Does Health Insurance Boost Subjective Well-being? Examining the Link in China through a National Survey

1 Introduction

Enhancing national welfare and fostering a heightened sense of access and happiness are crucial objectives in the trajectory of economic development. Since the initiation of reform and opening up in 1978, China has sustained prolonged and robust economic growth, witnessing a remarkable increase in GDP from RMB 364.5 billion in 1978 to RMB 121.02 trillion in 2021. However, despite continuous economic growth, the well-being of Chinese residents has not seen a corresponding rise in happiness; instead, it has faced a stagnation in happiness growth. Data from the World Values Survey between 1990 and 2018 reveal a concerning trend. The percentage of Chinese residents reporting being “very satisfied” with their lives declined from 27.5% in 1990 to 11.5% in 2001. Although there was a subsequent recovery to 27% in 2018, this figure remained below the 1990 level. This suggests a complex relationship between economic growth and the subjective well-being of the Chinese population, warranting a nuanced exploration of the factors influencing this phenomenon (Haerpfer et al., 2022). The World Happiness Index Report 2020, released by the United Nations, indicates a decline in China’s happiness ranking among 156 countries and regions globally. China’s position dropped from 79th in 2016 to 93rd in 2019, accompanied by a decrease in the happiness index from 5.273 to 5.124 (Helliwell et al., 2020). Certainly, the evolving focus of Chinese residents from material wealth to the quality of life, especially health protection, indicates a noteworthy shift alongside the economic growth. Investigating the factors that contribute to the improvement of national happiness becomes highly significant in this context.

A well-established social security system is widely believed to address social risks, alleviate societal pressures, and boost residents’ happiness. Presently, China grapples with emerging social conflicts and risks, including the challenges of economic transformation, an aging population, and soaring housing prices. These issues escalate life pressures and diminish perceived happiness. A robust social security system, however, can mitigate economic losses, counteract negative risk effects, alleviate impacts on physical and mental well-being, and ensure peace and stability, ultimately enhancing overall happiness (Tran et al., 2017).

As a vital component of the social security system, commercial insurance plays a crucial role in providing residents with personalized risk protection options, garnering increasing attention for its impact on well-being. Existing literature indicates that health shocks significantly diminish households’ total financial wealth. When allocating assets to mitigate risks, households prioritize life insurance (Berkowitz & Qiu, 2006; Cardak & Wilkins, 2009; Cavapozzi et al., 2013). Consequently, residents with life insurance are more inclined to invest in stocks or funds (Cavapozzi et al., 2013). Social health insurance or commercial health insurance coverage reduces the risk of medical expenses, enabling increased investment in risky financial assets (Jing et al., 2017).

Commercial insurance also caters to personalized protection needs, alleviating negative risk effects and contributing to residents’ happiness perception. For instance, commercial health insurance can boost individuals’ utilization of medical services, enhancing residents’ happiness through improved health levels (Cassells, 2019). Subjective well-being, rooted in positive emotional experiences, correlates with physical and mental happiness, leading to an overall increase in well-being for participants in commercial health insurance (Kim & Koh, 2022; Kobayashi et al., 2019).

However, the higher cost of commercial health insurance can limit residents’ disposable income for self-fulfillment, potentially diminishing well-being (Drake et al., 2017). The impact of commercial health insurance on residents’ subjective well-being remains challenging to consistently infer at the theoretical level. This study explores the relationship between commercial health insurance and residents’ well-being, delving into the perspective of commercial health insurance.

2 Literature Review

Happiness is a pivotal factor influencing a country’s overall well-being and economic development. The study of happiness originated in the mid-twentieth century, with scholars defining it as the subjective feelings of individuals seeking pleasure and avoiding pain, hence the term “subjective well-being”(Diener et al., 2018).

Previous research has explored the impact of macro-level factors, such as inflation, unemployment, economic globalization, gross national product, ecological quality, government governance, and social digitization on population well-being (Blanchflower et al., 2014; Prati, 2023). Summarily, increased inflation diminishes national happiness. Unemployment escalates economic losses and psychological stress, reducing overall well-being (Blanchflower et al., 2014). Environmental degradation lowers residents’ subjective well-being (Dong et al., 2022; Li et al., 2014), but effective environmental management enhances physical and mental health, significantly boosting well-being (Guo et al., 2020). Scholars emphasize that the level of government governance is crucial for improving residents’ well-being, serving as a universal condition for well-being (Ott, 2010, 2011). Enhanced government governance substantially elevates public subjective well-being, exerting a greater impact than economic growth (Helliwell et al., 2018). Micro-level factors like income, education, and gender also draw attention. While higher income positively correlates with residents’ well-being (Jachimowicz et al., 2021; Kushlev et al., 2015; Tsui, 2014), a widening social income gap negatively affects well-being (Ugur, 2021; Yang et al., 2019). However, conflicting findings exist, with some studies suggesting no correlation between income inequality and public happiness (Lam & Liu, 2014), and even indicating that widening income disparity increases happiness due to a strong “demonstration effect” (Knight et al., 2009; Ma & Chen, 2022). Similarly, there are contradictory findings regarding educational attainment and gender (Clark & Oswald, 1994; Frey & Stutzer, 2000; Layard, 2006; Moljord et al., 2011). Therefore, in the following analysis, income, education, and gender serve as control variables to further explore their effects on the subjective well-being of Chinese residents.

With the global enhancement of social and commercial insurance systems, scholars are increasingly delving into the correlation between insurance and residents’ well-being. Regarding social insurance, research indicates that participating in social insurance contributes to an overall improvement in residents’ well-being and enhances their sense of happiness (Tran et al., 2017). Social health insurance, in particular, exerts a significant positive impact by enhancing population health, reducing economic uncertainty due to health shocks, and addressing health inequalities through equal access to healthcare services (Keng & Wu, 2014). Analysis based on the 2014 Chinese Household Panel Study reveals a positive association between basic social health insurance and the well-being of the elderly (Han et al., 2022). Similarly, research utilizing China General Social Survey (CGSS) 2013 data confirms that basic health insurance enhances the well-being of low-income groups (Gu et al., 2017). Studies also find that residents with unemployment insurance, workers’ compensation insurance, health insurance, and pension insurance experience improved well-being (Chen et al., 2022). In summary, the extensive and comprehensive coverage of social insurance significantly elevates residents’ life satisfaction and well-being (Han et al., 2022). Expanding on the exploration of social security and residents’ well-being, the relationship between commercial insurance and well-being has gained increased attention. From a commercial insurance perspective, its unique advantages in service scope, depth, and mode make it a valuable complement to social insurance, enhancing residents’ protection and overall well-being. Scholars investigating this relationship have found that residents with commercial health insurance exhibit significantly higher health levels (Hullegie & Klein, 2010). Studies conducted in China indicate that both property and life insurance positively impact residents’ well-being (Hu & Su, 2014), and commercial pension insurance catalyzes improvements in residents’ well-being (Zhang et al., 2018).

China’s decentralized public health insurance system comprises various plans with differing premiums and benefit levels, leading to varying impacts on residents’ well-being (Yang & Hanewald, 2022). The two primary categories of health insurance in China are basic social health insurance and commercial health insurance. Basic social health insurance includes mandatory basic medical insurance for urban and rural residents, with an annual premium of 610 RMB per person in 2022. Those with formal employment are typically required to purchase employee health insurance, costing between RMB 2,670 and RMB 5,874 per person in 2022 (China National Healthcare Security Administration, 2023). Commercial health insurance is voluntary and incurs costs ranging from thousands to tens of thousands of yuan annually, with an average individual payment of 6,000 yuan in 2022 (Yan & Yu, 2022). Given the semi-mandatory or mandatory nature of basic social health insurance, as of the end of 2022, over 1.346 billion people were enrolled nationwide, boasting a stable participation rate exceeding 95%. In contrast, the participation rate for commercial health insurance stands at only 10% (China National Healthcare Security Administration, 2023).

The objective of social health insurance is to provide insured residents with fundamental medical protection, but it comes with limitations on benefits, medications, services, and diagnostic and therapeutic programs. For instance, lifestyle and medical cosmetic treatments are not covered by basic social health insurance. Employee medical insurance starts at 1,500 yuan, with a maximum benefit cap of 400,000 yuan. Basic medical insurance establishes annual payout limits for medical outpatient payments related to chronic diseases; for example, the maximum annual payout for diabetes mellitus (with serious complications) is 3,000 yuan, and for high blood pressure (Stage III at high risk and above), it is 2,000 yuan. Conversely, commercial medical insurance offers more comprehensive and flexible coverage in terms of benefits, medications, services, and diagnostic and treatment programs. With a lower “starting line” (typically referring to the initial amount that an individual is required to pay for their medical expenses) and a higher upper limit on benefits, it covers medical cosmetic treatments and outpatient expenses for chronic diseases such as high blood pressure (Stage III high risk and above) and diabetes (with serious complications) at 6,000 yuan or more annually (Yan & Yu, 2022).

However, in comparison to developed countries, the evolution of commercial health insurance in China has been relatively brief, and the system is not yet fully developed. The coverage rate of commercial health insurance in China is below 10% (Yan & Yu, 2022). Consequently, there is a dearth of research by Chinese scholars on the correlation between commercial health insurance and the well-being of Chinese residents. Therefore, this article addresses the following two research questions:

RQ1: Is there a happiness effect of commercial health insurance? Does the purchase of commercial health insurance help to improve the subjective well-being of Chinese residents?

RQ2: Is there heterogeneity in the health effects of commercial health insurance for different types of residents in China?

3 Methods

4 Results and Discussion

4.1 Propensity Score Matching Result

4.1.1 Calculating the Propensity Score

The propensity score, representing the conditional probability of residents purchasing commercial health insurance, is crucial for estimating its impact on subjective well-being. Table 2 presents the outcomes of logit regression model scores. Education, age, personal income, family income, ownership of a private car, and religious beliefs exhibit a statistically significant positive influence on residents’ likelihood to buy commercial health insurance. In the matching process, involving all control variables does not alter the results of the propensity score matching method. Instead, it enhances matching conditions, thereby refining the model accuracy (Caliendo & Kopeinig, 2008).

Table 2

Results of logit regression model scores

HeaInsuran	Coefficient	Std. err.	z	P > |z|	95% conf. interval
_cons	−7.596	0.876	−8.680	0.000	−9.312	−5.880
Family size	0.014	0.032	0.440	0.663	−0.048	0.076
Gender	0.100	0.107	0.930	0.351	−0.110	0.310
Education level	0.179	0.049	3.680	0.000	0.084	0.274
Age	−0.018	0.005	−3.930	0.000	−0.027	−0.009
Subsocequity	0.081	0.058	1.400	0.160	−0.032	0.194
Personincomln	0.348	0.069	5.020	0.000	0.212	0.483
House income	0.423	0.167	2.530	0.011	0.095	0.750
Private car	−0.338	0.114	−2.960	0.003	−0.561	−0.114
Ethnic	−0.073	0.214	−0.340	0.732	−0.493	0.346
Religibelief	0.530	0.195	2.730	0.006	0.149	0.912
Hukou	0.244	0.125	1.960	0.050	−0.000	0.489
Unmarried	0.011	0.323	0.030	0.974	−0.622	0.643
Married	0.030	0.273	0.110	0.912	−0.504	0.564
Divorce	0.646	0.355	1.820	0.069	−0.050	1.342
bereave	0.0000 (omitted)
LR chi2	423.630
Prob > chi2	0.000
Pseudo R 2	0.147

4.1.2 Calculating the Matching Effect

Nearest neighbor matching with a caliper (K = 4, caliper = 0.01) was employed for one-to-four matching (Austin, 2014). Table 3 presents the outcomes of propensity score matching regarding the impact of acquiring commercial health insurance on residents’ subjective well-being. The results indicate a statistically significant positive effect of purchasing commercial health insurance on residents’ subjective well-being at the 0.5% significance level. The acquisition of commercial health insurance demonstrates a validated increase in the coefficient of influence on the population’s subjective well-being by 1.0699.

Table 3

Results of ATT, ATU, and ATE

Variable	Sample	Treated	Controls	Difference	S.E.	T-stat
Sub well-being	Unmatched	4.108	3.986	0.121	0.038	3.220*
	ATT	4.108	3.038	1.070	0.242	4.417*
	ATU	3.991	4.050	0.059	—	—
	ATE			0.061	—	—

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

4.1.3 Balance Test, Common Support Domains, and Kernel Density Plots

In the context of propensity score matching, we aim to mitigate potential selection bias, ensuring that the two groups – those who purchased commercial health insurance and those who did not – exhibit similar characteristics on control variables. The balance test is conducted to verify if postmatching, the control variables between the two groups are sufficiently balanced, indicating a reduction in potential selection bias.

Table 4 demonstrates the results of the balance test for the control variables, and the absolute value of the standard deviation of all control variables after matching is less than 5%. Except for three variables, such as family size, gender, and ethnicity, all other variables have positive and large reduction values of deviation, which indicates that the selection bias between the group that purchased commercial health insurance and the group that did not purchase commercial health insurance was significantly reduced after matching and therefore passed the balance test.

Table 4

Results of the balance test for the control variables

Variable	U	Mean		% Reduct		t-test		V(T)/V(C)
	M	Treated	Control	% Bias	|bias|	t	p > |t|
Family size	U	3.303	3.294	0.500	−734.600	0.110	0.912	1.000
	M	3.306	3.226	4.400		0.770	0.443	1.380*
Gender	U	0.493	0.486	1.400	−94.800	0.290	0.770	—
	M	0.495	0.482	2.700		0.440	0.661	—
Education level	U	4.346	3.184	80.100	97.900	17.160	0.000	1.140
	M	4.341	4.366	−1.700		−0.270	0.790	0.990
Age	U	44.647	54.404	−65.500	97.100	−13.020	0.000	0.730*
	M	44.678	44.396	1.900		0.320	0.752	0.820*
Subsocequity	U	3.497	3.481	1.700	81.600	0.340	0.731	0.860
	M	3.503	3.506	−0.300		−0.050	0.958	1.060
Personincomln	U	10.984	9.971	83.000	99.400	16.350	0.000	0.680*
	M	10.970	10.964	0.500		0.090	0.931	1.060
Private car	U	0.372	0.618	−50.600	96.800	−10.570	0.000	—
	M	0.374	0.382	−1.600		−0.260	0.794	—
Ethnic	U	0.067	0.068	−0.300	−274.800	−0.050	0.957	—
	M	0.067	0.065	1.000		0.160	0.876	—
Religibelief	U	0.090	0.068	8.300	74.200	1.820	0.069	—
	M	0.089	0.083	2.100		0.330	0.740	—
Hukou	U	0.641	0.430	43.200	96.100	8.950	0.000	—
	M	0.640	0.648	−1.700		−0.280	0.780	—
Bereave	U	0.037	0.097	−24.400	96.800	−4.500	0.000	—
	M	0.037	0.039	−0.800		−0.160	0.870	—
Divorce	U	0.056	0.028	13.800	100.000	3.260	0.001	—
	M	0.056	0.056	0.000		0.000	1.000	—
Married	U	0.731	0.768	−8.600	89.400	−1.830	0.068	—
	M	0.732	0.728	0.900		0.140	0.887	—
Unmarried	U	0.177	0.106	20.200	97.200	4.580	0.000	—
	M	0.175	0.177	−0.600		−0.080	0.933	—

Note: If variance ratio outside [0.84; 1.19] for U (Unmatched) and [0.84; 1.19] for M (Matched); * represent 1, 5, and 10% significance levels, respectively.

Furthermore, Table 5 illustrates that after matching, the pseudo R ² is 0.0010, significantly smaller than the 0.1440 before matching. The likelihood ratio (LR) chi² value decreases from 415.4800 before matching to 1.5700 after matching, shifting from significant to nonsignificant. This change indicates the absence of statistical differences between the group that purchased commercial health insurance and the group that did not after matching, confirming the achieved balance (Austin, 2009). The overall balance test results demonstrate the effectiveness of propensity score matching in reducing selection bias and ensuring comparability between the groups.

Table 5

Results of the balance test for the control variables

Sample	Pseudo R 2	LR chi2	p > chi2	Mean bias	Med bias	B	R	% Var
Unmatched	0.1440	415.4800	0.0000	28.7000	17.0000	103.5000	0.7600	40.0000
Matched	0.0010	1.5700	1.0000	1.4000	1.3000	7.8000	1.0900	40.0000

Note: If B > 25%, R outside [0.5; 2].

In Figure 1a, the standardized deviations of control variables show a significant reduction after applying propensity score matching. The deviations closely align with the 0 line, indicating that the distribution of control variables has become more balanced between the treated (those who purchased commercial health insurance) and untreated groups. This reduction in deviations implies a successful mitigation of selection bias, enhancing the comparability of the two groups. Figure 1b provides a visual representation of the common support domain after matching. The majority of samples now fall within the shared range of values, demonstrating that there is substantial overlap in the distributions of propensity scores between the treated and untreated groups. This common support is crucial for ensuring that both groups are comparable and that the matching process has effectively addressed potential confounding variables (Rassen et al., 2012). Out of the initial 3,293 samples, 3,212 were successfully matched through propensity score matching. Among these, 519 belong to the treated group (those who purchased commercial health insurance), and 2,693 belong to the untreated group. This successful matching process contributes to the creation of a more balanced dataset, allowing for a more robust analysis of the impact of purchasing commercial health insurance on residents’ well-being.

Figure 1

(a) Standardized deviations of control variables and (b) common support domain after matching.

Figure 2 visually represents the kernel densities before and after applying propensity score matching, offering insights into the effectiveness of the matching process. Prior to matching, there is a noticeable disparity in propensity scores between the group that purchased commercial health insurance and the group that did not. The distributions exhibit a clear gap, suggesting a significant difference in the likelihood of purchasing commercial health insurance between the two groups. However, after implementing propensity score matching, the gap in scores between the two groups is markedly reduced. The kernel densities now overlap more closely, indicating a successful alignment of the propensity score distributions. This reduction in the gap signifies that the matching process has effectively balanced the observed covariates, making the treated and untreated groups more comparable in terms of their likelihood to purchase commercial health insurance. In essence, Figure 2 demonstrates that propensity score matching has worked well in minimizing the initial discrepancy in propensity scores, enhancing the comparability of the two groups and strengthening the validity of the analysis.

Figure 2

Kernel densities before and after matching.

5 Conclusion

This study investigated the effect of residents’ participation in commercial health insurance on their well-being using ordered logit model and propensity score matching method based on CGSS2021 Chinese national survey data. It was found that participation in commercial health insurance can significantly promote residents’ happiness, commercial health insurance had a greater effect on promoting happiness for middle-high income groups than for low-income groups.

Based on the aforementioned findings and discussions, this study proposes the following practical suggestions: First, the Chinese government should provide more policy support for the development of commercial health insurance. A feasible measure is to provide relatively preferential tax measures for universal commercial health insurance products and to guide commercial health insurance in both directions by economic means such as giving certain subsidies to residents who purchase commercial health insurance to combine with livelihood protection. Second, the increase in income and social trust will enhance the happiness effect of commercial health insurance. Therefore, while raising the income level of residents, the government should strengthen the standardized publicity of commercial health insurance, vigorously publicize and popularize the role of commercial health insurance for personal protection to the public, so that residents can have a comprehensive understanding of commercial health insurance and eliminate the doubts in people’s mind about commercial health insurance, so as to increase the public’s recognition of commercial health insurance; insurance companies should also actively launch different kinds of insurance products with various prices Insurance companies should also actively introduce different kinds of inclusive insurance products with various prices, so that people can choose suitable insurance items according to their own conditions, thus allowing commercial health insurance to better benefit low-income groups and low social trust groups to enhance their sense of well-being.