Impact of coal-to-gas transition on individual welfare considering the inverted U-shaped regulation effect of natural gas pipeline accessibility

Abstract

Amid the accelerated pace of industrialization and urbanization in China, severe air pollution has emerged as a pressing concern, posing substantial threats to public health and the overall quality of life. Given its pivotal role in the transition towards clean energy, it is of utmost importance to conduct a comprehensive and in-depth assessment of the micro-level impacts of the coal-to-gas transition (CTGT) on individual welfare (IW). This study employs data from the Chinese General Social Survey (CGSS) to scrutinize the effects of the CTGT on IW, and has uncovered three pivotal findings. Firstly, the CTGT significantly enhances IW through a combination of income substitution effect (ISE), capital aggregation effect (CAE), industrial upgrading effect (IUE), and environmental improvement effect(EIE). As a consequence, residents have witnessed significant improvements in their life happiness (LH) and physical health (PH) following the transition. Secondly, natural gas pipelines accessibility (NGPA) has been found to exert an inverted U-shaped regulation influence on the relationship between the CTGT and IW. Thirdly, the welfare-enhancing effects of the CTGT are particularly pronounced among females with lower consumption levels, with a junior high school education or below, and residents in urban areas boasting superior energy supply and infrastructure standards. The findings offer crucial theoretical foundations and policy recommendations for optimizing energy structures and enhancing individual welfare, providing practical guidance for advancing global clean energy transitions and achieving sustainable development goals.

Introduction

China’s rapid economic expansion has posed significant challenges in combating climate change and air pollution. As the world’s largest carbon emitter, China bears a substantial responsibility for attaining carbon neutrality, with a target date set for 2060. However, China’s energy consumption profile diverges notably from those of other economies, exhibiting a heavy reliance on fossil fuels, especially coal. Coal, an energy source directly extracted through industrial mining and processing, is widely employed in sectors such as steel, thermal power generation, and textiles, offering crucial industrial support and facilitating technological advancements in chemical production. Nevertheless, as environmental consciousness grows and expectations for people’s welfare and quality of life rise, the issues linked to coal combustion have become increasingly salient. The imperative for transformation and the attendant risks of substitution are intensifying, primarily due to the irreversible damage inflicted by the negative externalities of coal combustion on public well-being. In the Beijing–Tianjin–Hebei region of northern China, residential coal combustion accounts for up to 23.1% and 42.6% of PM_2.5 and SO₂ emissions, respectively¹. Elements like sulfur, arsenic, and fluorine not only aggravate air pollution² but also precipitate a range of respiratory ailments, including chronic obstructive pulmonary disease, asthma, and lung cancer, posing a grave threat to people’s lives, health, and overall quality of life³. China is confronted with an urgent imperative to transform its energy structure. By identifying clean energy alternatives to coal and devising a scientifically sound and rational energy transition strategy, it is anticipated that China can expedite the transition of its energy mix towards a cleaner, lower-carbon, and more efficient paradigm. This will foster a healthier and more habitable living environment for its citizens while ensuring sustainable economic growth.

Based on existing research, the literature pertinent to this study can be broadly classified into two categories. The first category delves into the factors influencing welfare. Despite being a topic of enduring and vigorous debate, perceptions of welfare have evolved over different historical epochs. Within the realm of economics, systematic inquiry into well-being can be traced back to Easterlin’s seminal work. From the vantage point of objective income, he elucidated the intricate relationship between material wealth and subjective well-being, demonstrating that the link between income and well-being is not straightforwardly linear. While income constitutes a significant determinant of well-being, it is not the exclusive one⁴. As global citizens’ aspirations for a sense of achievement, security, and happiness in life escalate, a plethora of studies have underscored that non-economic factors, such as the optimization of social policies⁵ and the attention to psychological states⁶, exert a more profound influence on augmenting residents’ happiness and ameliorating their physical health compared to the mere augmentation of wealth and income levels⁷. A quasi-natural experiment conducted by Guo et al. ⁸ in China’s industrial transformation zones revealed that industrial upgrading and transformation are instrumental in fostering green development and environmental enhancement. Moreover, the amelioration of residents’ living environments and the advancement of public services can further fortify physical health⁹. Additionally, scholars have progressively uncovered the impact of public participation patterns and community attachment on happiness¹⁰. When objective conditions, such as income levels, educational attainment, and leisure frequency, are fulfilled or stabilized, individuals’ quest for happiness progressively transitions from the material to the spiritual sphere, manifesting as an emphasis on dynamic subjective experiences. For instance, by dwelling in a fixed community environment for an extended duration, individuals can deepen their emotional connections through collective memories forged in public spaces, thereby spurring participation in community public activities to augment their sense of happiness and accomplishment¹¹.

The second category of literature delves into the welfare effects of energy transition. Scholars have predominantly investigated the nexus between the coal-to-gas policy and welfare, approaching it from the dual perspectives of air quality and environmental governance. Their focus has been on both the environmental benefits and health benefits^12,13,14 derived from the implementation of this policy. From an environmental benefits perspective, the strategic selection of low-carbon energy sources plays a pivotal role in fostering the development of a green economy within society¹⁵. For example, Lueken et al¹⁶ conducted a comprehensive study on the substitution of coal with natural gas in the United States in 2016. Their findings revealed that this substitution led to a reduction of over 90% in sulfur dioxide emissions and more than 60% in nitrogen oxide emissions. Similarly, Yu et al.¹⁷ examined the emission reduction impacts of China’s coal-to-gas policy on urban air pollution. They observed significant declines in atmospheric concentrations of PM_2.5, SO₂, and NOx following the policy’s implementation. Furthermore, the emission reduction effects were more marked in western regions than in eastern regions¹⁸. Regarding health benefits, the adoption of cleaner energy sources by households enhances the health protection standards for residents. This is most conspicuously evidenced by a substantial decrease in expenditures on medical treatments and healthcare services¹⁶. Additionally, scholars have progressively unveiled the potential role of natural gas pipeline infrastructure in promoting sustainable development during energy transitions^19,20. Yang et al. explored the feasibility of the coal-to-gas policy and its implementation by assessing the straight-line distance from coal-fired plants to nearby pipelines. However, their study indicated that only 43% of global coal-fired power generation is unencumbered by pipeline distance constraints. This suggests that the feasibility of large-scale CTGTs necessitates further investigation²¹.

However, existing literature may suffer from the following three deficiencies: First, the majority of the aforementioned studies adopt a macro perspective when examining the implementation of the coal-to-gas policy, neglecting to analyze the welfare effects and the underlying mechanisms of the CTGT from the residents’ micro perspective. In reality, as direct participants in this transition, residents’ micro behaviors, including their decision-making attitudes, economic capabilities, and changes in quality of life, are pivotal in determining whether the transition yields welfare benefits. Second, no study to date has thoroughly investigated the disparities in the welfare effects of CTGT across different levels of NGPA. In practice, NGPA directly impacts the feasibility, cost, and convenience of residents’ transition to gas, thereby serving as a critical determinant of the welfare effects associated with this transition. Third, the energy transition from coal to gas in China commenced relatively recently. Although existing research has demonstrated that the post-transition clean energy contributes positively to community environmental enhancement and residents’ physical health, it has failed to adequately consider the differences among demographic groups and between urban and rural facilities. This oversight may result in an incomplete understanding of the actual challenges, potential opportunities, and issues encountered during the CTGT. The questions that this paper endeavors to address are as follows: Does the CTGT, from the residents’ perspective, genuinely enhance welfare? Through which pathways does it achieve this enhancement? Does the NGPA play a pivotal role? Are there variations in welfare effects across different population group and urban and rural facilities?

The potential marginal contributions of this study to the existing body of literature are outlined below: Firstly, concerning research content, this paper offers a systematic analysis of the impact of the CTGT on residents’ welfare at the microeconomic level. It delves into the potential mechanisms through which this transition may exert its influence. This analysis not only provides fresh theoretical insights for government research on public environmental governance but also enriches the research corpus in the interdisciplinary domain of public governance and energy transition, thereby pushing the boundaries of public governance theory into new territories of application. Secondly, from a research standpoint, this study integrates the accessibility of natural gas pipelines into the analytical framework that examines how the CTGT contributes to an enhancement in residents’ welfare. It uncovers an inverted U-shaped moderating effect exerted by the accessibility of natural gas pipelines, a finding that adds nuance and depth to our understanding of the transition’s welfare implications. Thirdly, at the policy level, this research advocates for differentiated pathways for the CTGT, tailored to specific demographic groups, urban–rural infrastructure disparities, and the varying levels of natural gas pipeline accessibility. These pathways serve as actionable solutions for developing countries in dire need of energy structure transformation, facilitating their pursuit of sustainable economic development while paving the way for a more prosperous future.

Theoretical framework and research hypothesis

Coal-to-gas transition and individual welfare

Improving IW and enhancing their living standards are crucial objectives pertaining to people’s livelihoods. The CTGT, as a practical and viable energy program, is steadily and imperceptibly bringing about changes in residents’ daily lives.

• (1) The CTGT may enhance IW by generating ISE From the production standpoint, traditional coal combustion methods are frequently characterized by inefficiencies and significant pollution. Enterprises are compelled to allocate substantial resources to pollution control, which, to a certain extent, constrains their profit margins and may subsequently impact employee wages. However, following the CTGT, the high combustion efficiency and low pollution levels inherent in natural gas usage improve the production environment and streamline production processes for enterprises. This, in turn, facilitates a reduction in production costs and an enhancement in production efficiency²². With augmented profits, enterprises possess greater financial capacity to increase employee salaries or expand production to generate more employment opportunities, thereby directly elevating the disposable incomes of various demographic groups, particularly those engaged in relevant industries²³. From the consumption perspective, coal combustion pollution poses a threat to residents’ health, precipitating respiratory diseases and escalating medical expenditures. Upon transitioning to gas, air quality undergoes improvement, mitigating residents’ health risks and reducing medical expenses. This effectively translates into an indirect increase in disposable income²⁴. With regard to the urban–rural disparity, urban areas generally initiate the CTGT earlier and implement it more comprehensively, boasting well-developed infrastructure and expediting enterprise transformation. This enables them to promptly capitalize on the transition’s benefits, such as industrial upgrading and heightened job opportunities. Although rural areas lag behind in the CTGT process, they commence to reap the rewards as policy support and infrastructure enhancements materialize. Both urban and rural residents can diminish additional costs linked to traditional coal combustion following the optimization of their energy consumption structures. Furthermore, rural areas may garner increased subsidies or employment prospects during the energy transition process, gradually diminishing the income gap between urban and rural residents and enhancing overall residents’ welfare²⁵.

• (2) The CTGT may enhance IW by generating CAE The natural gas industry, a prototypical capital-intensive sector, demands substantial capital outlays across various stages, including exploration, extraction, pipeline installation, and storage facility construction. The CTGT fosters a conducive developmental milieu and policy framework for energy capital-intensive industries, acting as a powerful magnet to draw in significant social capital inflows.From a market dynamics perspective, amidst rising environmental consciousness and ongoing energy structure adjustments, natural gas, as a clean energy alternative, is experiencing a surge in market demand and boasts promising growth prospects. Investors, driven by the pursuit of long-term and stable returns and guided by precise market trend assessments, are inclined to channel capital into diverse segments of the natural gas industry chain²⁶. From a policy vantage point, the government has rolled out a series of initiatives to bolster the CTGT and the expansion of the natural gas industry, such as fiscal subsidies and tax concessions. These measures mitigate investment risks, amplify investment returns, and fortify investors’ confidence in the natural gas sector²⁷. Capital aggregation furnishes energy capital-intensive industries with abundant financial resources, expediting industrial advancement and technological innovation. On one hand, it propels the large-scale and modernized evolution of the natural gas industry, augmenting its competitive edge. On the other hand, it spurs the growth of ancillary supporting industries, including natural gas equipment manufacturing and energy services, thereby engendering an industrial cluster effect²⁸. The economic growth engendered by capital aggregation affords greater financial resources for public services and infrastructure development. The government can augment investments in healthcare, education, environmental protection, and other domains, enhancing residents’ medical conditions, elevating their health literacy, and safeguarding their physical well-being, ultimately enhancing their sense of happiness.

• (3) The CTGT may enhance IW by generating IUE The traditional coal industry, despite its substantial share in the economic system, is beset by challenges such as high energy consumption, severe pollution, and low efficiency, which significantly hinder the sustainable development of the economy. Following the CTGT, a profound transformation occurs in the energy structure, with natural gas-related industries emerging as a new driving force for economic growth. These industries predominantly belong to the tertiary sector or high-end manufacturing within the secondary sector, distinguished by their high technological sophistication, significant value-added, and minimal pollution emissions. Consequently, the transition towards natural gas leads to an increase in the proportion of the tertiary sector and a relative reduction in the dominance of traditional industrial sectors¹⁷.From the perspective of enhancing employment quality, the expansion of the natural gas industry has generated a substantial number of high-quality job opportunities that necessitate a high level of labor skills and qualifications. This has attracted talent from diverse backgrounds to regions associated with the natural gas industry²⁹. For highly skilled professionals, the natural gas industry provides an expansive platform for innovation and development, enabling them to fully harness their professional expertise. For less skilled workers, skill training programs facilitate their entry into suitable employment positions within the infrastructure construction, operation, and maintenance sectors of the natural gas industry. The industrial upgrading effect promotes the rational mobility and optimal allocation of labor, thereby improving the quality of employment and income levels for workers. This, in turn, bolsters residents’ economic security and sense of social belonging, ultimately elevating their LH and PH.

• (4) The CTGT may enhance IW by generating EIE The combustion of traditional coal releases substantial amounts of pollutants such as sulfur dioxide, nitrogen oxides, and particulate matter, which are primary sources of air and water pollution¹⁷. Air pollution can trigger a range of health issues, including respiratory diseases and cardiovascular conditions, thereby degrading residents’ quality of life. Water pollution, on the other hand, compromises the safety of residents’ drinking water, posing threats to their physical health³⁰. Following the CTGT, natural gas, as a clean energy source, primarily produces carbon dioxide and water upon combustion, significantly reducing pollutant emissions. The improvement in air quality directly lowers the incidence of diseases caused by inhaling pollutants, such as respiratory illnesses and lung cancer, thereby enhancing residents’ physical health. Simultaneously, the better air quality provides residents with more pleasant outdoor spaces, encouraging physical exercise and social activities, which in turn strengthen residents’ physical fitness and mental well-being. The reduction in water pollution safeguards residents’ drinking water safety, mitigating the risk of diseases stemming from water contamination. Moreover, the environmental improvement elevates the quality of residents’ living environments, enabling them to reside in fresher and more comfortable surroundings. This alleviates the anxiety and stress associated with environmental pollution, ultimately enhancing residents’ life satisfaction³¹. Based on the aforementioned analysis, the following hypothesis is proposed:

H1

The CTGT can enhance IW.

H2

The CTGT can enhance IW by generating ISE, CAE, IUE, EIE.

Natural gas pipeline accessibility, coal-to-gas transition and individual welfare

During the process of fuel transition, natural gas pipelines serve as a crucial bridge connecting natural gas energy sources with residents’ daily lives, playing a pivotal role in enabling the efficient transportation of natural gas from the market to end-users. As illustrated in Fig. 1, China’s natural gas production facilities exhibit a notable geographical distribution pattern, with a higher concentration in the western regions such as Xinjiang and Inner Mongolia, which are rich in natural gas resources and thus serve as the primary production hubs. In contrast, the eastern coastal regions act as the main gateways for natural gas imports, housing a significant number of liquefied natural gas (LNG) terminals and offshore natural gas landing points to meet the robust demand for natural gas in the economically developed eastern regions. This geographical concentration disparity directly contributes to the complexity and challenges associated with China’s natural gas pipeline layout³². Recent studies have highlighted that in the process of the CTGT, constructing natural gas pipeline infrastructure represents one of the major costs. Existing coal-fired power plants often face constraints in transitioning to natural gas due to the inadequate availability of natural gas transmission infrastructure, which limits the feasibility of residents’ coal-to-gas conversion²¹. The presence of natural gas pipelines in the vicinity of coal-fired power plants significantly enhances the implementation potential of coal-to-gas projects in those regions. Taking the United States as an example, its extensive natural gas pipeline network is approximately 6.5 times denser than its interstate highway system, a factor that has positioned the U.S. as a global leader in carbon emissions reduction. However, it is important to note that the construction of natural gas pipelines not only involves substantial fixed costs but also entails increased transportation expenses. For instance, while the Xinjiang region benefits from natural advantages and price competitiveness in energy supply due to the abundant resources of the Tarim Oilfield, the long-distance transportation of natural gas to regions such as Hunan and Jiangxi significantly drives up transportation costs, resulting in a substantial increase in costs for end-users.

Fig. 1

Preliminary analysis of NGPA. (a) Effective natural gas pipeline projects in 33 provincial administrative regions of China. (b) Effective LNG plant projects in 29 provincial administrative regions of China. (c) Number of effective projects of LNG import receiving stations in 17 provincial administrative regions of China. (d) Scale of LNG import receiving stations in 17 provincial administrative regions of China. (www.chinagasmap.com). It was drawn by ArcGIS 10.8(https://desktop.arcgis.com/zh-cn/desktop/index.html).

Therefore, NGPA may play a crucial role in enhancing residents’ welfare during the CTGT. The efficacy of this fuel transition is intricately linked to pipeline safety, cost considerations, and other factors^21,33. Specifically, when residential areas are situated in close proximity to natural gas pipelines, safety concerns become paramount. Ensuring the safe operation of pipelines mandates stringent spatial isolation and protective measures³⁴. In residential areas adjacent to pipelines, the repercussions of a pipeline leak or accident could be dire, posing a direct threat to residents’ lives and property³⁵. A survey conducted by Brodsky et al. across multiple residential areas near natural gas pipelines indicated that safety incidents, such as pipeline leaks and explosions, not only result in immediate casualties and property damage but also inflict long-lasting psychological trauma on residents³⁶. As the distance between residential areas and natural gas supply points along pipelines increases moderately, safety hazards gradually abate, and residents begin to reap the numerous benefits afforded by natural gas. The efficient utilization of natural gas also reduces residents’ energy expenditures and elevates their quality of life³⁷. However, when the distance between residential areas and natural gas supply points widens further, new challenges arise. Long-distance natural gas transmission not only exacerbates energy losses and diminishes transmission efficiency but also significantly inflates the costs associated with pipeline installation and maintenance. These costs may ultimately be transferred to residents, thereby affecting the acceptability and sustainability of the energy transition. A quintessential example of this phenomenon is India, which, due to its expansive territory and large population, has experienced relatively sluggish development in natural gas pipeline infrastructure. The exorbitant costs of pipeline installation and maintenance have deprived a substantial portion of the population of access to modern energy services³⁸. Overall, determining the “optimal distance” necessitates a comprehensive evaluation of the construction, maintenance, and transportation costs of natural gas pipelines, as well as the environmental and economic advantages derived from residents’ utilization of natural gas. Germany, as the largest natural gas market in the European Union³⁹, strategically plans the distance between pipelines and residential areas based on factors such as population density and energy demand in various regions. In newly developed residential areas surrounding certain cities, the distance between natural gas pipelines and residential areas is typically maintained within a range of 5–8 km, ensuring both the stability of energy supply and the mitigation of safety hazards and energy costs, thereby effectively enhancing residents’ welfare. Based on the aforementioned analysis, the following hypothesis is proposed:

H3

NGPA plays a regulation effect in the CTGT’s promotion of IW enhancement. The effectiveness of CTGT in improving IW may vary under different levels of NGPA.

In summary, the pathway diagram illustrating how the CTGT promotes IW enhancement is depicted in Fig. 2.

Fig. 2

Theoretical analysis framework of this paper.

Data source, variables and model selection

Data source

Chinese General Social Survey, initiated in 2003, is analogous to the General Social Survey conducted in the United States. The primary rationale for selecting data from the CGSS stems from two key factors. Firstly, the 2018 CGSS database meticulously captures comprehensive information regarding Chinese residents’ energy consumption patterns and diverse facets of their lives. Secondly, the CGSS, a collaborative effort between Renmin University of China and academic institutions nationwide, represents the first nationwide, comprehensive, large-scale social survey project in China. Utilizing a multi-stage probability sampling methodology, the survey samples are drawn across the country, covering 28 provincial administrative units in China (with Hong Kong, Macao, Taiwan, Xinjiang, Tibet, and Hainan excluded due to incomplete data). It systematically and comprehensively mirrors the fundamental conditions of various aspects of Chinese society and is widely acknowledged by the academic community as a representative dataset of substantial scientific research value. This study employs data from the 2018 CGSS, which adopted a stratified three-stage probability sampling approach. Depending on the specific stratum, the sampling units at each stage exhibit slight variations, as elaborated in Table 1. The 2018 CGSS data encompassed a total of 12,787 samples. After excluding missing and invalid data in accordance with the research theme, 3080 valid samples were retained. Furthermore, some of the macroeconomic variables utilized in this paper are sourced from statistical yearbooks of various regions across China.

Table 1 Sampling units in each stage.

Variable selection and explanation

• (1) Dependent variables IW serves as a comprehensive assessment that encapsulates residents’ subjective perceptions of life satisfaction and happiness. It is typically gauged using indicators such as life expectancy, health status, and self-reported happiness indices^6,40. In our dataset, the dependent variables of focus are residents’ LH and PH status, which collectively provide a measure of IW.For the question in the CGSS questionnaire, “Overall, do you perceive your life as happy?”, with response categories ranging from “Very unhappy”, “Somewhat unhappy”, “Neither happy nor unhappy”, “Somewhat happy”, to “Very happy”, we assign numerical values accordingly: “Very unhappy” is coded as 1, “Somewhat unhappy” as 2, “Neither happy nor unhappy” as 3, “Somewhat happy” as 4, and “Very happy” as 5. This results in an ordinal variable that reflects residents’ LH. Similarly, for the question “Overall, do you consider yourself to be in good health?”, with response options “Very unhealthy”, “Somewhat unhealthy”, “Neither healthy nor unhealthy”, “Somewhat healthy”, and “Very healthy”, we assign values as follows: “Very unhealthy” is coded as 1, “Somewhat unhealthy” as 2, “Neither healthy nor unhealthy” as 3, “Somewhat healthy” as 4, and “Very healthy” as 5. This yields an ordinal variable representing residents’ PH status.

• (2) Explanatory variable The key explanatory variable in this research is whether residents have experienced the CTGT⁴¹. The survey included the question, “Has your household participated in the coal-to-gas conversion project?” A “Yes” response was coded as 1, signifying participation, whereas a “No” response was coded as 0, indicating non-participation. Figure 3 depicts the spatial distribution of life happiness and physical health status among residents who have participated in the CTGT and those who have not. From this preliminary analysis, it is evident that residents in southern regions generally report higher levels of life happiness and physical health status compared to their counterparts in northern regions. Moreover, participation in the CTGT seems to have a substantial positive impact on residents’ LH and PH status.

• (3) Control variables Drawing from the existing body of literature^42,43, this study carefully selects control variables across four main dimensions: individual characteristics, household characteristics, cognitive characteristics, and regional characteristics. This strategic selection aims to minimize the potential bias introduced by omitted variables in the model estimation process. For individual characteristics, we incorporate variables such as the gender, age, educational attainment, marital status, and household registration of the household head. These variables provide insights into the demographic and socioeconomic backgrounds of the individuals under study. Regarding household characteristics, we include variables like total household income, household size, and total household expenditure. These factors are crucial for understanding the economic circumstances and living standards of the households. In the realm of cognitive characteristics, we consider variables related to perceptions, such as the awareness of varying pollution levels associated with different energy products and the willingness to make economic sacrifices to some extent for the sake of environmental protection. These variables shed light on the attitudes and beliefs of the respondents regarding energy and environmental issues. For regional characteristics, we introduce variables that encompass economic development level, infrastructure quality, energy prices, energy accessibility, and energy supply stability. These variables help capture the regional disparities and contextual factors that may influence the outcomes of interest. Additionally, this study incorporates regional dummy variables to categorize the sample into four distinct geographical regions: Northeast, East, Central, and West. The Northeast region is designated as the reference category, allowing for meaningful comparisons across regions. This categorization enables us to explore potential regional variations in the effects of the explanatory variables on the dependent variables of interest.

• (4) Mechanism variables Building upon the theoretical analysis outlined previously, and with the aim of gaining a more profound understanding of the impact of the CTGT on IW, this study takes cues from Yu et al¹⁷ and introduces four mechanism variables: the income substitution effect(ISE), the capital agglomeration effect(CAE), the industrial upgrading effect(IUE), and the environmental improvement effect(EIE).The income substitution effect is captured through four indicators: per capita disposable income for all residents, per capita disposable income for urban residents, per capita disposable income for rural residents, and the Theil index. These indicators offer insights into how the CTGT might influence income distribution and substitution patterns across various demographic segments.The capital agglomeration effect is gauged using three indicators: investment in the energy sector, investment in the natural gas extraction industry, and investment in the production and supply of electricity, steam, and hot water. These indicators reflect the degree to which capital is concentrated in sectors associated with CTGT.The industrial upgrading effect is evaluated through six indicators: the number of legal entities in the primary, secondary, and tertiary industries, industrial structure composition, labor force quality metrics, and the degree of industrial agglomeration. These indicators assist in assessing how the transition might drive industrial restructuring and upgrading processes. The environmental improvement effect is quantified using six indicators: sulfur dioxide emissions, nitrogen oxide emissions, and soot and dust emissions from exhaust gases, as well as chemical oxygen demand emissions, total nitrogen emissions, and total phosphorus emissions in wastewater. These indicators act as proxies for the environmental benefits that could potentially arise from the CTGT.

• (5) Regulation variable Informed by the research conducted by Yang et al. this study identifies natural gas pipeline accessibility as the moderating variable²¹. Utilizing the scenario posed in the CGSS questionnaire, which asks, “What is the distance between your home and the nearest piped natural gas supply point?”, and presenting respondents with options such as “Zero distance, supply directly to the household,” “Relatively close, 1–3 km (15–30 min’ walk),” “Somewhat far, 3–5 km (30–60 min’ walk),” “Quite far, 5–10 km (60–120 min’ walk),” and “Very far, over 10 km (more than 120 min’ walk),” we assign numerical values accordingly. Specifically, “Zero distance, supply directly to the household” is coded as 1, “Relatively close, 1–3 km (15–30 min’ walk)” as 2, “Somewhat far, 3–5 km (30–60 min’ walk)” as 3, “Quite far, 5–10 km (60–120 min’ walk)” as 4, and “Very far, over 10 km (more than 120 min’ walk)” as 5. This coding system effectively captures the varying levels of natural gas pipeline accessibility for residents. Recognizing that directly querying residents about the subjective distance to the natural gas supply point may introduce measurement inaccuracies, surveyors adopt a more nuanced approach. They inquire in detail about the approximate time residents typically spend walking to the supply point. By integrating this temporal information with specific contextual details, a more precise assessment of pipeline accessibility is attained. A comprehensive summary of all variables utilized in this study is presented in Table 2, as outlined in the subsequent section.

Fig. 3

Spatial distribution of LH and PH status of those who participated in and did not participate in the CTGT. It was drawn by ArcGIS 10.8(https://desktop.arcgis.com/zh-cn/desktop/index.html).

Table 2 Descriptive statistical results of main variables.

Model setting

In choosing the appropriate research methodology, we recognize that both residents’ LH and PH status are ordinal multinomial variables, comprising multiple ordered categories with a distinct sequential hierarchy. For example, residents’ LH is categorized into levels ranging from “extremely unhappy” to “extremely happy,” with intermediate categories of “unhappy,” “neutral,” and “relatively happy.” Similarly, PH status is divided into categories such as “extremely unhealthy,” “unhealthy,” “neutral,” “relatively healthy,” and “extremely healthy.” In these circumstances, traditional linear regression models are inadequate because they assume a continuous dependent variable and a normally distributed error term, which cannot accurately capture the relationships inherent in ordinal multinomial variables. Conversely, ordinal regression models are more suitable for handling such data. By estimating the intercepts between different categories and the impact coefficients of independent variables on each category of the dependent variable, ordinal regression models can elucidate the intrinsic connections between independent and ordinal dependent variables. Consequently, the model formulated is as follows:

$$\ln \left[ {\frac{{p(y_{i} \le j)}}{{1 - p(y_{i} \le j)}}} \right] = \alpha_{j} + \beta CGT_{i} + X_{i} \theta$$

(1)

In Eq. (1), $y_{i}$ represents an individual i’s LH and PH status. $p(y_{i} \le j)$ denotes the probability that an individual’s LH and PH status are less than or equal to a certain category j, $j = 1,2,3,4,5$. $\alpha_{j}$ is the intercept of the model. $CGT_{i}$ is the core explanatory variable in this study, representing the “CYGT”. $\beta$ is the regression coefficient corresponding to the explanatory variable $CGT_{i}$. $X_{i}$ indicates the vector of control variables, and $X_{i}$ is a vector of regression coefficients corresponding to each control variable in the set.

Meanwhile, the theoretical analysis section elucidates that the ISE, CAE, IUE and EIE function as mechanism variables through which the CTGT influences IW. The traditional Sobel test is predicated on a stringent normality assumption concerning the distribution of indirect effects, which is frequently unattainable with real-world data, potentially resulting in biased test outcomes. Although the Baron & Kenny test provides a coherent logical framework, it necessitates that the total effect of the independent variable on the dependent variable be statistically significant—a condition that may not be met in empirical research due to the existence of multi-path effects. Consequently, to streamline model assumptions and circumvent endogeneity issues that might emerge from incorporating new mechanism variables, this study adopts the methodology of Chen et al. by constructing mechanism variables and formulating a linear regression model⁴⁴ to investigate the impact of the CTGT on these mechanism variables. The specific model configuration is delineated as follows:

$$M_{i} = \eta + \beta CGT_{i} + X_{i} \theta + \gamma_{i}$$

(2)

Here, the definitions of variables $\beta$, $CGT_{i}$, $X_{i}$, and $X_{i}$ align with those presented in Eq. (1). Variable $M_{i}$ corresponds to the mechanism variable under scrutiny, $\eta$ represents the constant term, and $\gamma_{i}$ indicates the error term.

Drawing upon the ordinal regression model, and in order to assess the regulation role of NGPA in the impact of the CTGT on IW, as outlined in the theoretical analysis section, we adopt the methodology proposed by Shi et al. and establish the following regulation effect model⁴⁵:

$$\ln \left[ {\frac{{p(y_{i} \le j)}}{{1 - p(y_{i} \le j)}}} \right] = \alpha_{j} + \beta CGT_{i} + \varepsilon_{1} NGPA_{i} + \varepsilon_{2} NGPA_{i} \times CGT_{i} + X_{i} \theta$$

(3)

$$\ln \left[ {\frac{{p(y_{i} \le j)}}{{1 - p(y_{i} \le j)}}} \right] = \alpha_{j} + \beta CGT_{i} + \varepsilon_{1} NGPA_{i} + \varepsilon_{2} NGPA_{i} \times CGT_{i} + \varepsilon_{3} NGPA_{i}^{2} + \varepsilon_{4} NGPA_{i}^{2} \times CGT_{i} + X_{i} \theta$$

(4)

In this context, the definitions of variables $p(y_{i} \le j)$, $\alpha_{j}$, $\beta$, $CGT_{i}$, $X_{i}$, $X_{i}$ are identical to those specified in Eq. (1). Variable $NGPA_{i}$ serves as the regulation variable, whereas $\varepsilon_{{1}}$, $\varepsilon_{{2}}$, $\varepsilon_{3}$, $\varepsilon_{4}$ represent the estimated coefficients.

Empirical results

Benchmark regression result

Table 3 presents the benchmark regression results of the model. Columns (1) and (4) in Table 3 report the outcomes obtained without including any control variables. Columns (2) and (5) show the results after incorporating residents’ individual characteristics, household characteristics, and cognitive characteristics as control variables. Columns (3) and (6) display the findings when, in addition to the aforementioned characteristics, regional characteristics and regional dummy variables are also included in the analysis. It is evident that, irrespective of the model specification employed, the transition to coal-to-gas conversion exerts a statistically significant and positive influence on LH and PH status. This finding suggests that participation in the CTGT program can substantially enhance IW.

Table 3 Benchmark regression results of the impact of CTGT on IW.

Table 4 reports the marginal effects of the CTGT on residents’ LH and PH status. Examining the marginal impact of the CTGT, it is evident that this shift increases the likelihood of an enhancement in IW. There is a significant negative correlation between residents’ participation in the CTGT and their self-reported states of being “very unhappy,” “somewhat unhappy,” or “Neither happy nor unhappy” in terms of life satisfaction. Conversely, there is a significant positive correlation between participation and the states of being “somewhat happy” or “very happy.” Specifically, after participating in the CTGT, the probabilities of residents reporting being “very unhappy,” “somewhat unhappy,” or “Neither happy nor unhappy” decrease by 4.1%, 15%, and 14.8%, respectively, while the probabilities of reporting being “somewhat happy” or “very happy” increase by 11.8% and 22.0%, respectively. Similarly, a significant negative correlation exists between residents’ participation in the CTGT and their self-reported states of being “very unhealthy,” “somewhat unhealthy,” or “Neither healthy nor unhealthy” in terms of PH. In contrast, a significant positive correlation is observed between participation and the states of being “somewhat healthy” or “very healthy.” That is, after participating in the CTGT, the probabilities of residents reporting being “very unhealthy,” “somewhat unhealthy,” or “Neither healthy nor unhealthy” decrease by 10.1%, 32.0%, and 2%, respectively, while the probabilities of reporting being “somewhat healthy” or “very healthy” increase by 17.5% and 26.6%, respectively. Consequently, Hypothesis H1 is validated.

Table 4 Marginal effects.

Mechanism analysis

Income substitution effects

Equation (2) was utilized to investigate the mechanisms by which the ISW, CAE, IUE, and EIE contribute to the enhancement of IW resulting from the CTGT. As illustrated in Table 5, the CTGT has a statistically significant and positive impact on the disposable income of the overall population, urban dwellers, and rural inhabitants. Notably, this transition also alleviates the regional urban–rural income disparity, as evidenced by its substantial negative influence on the Theil index. This implies that the CTGT can augment the disposable income of various demographic segments through the ISE, while concurrently diminishing the urban–rural divide and thus fostering an improvement in IW.A reasonable explanation for this occurrence is that the CTGT optimizes the energy infrastructure in both urban and rural settings⁴⁶, leading to a reduction in living expenses. Furthermore, the integration of clean energy sources taps into latent consumption potential, stimulating heightened economic dynamism in rural areas. As a consequence, this facilitates a more equitable income distribution between urban and rural regions, bridging the income gap and contributing to an overall upliftment in IW.

Table 5 Income substitution effects.

Capital aggregation effects

As demonstrated in Table 6, CTGT significantly and positively impacts regional investment in the energy sector, encompassing natural gas extraction, as well as the production and supply of electricity, steam, and hot water. This finding suggests that the CTGT can stimulate investment in capital-intensive energy industries through the CAE, ultimately enhancing IW.A credible explanation for this phenomenon is that, propelled by policy measures, there has been a considerable surge in market demand for natural gas. This increased demand has attracted substantial capital inflows. In response, enterprises have intensified their investments in natural gas extraction, supply, and associated sectors, thereby fueling the growth of capital-intensive energy industries. As a result, this has contributed to an improvement in IW, particularly in terms of energy supply and related aspects⁴⁷.

Table 6 Capital aggregation effects.

Industrial upgrading effects

As illustrated in Table 7, the CTGT has a significant and negative impact on the primary and secondary industries in the region. However, it exerts a significant and positive influence on the tertiary industry, industrial structure, labor force level, and industrial agglomeration degree. This indicates that the CTGT can facilitate industrial upgrading by reducing the share of the primary and secondary industries while increasing that of the tertiary industry, thereby promoting the upgrading of the industrial structure. Additionally, it generates employment opportunities, fosters labor force agglomeration in the region, and ultimately enhances individual welfare. A plausible explanation for this phenomenon is that during the transition process, the primary and secondary industries face cost pressures related to equipment upgrades and technological transformations, which constrain their short-term growth. Meanwhile, the demand for clean energy spurs the development of related service industries, propelling the growth of the tertiary industry. Furthermore, the emergence of new industries creates job opportunities, attracts labor force agglomeration, facilitates industrial structure upgrading, and improves IW²⁹.

Table 7 Industrial upgrading effects.

Environmental improvement effects

As depicted in Table 8, the CTGT exerts a significant and negative impact on sulfur dioxide, nitrogen oxides, and particulate matter emissions in exhaust gases, as well as on chemical oxygen demand, total nitrogen, and total phosphorus levels in wastewater. This indicates that CTGT can enhance IW by leveraging the EIE to reduce pollutants in both the air and water bodies. A plausible explanation for this is that, compared to traditional coal, natural gas burns more completely and contains significantly fewer impurities such as sulfur and ash. Consequently, emissions of sulfur dioxide, nitrogen oxides, and particulate matter during combustion are substantially reduced^48,49. In terms of wastewater, the CTGT alleviates the pollution of water used in industrial production and heating processes, leading to a decrease in the discharge of pollutants such as chemical oxygen demand, total nitrogen, and total phosphorus³⁰. Additionally, the coal-to-gas project drives enterprises to upgrade their environmental protection equipment and production processes, further curbing pollutant emissions. The improved environment creates healthier air and water conditions for residents, reducing the risk of diseases caused by pollution and thereby enhancing their welfare. Consequently, Hypothesis H2 is validated.

Table 8 Environmental improvement effects.

Analysis of regulatory effect

Equations (3) and (4) were employed to examine the regulation effect of NGPA on the enhancement of IW resulting from the CTGT. In Table 9, columns (1) and (3) introduce NGPA and its interaction term with the CTGT, based on the benchmark regression. Columns (2) and (4) further incorporate the squared term of NGPA and its interaction term with the CTGT.

Table 9 Inspection of inverted U regulation effect on NGPA.

It can be observed that the interaction coefficients of the linear terms in columns (1) and (3) are significantly positive. Similarly, in columns (2) and (4), the interaction coefficients of the linear terms remain significantly positive, while the interaction coefficients of the quadratic terms are significantly negative. This suggests that NGPA exhibits an inverted U-shaped regulation effect on the impact of the CTGT on IW. Specifically, as the NGPA increases, the positive effect of the CTGT on IW initially rises and then declines.

A plausible explanation for this phenomenon is that when the distance is too short, short-term negative effects such as noise and land occupation from pipeline construction may offset some of the benefits of the transition. As the distance increases, these negative effects diminish, and an appropriate distance allows residents to enjoy the convenience of gas supply while being less directly disturbed, thereby enhancing the positive impact. However, when the distance becomes excessively long, the cost of gas transmission and the risk of supply instability rise, leading to a decline in the positive impact. This finding is consistent with the conclusions drawn by Yang et al.²¹. Consequently, Hypothesis H3 is validated.

Heterogeneity analysis

When evaluating the impact of the CTGT on IW, although preliminary findings suggest that participation in this transition contributes to enhancing IW, such conclusions primarily rely on the average effect across the entire sample and fail to fully account for the disparities among different demographic groups and varying levels of urban–rural infrastructure. To gain a more comprehensive and in-depth understanding of the CTGT’s effects, this paper analyzes the heterogeneous effects of the CTGT on improving IW across two levels and six dimensions: population groups (including gender, education level, and consumption level) and urban–rural infrastructure (household registration, infrastructure level, and energy supply stability).

Population group difference

As shown in Table 10, the CTGT yields more pronounced improvements in LH, PH for women with lower consumption levels and those who have completed junior high school or below. A plausible explanation for this is as follows: among groups with lower consumption levels, energy expenditures account for a higher proportion of their income. Following the transition, the reduction in fuel costs directly releases disposable income, thereby enhancing their life satisfaction. Women with a junior high school education or below may have a more limited awareness of the health hazards associated with traditional coal combustion. As a result, they are more likely to directly perceive the health benefits from improved indoor air quality after the transition. Moreover, given their high frequency of energy use in daily household chores, the reduction in labor intensity and health risks brought about by clean fuels is more significant for them, thus reinforcing the positive welfare effects of the transition⁵⁰.

Table 10 Population group difference.

Urban and rural facilities differences

As indicated in Table 11, the CTGT generally produces more significant improvements in LH and PH in urban areas with higher levels of energy supply and infrastructure development. A plausible explanation for this phenomenon is as follows: An adequate and stable energy supply ensures a continuous and uninterrupted supply of natural gas, minimizing disruptions to residents’ lives caused by supply outages and directly enhancing their life satisfaction. Moreover, well-developed transportation infrastructure facilitates the transportation of natural gas and the delivery and installation of related equipment, reducing logistics costs and time consumption, and accelerating the implementation efficiency of the transition. Additionally, regions with well-established transportation networks often have more comprehensive environmental monitoring and pollutant treatment systems in place, which can more effectively mitigate the emission risks associated with natural gas usage and strengthen the health improvement effects⁵¹. Furthermore, residents in urban areas have more accessible channels for obtaining energy services and health information, making it easier for them to perceive and benefit from the advantages of clean energy utilization.

Table 11 Urban and rural facilities differences.

Robustness test

Robustness test based on considering endogenous problems

The endogeneity issues in this study primarily stem from the following aspects: First, the participation of rural households in the CTGT contributes to enhancing their LH and PH. Meanwhile, the improvement in tangible benefits motivates these households to attract other residents to participate in the transition through their social networks. Consequently, there may exist a reciprocal causality between the two. Second, although this study comprehensively considers various factors influencing individual welfare, the factors affecting individual welfare are complex and diverse. Due to limitations in data availability, there remain numerous influential factors that are difficult to describe and measure. Therefore, this paper employs “residents’ level of understanding about the CTGT” as an instrumental variable for “residents’ participation in the CTGT.” The selection of this instrumental variable is primarily based on two considerations: On the one hand, the more residents understand about the CTGT, the stronger their willingness to participate in it, which satisfies the relevance condition of instrumental variables. On the other hand, understanding the CTGT does not directly affect their welfare, thus meeting the exogeneity condition of instrumental variables. Therefore, selecting “residents’ level of understanding about the CTGT” as an instrumental variable is theoretically reasonable.

Table 12 reports the Two-Stage Least Squares (2SLS) estimation results of the impact of the CTGT on residents’ LH and PH. The first-stage 2SLS estimation results indicate that the regression coefficient of "residents’ level of understanding about the CTGT" on the CTGT is significantly positive at the 1% level, suggesting a significant correlation between the instrumental variable selected in this paper and the core explanatory variable. Building on this, the validity of the instrumental variable is tested. The F-statistic of 197.25 exceeds 10, and the Cragg-Donald Wald F-value of 203.029 surpasses the critical value of 16.38 under the 10% bias in the Stock–Yogo weak instrumental variable F-test, indicating that the selected instrumental variable is not a weak one. Therefore, employing "residents’ level of understanding about the CTGT" as an instrumental variable for the CTGT is empirically appropriate. Further analysis of the second-stage 2SLS estimation results regarding the impact of the CTGT on residents’ IW, as shown in Table 12, reveals that the CTGT has a significantly positive impact on both LH and PH. After accounting for endogeneity, the benchmark results of this paper remain robust.

Table 12 Tool variable method.

Robustness test based on considering self-selection bias

• (1) Logit regression results The first step in applying propensity score matching (PSM) is to estimate the propensity scores. Selecting appropriate matching variables is crucial; these variables must simultaneously influence both residents’ participation in the CTGT and their IW, while also remaining unaffected by residents’ participation in the CTGT. Therefore, this paper selects all the control variables from the benchmark regression as matching covariates. The equation estimation results are presented in Table 13, with the model’s P-value being significant at the 1% level, indicating a good model fit.

• (2) Balance test The matching quality is assessed based on the distribution of the explanatory variable (whether residents participated in the CTGT) before and after propensity score matching, although this method cannot accurately estimate the probability of residents participating in the CTGT. Consequently, this paper conducts a balance test on the matching variables, the results of which are shown in Table 14. The results indicate that after propensity score matching, the standard deviations of most explanatory variables significantly decrease. The R² value drops from 0.217 before matching to 0.004–0.008 after matching, and the LR statistic decreases from 877.65 before matching to 13.86–25.77 after matching. According to the joint significance test, the significance levels of the explanatory variables have undergone substantial changes. Moreover, the mean bias of the explanatory variables is greatly reduced, from 30.9% before matching to 2.7%–3.8% after matching, while the median bias decreases from 26.2% before matching to 2.5%–3.6% after matching, indicating a substantial reduction in the total bias. This suggests that propensity score matching has reduced the differences in explanatory variables between the treatment and control groups, and that after matching, residents who participated in the CTGT and those who did not have similar characteristics in other respects.

• (3) Propensity score matching method Table 15 presents the estimated treatment effects of the CTGT on residents’ IW. The results obtained using three matching methods—K-nearest neighbor matching, caliper matching, and kernel matching—are similar and positively significant, to some extent reflecting the robustness of the benchmark regression results.

• (4) Placebo test Following the research of Cai et al.⁵², this paper generates 1000 random datasets and estimates the probability of the impact of participating in the CTGT on residents’ IW based on these false experiments. Figure 4 shows that 500 estimated coefficients are distributed around zero, with P-values concentrated above 0.1, indicating that the estimates are not severely biased due to omitted variables.

Table 13 Logit estimation results of whether residents participate in the transition tendency score of coal to gas.

Table 14 Balance test.

Table 15 Average treatment Effect.

Fig. 4

Placebo test results.

Robustness test based on endogenous switching regression models

In the aforementioned analysis, propensity score matching (PSM) has mitigated selection bias attributable to observable factors rather than unobservable ones. To further test the robustness of our results, this paper employs an endogenous switching regression (ESR) model, which accounts for both observable and unobservable factors. The ESR approach is particularly suitable for examining “the effects of selection decisions that allow for endogeneity, sample selection, and interactions between the adoption decision and other covariates influencing the outcome equation”⁵³. Specifically, the endogenous switching model is estimated in two stages: In the first stage, a selection estimation equation is established, which serves as a model to estimate whether rural households participate in the CTGT. The level of understanding about the CTGT is chosen as an identification variable to verify the influencing factors of rural households’ participation in the transition. (Due to space constraints, this part, which shares similarities with the preceding content, is not presented here but is retained by the author for reference upon request).In the second stage, an outcome estimation equation is constructed to estimate the equations for whether rural households participate in the CTGT, with the aim of verifying differences in their LH and PH under different scenarios. The results are shown in Table 16, where both the average treatment effect on the treated (ATT) and the average treatment effect on the untreated (ATU) are significantly positive at the 1% level. These findings indicate that the CTGT can enhance residents’ IW, thereby validating the robustness of our research conclusions.

Table 16 Average treatment effect of CTGT on IW based on endogenous switching regression models.

Conclusions and policy recommendations

This study employs data from the 2018 CGSS to examine the impact of the CTGT on IW and the underlying mechanisms at play. Additionally, it investigates the regulation role of NGPA and the heterogeneity across various demographic groups and urban–rural infrastructure. The key findings are outlined below: Firstly, the CTGT significantly enhances IW through ISE, CAE, IUE, and EIE. Following the transition, residents’ LH and PH have shown marked improvements. This conclusion holds robust even after addressing endogeneity concerns and utilizing methodologies such as propensity score matching, placebo tests, and endogenous switching regression models.Secondly, NGPA demonstrates an inverted "U"-shaped regulation influence on the relationship between the CTGT and IW. Specifically, as the NGPA increases, the positive impact of the CTGT on IW initially intensifies before subsequently diminishing.Lastly, the promoting effect of the CTGT on IW is not uniform. Due to disparities in demographic characteristics and urban–rural infrastructure, the promoting effect is more pronounced among female groups with lower consumption levels and education levels of junior high school or below, as well as in urban regions characterized by higher levels of energy supply and infrastructure development. Based on the aforementioned research findings, this paper puts forth the following policy recommendations:

Firstly, it is imperative to bolster the comprehensive effects of the CTGT to comprehensively enhance individual welfare. Regarding financial support, priority should be given to increasing investment in coal-to-gas projects, and a detailed and targeted subsidy policy framework should be devised to alleviate potential cost burdens. The following strategies can be implemented: ① Establish a dynamic subsidy adjustment mechanism that modifies subsidy standards and scopes in response to factors such as fluctuations in natural gas market prices and changes in residents’ income levels, thereby ensuring the pertinence and effectiveness of the subsidy policy. ② Explore diversified subsidy approaches. In addition to direct financial subsidies, tax incentives, credit support, and other measures can be utilized to alleviate cost pressures on residents and enterprises. ③ Strengthen the supervision of subsidy funds by establishing a rigorous subsidy fund management system to guarantee that funds are exclusively utilized for their designated purposes, enhance fund utilization efficiency, and prevent waste and misappropriation. In terms of capital aggregation within the energy industry, the government should assume a guiding role in establishing an efficient platform for industry-finance integration. On the one hand, through policy support, large energy enterprises should be enticed to increase their investments in natural gas infrastructure construction, fostering an industrial clustering effect. On the other hand, specific policies should be introduced to encourage the participation of social capital, such as guaranteeing a certain level of investment returns for social capital and granting it priority in subsequent project developments, thereby stimulating the vitality of social capital and jointly enhancing the stability and efficiency of energy supply. Regarding industrial restructuring and job creation, the government should formulate a development plan for the clean energy industry, clarifying the industry’s development direction and key areas. By establishing a dedicated industrial fund, support should be extended to enterprises engaged in clean energy technology research and development, equipment manufacturing, and other innovative endeavors, guiding the industrial structure towards a deeper adjustment towards clean energy. Simultaneously, cooperation with vocational colleges and training institutions should be fortified to provide vocational skills training specific to the coal-to-gas-related industries, creating more employment opportunities for residents in areas such as natural gas infrastructure construction, operation, and maintenance, achieving a win–win situation for economic transformation and upgrading and residents’ employment.In terms of environmental protection, sustained efforts should be made to reinforce environmental protection measures and establish a robust environmental monitoring and evaluation system during the implementation of the coal-to-gas policy. Regulatory and enforcement efforts against pollutant emissions should be intensified, with severe penalties imposed on enterprises and individuals that fail to comply with emission standards. Meanwhile, advanced energy-saving and emission-reduction technologies and equipment should be actively promoted and encouraged for adoption by residents and enterprises, further consolidating and expanding the achievements of the coal-to-gas policy in reducing pollutant emissions and improving air quality, thereby creating a more habitable environment for residents.

Secondly, it is crucial to optimize the layout of natural gas pipelines in a rational and strategic manner to ensure the comprehensive and effective implementation of the CTGT. Policy makers should adhere to the principles of scientific planning and precise layout to optimize the configuration of the natural gas pipeline network. To begin with, leveraging detailed resident distribution data and natural gas demand forecasting models, advanced technologies such as Geographic Information Systems (GIS) should be employed to conduct a scientific and data-driven planning of the natural gas pipeline network. During the planning phase, it is imperative to fully consider factors such as urban development plans and current land use patterns to ensure that pipeline construction maximizes coverage of residents while avoiding any dilution of policy effectiveness due to excessively long pipeline distances. For example, in newly constructed residential areas, pipeline interfaces should be pre-planned and reserved to facilitate immediate access to natural gas upon residents’ relocation. Furthermore, for residential areas situated far from existing natural gas pipelines, a two-pronged approach should be adopted. On one hand, branch pipelines should be reasonably planned and constructed based on actual conditions. During the construction process, close communication and coordination with local governments, communities, and residents should be maintained to optimize construction plans and minimize disruptions to residents’ daily lives and the surrounding environment. On the other hand, active exploration of alternative energy supply methods, such as the establishment of Liquefied Natural Gas (LNG) storage tanks or Compressed Natural Gas (CNG) filling stations, should be undertaken. During the site selection process for these alternatives, factors such as residents’ gas demand and transportation accessibility should be thoroughly considered to ensure the stability and reliability of energy supply. Last but not least, a comprehensive and robust safety supervision system for natural gas pipelines should be established and continuously improved. Safety supervision should be intensified across all aspects of pipeline construction, operation, and maintenance, with strict compliance to relevant safety standards and regulations. The frequency of routine inspections of pipeline facilities should be increased, and intelligent monitoring equipment should be deployed to monitor the operational status of pipelines in real-time, enabling the prompt identification and mitigation of potential safety hazards. Additionally, a comprehensive emergency response plan should be formulated, and regular emergency drills should be organized to enhance the capacity to respond to pipeline accidents, thereby ensuring the safe and reliable operation of pipelines and minimizing inconvenience and losses to residents’ lives caused by pipeline-related incidents.

Thirdly, it is of utmost importance to formulate differentiated policies to facilitate the energy structure transition and elevate individual welfare. For female populations characterized by lower consumption levels and an educational attainment of junior high school or below, the government ought to implement more targeted preferential measures. For example, a certain percentage discount on natural gas usage fees could be granted based on household gas consumption. The criteria for these discounts should be dynamically adjusted according to household income levels, with the aim of alleviating their economic burdens and stimulating their enthusiasm for participating in the CTGT. In urban areas boasting higher levels of energy supply and infrastructure, the government can prioritize enhancing residents’ subjective well-being. This can be accomplished by improving the urban environment and offering a diverse range of cultural and recreational activities to cater to their higher-level spiritual needs. Simultaneously, the government must persist in intensifying its focus and support for rural areas and low-income groups. It should increase investment in infrastructure, such as natural gas pipelines and gas storage facilities, in rural regions to enhance the stability and reliability of energy supply. Moreover, the government should establish a comprehensive feedback mechanism for the CTGT. By establishing dedicated feedback channels, including hotlines, online platforms, and suggestion boxes, it can systematically collect residents’ opinions and suggestions. A professional team should be assembled to analyze and organize the feedback information. Based on the feedback outcomes, policy directions and intensities should be promptly adjusted to ensure that policies can continuously align with residents’ needs and expectations. Additionally, the government should strengthen coordination and collaboration with other relevant policies, such as poverty alleviation policies, environmental protection policies, and energy policies, to create a policy synergy that collectively propels the energy structure transition and the improvement of individual welfare.

Discussion

This study, utilizing data from the Chinese General Social Survey (CGSS), examines the impact of the CTGT on IW. It unveils a range of effects, encompassing income substitution, capital accumulation, industrial upgrading, and environmental enhancement. Moreover, it identifies an inverted U-shaped moderating effect of natural gas pipeline accessibility, along with heterogeneity across different demographic groups and between urban and rural infrastructure. These findings contribute to the micro-level empirical evidence highlighting the pivotal role of the CTGT in enhancing residents’ well-being. Simultaneously, they offer robust policy insights for formulating scientifically sound and rational carbon neutrality policies, facilitating energy structure optimization, and improving residents’ well-being. Nevertheless, this study is not without certain limitations. First, concerning the dataset: This research relies on CGSS data, which exhibits insufficient representation of rural areas and regional disparities. This may undermine the comprehensiveness and generalizability of the research outcomes. Future studies could diversify data sources to bolster the representation of rural areas within the dataset and conduct more granular regional segmentation and analysis. Second, regarding the research content: Recent pertinent studies have demonstrated that Chinese residents’ well-being is intimately tied to changes in their activity patterns, and these changes exhibit regional variations¹¹. However, this study may not have offered an in-depth portrayal of the structural changes in residents’ activity patterns, thereby failing to fully elucidate the impact of the interplay between the CTGT and residents’ activity patterns on well-being levels. Future research could integrate changes in residents’ activity patterns to comprehensively and profoundly understand how the interaction between the CTGT and residents’ activity patterns jointly influences well-being levels.Lastly, at the policy discussion level: Although this paper has acknowledged the role of government subsidies, it has not adequately addressed the long-term sustainability of subsidy policies and the potential ramifications of carbon pricing policies. Future studies should incorporate scenarios involving subsidy reduction or elimination, as well as the implementation of carbon pricing policies, to furnish policymakers with more valuable references.