existing policies are pursued and … which policies will be pursued in the future.” 26 This approach helps to address the substantial disconnect between a theoretically ideal set of policies and what is actually implemented in practice. Under the positivistic perspective, the above-mentioned reference to Wagner’s Law is one example of an early theory of public expenditure. Some key positivistic perspectives on the choice of government expenditure allocation decisions are summarized below. Most positivistic theories of public expenditure invariably focus on political economy and “public choice.” Ultimately, because choices about policy directions are made and implemented by individuals, the most relevant positivistic theories on the role of government are those that seek to explain the behavior of individuals in a political setting. Public choice theory emphasizes 25. Gillingham 2013. 26. Musgrave 1959. 11 that individuals in the political arena are no different than people in any other market and are guided by their own self-interests. “Voters ‘vote their pocketbooks,’ supporting candidates and ballot propositions they think will make them personally better off; bureaucrats strive to advance their own careers; and politicians seek election or reelection to office.” 27 From this, one critical conclusion of public choice theory becomes apparent: it is the institutions and the rules associated with them that determine the nature of the incentives and constraints facing policy makers. 28 Whereas public finance argues for government interventions when market failures exist, positivistic theories of public choice help us understand government failure in terms of the inability or unwillingness of governments to act primarily in the interest of their citizens. 29 Applied to reprioritization in health, public choice theory suggests that bureaucrats and those in power support increases in government health spending when doing so advances their careers and/or enables them to stay in power. In a democratic setting, public choice theory predicts that a party’s policy choices can be seen as influenced (indirectly) by voting behavior. 30 The point is not that public policies are “…made through a voting mechanism but [that] choice procedures mirror some voting mechanism.” 31 This mechanism could be via preferences of the median voter; however, the real world applicability of the median voter theorem in understanding government allocation choices is limited. 32 Democracies are not perfect and voters lack full information, two prominent reasons that make it easy to see why in some low-income countries, even when the median voter is poor and may need public support to access health care, government allocations may not necessarily favor health and — even if they do — are not pro- poor in their benefit incidence. Regardless of the acceptability of the median voter theorem, there is general consensus that in a democracy, the electoral system can play an important role in determining how governments behave and, more specifically, about the size of the health sector. In addition, interest groups and institutions can and do play important roles. 33 The preceding discussion focuses heavily on theories relevant to democratic governments. However, health, education, and other policies are implemented in non-democracies — often with successful outcomes — and there is unequal participation even within the category of countries that are technically democratic. 34 The point of departure for autocratic regimes is whom the leader has an incentive to be responsive to and what type of goods that leader is motivated to provide. Autocrats can maintain favor by providing the “winning coalition” — a small group of individuals whose support is required to keep the incumbent in office — with private goods. 35 In summary, from the perspective of reprioritization, positivistic theories of public resource allocations generally imply that demonstrations of social welfare–enhancing aspects of additional government health spending will not necessarily be effective in increasing allocations to health. Enhancing democratization, improving citizen information, and increasing government 27. Ibid. 28. Touhy and Glied 2011. 29. Gruber 2011. 30. Roberts 1977. 31. Ibid. 32. Meltzer and Richard 1981. 33. Touhy and Glied 2011. 34. Dixit 2009. 35. Plümper and Martin 2003. 12 accountability may be more effective strategies to ensure that health is accorded the priority it merits from a public choice perspective. In addition, health sector analyses would need to take into account country-specific political economy considerations in assessing government allocations to the health sector from a fiscal space perspective. V. REPRIORITIZING HEALTH: CROSS-COUNTRY AND CASE STUDY EVIDENCE This section summarizes cross-country and country-specific case studies on the determinants of the government share of expenditures allocated to the health sector. Overall, empirical evidence related to the issue of health prioritization in developing countries is sparse and, in some cases, focuses on government health spending as a share of GDP and not as a share of total expenditure, obfuscating the link to the issue of prioritization per se. CROSS-COUNTRY EVIDENCE One key determinant in cross-country empirical studies focusing on health’s share of government expenditure is the level of democratization of a country. On average — likely a result of some of the factors discussed in the previous section — democratic societies and those with higher degrees of political liberty do tend to devote a larger share of government expenditure and GDP to health even after controlling for confounding factors. 36, 37, 38 In keeping with the theme of public choice issues as determinants of government allocations, the other variable to receive attention in empirical studies has been corruption. Most empirical evidence concludes that higher corruption levels are generally inimical to government allocations for health and favor spending on defense and energy at the expense of health, likely because of the higher possibilities of rent-seeking of the generally larger scale of contractual procurement amounts in the former sectors. 39, 40, 41, 42 The other prominent factor influencing government budgetary allocations for health, especially in low-income countries, is donor assistance. Development assistance for health (DAH), either on-budget or off-budget, can be fungible and may lead to unexpected shifts in government own- sourced revenue priorities, although the precise effect of the aid depends both on the composition of the aid itself and how government chooses to react to aid flow. 43,44 More broadly, there are mixed results in the literature on the fungibility of foreign aid. 45, 46 36. Ross 2006. 37. Baqir 2002. 38. Habibi 1994. The level of political liberty is measured by Gastil’s index of political rights. 39. Mauro 1997. 40. The author uses indexes of corruption from two resources: International Country Risk Guide (ICRG) and Business International (BI). The ICRG data is available for more than a hundred countries for 1982 and 1995. The BI data covers 67 countries between 1980 and 1983. The author’s corruption index is the simple average of the ICRG and BI indexes. 41. Delavallade 2006. 42. Gupta, De Mello, and Sharan 2001. 43. World Bank 1998. 44. Lu et al. 2010. 45. Feyzioglu, Swaroop, and Zhu 1998. 46. Devarajan, Rajkumar, and Swaroop 1999. 13 Though some of the empirical literature on government spending on health has looked at external aid inflows, it is worth noting that a considerable portion of foreign aid is tied to loans. Therefore, isolating the debt impact is important. One study finds that the debt-service burden shifts public spending away from health but that increases in external aid and the degree of constraints on the executive government lead to higher government spending on health as share of budget. 47 However, the positive impact of these two variables is not large enough to counteract the negative effect of the debt-service burden on public health spending. The study also implies that countries in sub-Saharan Africa would have spent even less on health without debt rescheduling. Other determinants of government allocations to health include the extent of ethnolinguistic heterogeneity in a country and female political representation. Empirical research on the impact of heterogeneity on government spending on public goods suggests that governments generally tend to spend less on health in ethnolinguistically diverse societies, controlling for the level of development, education, availability of public resources, and corruption. A study examining the relationship between female political representation and government spending on health as share of GDP in low-, middle-, and high-income countries finds that there is a positive association between female political representation and government spending on health as share of GDP, although not a strong one. 48, 49 To summarize, empirical work on determinants of government health prioritization has been sparse to date: cross-country econometric analyses suggest that factors such as democratization, lower levels of corruption, ethnolinguistic homogeneity, and more women in public office are correlated with higher shares of public spending on health. However, the numbers of studies on this issue are few, and some of the findings are not robust and are sensitive to model specification. This is an area that could benefit from further research and analysis. COUNTRY-SPECIFIC EXPERIENCES WITH REPRIORITIZATION This subsection summarizes experiences of selected country cases wherein attempts to reprioritize health in government spending have been made in recent years, some more successfully than others. In general, three sets of reprioritization efforts are evident (although these groupings are not mutually exclusive): (i) benchmarking of expenditures: this group includes countries that have focused on setting broad benchmarks for the share of health in government spending in an attempt to realize fiscal space for health; (ii) earmarking of revenues: countries that have earmarked certain taxes and other revenues for financing a larger share of health in government spending; or (iii) focus on outcomes: countries where the focus has been on improving health system outcomes, and the implications for the share of government health spending have been derived from this focus on outcomes. Countries in the latter two groups have generally been more successful in reprioritizing health than those in the former group. India is a recent and prominent example of a country that has tried to benchmark expenditures to reprioritize health. The prime minister pledged to increase public spending on health to 2 to 3 47. Fosu 2008. 48. Lie 2010. 49. Kujis 2000. 14 percent of GDP by 2012, up from about 0.9 percent of GDP in 2005. The pledge followed the 2004 election of a Congress Party–led alliance that initiated several social protection schemes aimed at benefiting the rural poor, taking advantage of an increase in overall fiscal space resulting from a period of sustained and robust economic growth. Although public financing for health has increased in the country in recent years, the 2 to 3 percent of GDP target has proved impossible to realize largely because of India’s decentralized federal structure, with health being a “state subject,” and aggregate state-level spending averaging about 75 percent of general (that is, central and state combined) public spending on health. For India to realize this target, state health spending would have had to increase by 22 to 38 percent per year since 2005, virtually an impossible scenario. 50 Since 1990, there was a steady increase in central health spending as a share of GDP, but this was offset by declining state allocations to health for most of the time period 1990 to 2010 (Figure 1.5). The decline in state-level allocations to health can be traced back to the fiscal crisis that impacted the states in the 1990s. An upward trend in state health spending began only around 2008. 51 The country remains far from attaining the 2 to 3 percent of GDP target for public spending on health: in 2011, India spent only 1.2 percent of its GDP on health. Figure 1.5 Trends and Income Elasticity of Government Health Expenditure in India, 1990–2009 Government health expenditure Income elasticity 1990-2009 1990-2009 1.2 50000 Total Government health expenditure (Rs crores) Total elasticity=0.96 .9 2010 . 15000 2008 States . Share of GDP (%) 2006 . 2004 .6 . 2002 . 5000 . 2000 States elasticity=0.87 1998 . .3 Central 1996 . 1994 Center elasticity=1.15 . 1992 . 1990 1000 0 1987 1990 1993 1996 1999 2002 2005 2008 2011 500000 1000000 2500000 7500000 Year GDP (Rs crores) Source: RBI; MOF Source: RBI; MOF; W EO India’s experience of reaching health-spending targets is not unique: Generally, countries everywhere have not realized benchmarking pledges. Other examples include Lao PDR, which plans to increase its spending on health to 9 percent of the budget (it is unclear whether the target was for total or recurrent expenditures, and whether it included externally financed government health spending) to improve access to care and reduce OOP payments. Also, Bhutan is considering whether to earmark 9 percent of government revenues for health. As mentioned 50. Berman et al. 2010. 51. La Forgia and Nagpal 2012. 15 earlier, the Abuja Declaration called for sub-Saharan African countries to allocate 15 percent of government spending for health, and WHO’s EMRO region countries have agreed to earmark 8 percent. In most countries, calls for benchmarking health’s share of government expenditures (or of revenues or GDP) have been largely aspirational. Reprioritization efforts have been more successful in some countries that have legally mandated how changes in GDP or government spending will affect the amounts allocated to health. In Vietnam, for example, in 2008 the National Assembly increased government allocations by passing Resolution no. 18/2008/NQ-QH12. According to article 2, the government would commit “…to increase the share of annual state budget allocations for health, and to ensure that the growth rate of spending on health is greater than the growth rate of overall spending through the state budget.” 52 As a result, the government share appears to have increased. 53 In Brazil, since the 1996 health financing reforms, states and municipalities have been responsible for managing and financing health care. 54 In 2000, a constitutional amendment was passed which committed budget resources at federal, state, and municipal levels. At the federal level, the amendment required a 5 percent increase in aggregate spending in 2000 in real terms, using the 1999 budget as a basis. This spending was to be adjusted according to the growth rate of nominal GDP from 2001 to 2004. Also, states and municipalities were obliged to earmark 12 percent and 15 percent of their revenues, respectively, to provide health care; and, the share of government spending on health increased from 4 percent in 2000 to 9 percent in 2011. 55 In fact, total health spending has increased over time (to about 8.8 percent of GDP in 2011), mostly due to the rise in public expenditures. This earmarking of revenues appears to have (a) increased the share of state and municipal governments in health financing; (b) provided incentives to decentralize primary care according to federal guidelines; and (c) reduced inequalities between municipalities in per capita health expenditures. Some countries also earmark specific taxes (that is, other than dedicated payroll tax earmarked for social health insurance), such as earmarked taxes on cigarettes and alcohol consumption. (See box 1.2 for a simple typology and some discussion related to earmarking.) For example, Ghana earmarks part of its VAT revenues for its national health insurance fund. A study by WHO indicates that more than 20 countries earmark tobacco tax revenue specifically for health. 56 Several countries earmark all of their tobacco tax revenue for health while other countries, such as Mongolia, Thailand, Qatar, and Bulgaria, earmark a small percentage (that is, 1 or 2 percent) of the total tobacco tax revenue to health. Still other countries, such as Tuvalu, earmark a fixed amount (that is, two cents) per cigarette for the health sector. Earmarking taxes on alcohol seems to be less common, although some countries (for example, Thailand) do have policies that allocate a portion of tax revenue from alcohol to health. After substantial opposition from interest groups, the Philippines Senate recently passed a Sin Tax Reform Law, which will earmark a portion of tax revenues for UHC and district and regional hospitals. 57 52. Grover 2011. 53. Van Tien et al. 2011. 54. WHO 2010a. 55. OECD 2005. 56. WHO 2009. 57. Tan 2012. 16 Box 1.2 On Earmarking Earmarking is a contentious issue. The basic idea behind it is using specific taxes for specific purposes, as defined by law. However, some define it more broadly. Before discussing its pros and cons, it is useful to look at the different forms as detailed by McCleary (1991), summarized in table 1B.2. 58 Type A includes “strong” earmarking, where a specific tax is levied and revenues are used for distinct purposes, including for the benefit of those paying the tax. Examples would include gasoline taxes to pay for highway investments and social health insurance premiums, and those on tobacco for health promotion activities. TABLE 1B.2 VARIETIES OF EARMARKING Type Revenue Expenditure Examples A Specific tax or Specific end use Social health insurance premium fee to provide specified coverage. B Specific tax or Broad end use ”Sin” taxes used to finance fee government health spending. C General tax Specific end use Fixed percentage of total revenue devoted to specific programs (e.g., for health). Source: McCleary 1991. Types B and C represent weaker forms of earmarking, as the connection between those paying the taxes and those receiving benefits is not direct. Type B includes taxes used for general purposes, for example, petroleum taxes used for general development. Type C general taxes earmark a part for a desired purpose, for example, if a government commits to spending a fixed percent of its revenue on health. Type D is the weakest form, with general taxes funding general uses. Source: Authors Earmarked allocations (of the type A, B, or C in box 1.2) for the health sector are generally supported by political rather than economic arguments except in the case of sin taxes, which are designed to offset negative externalities (for example, taxes on cigarettes to curb smoking). Earmarking can potentially protect certain forms of spending from political and macroeconomic vicissitudes. For example, if health spending is low or unstable, earmarking may be seen as a way to insulate health spending from other competing publicly funded activities. From an economic perspective, however, earmarking — either on spending or on revenues — is often viewed as an imposition of an unnecessary constraint on fiscal policy making, one that generally reduces a government’s macroeconomic policy flexibility and allocative efficiency. In theoretical terms, earmarking may be justified on the basis of the benefit principle in that those who are taxed get benefits from the revenue raised. Earmarking may also be favored if there is evidence of willingness to pay taxes if they are to be used for a well-defined purpose. However, in many instances, earmarking may lead to over- or under-funding of targeted activities. In addition, there are numerous examples of situations where earmarked funds have been diverted to other activities, especially in poor governance settings. 59 Earmarking certain sources of revenues may 58. McCleary 1991. 59. Prakongsai, Patcharanarumol, and Tangcharoensathien 2008. 17 also often be ineffective in being truly additional. For instance, mandating income-related contributions for social insurance have been ineffective in many countries with large levels of informality such as Indonesia, Philippines, and Vietnam. Hence, the implications of earmarking for reprioritizing health’s share of government spending are not always clear-cut and will vary significantly by the nature and magnitude of the earmarks and by country context. Over the past decade, Mexico and Thailand have both witnessed substantial increases in the share of government expenditure allocated to health. Triggered by positive political events, reforms in both countries expanded the extent of coverage and reprioritized health issues by focusing on outcomes. These two demonstrate that a combination of factors can be effective — such as where political will provided the impetus to reprioritize spending and to implement policies focused on results (see box 1.3). Box 1.3 Reprioritizing Health in Mexico and Thailand Mexico In 2003, Mexico embarked on a major health insurance reform process. The primary goal of the reform was to achieve universal health coverage by 2010. By the end of 2011, almost 98 percent of Mexico’s citizens were registered with one of the country’s three health insurance schemes: Instituto Mexicano del Seguro Social (IMSS), which covers salaried employees in the private sector; Instituto de Seguridad y Servicios Sociales de los Trabajadores del Estado (ISSSTE) for salaried workers in the public sector; and the Seguro Popular scheme for nonsalaried workers, self-employed, and families outside the labor force. 60 The financial architecture of each insurance scheme is similar, and contributions come from three sources: the federal government (social contribution), coresponsible contributor, and the beneficiary. The social contribution of the federal government is funded by general taxes, adjusted periodically for inflation, and is equal for all Mexican families.61 The second element of financing comes from co- responsible contributors (private employers for IMSS, public employers for ISSSTE, and a solidarity contribution split between the state and federal governments for Seguro Popular). The last source of funding is derived from beneficiary contributions. Contributions from employees in the formal sector for IMSS and ISSSTE are deducted from payroll. For Seguro Popular, family contributions are based on a family’s capacity to pay, with an upper limit of 5 percent of disposable income. Families in the bottom two deciles are exempt from contributions. In part due to the mobilization of resources for implementation of health reform, the Ministry of Health’s budget increased 142 percent between 2000 and 2010, while the IMSS and ISSSTE budgets increased 42 percent and 103 percent, respectively, in real terms. Some funding also comes from earmarked taxes on cigarette sales. On October 26, 2010, the Mexican Congress approved a tax increase of seven pesos on all tobacco products. This new law increased the price of a pack of 20 cigarettes by 37 percent. 62 Taxes are now 69 percent of the retail price. 63 Prior to the passage of this tax legislation, cigarette taxes in Mexico, and in Latin America more generally, were low compared to other regions of the world. In 2006 in Mexico, 60. Knaul, González-Pier, and Gómez-Dantés 2012. 61. Julio, González-Pier, and Gómez-Dantés 2006. 62. International Union Against Tuberculosis and Lung Disease 2011. 63. Framework Convention on Tobacco Control 2011. 18 taxes represented 59 percent of the total price of cigarettes compared to 75 percent or more in many high-income countries. 64 Revenues from the tobacco tax have increased 39 percent since the promulgation of the law and are expected to be used to treat tobacco-related illnesses. 65 Thailand In 2002, Thailand achieved universal health coverage. All citizens of Thailand are covered by one of the three health insurance schemes: the Social Security Scheme (SSS), which covers private sector employees (excluding their dependents); Civil Servant Medical Benefit Scheme (CSMBS), for government employees and dependents (parents, spouse, and two children under 20); and the Universal Coverage Scheme (UCS) for the rest of the population that is not covered by SSS and CSMBS. 66 The UCS reform was the most important factor in achieving universal health coverage. 67 The UCS scheme covers 75 percent of the population. This scheme is financed by general taxes and covers a comprehensive benefit package that includes clinical prevention and health promotion. Between 2002 and 2011, the UCS budget rose from $35.4 per capita to $78.8, which constituted a 71 percent real increase driven by increasing labor and material costs and increased utilization of services. Between 2002 and 2008, the general government expenditure on health increased 76 percent, from $2.7 billion to $7.4 billion. 68 The SSS, which covers 16 percent of the population, relies on payroll taxes with 1.5 percent of the salary equally contributed by the employee, employer, and government. Its per capita expenditure, at $71, is lower than with the USC. The CSMBS covers 9 percent of the population and spends $367 per beneficiary. The higher cost is due to the slightly higher level of coverage and the fact that CSMBS members have a free choice of public providers with no registration necessary. 69 Financial commitment to the UCS, increased utilization of health services, and production led to a substantial increase in government health expenditure between 2001 and 2010. Over this time period, general government expenditure on health as percentage of the general government budget increased from 9.0 to 12.7 percent. 70 In addition, general government expenditure on health as percentage of the total health expenditure increased from 56.4 percent in 2001 to 63.5 percent in 2002 and reached 75.0 percent in 2010. Within the first year of launching UCS, general government expenditure on health increased by 42 percent to US$2.7 billion in 2002 from US$1.9 billion in 2001. This upward trend has continued since then, and government spending on health reached US$7.4 billion in 2008, reflecting a 76 percent increase in real terms. In addition, the UCS budget increased 71 percent in real terms between 2002 and 2011. Even in 2009 when GDP of the country decreased by 2 percent, government expenditure on health were protected through spending cuts made by all ministries. Thailand has been earmarking tobacco and alcohol taxes since the government in 2001 passed the Health Promotion Foundation Act and established the ThaiHealth Promotion Foundation. Each year, 2 percent of the total national tax revenues from tobacco and alcohol products are earmarked to finance ThaiHealth. This amount is equivalent to US$35 million per year. 71 Thailand has experienced considerable political instability over the last decade. Between 2001 and 2011 there were seven different government, five general elections, as well as a coup d’etat in 2006. Between 2001 and 2010, the minister of health changed eight times, and there were six different Ministry of 64. Jimenez-Ruiz 2008. 65. Ibid. 66. McManus 2012. 67. Ibid. 68. Ibid. 69. Ibid. 70. WHO, Global Health Expenditure Database http://apps.who.int/nha/database/StandardReport.aspx?ID=REP_WEB_MINI_TEMPLATE_WEB_VERSION&COUNTRYKEY =84668. Accessed November 26, 2012. 71. WHO 2010b. 19 Public Health (MOPH) permanent secretaries. Despite political instability in the country, the UHC received considerable support from all political parties since the insurance scheme was highly popular among the people. Source: Authors One additional constraint to reprioritization that deserves note is the perception, especially among ministries of finance and planning, that the health sector is inefficient and unproductive. From a practical perspective, health ministries could do more to develop evidence-based cost- effective expenditure and implementation plans, ensuring that resource allocations are fully and effectively absorbed, and providing evidence of impact of interventions on health outputs and outcomes. 72 This can also be helpful in identifying unproductive or pro-rich spending in other sectors. For example, in a country such as Indonesia, health gets a small percentage of central government resources despite policy commitments to attaining UHC in the country by 2019: in 2011, health’s share of the central government budget was only 1.6 percent, compared to 18.2 percent for fuel subsidies, which disproportionately benefit the better-off segments of Indonesia’s population (Figure 1.6). 73 72. Powell-Jackson, Hanson, and McIntyre 2012. 73. Bi and Tandon 2013. 20 Figure 1.6 Share of Central Government Budget to Health, Education, and Fuel Subsidies, 1995–2011 Share of cenral government budget 1995-2010 100 % of central government budget 80 Other 40 60 Fuel subsidies 20 Education Health 0 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 19 19 19 19 19 20 20 20 20 20 20 20 20 20 20 20 20 Source: MOF (various years) VI. SUMMARY AND CONCLUSIONS The share of overall government expenditure devoted to health is often used as a metric to gauge the extent to which health is prioritized by governments. While there is a clear income gradient in health’s share of government expenditure across countries, significant variations persist even after controlling for income. This paper has provided a global overview of health’s share of government spending and summarized some of the key theoretical and empirical perspectives on allocation of public resources to health vis-à-vis other sectors. We have focused attention on this issue from the perspective of reprioritizing government spending on health, one of the modalities for realizing fiscal space for health. Although a government’s spending on health is only one contributing factor to health outcomes in any country, understanding why some governments allocate a higher share of their resources to health than others is important and merited for several reasons. First, additional public financing for health is a prominent policy issue for many low- and middle-income countries seeking to improve the levels and distribution of health-related MDGs, reduce OOP spending for health, and attain and sustain UHC. Second, government spending on health is ultimately a reflection of a decision to prioritize health vis-à-vis government resources, and thus serves as an intermediate link between a country’s overall macrofiscal environment and the health sector. Financing is just one of many ways governments can intervene in the economy. From a normative economic theory perspective, the rationale for government intervention in the health sector is based on “market failure” and equity-related arguments. Government interventions can — under certain conditions and in principle — be used to improve efficiency when market failures lead to suboptimal social welfare outcomes, and to improve equity when market allocations lead to outcomes that are perceived to be unfair. Three broad forms of market failure prominent in the health sector are frequently cited as the justification for government intervention in health, specifically: (i) the presence of externalities; (ii) the public good nature of certain health interventions; and (iii) the presence of extensive information asymmetries. 21 Unfortunate from the perspective of reprioritizing health, the same economic theory rationales used to justify public financing for health also apply to government interventions in many other sectors including defense, education, food, housing, water, sanitation, and infrastructure, to name a few. In terms of allocating resources across sectors, public finance theory argues that, given the real costs to society of raising revenues, the implications for public expenditure for any sector (including health) are that benefits from such aggregate expenditures should exceed aggregate costs of raising revenues, and that sectors should compete for allocation of scarce resources up to the point whereby the marginal benefit of an additional dollar of spending is equal across sectors. There are limits to using the normative approach to understand the discrepancies in health expenditure between governments: with its focus on the ideal or optimal role for policy intervention by a benign government that intends to maximize social welfare, the normative approach ignores the substantial disconnect between a theoretically ideal set of policies and achievable actions. Positivistic perspectives on the role of government help clarify how governments make choices, focusing heavily on the behavior of voters (in democracies), on private influences on public policy choices, and on inherent individual incentives of politicians and bureaucrats. Public choice theory argues that elected officials, bureaucrats, and voters are no different than people in any other market and are guided by their own self-interests. Applied to health, public choice theory predicts that bureaucrats and those in power will support increases in government spending on health when doing so would advance their careers or enable them to stay in power. In a democratic setting, voters will support candidates who call for increased government spending on health if and when voters perceive they will be better-off as a result. Theory argues that the median voter will be the decisive voter in a functional democracy. From the perspective of reprioritization, normative theory implies that the health sector would need to demonstrate that social benefits of additional public spending outlays exceed the costs of financing this increase in spending, both in terms of the additional costs of raising revenues and also in terms of foregone public spending in other sectors. On the other hand, positivistic theories of public resource allocations generally imply that demonstrations of social welfare–enhancing aspects of additional government health spending will not necessarily be effective in increasing allocations to health. Instead, enhancing democratization, improving citizen information, increasing government accountability, and reducing corruption may be more effective strategies to ensuring that health is accorded the priority it merits from a public choice perspective. Empirical evidence on factors associated with government preferences for public financing of the health sector is generally scarce. In the limited empirical literature that exists, a range of economic and noneconomic factors has been found to influence government allocations for health. Prominent among these are democracy, corruption, foreign aid, ethnolinguistic heterogeneity, and female representation in the polity of a country. On average, cross-country analyses suggest that democratic societies tend to devote a larger share of their public resources to health even after controlling for confounding factors, while more corrupt societies do the opposite. The effect of donor assistance for health on government own-sourced revenue priorities is a topic of great interest among donors and policymakers alike. The evidence suggests that the effect is mixed: in some countries, DAH is fungible and in others it is not. Where foreign aid is tied to loans, the evidence suggests that the debt-service burden shifts public spending away from health. Research on the impact of heterogeneity on government spending on public goods suggests that governments generally tend to spend less on health in ethnically diverse societies. 22 There are few empirical studies on the issue of health prioritization; furthermore, some findings are not robust and are sensitive to model specification. This is an area that could benefit from further research and analysis. Evidence from case studies on reprioritization suggests that country-specific political economy considerations are key and that results-focused reform efforts — in particular efforts to explicitly expand the breadth and depth of health coverage as opposed to efforts focused only on government budgetary targets — are more likely to result in sustained and politically feasible prioritization of health from a fiscal space perspective. 23 REFERENCES African Union. 2001. “Abuja Declaration on HIV/AIDS, Tuberculosis, and Other Related Infectious Diseases, African Summit on HIV/AIDS, Tuberculosis, and Other Related Infectious Diseases, 24-27 April: Abuja, Nigeria.” http://www.un.org/ga/aids/pdf/abuja_declaration.pdf. Accessed December 4, 2012. Atim, C., L. Fleisher, L. Hatt, S. Musau, and A. Arur. 2008. Health Financing in Africa Today: Challenges and Opportunities, Washington, DC: Africa’s Health in 2010, Academy for Educational Development, and Bethesda, MD: Health Systems 20/20 Project, Abt Associates Inc. Baqir, R. 2002. “Social Sector Spending in a Panel of Countries,” IMF Working Paper, Washington, DC. Berman, P., R. Ahuja, A. Tandon, S. Sparkes, and P Gottret. 2010. “Government Health Financing in India: Challenges in Achieving Ambitious Goals,” HNP Discussion Paper, Washington, DC: World Bank. Bi, X., and A. Tandon. 2013. “Fiscal Space for Jamkesmas: Implications for Universal Health Coverage in Indonesia,” Mimeo, Washington, DC: World Bank. Delavallade, C. 2006. “Corruption and Distribution of Public Spending in Developing Countries,” Journal of Economics and Finance, 30(2): 222-239. Devarajan, S., A.S. Rajkumar, and V. Swaroop. 1999. “What Does Aid to Africa Finance?” Policy Research Working Paper, 2092, World Bank, Washington, DC. Dixit, A. 2009. “Democracy, Autocracy, and Bureaucracy,” Draft for Discussion. November 27, 2009. Feyzioglu, T., V. Swaroop, and M. Zhu. 1998. “A Panel Data Analysis of the Fungibility of Foreign Aid,” World Bank Economic Review, 12(1): 29-58. Fleisher, L., A. Leive, and G. Schieber. 2013. “Taking Stock of Fiscal Health: Recent Trends in Global, Regional, and Country-Level Health Financing,” Mimeo, Washington, DC: World Bank. Fosu, A. K. 2008. “Implications of the External Debt-Servicing Constraint for Public Health Expenditure in Sub-Saharan Africa,” Oxford Development Studies, 36(4): 363-377. Framework Convention on Tobacco Control. 2011. Tobacco Taxes Success Story, Geneva. http://www.fctc.org/images/stories/success_Mexico_en.pdf. Accessed December 4, 2012. Gillingham, R. 2013. “Primer on Fiscal Policy for Health Policy Makers,” Mimeo, Washington, DC: World Bank. 24 Gottret, P. and G. Schieber. 2006. Health Financing Revisited, Washington, DC: World Bank. Gruber, J. 2011. Public Finance and Public Policy, New York: Worth Publishers. Grover, A. 2011. Report of the Special Rapporteur on the Right of Everyone to the Enjoyment of the Highest Attainable Standard of Physical and Mental Health, Geneva: Human Rights Council. Guerard, Y., M Weiner, C. Rokx, G. Schieber, P. Harimurti, and A. Tandon. 2011. “Acturial Costing of Universal Health Coverage in Indonesia: Options and Preliminary Results,” HNP Discussion Paper, Washington, DC: World Bank. Gupta, S., L. De Mello, and R. Sharan. 2001. “Corruption and Military Spending,” European Journal of Political Economy, 17(4): 749-777. Haacker, M. 2010. “Macroeconomic Constraints to Health Financing: A Guide for the Perplexed,” World Health Report Background Paper No. 49, Geneva: World Health Organization. Habibi, N. 1994. “Budgetary Policy and Political Liberty: A Cross-Sectional Analysis,” World Development, 22(4): 579-86. Heller, P. 2006. “The Prospect of Creating ‘Fiscal Space’ for the Health Sector,” Health Policy and Planning, 21(2): 75-79. International Union Against Tuberculosis and Lung Disease. 2011. Mexico Approves Very Significant Increase in Tobacco Tax, http://www.theunion.org/Regions/LatinAmerica/index.php/Union-Region-News/Mexico- approves-very-significant-increase-in-tobacco-tax.html. Accessed December 4, 2012. Jack, W. 1999. Principles of Health Economics for Developing Countries, Washington, DC: World Bank. Jimenez-Ruiz, J. A. 2008. “The Impact of Taxation on Tobacco Consumption in Mexico,” Tobacco Control, (17): 105-110. Julio, F., E González-Pier, and O. Gómez-Dantés. 2006. “Comprehensive Reform to Improve Health System Performance in Mexico,” Lancet, 368(9546): 1524-1534. Knaul, F. M, E González-Pier, and O Gómez-Dantés. 2012. “The Quest for Universal Health Coverage: Achieving Social Protection for All in Mexico,” Lancet, 380(9849): 1259-1279. Kujis, L. 2000. “The Impact of Ethnic Heterogeneity on the Quantity and Quality of Public Spending,” IMF Working Paper, Washington, DC. La Forgia, G. and S Nagpal. 2012. Government-Sponsored Health Insurance in India: Are You Covered?, World Bank, Washington, DC. 25 Lie, G. 2010. “Does Government Health Spending Increase When Women Decide?” Master Thesis. Department of Health Management and Health Economics, Faculty of Medicine University of Oslo. https://www.duo.uio.no/handle/123456789/30228. Accessed November 27, 2012. Liang, L., and Mirelman, A. 2012. “An Assessment of Sociopolitical Determinants, International Aid and Government Health Expenditure: A Cross-Country Analysis over 1995–2010,” HNP Discussion Paper, Washington, DC: World Bank. Lu, C., M. T. Schneider, P. Gubbins, and K. Leach-Kemon. 2010. “Public Financing of Health in Developing Countries: a Cross-national Systematic Analysis,” The Lancet, 375(9723): 1375- 1387. Mauro, P. 1997 “The Effects of Corruption on Growth, Investment, and Government Expenditure: Across-country Analysis,” in Corruption and the World Economy, edited by K.A. Elliot. McCleary, W. 1991. “The Earmarking of Government Revenue: A Review of Some World Bank Experience,” World Bank Research Observer, VI (1991): 81-104. McManus, J. 2012. Thailand’s Universal Coverage Scheme: Achievements and Challenges, http://www.gurn.info/en/topics/health-politics-and-trade-unions/development-and-health- determinants/development-and-health-determinants/thailand2019s-universal-coverage-scheme- achievements-and-challenges. Accessed November 26,2012. Meltzer, R. and A. Richard. 1981. "A rational theory on the size of government," Journal of Political Economy, 89: 914-927. Musgrave, R. A. 1959. Theory of Public Finance: A Study in Public Economy, New York: McGraw-Hill. Organization for Economic Co-operation and Development. 2005. OECD Economic Surveys Brazil, Paris: OECD. Plümper, T., and C. Martin. 2003. “Democracy, Government Spending, and Economic Growth: A Political-economic Explanation of the Barro effect,” Public Choice, 117: 27-50. Powell-Jackson, T., K. Hanson, and D. McIntyre. 2012. “Fiscal Space for Health: A Review of the Literature.” RESYST Working Paper, London. Prakongsai, P., W Patcharanarumol, and V Tangcharoensathien. 2008. “Can Earmarking Mobilize and Sustain Resources to the Health Sector?” Bulletin of the World Health Organization, 86(11): 898-901. Roberts, K. 1977). “Voting Over Income Tax Schedules,” Journal of Public Economics, 8: 329- 340. 26 Ross, M. 2006. “Is Democracy Good for the Poor,” American Journal of Political Science, 50(4): 860-874. Sachs, J (2001), Macroeconomics and Health: Investing in Health for Economic Development, Geneva: Report of the Commission on Macroeconomics and Health. _____. 2012. “Achieving Universal Health Coverage in Low-income Settings,” Lancet, 380: 944-947. Savedoff, W. 2007. “What Should a Country Spend on Health Care?” Health Affairs, 26:962- 970. Shelton, C. A. 2007. “The Size and Composition of Government Expenditure,” Journal of Public Economics, 91: 2230-2260. Taskforce on Innovative Health Financing for Health Systems. 2009. Working Group 1 Report: Constraints to Scaling Up and Costs, Geneva: World Health Organization. Tan, K. 2012. Senate Passes Sin Tax Bill Adding at least P12 per Cigarette Pack GMA News, November 20, 2012. http://www.gmanetwork.com/news/story/283090/news/nation/senate- passes-sin-tax-bill-adding-at-least-p12-per-cigarette-pack. Accessed November 27, 2012. Tandon, A. 2007. “Measuring Government Inclusiveness: An Application to Health Policy,” Asian Development Review, 24: 32-48. Touhy, C and S. Glied. 2011. “The Political Economy of Health Care,” in Oxford Handbook of Health Economics, edited by Sherry Glied and Peter Smith. United Nations. 2012. The Millennium Development Goals Report 2012, New York: United Nations. Van Tien, T., H. T. Phuong, I. Mathauer, and B. T. K. Phuong (2011), “A Health Financing Review of Viet Nam with a Focus on Social Health Insurance.” Geneva: World Health Organization. http://www.who.int/health_financing/documents/oasis_f_11-vietnam.pdf. Accessed November 26, 2012. World Bank. 1997. World Development Review: The State in a Changing World, Washington, DC: World Bank, Washington, DC. _____. 1998. Assessing Aid: What Works, What Doesn’t, and Why, World Bank, Washington, DC. World Health Organization. 2009. Health Financing Strategy for the Asia Pacific Region (2010- 2015), Manila: WHO Regional Office for the Western Pacific. _____. 2010c. Strategic Directions to Improve Health Care Financing in the Eastern Mediterranean Region: Moving Towards Universal Coverage 2011–2015, Regional Committee for the Eastern Mediterranean. 27 _____. 2010a. “Brazil's March Towards Universal Coverage,” Bulletin of the World Health Organization. http://www.who.int/bulletin/volumes/88/9/10-020910/en/index.html. Accessed November 26, 2012. _____. 2009. WHO Report on the Global Tobacco Epidemic 2009, Geneva: WHO. http://whqlibdoc.who.int/publications/2009/9789241563918_eng_full.pdf. Accessed November 27, 2012. _____. 2010b. The Political Economy of Tobacco Taxation, WHO Technical Manual on Tobacco Tax Administration. http://whqlibdoc.who.int/publications/2010/9789241563994_eng.pdf. Accessed November 26, 2012. _____. 2012. Global Health Expenditure Database, Geneva: WHO. http://apps.who.int/nha/database/standardreport.aspx?id=rep_web_mini_template_web_version& countrykey=84668. Accessed November 26, 2012 28 About this series... This series is produced by the Health, Nutrition, and Population Family (HNP) of the World Bank’s Human Development Network. The papers in this series aim to provide a vehicle for publishing preliminary and unpolished results on HNP topics to encourage discussion and debate. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations or to members of its Board of Executive Directors or the countries they represent. Citation and the use of material presented in this series should take into account this provisional character. For free copies of papers in this series please contact the individual authors whose name appears on the paper. Enquiries about the series and submissions should be made directly to the Editor Martin Lutalo (mlutalo@worldbank.org) or HNP Advisory Ser- vice (healthpop@worldbank.org, tel 202 473-2256, fax 202 522-3234). For more information, see also www.worldbank.org/hnppublications. The world bank 1818 H Street, NW Washington, DC USA 20433 Telephone: 202 473 1000 Facsimile: 202 477 6391 Internet: www.worldbank.org E-mail: feedback@worldbank.org