Summary: We compile the first large cross-country panel dataset of public sector performance and efficiency, encompassing 114 countries on all income levels from 1980 to 2006, with about 1,800 country-year observations for the education sector and about 900 observations for health. We regress these indicators on potential economic, institutional, demographic, and geographic determinants. Our most resounding conclusion is that higher government expenditure relative to GDP tends to be associated with lower efficiency in the respective sector. Moreover, we find that richer countries exhibit better public sector performance and efficiency, and that institutional and demographic factors also play a significant role.
From the introduction;
The efficiency of government spending has become one of the key issues in public finance. In the advanced economies and many transition countries, higher efficiency of spending seems to be the only way to avoid that public services are squeezed out between the opposing forces of age-related expenditure and rising tax competition (Heller and Hauner, 2006). In low-income countries, increased expenditure efficiency will have to complement increased social expenditure if the Millennium Development Goals are to be reached. Emerging markets, in turn, may seem under less pressure of this kind, given their rapid growth, but it is well-known that the demand for public services tends to rapidly increase as countries become richer (the so-called Wagner effect), and higher efficiency will be the only way to avoid a large increase in the tax burden. Moreover, good government is also of more general concern, as it has been shown, for example by Easterly and Levine (1997), that it is a crucial determinant of economic growth.
Against this background, government efficiency at the aggregate level has thus become the subject of a rapidly growing literature, including key contributions by Gupta and Verhoeven (2001), Tanzi and Schuknecht (1997, 2000), and Afonso et al. (2005). These studies typically measure public sector efficiency by relating government expenditure to socio-economic indicators that are assumed to be targeted by public spending, such as education enrolment ratios or infant mortality; the results of their cross-country examinations suggest substantial efficiency differences between countries, irrespective of their income level.
It is only logical that the more recent literature has started to examine the determinants of these efficiency differences: Afonso and Aubyn (2006) examine the differences in the efficiency of education spending in the OECD and find that income levels and parents’ education explain a large part of the variation. Afonso et al. (2006) examine public sector efficiency in the new member states of the European Union and conclude that security of property rights, income level, competence of the civil service, and the population’s education level affect efficiency. Hauner (2008) examines the determinants of expenditure efficiency for Russia’s regions. His results show that higher government efficiency tends to be associated, in particular, with higher per capita income, a smaller share of federal transfers in subnational government revenue, better governance, stronger democratic control, and smaller government expenditure. Examining government performance—but not efficiency as does this paper—and using a cross-section but not a panel, La Porta et al. (1999) find that countries that are poor, close to the equator, ethnolinguistically heterogeneous, use French or Socialist laws, or have high proportions of Catholics or Muslims exhibit inferior performance.
Here we make two main contributions to this literature. First, we compile the first large cross-country panel dataset of government efficiency, encompassing 114 countries on all income levels from 1980 to 2006, with about 1,800 country-year observations for education and about 900 observations for health. Previous papers have either studied a more limited number of countries or only a cross-section. Second, we are the first to examine the policy and environmental determinants of efficiency not only for such a large panel, but also for a much broader universe of regressors than previous studies, similar to the undertaking of Friedman et al. (2000) for unofficial economic activity. However, due to important but—at this stage— irresolvable data shortcomings that we will discuss, we see this paper only as a first step to a better understanding of broad global trends in expenditure efficiency and its determinants.
In the first part of the empirical analysis, we compute three types of scores per policy sector for each country: (i) public sector performance (PSP) that measures only outcomes; (ii) public sector efficiency (PSE) that relates outcomes to expenditure with a simple ratio; and (iii) Data Envelopment Analysis (DEA) scores that measure efficiency relative to a frontier. As most of the literature, we focus on the health and education sectors, because no useful data on outcomes for other sectors is widely available. The results broadly suggest that public sector performance and efficiency have been trending upward over time.
In the second part of the analysis, we relate these scores econometrically to a battery of possible determinants related to economics, demographics, geography, and institutions. As multicollinearity and endogeneity loom large, we run univariate as well as multivariate regressions and alternatively use fixed and random effects and system GMM estimators.
The most resounding conclusion is that higher expenditure relative to GDP tends to be associated with lower efficiency. Strikingly, in the education sector, even the relationship between spending and performance is tenuous, in stark contrast to the health sector. This means that citizens not only tend to get a declining marginal “bang for the buck” as their governments spend more on education, but possibly no additional “bang” at all. This finding has important policy implications, given that “throwing money at issues” is so often the first political reflex, whether it comes to Millennium Development Goals or the need of advanced economies to improve education in the face of challenges from globalization and technological progress.
We also find that institutions matter, as well as demographic and geographic factors. Public sector performance in education is stronger where governments are more accountable, while public sector performance and efficiency deteriorate as corruption becomes worse. More mundane demographic and geographic factors also play a role: for example, a relatively larger youthful population reduces performance and efficiency in the education sector, while higher population density (allowing for returns to scale) improves performance.