Deciphering the relationship between basic medical insurance and commercial health insurance from 2016 to 2020: a comprehensive analysis from the perspective of 31 provinces in China

Introduction

Background

In the field of healthcare coverage, there are significant differences among countries in the emphasis placed on basic medical insurance (BMI) and commercial health insurance (CHI). This is closely related to the goals pursued and values emphasized in health policies, which evolve and adjust with national conditions and policy changes (1). China implements a healthcare system centered around BMI, with CHI providing a beneficial supplement to meet deeper healthcare needs of the population (2). Against the backdrop of rapid economic and social development, aging demographics, and continuously rising medical costs, there is a rapid increase in demand for healthcare coverage. The “14th Five-Year Plan” for National Medical Insurance proposes to promote the development of a mature and distinctive Chinese healthcare system as the main objective, accelerating the construction of a multi-level healthcare system that covers the entire population, coordinates urban and rural areas, ensures fairness and unity, and is sustainable (3).

China’s BMI system comprises two core components, namely the urban employee BMI and the urban and rural resident BMI. By the end of 2025, the number of insured residents has reached 1.331 billion, with the national insurance participation rate consistently maintained above 95%. The financing mechanism of BMI embodies prominent characteristics of social mutual assistance and risk sharing. The urban employee BMI is jointly financed by employing units and individual employees, where employing units contribute approximately 6% of the total payroll and individual employees contribute 2% of their personal wages. The urban and rural resident BMI operates under a financing mode dominated by fiscal subsidies and supplemented by individual contributions. The BMI system is mainly implemented at the prefectural and municipal levels, resulting in inter-provincial and municipal differences in financing standards and benefit treatment policies. Nevertheless, the system adheres to the principles of fairness and universal benefit, and does not adopt differentiated premium rates based on individual health risks. Insurance premiums are not linked to personal health status or previous medical history. Despite the high coverage rate of BMI, its security scope remains limited. Residents still bear heavy out-of-pocket expenses for serious illnesses and large medical bills, leading to high levels of personal medical expenditure, which creates substantial room for the supplementary role of CHI. According to official data released by the State Administration of Financial Regulation and the National Health Commission, the indemnity expenditure of CHI in China reached 405.2 billion yuan in 2024. As two essential components of the unified national medical security system, BMI and CHI will inevitably interact and complement each other in the long run. Conducting in-depth research on the correlation between the popularization and expansion of BMI and the development of CHI, and exploring effective approaches to promote their coordinated development, will deliver significant academic value and practical implications. Such research facilitates the improvement and optimization of the national medical security system, better satisfies the multi-level medical needs of the people, and promotes the steady advancement of the Healthy China development strategy (4).

Rationale and knowledge gap

Foreign literature exploring the interaction and potential crowd-out effects between public and private healthcare coverage frequently focuses on the expansion of Medicaid under the Affordable Care Act (ACA) in the United States. Recent studies leveraging this post-2014 expansion provide updated insights into this dynamic, yet empirical conclusions remain divergent. For instance, Wagner [2015] (5) focused on Medicaid expansions for working-age disabled individuals and identified a high degree of private insurance crowd-out (ranging from 49% to 100%). Conversely, Courtemanche et al. [2017] (6) utilized data from the American Community Survey and found no evidence that the post-2014 Medicaid expansion crowded out private coverage. Given that academic conclusions remain inconsistent even with recent data, further exploration is warranted to better understand these interaction mechanisms, particularly within different institutional contexts.

Domestic research can be categorized into two main types: micro and macro. At the micro-level, analyses are primarily conducted using data from China Health and Nutrition Survey (CHNS) or China Health and Retirement Longitudinal Study (CHARLS), focusing on residents’ behavior in purchasing CHI to explore the impact of universal BMI on this behavior. Studies by Gan et al. [2010] (7) and Ding et al. [2013] (8) found that BMI effectively reduces the uncertainty of future medical expenses, decreases residents’ perception of medical risks and willingness to demand medical services, thereby reducing their precautionary savings for medical purposes. This reduction in risk and willingness to demand medical services also stimulates residents’ spending in non-medical consumption areas, thus increasing the likelihood of residents purchasing CHI. Cheng et al. [2012] (9) and Xie Bangchang et al. [2015] (10) discovered that BMI promotes resident consumption, increases the utilization of medical services, but does not reduce medical expenses. Therefore, in situations where the coverage of BMI is limited, residents tend to purchase CHI with higher coverage as a supplement to further alleviate the burden of medical expenses. In other words, BMI significantly promotes residents' demand for CHI.

At the macro level, the focus of research is on exploring the correlation between the income of BMI funds and the premium income of CHI in different countries or regions. Wang [2009] (11) examined the relationship between total premium income and total fund income nationwide and found a positive correlation, indicating a trend of simultaneous growth between BMI and CHI. Similarly, Wang [2011] (12) and Zhu et al. [2014] (13) used provincial-level data to reach similar conclusions, suggesting a positive promotional relationship between the two. However, Liu et al. [2010] (14) drew the opposite conclusion based on data from China between 1998 and 2009, suggesting a negative correlation between the increase in the level of social medical security and the income of CHI premiums in the short term.

The interaction mechanism between BMI and CHI shows notable complexity. To unpack this relationship more precisely, scholars have conducted further subgroup comparative studies and found that the association between the two is not simply linear. Zhang et al. [2015] (15) conducted a study using samples from 190 countries worldwide between 1995 and 2009, finding that public medical insurance generally crowded out the CHI market, with variations observed between different countries. Peng et al. [2017] (1) used panel data from 30 regions in China between 2004 and 2014 and found a significant “inverted U-shaped” relationship between BMI and CHI overall. They further discovered that in regions with higher levels of development, the relationship between the two was initially promotional and then crowding out, while in regions with moderate development, BMI promoted the development of CHI. In regions with lower levels of development, there was no significant relationship between the two. Li [2017] (16) used a vector autoregressive model and found that BMI had a certain degree of inhibitory effect on the demand for CHI. However, in the long run, residents would increase their demand for CHI to meet higher-level healthcare needs. Jiang [2019] (17) conducted an in-depth study on the interaction between BMI and CHI using a spatial Durbin model. The results showed that considering only the national scope, BMI significantly promoted the development of CHI.

In summary, the relationship between BMI and CHI manifests as a complex and heterogeneous process, characterized by both crowding-in and crowding-out effects. Some scholars have conducted micro-level analyses to unpack individual responses to BMI, as well as individuals’ decision-making regarding CHI enrollment and utilization. At the macro level, scholars have drawn on large-scale datasets to identify population-wide trends and correlations, delivering more objective and generalizable evidence on how BMI influences CHI. However, the selection of research methods is contingent on specific national contexts, and divergent or even conflicting conclusions have been reached across different analytical frameworks and sample choices. To date, no unified scholarly consensus has been established on this topic. Furthermore, further subgroup analyses by some researchers have documented a non-linear association between BMI and CHI.

China is a geographically vast country with marked disparities in economic development across its regions. Accordingly, regional characteristics and heterogeneity must be accounted for in research to explore the correlation between BMI and CHI across different localities. In addition, China’s medical security system takes BMI as its core pillar, with CHI functioning as a critical supplement that provides residents with differentiated, high-tier insurance coverage. The Chinese government has consistently prioritized the coordinated development of BMI and CHI. Nevertheless, CHI remains in an early developmental stage in China, so the priority of current research lies in unlocking the full supplementary potential of CHI.

For the Chinese context, Peng et al. identified an inverted U-shaped relationship and verified regional heterogeneity using data up to 2014, but their earlier sample cannot reflect the remarkable shifts in China’s medical security system and CHI market in recent years. Furthermore, most existing studies focus on linear relationship tests, lacking marginal effect analysis of the inflection point to fully validate the practical significance of this non-linear association. To address these gaps, our study employs 2016–2020 data, supplements marginal effect analysis to explore the real-world inflection point, and conducts up-to-date regional sub-sample comparisons.

Additionally, this study adopts a macro-level analytical approach, which offers three key advantages. First, macro-level research adopts a systemic perspective that situates both BMI and CHI within the full medical security system. This framework allows for an in-depth examination of the macro-level relationship between the two insurance schemes under the interplay of multiple social factors and variables, yielding a holistic understanding of the overall landscape. It also provides decision-making references for policymakers addressing issues including medical resource allocation, insurance market competition, and social equity. Second, macro-level analysis can be conducted using large-scale datasets and statistical data. Spanning a wide range of regions, population groups, and time periods, such data provide robust empirical support and enable granular trend analysis. Third, macro-level research facilitates cross-regional comparative analysis to explore similarities and differences in medical insurance systems across regions, providing referable insights and practical experience for policymakers and practitioners in other areas.

Objective

Given this backdrop, the primary objective of this study is to investigate the macro-level relationship between BMI and the development of CHI, as well as its regional heterogeneity in China. Specifically, this study aims to fulfill the following three interrelated empirical objectives:

First, to examine whether a non-linear relationship exists between BMI fund revenue and CHI premium income at the macro level, thereby identifying the fundamental nature of their interaction. Second, to investigate regional variations in this interaction by conducting sub-sample regressions across the eastern, central/northeastern, and western regions, explicitly accounting for China’s vast geographic and economic disparities. Third, to clarify the real-world practical significance and dynamic evolution of this relationship within the actual data distribution by performing a marginal effects analysis, thereby pinpointing specific empirical turning points across different regional contexts.

Methods

Research indicators

Following the approach of Peng et al. [2017] (1), this study processes the original data on a per capita basis to mitigate the impact of population size. Therefore, per capita BMI fund income and per capita CHI premium income are selected as indicators to measure the scale of development of both.

To ensure more precise control over the estimation results and reduce the influence of other potential factors, this study first reviews existing literature, comprehensively considers and preliminarily screens factors related to the development of CHI. Subsequently, this study conducts correlation analysis on these influencing factors, screens and removes control variables with significant collinearity, and finally selects five key control variables, including: (I) from the dimension of cultural habits: per capita healthcare expenditure (Indiv_Health) (17). This indicator reflects the local residents’ attitudes towards healthcare and the degree of emphasis on health. (II) From the dimension of economic level (18): consumer price index (CPI). It reflects the changes in retail prices of consumer goods and service prices locally, controlling for the impact of price factors on the results. (III) From the dimension of population structure: population mortality rate (death) (1), a core demographic indicator reflecting population health status and overall risk level. Higher mortality rates indicate greater overall health risks in a region, which will significantly affect residents’ willingness to purchase health insurance; elderly dependency ratio (old) (19). It refers to the proportion of the population aged 65 years and over to the total population. As the aging population increases, people tend to purchase insurance for the elderly to mitigate future health risks. (IV) From the perspective of healthcare infrastructure: government health expenditure as a percentage (Gov_Health) (20). Among the total healthcare expenses, the proportion of government health expenditure is one of the important indicators to measure the government's support for the health sector.

Data sources

This study employs administrative sources for its data to ensure accuracy, using the “China Statistical Yearbook” for BMI fund income, and complementing this with CHI premium income data from the China Banking and Insurance Regulatory Commission and the Wind Financial Database. Control variables’ data are sourced from the China Stock Market & Accounting Research (CSMAR) database. The analysis incorporates provincial panel data from 31 provinces in the mainland of China for the years 2016 to 2020, adjusting for the inclusion of maternity insurance data from 2019 onwards and integrating data from five separately planned cities to maintain consistency. To explore the influence of geographical factors on the development of medical insurance, the study categorizes China into eastern, central, western, and northeastern regions. Due to sample size considerations, the northeastern and central regions are merged, aiming to elucidate the varied impacts of BMI on CHI across different regions.

Model specification

Based on panel data, all regression analyses are performed using two-way fixed effects panel models controlling for province and year, as specified in Eq. [1]. To examine the nonlinear relationship between BMI and CHI, this study includes both the linear and quadratic terms of per capita BMI fund income. Additionally, this study logarithmically transforms the dependent variable, core explanatory variables, and non-ratio-form control variables during regression to reduce absolute differences between data and avoid the influence of individual outliers. All statistical tests are two-sided, with significance levels set at 0.05 and 0.01. Robustness checks are conducted by excluding the pandemic year [2020] and extreme observations (Beijing, Shanghai, Tibet) to ensure the stability of the results.

Commerciali,t=α+β1Basici,t+β2Basicⅈ,t2+δ′Xi,t+ui+θt+ϵi,t[1]

Where Commercial represents the logarithmically transformed per capita CHI premium income, Basic and Basic² represent the logarithmically transformed per capita BMI fund income and its squared term, respectively. X represents all control variables, including per capita healthcare expenditure, CPI, elderly dependency ratio, government health expenditure as a percentage, and population mortality rate. In the panel regression, this study controls for province and year, u represents province fixed effects, θ represents year fixed effects, and ε is the error term; subscripts i and t are used to identify changes in province and year, respectively.

Results

Descriptive statistics

As shown in Table 1, the eastern region has the highest mean per capita BMI fund income (2,456 RMB) and per capita CHI premium income (624.8 RMB), followed by the western region (1,319 RMB and 283.7 RMB) and the central & northeastern region (1,162 RMB and 354.0 RMB). Nationally, the average per capita basic fund income is 1,640 RMB, and the average commercial premium is 414.1 RMB over the sample period.

Regression analysis

The first column of Table 2 shows that the coefficient of Basic is 3.454 (t=5.07) and that of Basic² is −0.238 (t=−4.61), both significant at the 1% level. The negative quadratic term confirms an inverted U-shaped relationship. The turning point (peak) occurs at Basic =7.26, calculated as −β1/(2β2)=−3.454/(2×−0.238)=7.26, which lies within the observed range of per capita BMI fund income (5.5–8.5 in logarithmic form). This indicates that the crowding-out effect of BMI on CHI becomes dominant only after the coverage exceeds this threshold.

In the central & northeastern regions (the third column), the coefficients of Basic and Basic² are 5.552 (t=6.89) and −0.414 (t=−6.57), respectively, with a turning point at Basic =6.71. In the western region (the fourth column), the turning point is at Basic =6.95, which is almost identical to that of the central & northeastern regions. In the eastern region (the second column), the turning point is at Basic =8.47, implying a much longer promotion stage. These results consistently indicate that as BMI expands, CHI is first promoted and then crowded out across all regions.

Marginal effects analysis

In statistics, it has been verified earlier that both the total sample and each subsample exhibit a significant “inverted U-shaped” relationship. However, in real life, variables have their specific meanings and ranges of values, and statistical significance does not necessarily imply practical significance. Therefore, if the inflection point of the quadratic curve does not exist within the sample interval, it means that only one side of the curve, either the left or the right, has practical significance. In other words, there is only a monotonic relationship between per capita BMI fund income and per capita CHI premium income, indicating that only the marginal impact of BMI on CHI is changing. To clarify the relationship between variables in real life, this study calculates the derivative of per capita CHI premium income (commercial) with respect to per capita BMI fund income (Basic). The specific derivative equation is as follows:

Commerciali,t^=β1+2β2Basici,t^[2]

To interpret the practical significance of the quadratic relationship, this study estimates marginal effects using the margins-at-means (MEM) approach. Specifically, all control variables are fixed at their sample means, while the core explanatory variable (per capita BMI fund income) is varied across its observed log-transformed range in 0.5 increments. The marginal effect at each point is calculated as Eq. [2], with corresponding 95% confidence intervals derived via the delta method. This approach allows us to determine whether the inflection point lies within the actual sample range and to characterize the dynamic transition from promotion to crowding out.

Marginal effects analysis for the total sample

Figure 1 presents the marginal effects analysis based on the total sample. The x-axis represents the actual range of per capita BMI fund income, while the y-axis represents the values of the marginal effects of per capita CHI premium income. The upper and lower bounds of the curve represent the 95% confidence interval of the marginal effects at each point.

Figure 1 Total sample: marginal effect of BMI on CHI. The solid line refers to the estimated marginal effect of BMI on CHI, and the upper and lower boundary lines represent the 95% confidence interval of the marginal effect estimates. As per capita BMI fund income increases, the marginal effect shows a continuous downward trend and eventually changes from positive to negative, which verifies the significant inverted U-shaped relationship between the two variables in the national sample. Both the full-sample average and 2020 average fall in the negative marginal effect range, indicating that BMI has generally produced a crowding-out effect on CHI in China in recent years. BMI, basic medical insurance; CHI, commercial health insurance.

At the national average of Basic =7.4 (mean over the full sample period), the marginal effect of BMI on CHI is −0.07 (calculated as β1+2β2×Basic=3.454+2×(−0.238)×7.4=−0.07). At the 2020 average of Basic =7.6, the marginal effect further decreases to -0.16. These negative values indicate that, on average, BMI has already started to crowd out CHI nationally. The marginal effect crosses zero at Basic =7.26 (the turning point), confirming that the crowding-out effect dominates for approximately 60% of the observations.

Marginal effects analysis for the eastern region

Figure 2 depicts the marginal effect estimates for the eastern region. The marginal effect remains positive at the full-sample mean (Basic =7.8), with a value of +0.23. At the 2020 mean (Basic =7.9), the marginal effect is still positive at +0.20. The turning point occurs at Basic =8.47, which is above the 90th percentile of the observed distribution. This implies that for the vast majority of observations in the eastern region, BMI continues to promote the development of CHI without evidence of crowding out.

Figure 2 Eastern region: marginal effect of BMI on CHI. The marginal effect curve presents an obvious inverted U-shaped trend within the observed data range, and the inflection point where the marginal effect equals zero is approximately at X=8.5. The regional average of the logarithmic per capita BMI fund income over the full sample period is 7.8, and the average value in 2020 is 7.9. Both figures are located in the positive interval of marginal effects, which means BMI still plays a promotional role in the development of CHI in the Eastern region at the current stage. The crowding-out effect will only occur when the per capita BMI fund income exceeds the inflection point. BMI, basic medical insurance; CHI, commercial health insurance.

Marginal effects analysis for the central and northeastern regions

Turning to the central and northeastern regions, Figure 3 plots the corresponding estimation results. The marginal effect is already negative at the full-sample mean (Basic =7.0), with a value of −0.24. By 2020 (Basic =7.3), it further declines to −0.49. The turning point is at Basic =6.71, which is lower than the 25th percentile of the regional sample. This suggests that for most years and provinces in these regions, BMI has consistently crowded out CHI, with the effect intensifying over time.

Figure 3 Central & northeastern regions: marginal effect of BMI on CHI. The marginal effect gradually declines and crosses the zero line within the sample value range, confirming the inverted U-shaped relationship between the two variables in the central and northeastern regions. The regional average of the logarithmic per capita BMI fund income for the whole sample period is 7.0, and the average value in 2020 is 7.3. Both values correspond to negative marginal effects, demonstrating that BMI has formed a crowding-out effect on CHI in these regions, and the crowding-out degree keeps intensifying with the continuous expansion of BMI coverage. BMI, basic medical insurance; CHI, commercial health insurance.

Marginal effects analysis for the western region

For the western region, the corresponding marginal effect estimates are shown in Figure 4. The marginal effect is negative at both the full-sample mean (Basic =7.2, marginal effect =−0.13) and the 2020 mean (Basic =7.4, marginal effect =−0.23). The turning point is at Basic =6.95, which lies below the median of the regional sample. This indicates that the crowding-out effect has been the dominant pattern in the western region throughout most of the study period.

Figure 4 Western region: marginal effect of BMI on CHI. The marginal effect shows a sustained downward trend and turns negative in the sample interval, proving the existence of an inverted U-shaped relationship between BMI and CHI in the western region. The regional average of the logarithmic per capita BMI fund income during the sample period is 7.2, and the average value in 2020 is 7.4. All observed values are in the negative marginal effect zone, which indicates that BMI exerts a stable crowding-out effect on CHI in the western region, and the crowding-out effect deepens as the scale of BMI expands. BMI, basic medical insurance; CHI, commercial health insurance.

Robustness checks

Considering the impact of the coronavirus disease 2019 (COVID-19) pandemic on the data and the influence of extreme values on the regression results, this study excluded samples from the year significantly affected by the pandemic and samples with extreme values of core variables. Regression analyses were conducted again to verify the robustness of the results. The regression results are presented in Table 3.

After excluding the pandemic year (2020), the first column of Table 3 shows that the coefficients of Basic and Basic² are 3.259 (t=7.27) and −0.227 (t=−7.32), respectively, which are very close to the original estimates (3.454 and −0.238). The turning point shifts slightly from 7.26 to 7.18, still well within the sample range.

After excluding outliers (Beijing, Shanghai, Tibet), the second column shows coefficients of Basic =4.874 (t=14.26) and Basic²=−0.351 (t=−13.45), with a turning point at Basic =6.94. Despite a modest shift in magnitude, the inverted U-shape remains statistically significant and qualitatively unchanged, confirming robustness.

In the eastern region after excluding Beijing and Shanghai (the third column), the turning point moves from 8.47 to 6.67, suggesting that the prolonged promotion effect in the East is largely driven by these two mega-cities. This finding further highlights the importance of regional heterogeneity.

In the western region after excluding Tibet (the fourth column), the turning point is calculated as 6.93, which remains close to the original estimate of 6.95. The consistency of the coefficients and turning point before and after excluding the outlier confirms that the inverted U-shaped relationship in the western region is not driven by extreme observations and is robust to sample modifications.

Discussion

Key findings

Utilizing panel data from 31 provinces in the mainland of China (excluding Hong Kong, Macao, and Taiwan) spanning from 2016 to 2020, this study employs a two-way fixed effects model controlling for province and year to investigate the impact of BMI implementation on the development of CHI. The results demonstrate a significant “inverted U-shaped” relationship between BMI and CHI, which is statistically significant at the 1% level. To examine whether regional disparities alter the strength and direction of this relationship, a subsample regression analysis was conducted by dividing China into three major regions. The findings indicate that despite variations in development levels across different regions, a significant “inverted U-shaped” relationship between BMI and CHI persists.

Given that variables possess specific meanings and numerical ranges in real-world contexts, statistical significance does not inherently imply practical significance. To further explore the real-world relationships among these variables, this study conducts a detailed analysis of the marginal effects of BMI on CHI within their actual distribution ranges. The empirical results indicate that in the eastern region, from the perspective of the average level over the entire sample period, the marginal effect of BMI on CHI is positive, suggesting that the two are in a stage of coordinated development; from the perspective of the 2020 average level, the marginal effect remains positive, though the promotional effect is gradually weakening. Conversely, in the western, central, and northeastern regions, as well as on a national scale, the marginal effects of BMI on CHI are invariably negative, whether viewed from the sample period average or the 2020 average level. This reveals a crowding-out effect that continuously intensifies as the scale of BMI expands.

Strengths and limitations

Strengths

The primary academic value of this study lies in its integration of a non-linear trend framework with dynamic marginal effects analysis. By calculating marginal effects at specific sequential steps within the actual range of the core explanatory variable, this study translates abstract quadratic regression coefficients into distinct evolutionary trajectories that correspond directly to real-world regional financing levels, thereby offering a more precise quantitative basis for delineating the boundaries between government and market insurance sectors. Additionally, from a systemic macro perspective, the utilization of administrative panel data across 31 provinces effectively mitigates the localized biases inherent in micro-level behavioral data, ensuring high macro-generalizability and institutional relevance for policymakers.

Limitations

First, constrained by unavoidable publication lags in macro-level administrative data, the provincial panel sample is restricted to the specific temporal window of 2016–2020, which warrants a longer-term longitudinal tracking as subsequent statistical series become available to capture the long-term evolution of multi-tiered medical security systems. Second, the marginal effects analysis in this study focuses predominantly on the discussion of numerical trajectories within the actual sample intervals, without extensively expanding on the deep mathematical boundaries of statistical significance at specific point estimates, leaving room for more refined future verifications based on micro-data or spatial econometric models.

Comparison with similar research

As detailed in the preceding literature review, existing macro-empirical conclusions regarding the impact of BMI on CHI remain visibly divergent, primarily intertwining two opposing academic threads: “positive promotion” and “negative substitution”.

Distinguishing itself from traditional macro-level studies that rely heavily on linear specifications, this study examines their interaction within a non-linear framework. The results indicate that the impact of BMI funding scale on CHI is not a simple binary outcome but rather follows a dynamic non-linear trajectory that promotes initially before crowding out. This approach reconciles the conflicting findings in previous literature. Furthermore, by partitioning the sample by geographical regions, this study demonstrates that the crowding-out effects in the central, western, and northeastern regions and the synergistic growth in the eastern region simply reflect different developmental stages along the non-linear curve, thereby providing more targeted macro-empirical evidence for understanding the co-development of China’s multi-tiered medical security system.

Explanations of findings

Inverted U-shaped relationship between BMI and CHI

According to the regression results, there exists an “inverted U-shaped” relationship between per capita BMI fund income and per capita CHI premium income nationwide. This implies that as BMI develops and expands in China, CHI experiences initial promotion followed by crowding-out effects. This “inverted U-shaped” relationship is similarly validated in the sub-sample regressions across different regions.

Promotion effect of BMI on CHI

During the initial stages of the development of China's healthcare system, there emerged a synergistic relationship between CHI and BMI. This is attributed to the increased demand for insurance due to the widespread coverage of BMI (21). The extensive coverage of BMI enables residents to comprehend the essence of medical security, effectively enhancing individual risk avoidance and insurance awareness levels (22,23), thus increasing the demand for and acceptance of CHI. Additionally, during the initial phase of development, due to the limited coverage and protection capacity of BMI, residents actively seek CHI as additional protection to meet more diverse and comprehensive needs, thereby promoting the development of CHI. Moreover, BMI provides residents with a certain predictability of future medical expenses, enhances individual risk tolerance, reduces the incentive for precautionary savings, and ultimately increases residents’ purchasing power, stimulating demand for CHI (24).

Crowding-out effect of BMI on CHI

As China’s healthcare insurance system gradually matures, the quality and accessibility of basic medical security have significantly improved (25). At this stage, both BMI and CHI serve the functions of risk diversification and risk management through insurance channels (26,27). There is a certain degree of substitutability between the two, and the demand for medical insurance in the market is limited. Ultimately, the widespread expansion and improvement of BMI will lead to a gradual shrinkage of the market size of CHI. Despite the increase in demand for insurance markets with the improvement of local economic development, residents’ income levels, and education levels, if the government continues to increase the budget for medical security and vigorously develop and expand the coverage and protection capacity of BMI, the market demand for CHI will gradually decrease. This will weaken the promotion effect of BMI on CHI, gradually transitioning towards mutual competition and squeezing. Eventually, a “crowding-out effect” will occur.

Regional disparities in the marginal effects of BMI and CHI

Crowding-out effect in the western, central, and northeastern regions

According to the marginal effect analysis mentioned earlier, on average, the influence of BMI on CHI in China is transitioning from mutual promotion to competition and crowding-out effects. For the western, central, and northeastern regions, the situation is similar, with BMI exerting a crowding-out effect on CHI, and the degree of crowding-out deepening with the continuous expansion of BMI.

In China, there are regional disparities in healthcare resources, with the eastern regions relatively more developed compared to the western, central, and northeastern regions, which face shortages in healthcare resources and lag behind in various aspects of development (28-30). Additionally, the Chinese government has been committed to expanding the coverage of BMI, especially targeting low-income groups and rural areas, by increasing government investment and adjusting policies to achieve the goal of universal health insurance (31). These policy measures have led to a gradual increase in the beneficiaries of BMI, while reducing the market space for CHI. With government support and promotion, CHI faces greater pressure in competition.

Given the limited purchasing power, many people are inclined to choose BMI because it offers broader coverage and lower costs. Meanwhile, as competition intensifies with limited demand, CHI companies may choose to withdraw from the market in those regions. These factors interact with each other, ultimately leading to the crowding-out of the CHI market by BMI in these regions.

Promotion in the eastern region

In contrast to other regions, the eastern region, which is the most developed in the field of medical insurance, exhibits a positive marginal effect on average. This positive marginal effect persisted in 2020 without showing signs of market saturation. This phenomenon is believed to be correlated with the region’s economic development level (32). The eastern region leads in social development, with data from the four major first-tier cities (Beijing, Shanghai, Guangzhou, Shenzhen) significantly ahead. In a scenario where the economy is advanced, individual income levels and consumption capabilities are high, and BMI continues to expand while residents’ awareness of insurance strengthens, people are more willing to purchase CHI to further enhance medical security for themselves and their families (10). Specifically, in the eastern region, economic growth stimulates demand. However, the development of the insurance industry lags behind the rapid economic growth in the area. Thus, despite the stable expansion of medical insurance, the demand for medical security has not been fully met and covered. There still exists a certain insurance gap in the market, providing a favorable environment for CHI. Ultimately, BMI promotes the development of CHI.

Future development trajectories

Building on the identified inverted U-shaped relationship between BMI and CHI, it is essential to consider the potential impacts of recent healthcare reforms in China on this dynamic. In recent years, the Chinese government has implemented a series of pivotal reforms, including the nationwide rollout of diagnosis-related groups (DRG)/diagnosis-intervention packet (DIP) payment systems and the outpatient mutual aid security reform. While these reforms primarily target the expenditure side of the BMI fund and do not directly alter its revenue scale, they fundamentally reshape the conversion efficiency of fund income into actual healthcare protection. Specifically, the DRG/DIP reforms have effectively curbed excessive medical services and improved fund utilization efficiency, enabling the same level of BMI fund income to provide broader and deeper medical coverage. Concurrently, the outpatient mutual aid reform has extended BMI coverage from inpatient services to general outpatient care, filling a long-standing gap in basic medical security.

These changes are likely to modify the trajectory of the inverted U-shaped relationship in the foreseeable future. Nationally, the enhanced protection capacity of BMI may accelerate the arrival of the inflection point of the inverted U curve, potentially intensifying the crowding-out effect on the low-end CHI market in the short term. However, in the long run, as these reforms help clarify the functional boundaries of BMI and improve its long-term sustainability, the government is expected to focus more on guaranteeing basic medical needs, thereby creating more space for CHI to develop high-end and differentiated products. Regionally, the impacts may remain heterogeneous: in regions where the crowding-out effect has already been observed, the improved BMI protection may further squeeze the low-end CHI market; while in regions with stronger economic strength and higher demand for diversified healthcare services, the reforms may instead promote the transformation and upgrading of CHI towards high-end medical services and health management, reinforcing the synergistic development trend.

Implications and actions needed

Clarify the positioning and functions of BMI, and delineate the boundaries between government and market

Currently, China’s BMI system has provided coverage services to the vast majority of the population nationwide. However, there is still no clear definition of its functional positioning and market scope compared to CHI. To address the crowding-out effect issue, the government should further clarify the positioning and functions of BMI, delineate the boundaries between government and market, and avoid the gradual shrinkage of the CHI market due to the encroachment of BMI.

The government can set reasonable coverage standards, such as adjusting insurance payment levels appropriately, and focus on medical security for special populations to maintain the development space and market competitiveness of CHI. At the same time, the government should encourage CHI companies to provide higher-level and more diverse medical insurance plans for high-income groups, complementing BMI and forming a mutually supportive market structure.

Consider market segmentation and prioritize customer experience

CHI companies should focus on market segmentation: considering that the crowding-out effect of BMI on CHI is gradually strengthening in western, central, and northeastern regions, CHI companies in these areas can focus on innovating personalized products, such as high-end medical security or special disease insurance. By providing differentiated, high-value-added products and services, they can attract target customer groups and increase market share.

In addition, CHI companies need to pay attention to customer experience and satisfaction. For example, in claims services, CHI companies can use digital technology and automated processes to achieve fast settlement, risk assessment, and review, providing more convenient claims services, reducing claim time and process, and improving the efficiency and satisfaction of insurance claims.

Promote cross-industry cooperation and innovative development

To promote the healthy development of the CHI market and fully play its important complementary role, the government can promote cooperation between CHI companies and various parties in the medical industry to jointly establish health records, digitize and unify the management of patients’ health data and information. In addition, a health management information platform can be established to achieve collaboration and information sharing among medical institutions, doctors, and patients, providing customized, comprehensive, and full-cycle health management services, thereby achieving precision medicine, optimizing the process and resource allocation of medical services, and reducing patients’ medical expenses burden while improving patient satisfaction with medical services.

Conclusions

Overall, the improvement of China’s medical security system still faces multiple challenges that must be addressed jointly by the government and CHI insurers. First, the government should formulate targeted policies to guide CHI insurers in adopting sustainable operational models, thereby curbing excessive market competition and risks stemming from capital operations and safeguarding the legitimate rights and interests of policyholders. The sustainability of CHI products also warrants dedicated attention: insurers need to design more competitive and sustainable products to better respond to market demand. Second, the government may strengthen the supervision, regulation and guidance of the CHI market by establishing well-defined market entry and exit mechanisms as well as risk monitoring frameworks, in a bid to foster the sound development of the CHI sector. For its part, the CHI market needs to build information sharing platforms and deepen collaborative partnerships with medical institutions to improve the utilization efficiency of medical resources. Finally, CHI insurers should reinforce product outreach and risk management. They shall provide policyholders with more detailed product explanations and risk disclosures, helping them fully grasp the coverage and inherent risks of insurance products and reduce potential misunderstandings and dissatisfaction. Meanwhile, the government can scale up public education and outreach to enhance public understanding of CHI products, services and coverage provisions, and elevate the status and role of CHI within the medical security system. These efforts will give full play to the supplementary function of CHI in medical security and further promote the healthy development of the CHI market.

Acknowledgments

None.

Footnote

Peer Review File: Available at https://jphe.amegroups.com/article/view/10.21037/jphe-2026-0024/prf

Funding: This research was funded by the Guangzhou Key Research Base for Medical Ethics, China (Sui She Gui Ban [2024] No. 9).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jphe.amegroups.com/article/view/10.21037/jphe-2026-0024/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Peng H, Zheng Q, Guo Y. Does the expansion of social medical insurance in China promote the development of commercial health insurance? Journal of Financial Research 2017;(5):97-110.
2. Yip W, Fu H, Chen AT, et al. 10 years of health-care reform in China: progress and gaps in Universal Health Coverage. Lancet 2019;394:1192-204. [Crossref] [PubMed]
3. Poo MM. Innovation and reform: China's 14th Five-Year Plan unfolds. Natl Sci Rev 2021;8:nwaa294. [Crossref] [PubMed]
4. Tan X, Liu X, Shao H. Healthy China 2030: A Vision for Health Care. Value Health Reg Issues 2017;12:112-4. [Crossref] [PubMed]
5. Wagner KL. Medicaid expansions for the working age disabled: Revisiting the crowd-out of private health insurance. J Health Econ 2015;40:69-82. [Crossref] [PubMed]
6. Courtemanche Ch, Marton J, Ukert B, et al. Early Impacts of the Affordable Care Act on Health Insurance Coverage in Medicaid Expansion and Non-Expansion States. J Policy Anal Manage 2017;36:178-210. [Crossref] [PubMed]
7. Gan L, Liu G, Ma S. The impact of basic medical insurance on promoting household consumption. Economic Research Journal 2010;45:30-8.
8. Ding J, Ying M, Du Z. A study on rural household consumption behavior in China: An analysis from the perspective of health risk and medical security. Journal of Financial Research 2013;(10):154-66.
9. Cheng L, Zhang Y. The New Rural Cooperative Medical Scheme: Economic performance or health performance?. Economic Research Journal 2012;47:120-33.
10. Xie B, Han J. The impact of basic social medical insurance on household consumption. Journal of Business Economics 2015;(5):79-87.
11. Wang L. An empirical analysis of effective demand for health insurance in China. Journal of Technical Economics & Management Research 2009;(1):100-2.
12. Wang X. Social medical insurance, market structure and the development of commercial health insurance in China. Insurance Studies 2011;(7):35-41.
13. Zhu M, Gui Z. The relationship between basic medical insurance and commercial health insurance in the health financing system: A study based on provincial panel data from urban China (2003–2012). Insurance Studies 2014;(6):96-104.
14. Liu F, Wang X, Bian H. An empirical analysis of factors influencing the development of commercial health insurance in China. Chinese Journal of Health Policy 2010;3:38-44.
15. Zhang Y, Hu B, Chen J. A study on the crowding-out effect of public medical insurance on commercial health insurance: An empirical analysis from a dynamic perspective. Journal of Southeast University (Philosophy and Social Sciences) 2015;17:83-8.
16. Li Y. A study on factors affecting the demand for commercial health insurance in China based on the VAR model. Economic Forum 2017;(7):118-121,137.
17. Jiang L. A study on the developmental relationship between basic medical insurance and commercial health insurance in China: An analysis based on the spatial Durbin model. Journal of Financial Development Research 2019;(2):68-75.
18. Xie M, Li Q. Complementarity or substitution between commercial health insurance and social medical insurance in China: An analysis based on interprovincial panel data and threshold effects. Journal of Jiangxi University of Finance and Economics 2020;(3):52-60.
19. Cai X. A study on the crowding-out effect of basic medical insurance on commercial health insurance in China: A provincial dynamic panel analysis based on two-step system GMM estimation. Mathematics in Practice and Theory 2025;55:42-60.
20. Liu S, Liu X, Liu J, et al. The impact of basic medical insurance on commercial health insurance: Promotion or inhibition. China Soft Science 2020;(11):172-81.
21. Liu K, Liu W, Frank R, et al. Assessing the long-term effects of Basic Medical Insurance on catastrophic health spending in China. Health Policy Plan 2022;37:747-59. Erratum in: Health Policy Plan 2022 May 12;37(5):675. [Crossref] [PubMed]
22. Dong B. The promotion of pooling level of basic medical insurance and participants' health: impact effects and mediating mechanisms. Int J Equity Health 2023;22:113. [Crossref] [PubMed]
23. Liao F, Hu W, Zhang C. Does basic medical insurance promote public health? Evidence from China family panel study. Front Public Health 2023;11:1269277. [Crossref] [PubMed]
24. Tian L, Dong H. Study on the Dynamic Relationship between Chinese Residents' Individual Characteristics and Commercial Health Insurance Demand. Int J Environ Res Public Health 2023;20:4797. [Crossref] [PubMed]
25. Li Q, Zhang L, Jian W. The impact of integrated urban and rural resident basic medical insurance on health service equity: Evidence from China. Front Public Health 2023;11:1106166. [Crossref] [PubMed]
26. Wu J, Qiao J, Nicholas S, et al. The challenge of healthcare big data to China's commercial health insurance industry: evaluation and recommendations. BMC Health Serv Res 2022;22:1189. [Crossref] [PubMed]
27. Zhao T, Wanyan R, Suo L. Research on the Effect of Commercial Health Insurance Development on Economic Efficiency. Int J Environ Res Public Health 2023;20:5178. [Crossref] [PubMed]
28. Dong E, Xu J, Sun X, et al. Differences in regional distribution and inequality in health-resource allocation on institutions, beds, and workforce: a longitudinal study in China. Arch Public Health 2021;79:78. [Crossref] [PubMed]
29. Zhang T, Xu Y, Ren J, et al. Inequality in the distribution of health resources and health services in China: hospitals versus primary care institutions. Int J Equity Health 2017;16:42. [Crossref] [PubMed]
30. Yuan L, Cao J, Wang D, et al. Regional disparities and influencing factors of high quality medical resources distribution in China. Int J Equity Health 2023;22:8. [Crossref] [PubMed]
31. Wang J, Zhu H, Liu H, et al. Can the reform of integrating health insurance reduce inequity in catastrophic health expenditure? Evidence from China. Int J Equity Health 2020;19:49. [Crossref] [PubMed]
32. Zhou S, Zhou C, Yuan Q, et al. Universal Health Insurance Coverage and the Economic Burden of Disease in Eastern China: A Pooled Cross-Sectional Analysis From the National Health Service Survey in Jiangsu Province. Front Public Health 2022;10:738146. [Crossref] [PubMed]