More Health Expenditure, Better Economic Performance? Empirical Evidence From OECD Countries

Abstract

Recent economic downturns have led many countries to reduce health spending dramatically, with the World Health Organization raising concerns over the effects of this, in particular among the poor and vulnerable. With the provision of appropriate health care, the population of a country could have better health, thus strengthening the nation’s human capital, which could contribute to economic growth through improved productivity. How much should countries spend on health care? This study aims to estimate the optimal health care expenditure in a growing economy. Applying the experiences of countries from the Organization for Economic Co-Operation and Development (OECD) over the period 1990 to 2009, this research introduces the method of system generalized method of moments (GMM) to derive the design of the estimators of the focal variables. Empirical evidence indicates that when the ratio of health spending to gross domestic product (GDP) is less than the optimal level of 7.55%, increases in health spending effectively lead to better economic performance. Above this, more spending does not equate to better care. The real level of health spending in OECD countries is 5.48% of GDP, with a 1.87% economic growth rate. The question which is posed by this study is a pertinent one, especially in the current context of financially constrained health systems around the world. The analytical results of this work will allow policymakers to better allocate scarce resources to achieve their macroeconomic goals.

Introduction

The global financial crisis has led many countries to reduce spending dramatically, with health care being a common target of such reductions. However, the provision of health care is important for improving a population’s health, which in turn can lead to more productivity, better economic performance, and then more fiscal resources. However, could better economic performance be achieved through more health spending? The findings of the current literature with regard to the influences of health expenditure on economic growth are ambiguous.^1,2 How much should countries spend on health care?³ Previous research has not suggested the optimal level of health expenditure using empirical evidence. The current study thus aims to investigate whether raising health expenditure can effectively improve economic performance, and so it also aims to find the optimal level of health expenditure to maximize economic growth.

Methods and Specifications

Existing studies have examined various aspects of the health expenditure–economic growth nexus. However, the historical and institutional backgrounds of each country are unique, and thus the question is a rather complex one. An empirical methodology that controls for cross-country variations in certain political and economic institutional characteristics, and successfully deals with differences across countries, would be able to offer a better framework for statistical inferences. The use of simple ordinary least squares (OLS) estimation could lead to biased and inconsistent parameter estimates, due to regressor endogeneity. This article overcomes these issues by introducing the generalized method of moments (GMM) technique to derive the design of the estimators of interest.^4,5

Subjects

Health expenditure in developed countries has risen faster than gross domestic product (GDP). Organization for Economic Co-Operation and Development (OECD) countries are all developed nations. The well-established social security systems in these counties have contributed to improvements in the health of their populations, although this has also caused a fast rise in health expenditure. This research provides rigorous scientific examinations to examine the following issues: (1) how health expenditure affects an economy’s GDP growth, (2) whether there is an optimal ratio of health expenditure relative to GDP, and (3) whether and how the other factors influence economic performance. The results could be used to derive the appropriate health expenditure level to obtain better economic performance and social well-being.

Analytical Framework

The Grossman model has yielded considerable insights into the determinants of health, as well as into the allocation of time and money into health production.⁶ The working hours and productivities of workers in poor health both decline, and this is one economic side effect of a health care policy.^7,8 A key premise of the literature is that good health enhances worker productivity and improves economic outputs.^9,10

Human capital can only be productive if it is healthy. Health expenditures thus influence economic growth by improving and enhancing productivity. The production function of per capita GDP(y) could be a function of health expenditure (h) and a vector (X) of public services and socio-economic variables that affect real per capita GDP:

$$y=f(h,X).$$

Empirical Tests

Economic growth rates depend on the combined influences of several factors, rather than just one or two key elements. Provision of health services requires the use of public resources. At the macroeconomic level, there is a potential trade-off between health and other public expenditures that may affect economic performance. An increase in the ratio of health expenditure to GDP affects economic performance through two channels. The first is due to the crowding out effect, whereby an increase in health expenditure reduces expenditure on other public spending, which might also be able to improve productivity. This channel tends to reduce economic growth. The second channel is related to the effects of improvements in public health, whereby an increase in health spending tends to improve the productivity of workers. This channel leads to better economic performance. The net effect of a rise in health expenditure on economic growth depends on the relative strength of these two channels. Obviously, a rise in the share of health expenditure favors (deters) the economic growth rate if it improves (deteriorates) workers’ health. The relationship between health expenditure and economic growth may be non-linear, and this issue needs to be examined. This research then tries to find the optimal level of health expenditure. The model used here is typical of economic growth determinant models, and is as follows: ¹¹

$$\begin{array}{l}\log y_{it}={\omega }_1\log y_{it-1}+{\omega }_2{h}_{it}+\\ {\omega }_3{h}_{it}^2+{\omega }_4{X}_{it}+{\epsilon }_{it}t=1,2,\dots ,\mathrm{T};i=1,2,\dots ,N,\end{array}$$

where log y_it represents per capita real GDP in logarithmic form, h_it represents the ratio of health expenditure relative to GDP, $h_{it}^2$ represents the square of the ratio of health expenditure relative to GDP, ε_it represents an idiosyncratic error term, and i and t represent country and time period, respectively. The vectors of X_it are composed of socio-economic and demographic explanatory variables. Our primary interest is how health expenditure affects economic growth, and whether other factors influence this relationship.

Econometric Method

GMM has been widely applied by researchers, to examine subjects as diverse as financial and economic issues to public health studies on the effects of AIDS deaths on households since the 1990s.⁵ This econometric technique is specifically designed to extract causal lessons from data or observations (whether countries, hospitals, or patients), each of which is observed only annually over 5 or 10 years, as well as to resolve the issue of multicollinearity between explanatory variables.

Data Description

Applying the experiences of OECD countries, the panel datasets of the dependent and explanatory variables are observed over the period from 1990 to 2009, thus excluding the global financial crisis that arose after this. The following socio-economic and demographic explanatory variables are examined: the ratio of health expenditure relative to GDP; the square of the ratio of health expenditure relative to GDP; the ratios of military expenditure, education expenditure, and government consumption expenditure to GDP; the tax rate; the growth rate of the college enrollment rate; the inflation rate; the openness and the corruption indexes; the development index; the aggregate birth rate; the population; the ratio of young people in the population; the ratio of elderly people in the population; and the per capita GDP growth rate for the last period in logarithmic form. The survey data are collected from World Development Indicator for the period from 1990 to 2009. Table 1 summarizes the descriptive statistics and Figure 1 presents a scatter plot showing the relationship between government health expenditure and the economic growth rate over the period 1990 to 2009.

Table 1.

Summary Statistics.

Variables	Observations	Mean	SD	Minimum	Maximum
Growth rate of per capita GDP (%)	677	1.87	3.43	−19.73	12.07
Health expenditure/GDP (%)	648	5.48	1.5	0	8.98
Military expenditure/GDP (%)	650	2.10	1.58	0.05	15.50
Education expenditure/GDP (%)	524	5.18	1.26	2	8.4
Government Consumption expenditure/GDP (%)	679	19.1	4.46	8.38	30
The tax rate (%)	655	21.12	6.58	8.03	38.34
The growth rate of college enrollment rate (%)	622	50.63	19.92	9.81	103.87
Openness	677	82.31	46.7	16.01	319.55
Inflation rate	677	8.25	36.81	−5.56	873.64
The population	680	3.38e+07	5.28e+07	254 800	3.07e+08
Per capita GDP	680	18 947.59	11 244.24	2744.22	56 388.99
The youth/the population (%)	680	19.82	4.74	13.44	38.46
The elderly/the population (%)	680	13.43	3.49	3.79	22.05
The corruption index	680	6.98	1.92	2.66	10
The developing index	680	0.79	0.41	0	1
The aggregate birth rate	680	1.72	0.42	1.08	3.4

Note. GDP = gross domestic product.

Figure 1.

Health expenditure and economic growth rate in OECD over the period 1990-2009.

Note. GDP = gross domestic product; OECD = Organization for Economic Co-Operation and Development.

After identifying the factors that might affect the per capita GDP growth rate in log form and by applying the GMM estimators in STATA 10.0 to the survey data, this research further estimates the significance of the coefficients corresponding to each of the above-mentioned explanatory variables.

Results

The GMM estimated results for per capita GDP growth rates in log form, as well as for the other variables, are presented in Table 2. The results for AR(1), AR(2), and the Sargan test are 0, 0.876, and 0.051, respectively, and indicate that the selection of variables is valid, the estimation equation is correctly specified, and the estimation results are robust.

Table 2.

System GMM Regression Estimates of Per Capita GDP Growth Rates in Log on Selected Variables.

Period of test = 1990~2009	AR(1) = 0
No. of observations = 432	AR(2) = 0.876
No. of groups = 33	Sargan test = 0.051
Per capita GDP growth rate in log	Coefficient	SE
Health expenditure/GDP (%)	0.2197	0.0936**
The square of health expenditure/GDP (%)	−0.0291	0.0092***
Military expenditure/GDP (%)	0.0241	0.0203
Education expenditure/GDP (%)	−0.0173	0.0225
Government consumption expenditure/GDP (%)	0.0094	0.0078
The tax rate (%)	−0.0028	0.0050
The growth rate of college enrollment rate (%)	0.0032	0.0015**
The population	0.0039	0.0658
The youth/the population (%)	0.0184	0.0216
The elderly/the population (%)	−0.0041	0.0191
The inflation rate	0.0010	0.0034
Openness	0.0012	0.0008
The corruption index	0.0331	0.0166**
The developing index	−0.1595	0.0682**
Aggregate birth rate in log	−1.2221	0.5600
Per capita GDP growth rate in log of the last period	0.1645	0.0491***
Constant	−0.3926	0.8902

Note. GMM = generalized method of moment; GDP = gross domestic product.

^*Significant at 10%. **Significant at 5%. ***Significant at 1%.

Based on Table 2, the growth rate of the college enrollment rate, the corruption index, the development index, the aggregate birth rate, and per capita GDP growth rate in log form of the last period all statistically and significantly influence per capita GDP growth rate. Multiplying the coefficients of these socio-economic explanatory variables by the respective mean values from the summary statistics in Table 1 and adding them up to the constant term yields the specifications of Equation (3), and this can then be used to explore the exact relationship between health expenditure and economic performance. The main concern of this study is to explore how health expenditure affects economic performance. The impact of health expenditure on economic performance is estimated as follows:

$$\mathrm{Per\; capita\; GDP\; growth\; rate\; in\; log}=0.4772+0.2197h-0.0291h^2,$$

where h presents the ratio of health expenditure to GDP.

Figure 2 presents the estimated graph, and shows the relationship between health spending and economic growth is non-linear. When the ratio of health spending to GDP is less (more) than the optimal level of 7.55%, increases in health spending are associated with faster (slower) economic growth. The maximum economic growth is 3%. The real level of health spending is 5.48% of GDP in OECD countries, and the economic growth rate is 1.87% over the period 1990 to 2009.

Figure 2.

The relationship between health expenditure and economic growth rate.

Note. GDP = gross domestic product.

Discussion

In general, people with better health are capable of producing more goods and services than those in poor health, leading to faster economic growth. Nevertheless, an increase in health expenditure, which could improve the health of the workforce, might crowd out other expenditure, such as spending on infrastructure inputs, that could also stimulate economic growth. The impact of increases in health expenditure on economic growth is thus ambiguous.

Over the coming decades both developing and industrialized countries will face sharp rises in health expenditure, as well as other long-term health care challenges, because of their aging populations. Our findings indicate that it is worthwhile for governments to increase investment in health until it reaches an optimal level to achieve greater economic development. The Dartmouth Atlas project and researchers agree that more spending on health does not necessarily equate to better care.^12-14 The current study challenges the assumption that increased spending in health care is inevitable and unavoidable, and tries to find the effective level of health spending for stimulating economic growth, while also improving the population’s well-being. Based on empirical evidence from OECD countries, when the ratio of health spending to GDP is less than the optimal level of 7.55%, increases in health spending effectively lead to better economic performance. Otherwise, such spending does not cause improvements in care. With the provision of appropriate health care, the population could have better health, and thus a nation’s human capital would be stronger and better able to contribute to economic growth through improved productivity.¹⁵

Recent economic downturns tend to have the greatest effects on working age adults. The World Health Organization has raised concern over the financial crisis’ impact on global health, in particular among the poor and vulnerable.^7,8 Furthermore, this article finds that increases in the ratio of the elderly people in the population do not affect economic growth significantly. Our findings reflect that appropriate spending on health care supports economic development. The overarching empirical question that is posed by this research is a pertinent one, especially in the context of the financially constrained health systems that now operate around the world, which are being squeezed by multiple developments, including (1) a financial squeeze due to the global economic downturn, (2) rapidly expanding and increasingly costly treatment regimes, and (3) rapidly aging populations. It is thus more crucial than ever to carry out an appropriate policy analysis at the macroeconomic level, which will allow policymakers to better allocate scarce resources in the public sector. In future work, we shall investigate the impact of preventative health care on a nation’s health and economic performance, through its effects on improved health and productivity, as well as reduced future demand for health care and possible reductions in health expenditure. This issue is important, as the knowledge gained from such research could help reduce the deficits in public budgets caused by expensive health systems.