Asymmetric Effects of Human Health Capital on Economic Growth in China: An Empirical Investigation Based on the NARDL Model

Abstract

Human health capital is an important factor that affects countries’ economic development. This research explores the nonlinear effect of human health capital on economic growth and assesses the asymmetry over time. We used annual data from 1978 to 2021 in China and the nonlinear autoregressive distributed lag (NARDL) model to examine the long- and short-term effects of positive and negative human health capital shocks on economic growth. Human health capital is measured by personal health expenditure (PHE), government health expenditure (GHE), and social service expenditure (SSE). A reduction of one unit in short-term private health expenditure leads to a 7.48% decrease in GDP per capita. An increase in private health expenditure leads to an increase in GDP per capita of 3.51%. The positive (${\beta }_{PHE}^{+}$) and negative (${\beta }_{PHE}^{-}$) coefficients of change in long-term private health expenditure are 1.31 and 3.87, respectively. A reduction in short-term government expenditure on health leads to a 10.99% decline in GDP per capita. The positive (${\beta }_{GHE}^{+}$) and negative (${\beta }_{GHE}^{-}$) coefficients of long-term government health expenditure are −4.33 and 1.99, respectively. A one-unit reduction in short-term social service spending leads to a 5.56 percent drop in GDP per capita, while an increase in social service expenditure leads to a 5.97 percent increase in GDP per capita. The positive (${\beta }_{SSE}^{+}$) and negative (${\beta }_{SSE}^{-}$) coefficients of change in long-term social service expenditure are 5.76 and 4.62, respectively. Both private health expenditure and government health expenditure have shown significant asymmetry in their long- and short-term effects on economic growth. Human health capital that is rationally distributed can effectively enhance economic growth.

1. Introduction

The novel coronavirus disease 2019 (COVID-19) has cost humanity a great deal of money. It has greatly hindered economic development and slowed the economic growth of many countries. The pandemic has damaged human health and increased the burden of individuals’ medical treatment [1]. The COVID-19 pandemic is still affecting economies today. How to safeguard human lives and sustain economic growth in this uncertain environment has become an urgent problem for governments. As an important part of human capital, human health capital is closely related to the level of medical expenditures and medical services enjoyed by economic individuals [2]. The accumulation of human health capital can provide support for disease prevention and the recovery of individuals. Human health capital is one of the basic factors for countries to maintain long-term economic growth and social stability and development. In the World Development Report (1993), the World Bank clearly pointed out that “good health plays a positive role in improving individual labor productivity and economic growth rates in various countries” [3]. Healthy populations are more economically productive. The increase in health investment can effectively improve the physical quality of workers and increase the number of years of work, thus improving the enthusiasm and efficiency of production, further improving the future productivity of human capital and increasing the accumulation of human capital, playing a positive role in human capital [4]. The most basic investment in human capital is in health. Human capital is incomplete without health [5]. Human health capital is dependent on the individual and represents the individual’s health level. As a basic kind of human capital, human health capital can affect other forms of human capital, such as human education capital, and is affected by other human capital. The ways in which human health capital affects economic growth have attracted extensive attention. In our study, we investigate the asymmetric effects of increased and decreased human health capital on economic growth using the NARDL model.

Economic growth is driven by many factors, including the labor force, material capital, human capital, and technological progress. Among the different elements that drive economic success, human health capital is particularly important. China’s annual GDP growth rate from 2000 to 2019 was above 6%. After the COVID-19 pandemic began, China’s GDP growth rate in 2020 was still 2.2%. The total health expenditure in China in 1990 was approximately USD 10 billion. By 2019, health expenditure was approximately USD 925.2 billion, with an average annual growth rate of 16.7%. This means that the development rate of medical expenditure affecting human health capital is quite high. On the one hand, as the basis for individuals to participate in productive labor and receive education, health has gradually become included in the study of economic growth; on the other hand, individuals are increasingly paying attention to health as a kind of human capital that is as vital as education [6,7].

Some studies focus on the connection between human capital and economic development. It has been found that health spending has a favorable and considerable impact on the accumulation of health capital and then has a positive and indirect impact on economic growth [8]. The level of human health capital is related to the work enthusiasm and labor efficiency of the individuals in the labor force. An increase in human health capital can not only improve individual work efficiency but also improve a country’s efficiency and total factor productivity. Barro (1996), when studying the impact of human capital on economic growth, expanded the three-sector neoclassical model by introducing human health capital, human educational capital, and material capital and was able to conduct a more in-depth study and analysis of human health capital at the macroeconomic level [9]. It has been proposed that human health capital can affect social and economic growth by changing labor productivity, which proves that human health capital can promote economic growth [10]. Health is not only a type of human capital in and of itself but also involves the input of other types of human capital [11,12]. Therefore, through the accumulation of human health capital, we can improve labor productivity and promote the sustainable development of the economy [6]. On a macro level, the proportion of health expenditure to GDP in high-income countries is generally higher than that in low-income countries. It seems that the increase in health expenditure is only the “result” of economic growth. If the increase in health expenditure can become the “driving force” of economic growth, it is undoubtedly very important for a country to increase its health expenditure, whether such initiatives proceed from the perspective of promoting economic growth or protecting the health of the population. Financial expenditure on health is an important part of government health expenditure. With the increasing demand for medical and health services and the diversity of medical service demand, the total scale of health expenditure continues to expand, resulting in a certain contradiction between public health spending and sustained and stable economic growth [13]. Leaders must formulate policies for the rational distribution of medical benefits to reduce and reverse the negative economic impact of unnecessary medical spending. This indicates that the information revealed by the linear model in previous research may not be rich enough to make strong inferences or reliable predictions [13,14]. Therefore, for a country to establish a long-term, sustainable system of social security, it is very important to elucidate the nonlinear impact of health spending on economic growth and its primary mechanism.

The rest of the paper is organized as follows. The second section is a literature review and our hypothesis development. The next section introduces the data and methodology. The subsequent section presents the empirical results, and the last section concludes the paper.

2. Literature Review and Hypothesis Development

2.1. Literature Review

Previous studies have found that since health is a form of capital, investing in it can boost the accumulation of both human and material capital, resulting in overall economic growth [15,16]. Health was once regarded as the product of economic growth: as the economy developed, people obtained higher incomes and were better able to ensure their health. However, when health is regarded as a form of human capital, its level will affect the economic level of individuals and even countries [17]. An increase in poor personal health may lead to the loss of labor and productivity [18,19]. Good health can prolong life expectancy and encourage personal savings and incentives for more business investment, which are all activities beneficial to economic performance. Population health is an important part of healthcare, and its impact should be considered. A healthy population can reduce national healthcare expenditure and increase income potential [16]. Increasing health expenditure can provide people with better healthcare by improving their health status [20]. The efficiency and productivity of human resources may be improved by increasing investment in both physical and human capital. Investment in health and healthcare can promote economic growth by improving the health of the population. Healthier populations tend to have higher labor force participation rates, higher productivity, and longer working lives [4]. Increased private and state spending boosts the positive impact of health investment on economic growth, as the productivity of the workforce depends on their health status [6,21]. The healthier the labor force is, the higher the economic efficiency of the country [10,22,23,24]. Health is foundational for people to be able to work, and it affects the economy by increasing potential human health capital. It has been proven that health has a significant positive impact on a country’s economic growth [25,26], and scholars have found that human health capital can have a similar impact [27,28,29,30]. Yang used a panel threshold model to analyze the relationship between health expenditure and economic growth in 21 developing countries under different levels of human capital. The study found that different levels of human capital will produce a significant interval effect of health expenditure on economic growth [10]. Tanzila Sultana et al. used the SGMM method to study whether human capital, including health, has a positive impact on the economy of developing countries and, in contrast, whether it inhibits the economic growth of developed countries [31]. Many studies have examined the impact of healthy human capital on economic growth, but there seems to be no unified consensus on the matter.

In order to improve health, a country must provide adequate health resources. Funding is needed to provide these resources, so health expenditure is needed to achieve a healthy country [32,33]. Increased health spending and improvements in the health sector have improved the quality of human capital. Total expenditure on health can be seen as the sum of public and private expenditures on health. In addition, public health spending can be broken down into government health spending and social service spending. Private health expenditure mainly refers to the expenses directly borne by the individual or family when receiving health services and the costs of participating in various social and commercial medical insurances. The level of private health expenditure can be seen as a individual’s ability to protect their own health [34]. The increase in out-of-pocket health spending, such as private health spending, increases the amount of catastrophic spending and can lead to more poverty [35]. Government expenditure on health generally refers to government expenditure on various undertakings, such as medical and health services, medical security assistance, and health and medical administration. Therefore, government health expenditure is an important factor in the accumulation of human capital. An increase in government health expenditure will significantly increase human health capital [36,37]. However, as the percentage of overall government health expenditure in GDP grows, the impact of health expenditures on GDP growth decreases dramatically [8]. From the perspective of social welfare, the scale of government health expenditure is reasonable, which affects the health level of the people and the economic development of the country. Mpundu et al. (2019) used the ARDL method for time series data from 1980 to 2017 to discuss the fundamental changes in public expenditure and its impact on GDP [20]. Their control variables included foreign direct investment and current account balances, and the purpose of their study was to determine the changes in the performance of GDP since 1980. They found that increasing government expenditure may not be ideal for economic growth [20,38]. Public health services play an important role in promoting human capital accumulation to improve a country’s overall level of health [39]. Linden analyzed 34 OECD countries using an econometric panel time series approach and found a significant positive correlation between public health spending and life expectancy. When spending on social services increases, human capital can be indirectly increased by improving people’s health [40]. Eggoh et al. (2015) reported a negative relationship between public health expenditure and economic growth in 49 African countries from 1996 to 2010 [21]. Liu (2016) believes that medical insurance expenditure can reduce the adverse impact of external health shocks on families, reduce household income fluctuations, and improve household consumption level and investment in human capital [41]. Consumption plays an important role in economic growth. According to precautionary saving theory, if consumption can reduce consumers’ negative feelings about the uncertainty of future income and expenditure, it can affect individuals’ behavioral choices regarding current consumption and savings to a certain extent. The impact of fiscal health expenditure on individuals’ consumption propensity has an income effect and a substitution effect. From the perspective of the income effect, when the government increases its expenditure in the field of medicine and healthcare by means of financial investment or subsidies, it will reduce the individual medical consumption expenditure of individuals to a certain extent so that the saved expenditure will be income for individuals and will increase relative income. From the perspective of the substitution effect, when the government increases the expenditure in the field of medicine and healthcare by means of financial investment or subsidies, correspondingly, the part of the medical expenses that consumers originally spent money on will be reduced, thus leading to a reduction in the consumption tendency in the field of medical care. Therefore, whether fiscal expenditure on medicine and healthcare promotes an increase or decrease in consumption propensity depends on the comprehensive results of the substitution effect and income effect. Zhao (2017) studied the impact of social insurance on personal behavior and welfare under the framework of the general equilibrium model and found that the existence of social insurance reduced personal willingness to save money and work, resulting in a substantial economic crowding out effect [42]. In another interesting report, Yang (2020) found that when the human capital level of 21 developing countries was low between 2000 and 2016, health expenditure had a negative impact on economic growth [10]. Maruthappu (2015) proposed that both the input of human health capital and the stock of social material capital will be affected by the total amount of social investment, and the input of human health capital has an uncertain impact on social and economic growth [43].

The findings of the current study do not agree with the above-mentioned studies regarding how health expenditures affect economic growth. Overall, health expenditures affect the accumulation of human health capital, so the effect of health expenditures on economic growth seems evident. However, the asymmetries of the influence of human health capital on economic development have not been studied.

2.2. Hypothesis Development

Based on the human capital effect, high-quality human capital is an important condition for economic growth. Health investment that is too low affects the accumulation of health capital, is not conducive to improving the quality of human capital, and thus affects economic growth. Health capital and education capital are important components of human capital [6,7,8,9,10]. Health expenditure can prolong the life span of residents, improve their physical quality, increase their working years, increase their enthusiasm for production, and increase human capital [4,6]. In the case of poor health and nutrition status, residents are passive in investing in health, reduce education investment, and are even unable to invest in education. Due to the lack of necessary knowledge and technology, ordinary workers cannot easily develop high-quality labor skills, which affects the optimization of the human capital structure, causes the quality of human capital to decline, and has a negative impact on improving the quality of economic growth [44]. The impact mechanism of medical and health expenditure on economic growth can be examined from both short-term and long-term perspectives: in the short term, medical and health expenditure is a type of purchasing expenditure and government public expenditure. Therefore, in the macroeconomy, the increase and decrease in financial, medical, and health expenditure will cause changes in the total national economy through the multiplier effect [45]. In addition, medical and health expenditure itself has the special nature of being a public good; that is, it will impact macroeconomic growth through the mechanism of positive externalities. In the long run, the increase in financial, medical, and health expenditures can help individuals share some medical and health costs to resist the economic risks caused by diseases. However, if a certain scale of the medical system and health services is already present, excessive government health expenditure cannot be converted into high-quality health output. The further expansion of the health sector will increase the financial burden on the government, which will crowd out the government’s investment in other projects that are more conducive to development in the long run, leading to negative economic growth [6]. Based on this, this study proposes the following preliminary hypothesis:

H1. 

Health expenditure has an asymmetric impact on economic growth in the long- and short-term.

3. Materials and Methods

3.1. Data

This study investigates the asymmetric relationship between economic growth and human health capital. Human health capital is measured by personal health expenditure (PHE), government health expenditure (GHE), and social service expenditure (SSE). The data for all variables have been compiled from the China National Bureau of Statistics. The time range for data is 1978–2021. Additionally, summary statistics of the variables are reported in

Table 1. Summary statistics.

Variables	Obs	Mean	Std. Dev.	Min	Max	Skewness	Kurtosis
PGDP	44	19,423.14	23,714.88	385.00	81,370.00	1.191052	3.120067
PHE	44	4995.84	6127.359	22.52	21,205.67	1.258606	3.465875
GHE	44	4192.35	6337.955	35.44	21,941.90	1.505452	3.970711
SSE	44	6038.36	9440.012	52.25	34,963.26	1.740078	4.863360

. PGDP in Table 1 represents GDP per capita. The unit of PGDP is yuan, and the unit of other variables is 100 million yuan.

3.2. The Model and Methods

In terms of methods, the nonlinear autoregressive distributed lag (NARDL) model is chosen for empirical research in this paper. Shin et al. proposed the NARDL model, which is an asymmetric nonlinear cointegration method [46]. This approach is particularly beneficial for analyzing short-term and long-term nonlinearity by splitting explanatory factors into positive and negative portions.

The positive and negative variances in the explanatory variable $x_t$ are first separated. $x_t^{+}$ and $x_t^{-}$ represent the positive impact and negative impact of $x_t$, respectively. $x_0$ is the initial value at time $t=0$.

$$x_t=x_0+x_t^{+}+x_t^{-}$$

(1)

$$x_t^{+}={\displaystyle \sum _{j=1}^t\Delta x_j^{+}}={\displaystyle \sum _{j=1}^t\max (\Delta x_j,0)}$$

(2)

$$x_t^{-}={\displaystyle \sum _{j=1}^t\Delta x_j^{-}}={\displaystyle \sum _{j=1}^t\min (\Delta x_j,0)}$$

(3)

The long-term asymmetric relationship between the impact of health expenditures and economic growth can be represented as follows:

$$y_t={\beta }^{+}{x}_t^{+}+{\beta }^{-}{x}_t^{-}+{\mu }_t$$

(4)

$y_t$ is the dependent variable. A long-term relationship is represented by the parameters ${\beta }^{+}$ and ${\beta }^{-}$.

Shin et al. built a dynamic parameter framework that may encompass both long-term and short-term asymmetry on the basis of prior studies to examine the long-term and short-term asymmetry relationship [46]. The linear error correction model (ECM) is extended to a nonlinear ARDL model with larger cointegration under this parameter framework:

$$\Delta y_t=\rho y_{t-1}+{\theta }^{+}{x}_{t-1}^{+}+{\theta }^{-}{x}_{t-1}^{-}+{\displaystyle \sum _{i=1}^{p-1}{\gamma }_i\Delta y_{t-i}}+{\displaystyle \sum _{i=0}^{q-1}({\pi }_i^{+}\Delta x_{t-i}^{+}+{\pi }_i^{-}\Delta x_{t-i}^{-})}+{\epsilon }_t$$

(5)

This model can also be expressed as

$$\Delta y_t=\rho {\xi }_{t-1}+{\displaystyle \sum _{i=1}^{p-1}{\gamma }_i\Delta y_{t-i}}+{\displaystyle \sum _{i=0}^{q-1}({\pi }_i^{+}\Delta x_{t-i}^{+}+{\pi }_i^{-}\Delta x_{t-i}^{-})}+{\epsilon }_t$$

(6)

The maximum value of the lag order of the explanatory variable and the explained variable in the model are represented by p and q, respectively., and ${\epsilon }_t\sim iid\left(0,{\sigma }_{\epsilon }^2\right).$

Then we looked at the impact of personal health spending, social health spending, and government health spending on per capita GDP using the NARDL (p, q) model, which can be written as follows:

$$\begin{array}{ll}\Delta GD{P}_t=& c+\rho GD{P}_{t-1}+{\theta }_1^{+}PH{E}_{t-1}^{+}+{\theta }_1^{-}PH{E}_{t-1}^{-}+{\theta }_2^{+}GH{E}_{t-1}^{+}+{\theta }_2^{-}GH{E}_{t-1}^{-}\\ & +{\theta }_3^{+}SS{E}_{t-1}^{+}+{\theta }_3^{-}SS{E}_{t-1}^{-}+{\displaystyle \sum _{i=1}^{p-1}{\gamma }_i\Delta GD{P}_{t-i}}\\ & +{\displaystyle \sum _{i=0}^{q-1}({\pi }_{1i}^{+}\Delta PH{E}_{t-i}^{+}+{\pi }_{1i}^{-}\Delta PH{E}_{t-i}^{-})}+{\displaystyle \sum _{i=0}^{q-1}({\pi }_{2i}^{+}\Delta GH{E}_{t-i}^{+}+{\pi }_{2i}^{-}\Delta GH{E}_{t-i}^{-})}\\ & +{\displaystyle \sum _{i=0}^{q-1}({\pi }_{3i}^{+}\Delta SS{E}_{t-i}^{+}+{\pi }_{3i}^{-}\Delta SS{E}_{t-i}^{-})}+{\epsilon }_t\end{array}$$

(7)

We estimated Equation (7) and then examined cointegration and asymmetries in the long and short term. In the short term, we used ${{\displaystyle \sum }}_{i=0}^{q-1}{\pi }_{ji}^{+}={{\displaystyle \sum }}_{i=0}^{q-1}{\pi }_{ji}^{-}$(j = 1,2,3) to test the asymmetry. The long-term asymmetry is measured by ${\beta }_j^{+}=-{\theta }_j^{+}/\rho$ and ${\beta }_j^{-}=-{\theta }_j^{-}/\rho$.

4. Results

The ADF test and PP test are used in the first phase to check the integration of the variables.

Table 2. Unit root test results.

Variables	ADF Test		PP Test
	Level	First Difference	Level	First Difference
PHE	4.075927 (0.9999)	−6.613411 *** (0.0000)	14.18989 (0.9999)	−12.54145 *** (0.0000)
GHE	−2.116825 (0.2396)	−5.683656 *** (0.0000)	4.957534 (0.9999)	−13.35307 *** (0.0000)
SSE	−1.825528 (0.3623)	−3.185239 * (0.0289)	12.60500 (0.9999)	−10.11902 *** (0.0000)
PGDP	6.262191 (0.9999)	−7.761353 *** (0.0000)	9.773890 (0.9999)	−5.308075 *** (0.0001)

Note: *** and * indicate rejection of the null hypothesis of symmetry at the 1% and 10% levels. The numbers in ‘‘()’’ are p-values.

shows that none of the variables are I(2) in the unit root test. This indicates that all of the variables meet the requirements for using the nonlinear ARDL model.

Then we studied long-run nonlinear cointegration connections among private health expenditure, social service expenditure, government health expenditure, and GDP per capita.

Table 3. Bounds tests for asymmetric cointegration.

Test Statistic	Value	Significant Level	I(0)	I(1)
F-statistic	36.47189 ***	10%	1.99	2.94
k	6	5%	2.27	3.28
		2.5%	2.55	3.61
		1%	2.88	3.99

Note: *** indicates the 1% significance level.

shows that the result of the F-statistic is 36.11846 at the 1% level of significance. The results show that the relationship between the four variables is stable and is above the critical upper bound; furthermore, the variables have cointegration linkages.

After Equation (7) was estimated by OLS, we determined p = 4 and q = 4 to build an appropriate model. Table 3 and

Table 4. Results of the short-run test.

Regressors	Coefficient	t-Statistic	Probability
$\Delta PGD{P}_{t-1}$	2.864623 ***	10.87404	0.0001
$\Delta PGD{P}_{t-2}$	0.457396	1.923363	0.1124
$\Delta PGD{P}_{t-3}$	1.280368 ***	5.630073	0.0024
$\Delta PH{E}_t^{+}$	3.514629 **	3.929308	0.0111
$\Delta PH{E}_{t-1}^{+}$	0.218079	0.164822	0.8755
$\Delta PH{E}_t^{-}$	−7.484464 ***	−4.562640	0.0060
$\Delta PH{E}_{t-1}^{-}$	−14.95590 ***	−7.121060	0.0008
$\Delta GH{E}_t^{+}$	−0.427581	−0.484431	0.6486
$\Delta GH{E}_{t-1}^{+}$	3.138878 *	2.423853	0.0598
$\Delta GH{E}_t^{-}$	−10.98566 ***	−6.980852	0.0009
$\Delta GH{E}_{t-1}^{-}$	−12.00262 ***	−6.901572	0.0010
$\Delta SS{E}_t^{+}$	5.975420 ***	5.718229	0.0023
$\Delta SS{E}_{t-1}^{+}$	−8.084124 **	−3.464927	0.0179
$\Delta SS{E}_t^{-}$	−5.557417 *	−2.227551	0.0764
$\Delta SS{E}_{t-1}^{-}$	3.124636	0.429227	0.6856

Note: ***, **, and * indicate the 1%, 5%, and 10% significance levels, respectively.

present the estimation results and show how private health spending, social service spending, and government health spending affect GDP per capita.

Table 4 shows that increases in private health expenditure, social services expenditure, and government health expenditure have asymmetric effects on GDP per capita in the short term. A positive 1% change in private health spending would lead to a 3.51% increase in GDP per capita. The coefficient for the negative private health expenditure change ($\Delta PH{E}^{-}$) is significantly negative and equal to −7.48. This means that negative changes in private health expenditure in the short term will lead to a reduction in GDP per capita. The decline in GDP due to the decline in private health spending is more than twice the corresponding increase in GDP. As a component of individuals’ consumption expenditure, it is reasonable to promote the growth of the market economy by consuming medical and health products and services. As economic growth continues, the income of individuals also increases, and people become increasingly inclined to improve their quality of life, so they also have higher requirements for medical services. As economic growth continues, the income of individuals also increases, and people become increasingly inclined to improve their quality of life, so they also have higher requirements for medical services. From the perspective of the consumption demand effect, the demand for medical and health services and products is rigid. Because of the economic multiplier effect, the consumption demand for health will amplify the effect of health expenditure on the economy [19]. People with a poor health status usually have low incomes and less investment in health and are unable to improve their health status. Once a health shock occurs, it is easy for people to fall into the dilemma of “poverty due to illness”, thus forming a vicious circle [8,47]. After individuals consume healthcare services, they can improve their health level, which is conducive to the accumulation and development of human health capital, thus promoting the growth of the national economy. The positive ($\Delta GH{E}^{+}$) government health expenditure coefficient is 3.14. In the short term, the increase in government health spending significantly increases the human capital of health, thereby increasing GDP per capita. The coefficients for negative ($\Delta GH{E}^{-}$) government health expenditure are −10.99 and −12.00. A negative change in government health expenditure causes a negative change in GDP in the short term. The impact of falling government spending on health is far greater than the impact of increasing spending. This result is supported by public economic theory. According to the theory of the public economy, government public expenditure is an important driving factor in economic and social development; it not only directly affects economic development but also plays an important role in all aspects of society, that is, the effect of government public expenditure. On the one hand, the growth in government health expenditure can directly increase the labor supply by improving the health level of workers, thus affecting economic growth; on the other hand, government health expenditure can indirectly act on social scientific and technological innovation factors by promoting the quality of workers and thereby further act on economic growth. Government health spending accounts for a significant portion of human health capital [48]. The government increases the financial expenditure in the field of health, which is equivalent to bearing part of the expenses that should be borne by the residents themselves for the consumption of medical products [49]. The government’s increased investment in the field of health can improve the level of medical services, reduce the medical burden of residents, and enable individuals to have more funds for consumption beyond medical care. At this time, the growth in government health expenditure has a multiplier effect [45]. It promotes the development of medical and health undertakings and drives the development of surrounding industries, thus promoting the economic growth of the whole society. A reduction in government health expenditure in the short term means that human health capital is insufficient, which slows economic development [50]. The positive change ($\Delta SS{E}^{+}$) coefficient of social service expenditure is 5.98. The negative ($\Delta SS{E}^{-}$) coefficient of change in social service expenditure is −5.56. This means that increasing social service expenditure in the short term will improve the economy, while insufficient social service expenditure will hinder economic development. Social service expenditure has become an important part of health expenditure. Social service expenditure also promotes economic growth through investment in public health infrastructure construction [51]. When social health expenditure is insufficient, medical security for individuals is also insufficient, which leads to an economic downturn.

Table 5. Results of the long-run test.

Panel A: Estimated Long-Term Coefficients
Regressors	Coefficient	t-Statistic	Probability
$\Delta GD{P}_{t-1}$	−3.059864 ***	−10.11325	0.0002
$\Delta PH{E}_{t-1}^{+}$	4.016686 ***	4.496772	0.0064
$\Delta PH{E}_{t-1}^{-}$	11.85568 ***	4.416770	0.0069
$\Delta GH{E}_{t-1}^{+}$	−13.23850 ***	−6.376866	0.0014
$\Delta GH{E}_{t-1}^{-}$	6.112802 **	2.743920	0.0406
$\Delta SS{E}_{t-1}^{+}$	17.63887 ***	5.521861	0.0027
$\Delta SS{E}_{t-1}^{-}$	14.15027 **	2.835459	0.0364
Panel B: Long Term Influence Coefficient
${\beta }_{PHE}^{+}$	1.312701 ***
${\beta }_{PHE}^{-}$	3.874577 ***
${\beta }_{GHE}^{+}$	−4.326499 ***
${\beta }_{GHE}^{-}$	1.997737 **
${\beta }_{SSE}^{+}$	5.764593 ***
${\beta }_{SSE}^{-}$	4.624477 **

Note: *** and ** indicate the 1% and 5% significance levels, respectively.

shows the asymmetric influence of health expenditure on economic development in the long run. The coefficients for the influence of positive (${\beta }_{PHE}^{+}$) private health expenditure and negative (${\beta }_{PHE}^{-}$) private health expenditure changes on GDP per capita equal 1.31 and 3.87, respectively, in the long term. These results mean that in the long run, the positive and negative changes in private health expenditure will lead to an increase in GDP per capita and that the negative change in private health expenditure has a stronger impact on GDP. While short-term increases in private health spending can spur per capita GDP growth, increases in long-term private health spending are detrimental to other aspects of household savings and consumption, resulting in a slowdown in economic growth [44,51]. The positive (${\beta }_{GHE}^{+}$) and negative (${\beta }_{GHE}^{-}$) change coefficients of government health expenditure are −4.33 and 1.99, respectively. In the long run, a change in government health expenditure will result in a change in GDP in the opposite direction, and a drop in government expenditure will raise GDP to some extent [52,53]. This is consistent with the theory of Zong and Muysken; that is, due to the limitations on total output, there will be a dilemma between health investment and material capital investment, so health investment may have a crowding-out effect on economic growth, leading to negative and uncertain impacts [54]. The influence coefficients of positive (${\beta }_{SSE}^{+}$) and negative (${\beta }_{SSE}^{-}$) changes in social service expenditure are 5.76 and 4.62, respectively. This means that social service expenditure is not the main factor determining the change in GDP on a long-term basis [55]. However, its increase will increase the level of health services in society as a whole and improve the health of individuals. Public expenditure invested in health projects acts as a macroeconomic stabilizer and benefits economic growth in the long run [8,56].

We also use the Wald test to look at asymmetries over the short- and long-term. The Wald test findings are given in

Table 6. Results of asymmetry tests.

	PHE	GHE	SSE
Wlr	−3.121189 ** (0.0262)	−5.092586 *** (0.0038)	0.985638 (0.3696)
Wsr	4.485434 *** (0.0065)	7.368960 *** (0.0007)	3.859679 ** (0.0119)

Note: *** and ** indicate the 1% and 5% significance levels, respectively. The numbers in ‘‘()’’ are p-values.

. The results show that the impact of three health expenditures (except long-term social service expenditures) on per capita GDP is asymmetrical in both the long and short term. Our hypothesis 1 is verified.

In the last step, we use Equation (7) to describe the asymmetric line, and the results are shown in Figure 1, Figure 2 and Figure 3. The unmarked red dotted lines in figures represent a 95% confidence interval. The horizontal axis is the time period, and the vertical axis is the monetary unit. The cumulative dynamic multiplier effect defines the symmetrical or asymmetric adjustment that occurs as a result of a unit’s positive or negative influence on explanatory variable changes. Figure 1, Figure 2 and Figure 3 describe the symmetric or asymmetric adjustment of per capita GDP from the original equilibrium to the new equilibrium with the positive or negative impact of a unit of personal health expenditure, government health expenditure, and social service expenditure. They represent the asymmetry of the adjustment degree and speed of the positive and negative changes in personal health expenditure, government health expenditure, and social service expenditure on the per capita GDP. The estimation parameters of the time series model may shift over time, which might be a concern. In order to prevent doubt about the model’s reliability owing to parameter instability, this paper analyses the stability of parameters with CUSUM and CUSUMSQ. The significance level of the two tests is 5%. Figure 4 and Figure 5 show that the CUSUM statistics do not deviate from the boundary range. CUSUMSQ, on the other hand, is slightly stable within a significant cutline of 5%. Kisswani (2021) argues that once one of the CUSUM and CUSUMSQ tests shows evidence of stability, it is often possible to determine that the parametric model is stable [57]. The results show that our NARDL model and its results can be considered reliable.

5. Discussion and Conclusions

Based on annual data from 1978 to 2020 in China, this research investigates the long- and short-term asymmetric effects of human health capital on economic development using the NARDL model. The findings show that all three types of health expenditure, which measure the level of human health capital, have an asymmetrical effect on GDP per capita. We demonstrate that the reaction of economic growth to the positive and negative shocks of the three types of health expenditure differs significantly. Both positive and negative changes in health expenditure will significantly affect economic growth. In the short term, the increase in private health spending, social service spending, and government health spending will stimulate an increase in per capita GDP, and their decline will hinder economic growth. This result is consistent with previous studies, which also identified the positive impact of public health expenditure [58,59,60,61,62,63,64,65]. The economic growth effect of medical and health expenditure is mainly realized through the accumulation of material capital and human capital [65]. Short-term negative changes in private health expenditure have a greater impact on GDP per capita than positive changes do. The lack of individuals’ awareness of their own health will lead to a shortage of personal health expenditures in the short term. This weak awareness may be caused by the surrounding environment, so it is difficult to eliminate this negative impact in the short term [66]. In contrast, positive changes in short-term social service spending have a greater impact on per capita GDP than negative changes do. In the short run, the negative change in private health expenditure, social service expenditure, and government health expenditure will all lead to negative changes in per capita GDP [8,67]. All three types of insufficient health spending in the short term, especially insufficient government health expenditure, will inhibit economic growth to some extent [68]. This inhibitory effect is far greater than the catalytic effect of increased government health expenditure on economic growth. The above results suggest that human health capital will show the same trend of change as economic growth in the short term. On the one hand, the growth of health expenditure can directly increase the labor supply by improving the health level of workers, thus acting on economic growth. On the other hand, health expenditure can indirectly act on social scientific and technological innovation factors by promoting the quality of workers and then act on economic growth. In the long run, no matter how personal health expenditure and social service expenditure change, the per capita GDP will increase. However, a reduction in personal health expenditure has a greater positive impact on the economy than an increase in such expenditure. In contrast, an increase in social service expenditure has a greater impact on economic growth. A positive change in government health expenditure has more impact on GDP per capita than a negative change, and the change in government health expenditure is the opposite of that in GDP per capita. From a long-term perspective, an increase in government health spending stifles economic growth more than a decrease in government health spending stimulates it. Excessive health investment may crowd out physical capital investment and then have negative effects on economic growth. On the one hand, the continuous expansion of government health expenditure can improve the health level of individuals, enrich the medical and health resources owned by individuals, promote the accumulation of human health capital, and thus promote economic growth. On the other hand, it also brings a substantial economic burden to the government’s public financial expenditure, which has a negative impact on economic development [69,70,71,72,73,74]. The influence of positive and negative changes in social service expenditure on GDP will lead to economic growth. However, long-term spending on social services will ease the government’s financial pressure on health and provide individuals with free healthcare. This means that long-term social service spending can indirectly increase human health capital, thereby affecting economic growth. Furthermore, the developed and underdeveloped regions in China have differences in the distribution of government, medical, and health expenditures. The medical and health expenditure in developed regions is generally higher than that in underdeveloped regions [75]. The effective use of health expenditure can have a positive impact on economic growth through labor productivity. However, if the use of health expenditure is not efficient or cannot improve health outcomes, it may lead to a waste of resources [76]. This uneven distribution of medical and health expenditure leads to inefficient use of health capital and inhibits economic growth to a certain extent. Therefore, to maintain stable GDP growth, the government and related agencies should enhance their health expenditure in the short term, control the scale of government health expenditure in the long term, and improve the health level of the whole population, which can significantly promote economic growth.

Useful policy implications can be identified from the results of this research. Considering the asymmetric effects of three different types of health expenditure on economic growth, authorities should balance health expenditure and the social security system. Individual health is affected by a variety of environmental factors, such as the overall social education level, that is, the government and individuals’ awareness of healthcare expenditure. Additionally, individual health determines the size of the government and individual investment in healthcare and the level of health that individuals can obtain from the individual health function and, thus, affects the level of output that individuals can achieve. Therefore, the improvement in government and individual awareness of healthcare expenditure is a very important factor in promoting China’s sustained economic growth. Government spending on health is not always better at higher levels. A reasonable health expenditure system will promote a country’s economic growth [76]. When institutional quality interacts positively with health capital, the result is more effective resource allocation and economic growth [77]. Faced with the financial expenditure pressure caused by the government’s expanding medical and health expenditure, China should handle the relationship between the government and the market, let the government play a leading role in the development of medicine and healthcare, and introduce social forces to assist in the development of medical initiatives. This can not only expand the medical and health industry and reduce the inefficiency caused by complete financial expenditure but also reduce the financial pressure on government medical and health expenditure to achieve sustainable development. The scale of long-term health government spending should be controlled to achieve the best co-ordination between human health capital and economic growth.

This paper is innovative in the following ways. We use the NARDL model to find the asymmetric impact of human health capital on economic growth. Although previous research has studied the link between human health capital and economic growth, most studies analyze only the linear relationship between them. We find that the positive and negative impacts of the three kinds of health expenditure have different effects on economic growth. (1). The impact of private health expenditures on per capita GDP differs in the short and long terms. The decline in per capita GDP brought about by reduced health expenditures is greater than the increase in per capita GDP brought about by increased health expenditures. (2). The impact of government expenditures on health spending on economic growth in the short and long terms also varies. Changes in short-term government spending on health will lead to a change in the direction of economic growth. The impact of the reduction in short-term government spending on per capita health far outweighs the economic benefits of its growth. In the long run, the inhibitory effect of increased government health expenditure on economic growth is greater than the pro-growth benefit of decreased government health expenditure. (3). The impact of public service expenditures on GDP per capita does not appear to be asymmetrical in the long term. However, its long-term increase can still affect economic growth by accumulating human health capital. We demonstrated that the reactions of economic growth to the positive and negative shocks of the three types of health expenditure differ significantly.

However, there is no doubt that this study also has the following shortcomings. First, this paper studies only the asymmetric impact of human health capital on economic growth. Education, another important component of human capital, is not included in this study. Second, we focus only on the impact of China’s human health capital on economic growth. Due to the availability of the data, we use annual data for empirical analysis. This also provides opportunities for future research.