Introduction

Advancing human well-being is the ultimate goal of measuring societal progress. In the early stages of industrialization, people worldwide equated economic development with the enhancement of human well-being. Therefore, promoting economic development has become a widely discussed topic. However, with the acceleration of industrialization, issues such as air pollution have gradually emerged, leading people to realize that economic development does not necessarily equate to the improvement of social well-being. Consequently, in recent years, scholars have shifted their focus towards improving welfare. As the largest developing country, China has improved the quality of life and achieved key goals related to shared prosperity and governance. Increasing the happiness index has not only aroused widespread public interest but also become a crucial topic in economic and sociological research.

In the early stages, the focus on people’s well-being was primarily within the realms of psychology and sociology, whereas economic studies on happiness can be traced back to the 1970s. However, traditional economic research has predominantly focused on the influence of macroeconomic conditions, income levels, and consumption patterns on people’s sense of happiness (Ferrer-i-Carbonell, 2005; Chan and Mogilner, 2017; Wu and Li, 2017). Although improvements in public infrastructure—including healthcare, education, transportation, and social security systems—have been identified as important factors in enhancing convenience and well-being, less attention has been paid to the significant role of infrastructure development in increasing happiness (Ettema et al. 2012). Moreover, there is currently a lack of understanding regarding the mechanisms through which infrastructure enhances happiness.

As a key aspect of transportation infrastructure, the opening of high-speed rail (HSR) has a profound effect on medical care, education, income, and other factors, thereby significantly influencing residents’ subjective well-being (SWB) in China. By the end of 2021, the HSR coverage rate of county-level administrative units exceeded 81%. In particular, in cities with populations exceeding one million, HSR coverage has reached 95%. The development of HSR networks significantly improves access to medical treatment for patients. Due to its efficiency, convenience, and resilience to weather conditions, HSR promotes patient mobility between different levels of cities and between urban and rural areas (Li, 2020; Xia et al. 2022), thereby promoting health equity and improving residents’ SWB (Angner et al. 2013). In addition, the opening of HSR stimulates regional economic development in underdeveloped areas by attracting interprovincial investment and improving local employment and income levels (Di Matteo and Cardinale, 2023; Liu and Yang, 2023). This reduces household income inequality and improves residents’ SWB (Wu and Li, 2017).

Furthermore, the environment is an important factor that affects residents’ SWB. Compared to other forms of transportation infrastructure, HSR offers advantages such as convenience (Zhou et al., 2018), rapid commuting, and environmental friendliness (Lin et al. 2021). Unlike intercity cars, airplanes, and regular trains, which emit high levels of pollutants, HSR significantly enhances the efficiency of public transportation, reduces regional air pollution, improves the residential environment in connected areas (Yang et al. 2019; Zhao et al. 2021; Zhang et al. 2023), and increases overall happiness (Sun et al. 2020; F. Chen and Chen, 2023). Additionally, HSR enhances the locational advantages of the regions it connects. To further capitalize on the benefits of HSR and attract talent, local governments are motivated to improve public infrastructure in connected areas (Zhao et al. 2022). To some extent, this enhances the convenience of life for residents in these areas, thereby increasing their overall happiness.

Although previous studies have explored how HSR affects residents’ population mobility and convenience from the perspective of microeconomic behaviour (Liu and Zhang, 2018; Wang et al. 2019), the influence of HSR on residents’ psychological well-being and SWB is often overlooked, primarily due to limitations in individual-level data. To fill this gap, this study utilises data from the China Family Panel Studies (CFPS) to examine the impact of HSR opening on residents’ SWB. The results indicate that HSR significantly enhances residents’ SWB in connected areas. This finding remains robust across various statistical tests. Moreover, compared to other cities, the impact of HSR openings on SWB is more pronounced in central and eastern cities. We also analysed the pathways through which HSR affects SWB. Through the exploration of these pathways, we found that HSR can elevate residents’ SWB levels by improving medical, educational, environmental, and health-related conditions in connected areas.

This study contributes to the literature in two ways. First, it adds to the literature on the impact of HSR by utilising individual-level data. Previous research has predominantly examined the positive externalities of HSRs from macroeconomic and sociological perspectives. Residents’ SWB is a fundamental goal of social development. However, owing to data limitations, previous studies have conducted relatively few investigations into the impact of HSR on residents’ SWB. Second, this study explores the relationship between HSR opening and individual SWB by investigating the channels through which it affects SWB, thereby providing policy support for local governments to enhance local residents’ quality of life. Specifically, a dynamic model was constructed to reveal both the immediate and delayed impacts of HSR on residents’ SWB. Additionally, the HSR opening incentivises local governments to expand public infrastructure, such as medical facilities and educational services, while also improving local air quality. These findings can assist local governments in better understanding the relationship between infrastructure development and individual SWB.

The structure of the study is as follows: Section Two reviews the relevant literature; Section Three describes the data selection and model framework; Section Four presents the empirical analysis, examining the impact of HSR on residents’ SWB, its variation across demographics, the immediate and delayed effects, robustness checks, and the mechanisms through which HSR influences well-being in connected areas. Finally, Section Five concludes the study with a summary of the research findings.

Theoretical analysis and research hypotheses

Research background

HSR in China

China’s large population and high frequency of internal mobility have significantly accelerated the development of its HSR network, which is currently the largest in the world. The HSR system has played a pivotal role in enhancing the country’s transportation infrastructure, particularly with regard to travel speed and the breadth of urban coverage. Undeniably, China’s HSR has substantially improved the convenience and efficiency of short- and medium-distance travel.

HSR in China is classified into three categories based on operating speed: HSR with speeds ranging from 300 to 350 km/h (high-speed electric multiple unit, or EMU, trains); HSR operating at 200 to 250 km/h using both EMU trains and conventional passenger cars; and mixed-use railways that accommodate both passenger and freight transport, also at speeds of 200 to 250 km/h. Functionally, the HSR system is further categorized into two types: intercity railways and passenger-dedicated lines. Intercity railways primarily serve commuting needs and are typically located in regions characterized by high population density, developed economies, and strong population mobility—such as the Yangtze River Delta and the Guangdong-Hong Kong-Macao Greater Bay Area. Their operational speeds are comparable to those of standard HSR. Passenger-dedicated lines refer to railway lines exclusively used for passenger transport; however, since December 2009, the term “high-speed rail” has largely supplanted the concept of passenger-dedicated lines in both policy and practice.

By the end of 2023, China’s HSR network had extended to 31 provincial-level administrative regions, including the Hong Kong Special Administrative Region. Among these, only Macao and the Tibet Autonomous Region had not yet implemented HSR services operating at speeds exceeding 200 km/h. Based on the temporal scope of our study (2010–2018), we visualized the spatial distribution of cities with HSR service openings in 2010 and 2018, respectively.

As shown in Fig. 1, in 2010, China’s HSR lines were primarily concentrated in the southeastern region, the Beijing-Tianjin-Hebei area, and the Bohai Rim. Only a small number of lines were located in the central region, with most HSR services concentrated in provincial capitals. By 2018, HSR coverage had expanded significantly, extending into the southwestern and northwestern regions. The southeastern region continued to exhibit a relatively dense HSR network compared to other areas.

Fig. 1: Cities with HSR Openings in 2010 and 2018.
Fig. 1: Cities with HSR Openings in 2010 and 2018.
Full size image

Notes: (1) HSR lines that commenced operation in the first half of 2010 and 2018 are classified as lines opened in 2010 and 2018, respectively; (2) HSR lines that began operation in the second half of 2009 and 2017 are also considered as having opened in 2010 and 2018, respectively.

The impact of HSR on the residents’ SWB in China

Residents’ SWB refers to an individual’s psychological evaluation of their living conditions and circumstances. In contemporary China, the issues of greatest concern to residents include housing, healthcare, food safety, price levels, education, employment, income, social security, corruption, and environmental quality. These concerns can be broadly categorized into three dimensions: individual and family life, social life, and the quality of the living environment. Addressing these key issues has a significant and far-reaching influence on residents’ SWB.

The expansion of HSR across various cities not only enhances regional connectivity but also signifies a broader advancement in China’s transportation infrastructure. This development has exerted significant positive effects on the three aforementioned dimensions, thereby contributing to the improvement of residents’ SWB. Specifically: (1) HSR has largely replaced conventional medium- and long-distance road transport as well as low-speed rail services, substantially improving transportation efficiency and alleviating the issue of slow travel for residents; (2) The introduction of intercity railways has facilitated greater convenience and efficiency for residents living in the suburban areas of developed cities, meeting their commuting, healthcare, and educational needs; (3) Compared to traditional modes of transportation, HSR is more environmentally friendly. Its substitution effect helps reduce environmental pressure and contributes to an improved living environment; (4) Strengthened intercity connectivity has also facilitated the inflow of capital, technology, and human resources into less developed cities, potentially enhancing employment opportunities and income levels for local residents within HSR-served regions.

Research hypotheses

Regardless of whether in underdeveloped or developed cities, or in urban versus suburban areas, the direct impact of HSR openings on residents’ SWB within service areas primarily stems from improvements in travel efficiency and convenience. These improvements constitute the core factors through which residents experience enhanced SWB. Furthermore, it is precisely this increased efficiency and convenience that enables HSR to influence residents’ SWB through three key mechanisms: the medical environment, the educational environment, and the natural environment. This section will examine these three mechanisms in detail.

The function of the medical environment between HSR and the residents’ SWB

HSR contributes to the enhancement of residents’ SWB primarily through its role in improving the medical environment, by reshaping healthcare accessibility, optimizing resource allocation, and accelerating institutional reforms. First, HSR significantly improves the physical accessibility of healthcare services by compressing time–space distances between patients and providers. In contemporary China, cross-regional patient flows have become increasingly common, driven largely by the strong demand for high-quality healthcare from residents in remote and suburban areas (Bentham and Haynes, 1985; Balia et al. 2018). Compared to underdeveloped regions, major urban centers generally possess more advanced medical infrastructure, highly trained professionals, and comprehensive specialty services. The expansion of HSR networks reduces spatial barriers, allowing patients in less developed or satellite cities to conveniently access superior healthcare services in core urban areas, thereby mitigating structural inequalities in medical access and strengthening individuals’ sense of health security (Xia et al. 2022).

Second, beyond facilitating patient mobility, HSR catalyzes the downward diffusion of healthcare resources by enabling medical experts from central cities to travel regularly to grassroots and remote areas (Miwa et al. 2022; Zhao et al. 2022). This mechanism is exemplified by the cross-provincial medical alliance between Sun Yat-sen Memorial Hospital and the Third People’s Hospital of ChenzhouFootnote 1, where HSR has significantly improved the cross-regional mobility of medical personnel, allowing local doctors to engage in frequent on-site training at leading hospitals while enabling specialists to conduct routine outreach for clinical guidance and surgical instruction. This, in turn, reinforced the operational effectiveness of institutional arrangements such as medical alliances and specialty networks, and facilitated a more equitable spatial distribution of high-quality medical services. The “time–space compression” effect not only reduced the travel burden and opportunity costs for patients but also lowered the risk of delayed diagnoses and adverse health outcomes, contributing to higher levels of medical satisfaction.

More importantly, the growing demand for cross-regional medical treatment, induced by the expansion of HSR, has triggered reforms in the healthcare insurance system. Under the previous reimbursement model, patients had to prepay for medical expenses when seeking care outside their registered locality and undergo lengthy reimbursement procedures afterward, which imposed both economic and psychological burdens. In response, the Chinese government launched a direct settlement scheme in 2016 for outpatient expenses under the national social health insurance system. This reform shortened the reimbursement cycle and alleviated the financial burden of prepayment. According to the National Healthcare Security Administration, the number of cross-provincial direct settlements reached 130 million in 2023, reducing patients’ upfront payments by a total of 153.67 billion CNY (21.81 billion USDFootnote 2)—representing year-on-year increases of 238.7% and 89.9%, respectivelyFootnote 3.

Taken together, HSR not only facilitates efficient access to healthcare and improves service quality but also stimulates deeper institutional adaptations that enhance the equity and responsiveness of the healthcare system. These synergistic effects ultimately translate into tangible improvements in residents’ SWB by meeting both their physical health needs and institutional expectations (Ortuzar et al. 2021).

The function of the educational environment between HSR and the residents’ SWB

First, the development of HSR helps reduce regional disparities in educational resources, thereby contributing to the improvement of residents’ SWB. Under traditional transportation systems, high-quality educational resources tend to be heavily concentrated in metropolitan areas, while underdeveloped regions often suffer from a lack of qualified teachers and outdated pedagogical practices. The “time-space compression” effect brought by HSR significantly lowers the institutional and physical barriers to cross-regional teacher training and educational exchange. As a result, cities with relatively limited educational capacity can more easily send teachers to more advanced regions to acquire modern teaching philosophies and methods, facilitating the localized adaptation and internalization of external educational concepts.

In parallel, HSR enhances the spatial accessibility and labor market attractiveness of small and medium-sized cities, making them more appealing destinations for highly qualified teaching personnel (Henke et al. 2023). This dynamic helps mitigate long-standing issues of talent outflow and facilitates the return and redistribution of human capital in education. Moreover, the improved mobility afforded by HSR offers students and their families—particularly those in rural or remote areas—greater access to diverse educational opportunities and information, expanding their academic horizons and fostering a stronger sense of future possibilities. Collectively, improvements in both educational equity and quality enhance families’ confidence in their children’s upward mobility and reinforce individuals’ belief in the fairness of social advancement channels, thus elevating their perceived life satisfaction (Cuñado and de Gracia, 2012; Ruiu and Ruiu, 2019).

Second, HSR-induced urban spatial restructuring has also triggered renewed investment in educational infrastructure, thereby strengthening the material foundation for educational improvement. For instanceFootnote 4, following the opening of Changsha South Railway Station in late 2009, population inflows and business activities rapidly concentrated in the surrounding Yuhua District, forming a new urban growth pole. In response, local governments prioritized the development of public services in the newly emerging HSR zones, among which education infrastructure was often given precedence. In Yuhua District, the government invested in building a new junior high school to meet the rising demand for basic education from incoming populations. This example illustrates a broader pattern in which local authorities, in the context of HSR-driven urban expansion, tend to allocate educational resources preferentially through the construction and expansion of schools, the introduction of high-quality educational organizations, and the upgrading of campus facilities (Haughwout, 2002; Meng et al. 2023). Such measures have not only enhanced the region’s educational service capacity but also reduced the time and cost for households to access schooling, improving the overall accessibility and quality of education services and thereby increasing residents’ sense of educational fulfillment.

In summary, HSR improves educational environments through a dual mechanism: it facilitates the inflow of advanced pedagogical ideas and teaching talent, while also promoting infrastructure investment and public resource allocation. Together, these processes contribute to the creation of a more equitable, high-quality, and accessible educational ecosystem, which plays a vital mediating role in enhancing residents’ SWB.

The function of natural environment between HSR and the residents’ SWB

The positive impact of HSR on SWB through the lens of the natural environment operates mainly via two pathways: (1) direct environmental effects resulting from the substitution of HSR for high-pollution, high-carbon modes of transport; (2) structural improvements driven by the transformation and upgrading of the industrial structure.

The direct environmental effects of HSR include improvements in air quality and reductions in carbon emissions. Taking carbon dioxide emissions as an example, data from the International Energy Agency indicate that since 2000, the transportation sector has consistently ranked as the third-largest source of carbon emissions among all sectors, with emissions exhibiting a slow upward trend (see Fig. 2). Further, Tian et al. (2023) finds that in 2019, the majority of carbon emissions from China’s transportation sector originated from road transport, accounting for 79.15%, followed by air transport (9.13%), waterway transport (7.06%), rail transport (4.39%), and pipeline transport (0.26%) (see Fig. 3). Evidently, among various modes of transport, HSR—representing rail transport—possesses a significant low-carbon advantage. It has the potential to effectively substitute for traditional energy-intensive modes such as long-distance buses and short-haul flights, thereby substantially reducing environmental pressure (Jia et al. 2021; Lin et al. 2021). Moreover, existing literature has shown that the continued optimization of the transport structure contributes to a sustained decline in regional environmental pollution levels, particularly in terms of improved air quality (Yang et al. 2019; Jia et al. 2021; Zhao et al. 2021; Zhang et al. 2023). Improvements in air quality not only enhance public health and environmental satisfaction but also help to alleviate health-related anxiety and climate-related psychological distress induced by environmental degradation.

Fig. 2: CO2 emissions by sector in China from 2000 to 2022.
Fig. 2: CO2 emissions by sector in China from 2000 to 2022.
Full size image

Data source: International Energy Agency.

Fig. 3: Share of CO2 emissions by transport type in China in 2019.
Fig. 3: Share of CO2 emissions by transport type in China in 2019.
Full size image

Data source: Tian et al. (2023).

In terms of indirect environmental effects, HSR also facilitates industrial restructuring and upgrading (Yang et al. 2019; Jia et al. 2021). According to recent findings, the opening of HSR has led to a 3.75–4.84% change in urban industrial structureFootnote 5. From the perspective of industrial relocation, HSR-induced urban expansion typically brings about a significant revaluation of production factor costs, especially land prices, which exerts a crowding-out effect on resource- and land-intensive industries. At the same time, HSR substantially lowers the access costs of high-end production factors—such as technology, capital, and skilled labor—making HSR-connected cities, particularly those with high centrality, more attractive for the development of high value-added, low-carbon, and knowledge-intensive industries. As traditional industries lose their comparative advantage, they gradually shift to less central or non-HSR cities, thereby freeing up spatial and resource capacity in core cities for green industrial upgrading. Meanwhile, receiving cities benefit from technology spillovers, job creation, and improved industrial support systems, enabling them to enhance the technological content and environmental performance of local industries, and transition from resource-dependent development to a cleaner and smarter growth model (Shi and Wang, 2024).

Based on the above, the following hypotheses are proposed and illustrated in Fig. 4.

Fig. 4: Diagram of the theoretical mechanism.
Fig. 4: Diagram of the theoretical mechanism.
Full size image

Notes: Theoretical mechanisms are categorized into three dimensions: medical environment, educational environment, and natural environment. Improvements in the medical environment are reflected by increases in the number of doctors and hospital beds; improvements in the educational environment are indicated by a higher proportion of government investment in education; and enhancements in the natural environment are measured by reductions in air pollution.

Hypothesis 1: The opening of HSR improves the residents’ SWB within its service city.

Hypothesis 2: The opening of HSR improves residents’ SWB primarily through enhancements in the medical environment, educational environment and natural environment.

Methodology

Data. The data on the opening dates of HSR lines are obtained from publicly available sources released by the China Railway Corporation. For ease of analysis, we manually compiled this information. Specifically, HSR lines that commenced operation in the first half of a given year are classified as lines opened in that year, whereas those that began operation in the second half are considered as lines opened in the following year.

In addition, data on SWB are drawn from the China Family Panel Studies (CFPS), a nationally representative longitudinal survey launched in 2010 by the China Social Science Survey Center at Peking University. The baseline survey covered 25 provincial-level regions (including provinces, autonomous regions, and municipalities) and sampled approximately 15,000 households, including all members within each household. Since then, the CFPS has conducted biennial follow-up surveys to track respondents over time. To ensure the confidentiality of respondents’ information, the CFPS has implemented strict data protection protocols. During the survey process, access to sensitive information is granted only to authorized personnel and strictly on a need-to-know basis. In the publicly shared dataset, personally identifiable information – such as names, addresses, and employers – is removed, and location codes below the provincial level are re-coded to enhance privacy protection. In this study, we use data from five CFPS waves – 2010, 2012, 2014, 2016, and 2018 – to construct an unbalanced panel dataset at the city level. Due to the inclusion of sensitive personal and household information, the dataset is subject to strict confidentiality protocols and is not publicly available.

Variable

Dependent variable

SWB of residents. We measure residents’ SWB using the CFPS survey question: “How satisfied are you with your life?” Responses range from 1 (very dissatisfied) to 5 (very satisfied).

Independent variable

The key explanatory variable is whether HSR service is available. If HSR service is operational in city i in year t, the variable takes a value of 1; otherwise, it is assigned a value of 0.

Control variable

Following previous studies, we include a set of individual-level control variables: gender, age, domicile (urban or rural), marital status, education level, self-rated health, employment status, income, insurance coverage, and perceived social status.

Mechanism variables

Based on the theoretical framework discussed above, the potential mechanisms through which HSR influences SWB include the medical environment, educational environment, and natural environment. In this study, we use the number of doctors and hospital beds to proxy the medical environment, local public expenditure on education to capture the educational environment, and the annual average PM2.5 concentration to represent the natural environment. Detailed definitions are provided in Table 1.

Table 1 Variable Description.
Full size table

Descriptive statistics

Descriptive statistics for the relevant variables are presented in Table 2. The mean value of residents’ SWB is 3.688, suggesting that the majority of respondents report a relatively high level of life satisfaction. The average value for HSR operation is 0.415, indicating that 41.5% of the sampled individuals lived in areas with HSR service during the sample period. The mean value for gender is 0.47, which is close to 0.5, suggesting a relatively balanced distribution of male and female respondents. The average age of respondents is 50.446 years, and 44.3% are urban residents. The mean value for marital status is 2.148. The average years of education attained is 6.435. The mean self-rated health score is 2.933. In terms of employment status, 67.8% of respondents are employed. The mean value for self-rated income is 2.491, and 91.4% of respondents reported having insurance coverage. Lastly, the average self-rated social status is 2.931.

Table 2 Descriptive Statistics.
Full size table

Model design

Baseline regression

To empirically evaluate the impact of HSR introduction on residents’ SWB, we employ a baseline regression model as specified in Eq. (1).

$${satisficatio}{n}_{{itk}}={a}_{0}+{a}_{1}{{HSR}}_{{itk}}+\beta {contro}{l}_{{itk}}+{\gamma }_{t}+{\lambda }_{i}+{\rho }_{k}+{\varepsilon }_{{itk}}$$
(1)

Where \({\boldsymbol{satisficatio}}{{\boldsymbol{n}}}_{{\boldsymbol{itk}}}\) denotes the SWB of individual i in city k in year t. The binary variable\(\,{{\boldsymbol{HSR}}}_{{\boldsymbol{itk}}}\) represents the operational status of HSR; It takes a value of 1 if HSR is operational in city k where individual i resides in year t, and 0 otherwise. \({{\boldsymbol{control}}}_{{\boldsymbol{itk}}}\) includes a set of control variables that capture various factors potentially influencing residents’ SWB. \({{\boldsymbol{\gamma }}}_{{\boldsymbol{t}}}\) denotes time fixed effects, while \({{\boldsymbol{\lambda }}}_{{\boldsymbol{i}}}\) and \({{\boldsymbol{\rho }}}_{{\boldsymbol{k}}}\) represent individual and city fixed effects, respectively. The term \({{\boldsymbol{\varepsilon }}}_{{\boldsymbol{itk}}}\) captures the random error.

Mechanism tests

Building on the theoretical analysis, we posit that the impact of HSR on residents’ SWB operates through improvements in the medical environment, educational environment, and natural environment within the cities it serves. To empirically identify these mechanisms, we employ a three-way fixed effects model at the prefecture-level city scale, as specified in Eq. (2).

$${mechanis}{m}_{{itk}}={a}_{0}+{a}_{1}{{HSR}}_{{itk}}+\beta {contro}{l}_{{itk}}+{\gamma }_{t}+{\lambda }_{i}+{\rho }_{k}+{\varepsilon }_{{itk}}$$
(2)

Where \({\boldsymbol{mechanis}}{{\boldsymbol{m}}}_{{\boldsymbol{itk}}}\) denotes the dependent variable, representing one of the three potential channels: the medical environment, educational environment, or natural environment. \({{\boldsymbol{HSR}}}_{{\boldsymbol{itk}}}\) is the key independent variable, indicating the operational status of HSR in the city where individual i resides. \({\boldsymbol{contro}}{{\boldsymbol{l}}}_{{\boldsymbol{itk}}}\) refers to a set of control variables that may influence the respective mechanisms. \({{\boldsymbol{\gamma }}}_{{\boldsymbol{t}}}\) captures time fixed effects, while \({{\boldsymbol{\lambda }}}_{{\boldsymbol{i}}}\) and \({{\boldsymbol{\rho }}}_{{\boldsymbol{k}}}\) represent individual and city fixed effects, respectively. \({{\boldsymbol{\varepsilon }}}_{{\boldsymbol{itk}}}\) denotes the random error term.

Dynamic and lagged effects

To further examine the dynamic and lagged effects of HSR openings, we modify Eq. (1) into Eq. (3), as presented below.

$${satisficatio}{n}_{{itk}}={a}_{0}+\mathop{\sum}\limits_{n=1}^{5}{\alpha }_{n}{{HSR}}_{{itk}}^{n}+\beta {contro}{l}_{{itk}}+{\gamma }_{t}+{\lambda }_{i}+{\rho }_{k}+{\varepsilon }_{{itk}}$$
(3)

Where \({{\boldsymbol{HSR}}}_{{\boldsymbol{itk}}}^{{\boldsymbol{n}}}\) denotes a set of five year-specific dummy variables. Specifically, \({{\boldsymbol{HSR}}}_{{\boldsymbol{itk}}}^{{\boldsymbol{n}}}\) takes the value of 1 in a given year (e.g., 2010) if HSR service was operational in that year in city k, where individual i resides, and 0 otherwise. The same coding rule is applied to the remaining survey years. The coefficient \({{\boldsymbol{\alpha }}}_{{\boldsymbol{n}}}\) captures the effect of HSR openings on residents’ SWB in each respective survey year (2010, 2012, 2014, 2016, and 2018). \({\boldsymbol{contro}}{{\boldsymbol{l}}}_{{\boldsymbol{itk}}}\) represents a vector of control variables. \({{\boldsymbol{\gamma }}}_{{\boldsymbol{t}}}\) denotes time fixed effects, while \({{\boldsymbol{\lambda }}}_{{\boldsymbol{i}}}\) and \({{\boldsymbol{\rho }}}_{{\boldsymbol{k}}}\) represent individual and city fixed effects, respectively. The term \({{\boldsymbol{\varepsilon }}}_{{\boldsymbol{itk}}}\) captures the random error.

Empirical estimation results

This section investigates the impact of HSR on residents’ SWB in cities where it has been implemented. First, a baseline regression model is estimated. Second, regression analyses are conducted to test the three proposed potential mechanisms. Third, the dynamic effects of HSR on residents’ SWB are examined. Fourth, a series of robustness checks are performed to validate the baseline results. Finally, sub-sample regression analyses are carried out to explore the heterogeneity of HSR’s effects on residents’ SWB.

The impact of HSR opening on residents’ SWB

As shown in Table 3 (with the key coefficients visualized in Fig. 5), the estimated coefficient of HSR on residents’ SWB in cities served by HSR is 0.209 and statistically significant at the 1% level, even in the absence of control variables. This suggests that the introduction of HSR has a substantial positive effect on residents’ SWB. When individual-level control variables are included, the coefficient remains significantly positive at the 1% level, although its magnitude decreases to 0.142. Furthermore, after incorporating control variables and systematically accounting for year, city, and individual fixed effects, while clustering standard errors at the individual level, the results consistently indicate that HSR has a significantly positive impact on residents’ SWB at the 1% level. These findings support Hypothesis 1, confirming that the opening of HSR significantly enhances residents’ SWB. This result is consistent with previous studies (Chen and Chen, 2023; Deng et al. 2024).

Table 3 Baseline Results.
Full size table
Fig. 5: Key coefficients of baseline results.
Fig. 5: Key coefficients of baseline results.
Full size image

Notes: The vertical axis columns (1) – (5) correspond to the regression coefficients in columns (1) – (5) of the first row in Table 3.

Potential factors of the impact of HSR opening on residents’ SWB. For medical environment, as outlined in the earlier hypothesis section, the improvement in the medical environment brought about by the introduction of HSR primarily stems from the enhanced mobility of medical resources – such as cross-regional patient flows and the redistribution of high-quality healthcare services. In large urban hospitals, meeting the growing demand from non-local patients without crowding out local residents requires a substantial expansion of medical resources – particularly an increase in the number of physicians and hospital beds. Meanwhile, primary and local hospitals often experience significant patient outflows, leading to deficits in local medical insurance funds. To mitigate these issues and reduce cross-provincial patient mobility, it is essential to attract qualified medical professionals and expand the number of hospital beds. For instance, in Henan Province, the construction of 12 national medical centers has added approximately 11,500 new beds and introduced 712 high-level medical personnel, which has significantly reduced the proportion of patients seeking treatment outside the province. Following this mechanism, we use the number of physicians and hospital beds as proxy indicators for the quality of the medical environment.

Educational environment

The opening of HSR not only enhances regional accessibility but also intensifies intercity competition for public resources. In response, local governments often increase investment in educational infrastructure as a strategic effort to improve the educational environment and attract high-quality human capital. Measures such as expanding primary and secondary schools or introducing reputable educational groups require substantial fiscal support. Therefore, government expenditure on education serves as a meaningful reflection of efforts to enhance the local educational environment. Based on this rationale, we use government education investment as a proxy indicator for the education mechanism to assess whether the launch of HSR has facilitated the reallocation and improvement of educational resources.

Natural environment

In the research hypothesis section above, we have pointed out that the most direct environmental impact of HSR lies in its substitution effect on high-pollution and high-carbon travel modes. According to statistics from the World Health Organization, in 2019, approximately 4.2 million people worldwide died prematurely due to exposure to fine particulate matter (PM2.5), which can cause cardiovascular diseases, respiratory illnesses, and cancer. In China, PM2.5-induced haze poses challenges to residents’ mobility and exerts a detrimental impact on their mental well-being. As such, this paper selects PM2.5—the air pollutant with the most significant health impact—as a proxy indicator for measuring air pollution levels. Given that different PM2.5 concentrations correspond to different levels of the Air Quality Index (AQI)—namely excellent, good, mildly polluted, moderately polluted, heavily polluted, and severely polluted—and considering that the maximum PM2.5 concentration in our sample is 150 μg/m³, we classify PM2.5 into four levels: excellent, good, mildly polluted, and moderately polluted. These are assigned values from 1 to 4, respectively, and used as a proxy variable for environmental improvement.

Columns (1) and (2) of Table 4 report coefficients that are significantly positive at the 1% level, suggesting that the introduction of HSR has increased the number of doctors and hospital beds, thereby improving the healthcare environment in the cities served (key coefficients of the mechanisms are visualized in Fig. 6). Column (3) shows that HSR has incentivized local governments to increase education spending, thus enhancing the educational environment. Column (4) reveals a significant negative association between HSR operation and city-level PM2.5 concentrations, indicating that the introduction of HSR has contributed to improved air quality. Taken together, these findings underscore the importance of improvements in the medical environment, educational environment, and natural environment as key mechanisms through which HSR influences residents’ SWB in the cities it serves. This result supports Hypothesis 2.

Table 4 Mechanisms.
Full size table
Fig. 6: Key coefficients of mechanisms.
Fig. 6: Key coefficients of mechanisms.
Full size image

Notes: The vertical axis columns (1) – (4) correspond to the regression coefficients in columns (1) – (4) of the first row in Table 4.

Dynamic and lagged effects of HSR opening on residents’ SWB

An analysis of Fig. 7 reveals a rising trend in the coefficients of HSR_2010 through HSR_2018, indicating an increasingly positive influence of HSR on residents’ SWB in the cities it serves. This pattern reflects the dynamic nature of HSR’s impact over time. Moreover, while the coefficients for HSR_2010, HSR_2012, and HSR_2014 are not statistically significant at the 5% level, those for HSR_2016 and HSR_2018 are significant. This suggests the presence of a lagged effect, whereby the positive impact of HSR on residents’ SWB becomes more pronounced over time.

Fig. 7: Results on the dynamic and lagged effects.
Fig. 7: Results on the dynamic and lagged effects.
Full size image

Notes: This figure presents the regression results from Eq. (3). The vertical axis indicates the year-specific effects, with HSR_2010 to HSR_2018 corresponding to the years 2010 to 2018.

Robustness tests

To ensure the reliability of the above regression results, we conducted a series of robustness checks, including the following four approaches: (1) testing the parallel trends assumption; (2) testing for policy exogeneity; (3) applying the PSM-DID method; and (4) incorporating local economic characteristic variables.

First, we examine the parallel trends assumption, a necessary condition for the validity of the DID model. This assumption requires that, prior to policy implementation, the treatment and control groups exhibit similar trends, while significant differences emerge only after the policy is enacted. Specifically, we select 2014 – a year with a large number of HSR openings – as the intervention point, and examine the trend differences between the treatment and control groups during the periods 2010–2014 and 2014–2018. As shown in Fig. 8, the trends in SWB were similar between the two groups before the introduction of HSR, but diverged significantly afterward. This supports the validity of the DID design for further empirical analysis.

Fig. 8: Parallel trends test results.
Fig. 8: Parallel trends test results.
Full size image

Notes: The vertical axis represents respondents’ self-rated levels of SWB, and the horizontal axis indicates the year of the interview. The green line shows the well-being levels of respondents in cities with HSR service, while the brown dashed line indicates the well-being levels in cities without HSR service.

Moreover, to further validate the parallel trends assumption, we employ an event study approach. Specifically, we designate 2014 as the treatment (event) year and 2012 as the reference year. Interaction terms between the year-specific dummy variables and the treatment group indicator are incorporated into the baseline regression model. The results, presented in Fig. 9, indicate that prior to 2014, the effect of HSR on residents’ SWB was statistically insignificant. However, after 2014, HSR significantly enhanced residents’ SWB, with the estimated coefficients falling within the 95% confidence interval.

Fig. 9: Event study results.
Fig. 9: Event study results.
Full size image

Notes: Although the policy was implemented in multiple phases, we select 2014 – the year of large-scale HSR expansion – as the policy shock year, using 2012 as the baseline year for the parallel trends test.

Second, we test for policy exogeneity. During the DID estimation process, treatment and control groups may be affected by unobserved or random policy shocks, which could bias the estimation results. To assess whether the treatment assignment is confounded by such random interventions and to enhance the robustness and credibility of the findings, this study adopts a counterfactual strategy. Specifically, we hypothetically advance the HSR opening by one and two years, respectively, and re-estimate the baseline model (Eq. 1). If the estimated coefficients of the HSR variable remain statistically insignificant under these placebo tests, it suggests that the observed effects are unlikely to be driven by random policy shocks. As shown in columns (1) and (2) of Table 5 (first and second lines of Fig. 10), the coefficients for the placebo variables (HSR_time1 and HSR_time2) are not statistically significant. This indicates that the observed improvement in residents’ SWB is unlikely to result from random policy interference, thus reinforcing the validity of the causal relationship.

Table 5 Robustness test results.
Full size table
Fig. 10: Key coefficients of robustness tests.
Fig. 10: Key coefficients of robustness tests.
Full size image

Notes: The vertical axis column (1) corresponds to the regression coefficient in column (1) of the first row in Table 5; column (2) corresponds to column (2) of the second row in Table 5; columns (3) and (4) correspond to columns (3) and (4) of the third row in Table 5, respectively.

Third, we employ the PSM-DID approach to address potential selection bias. Given the heterogeneity in economic characteristics and development levels across regions and city tiers, sample selection bias may arise in DID estimations, potentially resulting in endogeneity and misestimation of effects. To mitigate this concern, we adopt a Propensity Score Matching Difference-in-Differences (PSM-DID) method using nearest-neighbor matching. This procedure eliminates control group samples whose economic profiles significantly differ from those of the treated group. The results, presented in column (3) of Table 5 (third line of Fig. 10), show that HSR inauguration continues to significantly improve residents’ SWB, providing additional support for Hypothesis 1.

Fourth, we incorporate local economic characteristics into the regression model. Previous studies have highlighted that regional economic features—such as industrial structure and development level—may significantly influence residents’ SWB. Omitting these variables may lead to biased estimates. To improve the robustness and accuracy of the analysis, we augment the baseline model with additional control variables, including the natural logarithm of per capita GDP, the share of secondary and tertiary industries in GDP, GDP growth rate, the natural logarithm of foreign direct investment, and population growth rate. The results, shown in column (4) of Table 5 (fourth line of Fig. 10), confirm that HSR introduction significantly enhances residents’ SWB in the cities it serves.

Heterogeneity analysis

The impact of HSR services on residents’ SWB varies across cities of different tiers, reflecting substantial differences in their functional roles, administrative responsibilities, and stages of economic development. Additionally, regional disparities in the effect of HSR are evident, owing to significant variations in economic advancement and the distribution of cultural resources. To explore these heterogeneous effects, we conduct sub-sample regression analyses by city tier and geographic region. The corresponding empirical results are presented in Fig. 11.

Fig. 11: Heterogeneity analysis results.
Fig. 11: Heterogeneity analysis results.
Full size image

Notes: The top-left and bottom-left panels display the heterogeneity analysis by city tiers, while the top-right and bottom-right panels present the results by region. All four panels show 95% confidence intervals.

City heterogeneity

From the perspective of public service provision, central cities typically possess more comprehensive urban functions, including well-developed healthcare, education, and cultural facilities. These advantages enable them to more effectively absorb the knowledge spillovers and human capital inflows facilitated by HSR, thereby improving the accessibility and quality of local public services. As a result, residents in central cities are more likely to experience tangible improvements in their quality of life and SWB. In contrast, although non-central cities may benefit from enhanced connectivity and certain resource inflows due to HSR, their relatively underdeveloped infrastructure and weaker institutional capacity hinder the efficient conversion of these resources into public service improvements. Consequently, the well-being effects of HSR tend to be more pronounced in central cities than in their non-central counterparts.

Moreover, in terms of industrial green transformation, the launch of HSR services tends to crowd out resource- and land-intensive industries while simultaneously reducing the access costs for technology, capital, and skilled labor. These changes create favorable conditions for central cities to develop high-value-added, low-emission industries that are technology- and knowledge-intensive. Such transitions not only contribute to improved environmental quality but also enhance residents’ SWB by fostering a cleaner, more innovative urban economy. In contrast, while HSR improves transportation accessibility in non-central cities, it does not necessarily lead to the agglomeration of emerging industries. On the contrary, the strong siphon effect of central cities may result in a net outflow of resources and talent from less developed areas, thereby weakening their development potential and exacerbating regional disparities in well-being outcomes.

Therefore, in categorizing city tiers, we classify municipalitiesFootnote 6, provincial capitalsFootnote 7 and sub-provincial citiesFootnote 8 as central cities, whereas all other prefecture-level cities are designated as non-central cities (for details on city classification, see Fig. 12). Subsequent sub-sample regressions are conducted based on this classification. The empirical results for central and non-central cities are presented in Fig. 11. Prior to the inclusion of regional economic control variables, both central and non-central cities exhibit significantly positive effects at the 95% confidence level, although the effect size in non-central cities is comparatively smaller. After incorporating regional economic control variables, the effect remains significantly positive for central cities but becomes statistically insignificant for non-central cities. These findings suggest that the well-being effect of HSR is more pronounced in central cities.

Fig. 12: The city classification.
Fig. 12: The city classification.
Full size image

Notes: To improve visual clarity, only central cities are shown on the map due to the large number of non-central cities.

Regional heterogeneity

Regional heterogeneity in the well-being effects of HSR can largely be attributed to differences in both natural geography and socioeconomic conditions. From a geographical perspective, the eastern region of China is predominantly composed of plains and low hills, which are conducive to the construction and expansion of HSR infrastructure. In contrast, the central and western regions are characterized by plateaus, deserts, and mountainous terrain, which pose significant challenges to large-scale transportation development. Consequently, the eastern region enjoys the most extensive and well-connected HSR network in the country. Moreover, the dense urban distribution in the east, compared with the more dispersed cities in the central and western regions, further reinforces the concentration of HSR lines in the east. This results in higher transportation accessibility and greater convenience for residents, allowing them to benefit more fully from the social welfare improvements brought about by HSR, including reduced travel time, enhanced public service access, and improved life satisfaction.

From a socioeconomic perspective, the eastern region is home to many of China’s most economically advanced cities, including key hubs within the Beijing-Tianjin-Hebei, Yangtze River Delta, and Pearl River Delta metropolitan clusters. These areas boast robust economic foundations, comprehensive industrial systems, and well-developed public service infrastructures. In contrast, the central and western regions generally lag behind in economic development and the provision of public goods. In the eastern region, HSR not only serves as a transportation mode but also facilitates the efficient flow of labor, capital, and knowledge between cities. For example, in the Yangtze River Delta, residents of smaller cities near Shanghai can use HSR for daily commutes, thereby accessing superior employment opportunities and high-quality healthcare and education services offered by mega-cities, while maintaining residence in more affordable locations. This ability to spatially optimize work and living arrangements significantly enhances individuals’ quality of life and contributes to higher levels of SWB in the eastern region.

Therefore, given the notable disparities in regional development across China, and in line with classifications used in previous studies (Chen and Haynes, 2017), we divide China into the Eastern and Central-Western regions (for detailed classification, see Fig. 13). Prior to the inclusion of urban economic control variables, the well-being effect in the Eastern region is significantly positive at the 95% confidence level, whereas the effect in the Central-Western regions is statistically insignificant. This pattern remains unchanged after the inclusion of regional economic control variables. These findings indicate that the well-being effect of HSR operation is significantly stronger in the Eastern region than in the Central-Western regions.

Fig. 13: The region classification.
Fig. 13: The region classification.
Full size image

Notes: The dark green areas represent provinces in eastern China, while the light green areas represent provinces in central and western China. White areas are not included in the regional classification used in this study.

Conclusion and Implications

This study primarily utilizes micro-level data from the 2010–2018 China Family Panel Studies and employs a DID approach to examine the impact of HSR introduction on residents’ SWB from a micro-level perspective. The empirical results demonstrate that the introduction of HSR significantly improves residents’ SWB in the cities it serves. To ensure the robustness of these findings, a series of checks – including tests of the parallel trends assumption, policy endogeneity evaluation, and the application of the PSM-DID method – are conducted, thereby enhancing the study’s scientific rigor and credibility. Furthermore, the mechanism analysis reveals that HSR contributes to improvements in healthcare, education, and environmental conditions in the affected areas. The study also explores heterogeneity across city tiers and regions. Specifically, HSR exerts a more pronounced positive effect in central cities than in non-central cities, and its impact on residents’ SWB is more significant in eastern regions compared to central and western regions.

Therefore, we draw the following three implications:

First, we find that the opening of HSR significantly enhances residents’ SWB, indicating that modern transportation infrastructure not only yields substantial economic returns but also generates meaningful social welfare effects. Accordingly, policymakers should regard HSR development as a strategic instrument for improving individual well-being and promoting holistic societal progress. Future transportation planning should continue to optimize the spatial layout of the HSR network, with a particular focus on strengthening connectivity across regions. In particular, expanding HSR coverage in central, western, and peripheral areas may enhance accessibility and unlock the potential of HSR to foster regional integration, improve public service delivery, and ultimately raise the overall quality of life.

Second, Further analysis reveals that HSR enhances SWB indirectly by improving the medical, educational, and natural environments, underscoring its multifaceted externalities as public infrastructure. Therefore, alongside continued investment in HSR, governments are advised to treat HSR node cities as key zones for functional upgrading. Targeted policies should be implemented to facilitate the rational allocation and efficient flow of high-quality medical and educational resources, thereby enhancing residents’ access to and satisfaction with local public services. Moreover, by leveraging the centrality of HSR stations, regional industrial integration should be promoted, fostering the formation of rail-linked urban clusters. These clusters can drive green industrial transformation, improve environmental quality, and provide ecological foundations for long-term improvements in SWB.

Third, we also find that the well-being-enhancing effects of HSR are more pronounced in core cities and eastern regions, suggesting that local economic development levels, urban hierarchy, and public service capacity play important moderating roles in the effectiveness of HSR. This highlights the need for regionally differentiated planning in future HSR expansion. Specifically, for non-core cities in central and western China, efforts should focus on strengthening supporting infrastructure that connects them to the broader HSR network, such as intercity express links and multimodal transport hubs. Such initiatives can amplify the spillover benefits of HSR, prevent marginalization in regional development dynamics, and contribute to more inclusive and balanced regional growth.

Future research directions

We investigate the impact of improvements in transportation infrastructure—specifically, the introduction of HSR—on residents’ SWB. While the findings provide valuable insights, there remains substantial scope for further exploration. Future research may consider the following directions:

First, beyond transportation infrastructure, improvements in other types of infrastructure may also generate significant welfare effects. For instance, advances in information infrastructure—particularly those enabling telemedicine and online education via digital platforms—and energy infrastructure, such as clean energy generation projects and centralized biogas supply systems in rural areas, may substantially enhance public service accessibility and environmental quality. Future studies could examine how these forms of infrastructure contribute to social welfare and individual well-being through various channels.

Second, more attention should be given to the heterogeneous effects of HSR on the SWB of different population groups. In the context of China’s rapidly aging population, elderly individuals are becoming increasingly sensitive to access to healthcare and environmental quality. Future research could focus on whether and how HSR meets the needs of older adults, thereby improving their quality of life and contributing to more inclusive infrastructure policy design.

Third, although HSR appears to have a positive impact on the natural environment by substituting high-pollution transportation modes and promoting industrial transformation, its construction phase may have adverse ecological consequences. These include the occupation of farmland, destruction of vegetation, and disruption of biodiversity. Future research could employ cost–benefit analysis to assess the complex environmental effects of HSR, providing a more balanced evaluation of its ecological trade-offs.

Fourth, SWB is influenced by a broad set of multidimensional factors. Future work could explore the deeper mechanisms through which HSR affects residents’ daily lives, such as changes in lifestyle rhythms, time allocation, and psychological expectations. Such inquiry would contribute to a more comprehensive and systematic understanding of the relationship between infrastructure development and social welfare.