1. Introduction
Population aging is becoming more widespread and accelerated. The global population aged 60 years and over was only 205 million in 1950, increasing to 810 million in 2012, and it is estimated to reach two billion by 2050. By then, the proportions of people aged 60 and over will be 10% in Africa, 24% in Asia and Oceania, 25% in Latin America and the Caribbean, 27% in North America, and 34% in Europe [
1]. On the other hand, the proportion of people aged 65 and over increased from 7% to 14% in France in the space of 116 years, compared to only 68 and 46 years in the United States and the United Kingdom, respectively. Surprisingly, it is expected to take only 26 and 21 years in China and Brazil, respectively [
2]. Population aging is often associated with health and medical services and is a major driver of growth in social welfare spending [
3]. Due to the natural situations of physiological decline and decreased resistance, the elderly inevitably behaves more demand in healthcare and medical services [
4]. As known to all, pool health and disease burden can bring a multitude of unfavorable macroeconomic consequences [
5]. Along with the rising population aging and healthcare expenditure, many governments have been under more and more fiscal pressure. Japan, one of the most aging countries in the world, has nearly tripled its share of spending on medical care, nursing care, healthcare, and other social security areas from 17.5% in 1990 to 32.7% in 2014 [
6]; a similar situation as in Europe [
7]. As can be seen, the health conditions of the increasing elderly population are expected to be a challenge to public finance, and thereby it should be attached importance to figure out the fiscal effects and mechanisms for policy-making.
The macroeconomic impacts of the demographic structure have been extensively studied in many empirical and theoretical literature [
8]. The fiscal implications of population aging have been one main subject of considerable academic attention. Past research can be broadly divided into three categories. The first focuses on exploring whether population aging affects public finance. Hondroyiannis and Papapetrou [
9] found the low fertility and high old-age dependency ratios that characterized the demographic structure of Greece between 1960 and 1995. The “double aging” trend found that population aging increases public debt and fiscal expenditure, leading to a deterioration in fiscal development. The second category focuses on predicting how population aging will affect public finance in the future. Demographic transition is associated with the future fiscal impact of population aging in terms of significant tax increases, social security increases, and welfare cuts [
10]. The public debt ratio in Germany will continue to rise along with population aging, rapidly reaching a threshold of 100% [
11]. Some argued that public finance needs to be estimated in advance, and forward-looking strategies need to be adopted to restructure fiscal revenue and expenditure and reduce the debt burden to effectively cope with the full impact of population aging [
12]. The third category focuses on dissecting the underlying mechanisms of the fiscal effects already caused by population aging. The linkage mechanism has been identified between aging and social security from the trend of social security benefits continuously increasing with the proportion of the elderly population in the United States since the 1970s. Since a more significant proportion of the elderly population means that the proportion of older voters will also increase, and since older voters will tend to vote for motions to increase social security benefits, the trend of continuously rising social security benefit spending is difficult to reverse [
13]. Empirical evidence from the static panel data of the United States and 12 eastern European has shown the dependency ratio and its significant negative relationship with the level of fiscal social security spending [
14].
Furthermore, literature shows that health is productive to promote economic growth and human development via improving the efficiency of individual labor and expanding labor supply [
15]. In contrast, disease and poor health mean sickness, absenteeism, and less productivity, which may fail to posse human capital accumulation [
16]. It is implied that the disease burden of the population is possible to induce an overwhelming fiscal burden. However, the elderly population does not always create a burden. In terms of the healthy elderly population, they can also play a positive role in the labor market or household production [
17]. It is evident that disease and poor health of elderly people need public finance to secure health care services, so the health status of the elderly population has far-reaching fiscal implications.
Most discussions and analyses in previous studies have referred to the impact of population aging and population health, respectively, on public finance, but they have paid less attention to the fiscal effects of the elderly health. In fact, elderly people are not the determinant of fiscal deficits. Poor health and disease burden ought to be the real reason for putting governments immersed in debt. With these facts in mind, we find this is a significant gap in the literature and a lack of empirical evidence. It is also practical significant given that even advanced countries are still struggling to find money to pay for the health care costs for the rising elderly population [
18].
The purpose of this study is to empirically identify the fiscal effect and mechanisms of the elderly health burden using the panel data from 45 countries from 2000 to 2019. The remainder of this paper is organized as follows:
Section 2 provides a theoretical framework;
Section 3 introduces the empirical strategy and study design;
Section 4 reports the results;
Section 5 presents associated analysis and discussion;
Section 6 proposes the conclusions and policy suggestions.
4. Results
4.1. Basic Regression
During the basic regression analysis, changes in the relationship between the elderly health burden and the fiscal balance were observed by gradually adding control variables, and the results are reported in
Table 3. Column 1 shows that the estimated coefficient of the elderly health burden without the inclusion of any control variables was −0.065, which had statistical significance at the 1% level. The results from Columns 2 to 6 show that the estimated coefficients of the elderly health burden ranging from −0.128 to −0.114 did not change much when control variables were gradually added, all of which were significantly negative at the 1% statistical level. It is indicated that the elderly health burden had a significant dampening effect on the fiscal balance and that countries or regions with a higher elderly health burden had a weaker fiscal balance. Considering the control variables, the results from Column 6 show that the estimated coefficients of all control variables were significant at the 1% or 5% statistical level. Among them, urbanization, openness, forestation, and GDP per capita growth had a significantly positive effect on the fiscal balance, indicating that a higher urbanization rate, more open trade, a greener environment, and faster GDP per capita growth lead to a better impact on public finance. In contrast, the effect of foreign direct investment on the fiscal balance was significantly negative at the 5% statistical level, indicating that a higher share of foreign investment leads to a lower fiscal balance. In addition, after controlling for the time and geographic effects of the sample countries, the effect of the elderly health burden on the fiscal balance remained significantly negative.
4.2. Endogeneity Concerns
To control for endogeneity issues of the model, this paper used lagged term as the instrumental variable for further dynamic regression analysis by system generalized method of moments (GMM) [
39].
Table 4 shows the regression results of the system GMM estimation. Column 1 shows that the estimated coefficient of the first-order lagged term of fiscal balance is significantly positive at the 1% significant level, indicating that the current period of fiscal balance is positively correlated with the previous period. Moreover, the regression coefficient of the elderly health burden was significant at the 5% level, which indicated a negative-driving effect on fiscal balance. In addition, to avoid measurement error, we replaced the variable measurements to re-run the regressions by taking the ratio of net borrowing and lending to GDP (
Finance) and the old-age dependency ratio (
Aging) as an alternative. As the results are shown in Columns 2 to 4, the core explanatory variable remained significant except in Column 4. Finally, according to model specification tests, it is also demonstrated that moment conditions employed in the system GMM are appropriate—errors are uncorrelated of order 2 (AB test
p-value > 0.1), and there are no overidentifying restrictions (Hansen test
p-value > 0.1), thus indicating that the selection of instrumental variables and the setting of the model are reasonable.
4.3. Robustness Tests
The results of various robustness tests are reported in
Table 5. First, given the cyclical nature of the fiscal budget and the possibility of inter-year effects on the fiscal balance, it was necessary to examine whether the current period’s level of the elderly health burden had an effect on the next period’s fiscal balance (
Balancet+1). The results of Column 1 show that the regression coefficient of the elderly health burden was significantly negative at the 1% statistical level, consistent with the conclusion above.
Second, we replaced the OLS regression model in the basic regression with a logistic regression model, redefining the explained variable fiscal balance as a dummy variable (Balance_dummy), whereby a value of one was assigned for a fiscal surplus, and the value of zero was assigned for a fiscal deficit. According to the logistic regression results in Column 2, the findings did not change substantially from the conclusions above.
Third, selecting the restricted samples. In order to exclude the influence of sample outliers on the research findings, we conducted robustness tests by adjusting the sample groupings. On the one hand, the public finance situation was not only affected by the domestic economy but was subjected to shocks from global macroeconomic events. Thus, we retested the regression results after excluding the sample data during the financial crisis of 2008–2010, as shown in Column 3. On the other hand, sample data with too extreme fiscal surplus or deficit rates were excluded. The regression results are reported in Column 4 using the sample data with a fiscal balance rate in the range of ±15%. The results show that the regression coefficients of the elderly health burden were both significantly negative at the 1% statistical level, indicating that an increase in elderly health burden still significantly reduces the fiscal balance, posing a negative impact. The results remained robust.
4.4. Heterogeneity Analysis
Considering that country heterogeneity possibly affects the basic result, we have performed a heterogeneity analysis from grouping by OECD member, aged society, and life expectancy. The results are reported in
Table 6. First, from the subgroup results of whether OECD members in Columns 1 and 2, we can see that the elderly health burden has a significantly negative effect on fiscal balance at the 1% level, in both groups. Second, according to the United Nations, an “aged society” is defined as a country whose share of people aged 65 years or more has reached 14% of the total population. Thus, we classified the countries with the elderly population proportion less than 14% as being in the Pre-aged group, and more than 14%, the Aged group [
40]. From the results presented in Columns 3 and 4, it is clear that the elderly health burden has a significant negative fiscal effect at the 1% level in aged society countries, but not in pre-aged society. Third, according to the rank of the sample countries by the mean value of life expectancy from 2000 to 2019, the median is 79.41 years old. Countries with an average life expectancy below 79.41 years old are classified as being in the Normal group, and countries with an average life expectancy above 79.41 years old are classified as being in the Longevity group. The results presented in Columns 5 and 6 show that the elderly health burden has negatively affected the fiscal balance in the Longevity group at the 1% level, but it is not significant in the Normal group. Overall, the reason for the significant fiscal impact of the elderly health burden in aged society and longevity countries may be due to the deep population aging and high life expectancy, which means a larger proportion of the elderly population over 75 or 80 years old, who are likely to incur greater costs in dying [
41]. This is because the two years before dying is a period of high healthcare costs for the elderly [
42,
43], with the greater probability of higher spending on healthcare as death draws nearer [
44].
4.5. Mediating Effect
Further, to recognize the mechanism of healthcare resources, the mediating effect was further analyzed. According to the research design, healthcare resources were measured in various forms, including funds, labor, and facilities. They were tested one-by-one in accordance with the model and analysis steps of the mediating effects to verify the existence of the three transmission mechanisms, and the results are reported in
Table 7.
First, in terms of healthcare funds, the results in Columns 1 and 2 show that the regression coefficients of the elderly health burden on healthcare funds, and of healthcare funds on the fiscal balance, were statistically significant at the 1% level. Thus, it is straightforward to conclude that a mediating effect exists without the need for Sobel testing. Column 1 shows that the effect of the elderly health burden on health funds was significantly positive, while Column 2 shows that the effect of healthcare funds on the fiscal balance was significantly negative, indicating that a rise in the elderly health burden increases the health expenditure and, thus, reduces the fiscal balance.
Second, in terms of healthcare labor, the estimates in Columns 3 and 4 show that the two-stage regression coefficients of the elderly health burden on healthcare labor and of healthcare labor on the fiscal balance were statistically significant at the 1% level. Thus, a mediating effect existed. Column 3 reveals a significant positive effect of the elderly health burden on healthcare labor, while Column 4 reveals a significant negative effect of healthcare labor on the fiscal balance, demonstrating that an increase in the elderly health burden necessitates increased healthcare labor, thus reducing the fiscal balance.
Third, in terms of healthcare facilities, the regression coefficients of the elderly health burden on healthcare facilities and of healthcare facilities on the fiscal balance in Columns 5 and 6 were statistically significant at the 1% level; hence, a mediating effect existed. Column 5 shows a significant positive effect of the elderly health burden on healthcare facilities, while Column 6 shows a significant negative effect of healthcare facilities on the fiscal balance, indicating that an increase in the elderly health burden necessitates increased healthcare facilities for support, which decreases the fiscal balance.
5. Discussions
This paper empirically investigated the fiscal effect of elderly population health and its mediation mechanism through cross-country panel data, and the main findings were as follows: (1) The regression results of the fixed effects model showed that the impact of the elderly health burden on the fiscal balance was significantly negative, suggesting that an increase in the elderly health burden adversely affects the fiscal balance; (2) the heterogeneity effects have been identified in different countries, the elderly health burden performed a significantly negative influence on the fiscal balance in aged society and longevity countries; and (3) the results of the mediation effect tests showed that a rising elderly health burden leads to an increase in healthcare resources, which in turn has a negative impact on the fiscal balance.
Taken together, such results keep consistent with most of the previous studies [
45,
46]. It is explainable that when the disease burden of the elderly population in society rises, the social demand for healthcare increases. Healthcare is the main aspect of public services, which is related to the livelihood of the country; thus, the government has a responsibility to provide basic healthcare services. Accordingly, the budget structure is adjusted to increase healthcare resources, forming a crowding-out effect [
47]. This can also be analyzed from the perspective of supply and demand, whereby an increased elderly health burden leads to a rise in a society’s healthcare on the demand side; hence, the government has to increase the fiscal investment to meet the growing health demand on the supply side [
48]. This analysis confirms that an increase in the elderly health burden leads to an increase in healthcare resource consumption, which in turn leads to a decline in the fiscal balance.
The limitation of this study is mainly that it lacks a detailed analysis of healthcare resources. Actually, healthcare resources encompass a wide range of components, and in this paper, only three terms of healthcare resources were discussed, which may not be fully representative, so the final research conclusions have less completeness and guidance for fiscal policies on population aging. Future research is expected to cover this limitation and strengthen the conclusions of this study.
6. Conclusions
International experience shows that rising healthcare expenditure can be a dangerous shock to fiscal sustainability, whereby higher welfare is often associated with higher debt [
49]. Previous studies have focused on the impact of the elderly population quantity on public finance, but the elderly health quality is rarely considered. This study, from a new perspective of disease burden, has demonstrated that the elderly health burden, which could be a new indicator of population aging, has a negative effect on fiscal balance, and the mediation mechanism of healthcare resources on the fiscal effect of elderly health burden has been confirmed. Therefore, it is beneficial for governmental fiscal balance to reduce the disease burden of the elderly population and improve the utilization efficiency of healthcare resources.
Based on the above findings, some policy recommendations could be considered. The first one is to consider “aging and health” as the top priority in fiscal policies. It is necessary to encourage the innovation of health financial products such as pension funds, chronic disease medical funds, and long-term care insurance to cover health shocks and risks. The second is to improve the efficiency of healthcare resources to respond to population aging. “Value-based healthcare” [
50] is supposed to be effective for elderly healthcare to optimize higher health outcomes and lower health costs. The third is to promote to elderly people a sense of “active health” in order to keep healthy and reduce disease burden. Health literacy for the elderly should be improved in order to inform them on how to conduct health management correctly and actively.