China's RLE governance is built on the national strategy of Ecological Civilization Construction, organizes and utilizes existing resources to help rural communities overcome pollution problems, create a clean and beautiful rural environment, and promote sustainable development (Zhou et al., 2024). Currently, air, water, and soil pollution are affecting the health and quality of life of China's rural inhabitants. As such, the RLE governance in China has been focusing on rural household garbage disposal, sewage and human waste discharge, and ecological environment restoration (Deng et al., 2022; Liu and Liu, 2020). Studies have shown that the implementation of RLE governance is effective in controlling rural pollution (Deng et al., 2022), lowering illness rates and healthcare costs (Liang et al., 2023), and enhancing the overall wellbeing of rural inhabitants (Zhang et al., 2023).

Socioeconomic conditions, such as the economic growth and the living patterns of the village community, impact the quality of the RLE. For example, the relationship between rural household waste and households' disposable income looks like an inverted U-shape. Waste generation increases as income rises, but it starts to decrease once disposable income per capita surpasses $2,500 (Han et al., 2018). Currently, empirical studies based on rural China have found that the RLE often deteriorates as the rural economy develops. For example, the enhancement of rural living conditions (Liu et al., 2015), the growth of the rural economy (Peng and Zhang, 2019; Zhang F. et al., 2024), and the extensive use of express delivery service (Liu and Huang, 2014) has increased the rural household solid waste and sewage and damaged the RLE.

Governments try to alleviate or solve the pollution problems of rural China by providing public services, such as improving the infrastructure (Li et al., 2021), allocating financial resources to local communities (Li and Hu, 2023), and incentivizing environmental services (Zhang et al., 2023). According to public service motivation theory, it is essential to increase the public service motivation of local government officials to address pollution problems, improve rural living conditions, and protect the interests of rural residents (Liu et al., 2022). Environmental concerns from the public can exert supervisory pressure to motivate public servants to leverage more public services to reduce pollution (Su Y. et al., 2023) and improve the RLE. A grassroots-based governance model is a critical factor contributing to better RLE (Xiao et al., 2022; Su Y. et al., 2023). Including rural residents' participation in environmental governance (Du and Jiao, 2023; Su Y. et al., 2023) and the improvement of rural infrastructure Li Y. et al. (2022) can enhance the quality of RLE.

The central government in China has primary authority over environmental governance, with local governments tasked with carrying out its policies and meeting its goals (Cheng and Yu, 2023). The central government establishes governance structure and performance criteria in accordance with existing environmental laws, while local governments develop specific action plans based on central government policies, taking into account local factors such as local economic development (Du and Ullah, 2024). This top-down governance system can effectively monitor large-scale environmental pollution sources but is less effective in regulating a large number of scattered, small-scale pollution sources due to the high costs of regulation (Tian et al., 2020). Public engagement can fill these regulatory gaps in environmental governance (Yu et al., 2023); public participation in environmental governance may address these gaps by monitoring pollution and waste disposal behaviors of enterprises in the private sector and pushing local authorities to fulfill their environmental governance obligations (Li X. et al., 2022).

Public environmental concerns (PEC) have caught the attention of scholars and are primarily categorized into three types. The first type includes public actions such as complaints, petitions, and proposals. Public expression of environmental concerns through these methods contributes to improving wastewater treatment (Langpap and Shimshack, 2010), reducing industrial emissions (Zhang M. et al., 2024), and enhancing technological efficiency in enterprises (Cao and Chen, 2024). Moreover, these concerns have been found to discourage foreign direct investment, leading to a decrease in pollution in the host nations (Xu et al., 2024), and it is particularly effective in developed but polluted cities (Zhou et al., 2024). Conversely, Zhang et al. (2019) propose that public environmental complaints have no impact on the living environment. The second form involves environmental incidents reporting in news media, such as newspapers, radio, and television, which has demonstrated positive impacts on water pollution control in India (Kathuria, 2007) and on the environmental performance of enterprises in South Korea (Mamingi et al., 2008). The third approach entails using online media (search engines, microblogs, Twitter) to express environmental concerns. Raising PEC through online media has contributed to alleviating smog pollution (Wang J. et al., 2023), improving air quality (Zhang et al., 2018), and driving green technological innovation in heavily polluting industries (He et al., 2022). Furthermore, research on measuring PEC based on statistical data from search engines such as Google and Baidu have shown that an elevated PEC can promote government pollution control (Zheng et al., 2013), reduce rural-urban environmental inequality (Long et al., 2022), alleviate air pollution (Yu et al., 2023), and reduce greenhouse gas emission (Wang Y. et al., 2024). The impact of the PEC on emissions control is particularly pronounced in North China and cities that focus on resource recovery and sustainability (Wang Y. et al., 2024). Meanwhile, the PEC can also encourage corporate environmental investments (Li L. et al., 2023) and green innovation (Geng et al., 2023) and suppress the share yields of heavily polluting firms (Guo et al., 2020). PEC not only obstructs the market entry of polluting firms (Du et al., 2023) but also encourages the outward migration of such enterprises, leading to the phenomenon of pollution transfer (Wang Z. et al., 2024). Ren and Ren (2024) argue that PEC enhances enterprise ESG performance, whereas Chen et al. (2024) assert that these concerns amplify enterprise risk exposure, adversely impacting enterprise ESG performance.

Previous research focuses on the impact of PEC on industrial environmental performance and urban environmental governance. However, few research examines the effects of PEC on rural environmental governance. This gap provides an opportunity for this research to investigate the association between PEC and RLE.

Public concerns can motivate people to participate in governance and potentially impact policies through two models. In a classical “voting with feet” model, the public has the right to move, and they can express their concerns with political, economic, and social conditions by choosing to leave the jurisdiction in which they reside, work, or invest (Tiebout, 1956). The outflow of taxpayers, talent, or capital will significantly impact the local economy, and the government may be motivated to improve social conditions. Thus, “voting with feet” puts pressure on local governments to improve public services and force them to tighten environmental regulations (Tiebout, 1956). Hirschman (1972) described another model of public participation, “voting by hand,” which holds that the public expresses their concerns about political, economic, and social policies and public services through formal elections, voting, or appeals. This public participation model encourages governments to be more responsive to public interests, consider public satisfaction, and assess the real-world impact of policies during implementation. With the rising living standards in China's rural communities, rural residents' environmental concerns and desires for a better living environment are also increasing (Su M. et al., 2023), and the public has demonstrated a strong willingness to participate in the governance of the RLE (Zheng et al., 2013). Meanwhile, with the democracy development, Chinese governments increasingly accept more forms of public participation in governance, and online media channels are widely used in the discussion of social and environmental problems and empower the public to voice their concerns (Wang J. et al., 2023). Partly because online platforms are convenient for stakeholders to openly express their opinions anywhere and anytime (Li X. et al., 2022), public participation in environmental governance through online media channels has made a more significant impact than traditional channels of complaint (Tu et al., 2024).

PEC is becoming increasingly important in environmental governance (Liu and Mu, 2016). PEC encompasses various aspects, including attitudes toward environmental pollution and support for environmental protection initiatives (Jones and Russo, 2024). This concern prompts individuals to advocate for upgrades in pollution control facilities, the optimization of clean production technologies, and enhanced environmental management in rural areas. Furthermore, PEC enables effective monitoring of pollution activities by enterprises and the private sector in rural areas, urging local governments to fulfill their environmental governance tasks (Li X. et al., 2022), thereby boosting the performance of RLE governance. As a result, PEC can encourage local governments to implement measures to address deficiencies in rural environmental governance (Long et al., 2022), leading to improvements in the RLE. Based on this, the study presents its first research hypothesis.

While research indicates that public engagement can improve environmental governance (Long et al., 2022; Yu et al., 2023), the intrinsic mechanisms of PEC affecting the RLE warrant further exploration. This paper proposes that local governments' attention is a critical mechanism connecting public concern and improving the rural environment. Attention is the process through which managers are focused on specific pieces of information and ignore other information (Simon, 1947). Local governments are more likely to act on issues and problems that are deemed to be high priorities, and the effectiveness of action is determined by how much of the government attention is allocated to the issue (Flavin and Franko, 2017; Meng et al., 2024). The attention management theory states that attention, as a scarce resource, determines the content and focus of organizational decision-making (Ocasio, 1997). Governments must be selective in their concerns (Fan et al., 2022), as an organization's attentional focus reflects how the organization prioritizes the objectives and allocates time, finance, workforce, and other resources (Ocasio, 2011). Government attention allocation is a process in which authorities assign attention and resources to selected problems and solutions (Jones and Baumgartner, 2005). Therefore, the allocation of government attention to environmental problems is important for local governments to reach the RLE-related policy objectives. However, the government's attention does not always focus on social and environmental issues. Local governments in China tend to focus on economic growth. Such a “mismatch of attention” between environmental, societal, and economic growth (Tu et al., 2024) could adversely affect improvements in the RLE.

External pressure can sometimes push local authorities to pay attention to matters of public concern (Pan and Fan, 2023; Zhang M. et al., 2024). As the stakeholders of the RLE, the public or social groups often try to communicate with relevant governments through social media and other channels to report pollution issues and demand a better RLE. The dissemination of opinions or attitudes through popular online media can sometimes quickly form a salient Internet public opinion (Wang Z. et al., 2023). Such publicly expressed environmental concerns can exert external pressure on local governments and force them to focus more on improving the RLE. Since 2011, China's central government has launched a number of policies requiring governments at all levels to develop mechanisms to collect, analyze and respond to Internet public opinion, as well as to actively address the public's concerns on key issues. The Office of Cybersecurity and Information of local governments is responsible for tracking and responding to Internet public opinion (Yuan et al., 2023), including public concerns for the RLE. Moreover, public opinions and attitudes, if expressed through the Internet, can form social phenomena and adversely affect the reputation of local governments. If local governments fail to adequately address significant public concerns, it may prompt accountability inquiries from higher levels of government (Sun et al., 2024). As such, when public concern is visible on the Internet, local governments may be forced to pay attention to issues related to the RLE. The increased environmental attention may facilitate the government departments to identify the problems, find solutions, and allocate resources (Li S. et al., 2023). Therefore, the study proposes a mechanism: PEC can pressure local governments, affecting their attention allocation and consequently improving the RLE.

This research analyzed the panel data from prefectural cities in China over the period 2012 to 2021. In China's administrative structure, a prefecture-level city is a jurisdictional division below the province level and has regional-level administrative authority and functions. A prefectural-level city can include several counties, autonomous counties, or county-level cities. A prefectural city typically contains multiple cities, towns, and rural areas. The governments of prefectural-level cities in China are responsible for overseeing and managing environmental issues and rural development within their jurisdictions (Ran, 2017). There are 294 prefecture-level cities in China, while 280 prefecture-level cities with 10-year panel data were identified through an initial scan of databases needed for this study. Some prefecture-level cities identified in the initial scan had missing data. After removing these jurisdictions, the data of 245 cities were used in the analyses. The data on PEC came from the Baidu Index, while the frequency of environment-related keywords was extracted from local authorities' yearly publications. Other data were derived from various statistical yearbooks, particularly at the prefectural level. The study employed data from the rural regions of each prefecture-level authority to analyze the RLE. Data pertaining to rural areas were collected from each prefecture-level city's statistical yearbooks and bulletins.

This study employs panel data for empirical analysis, incorporating a cross-sectional (245 prefecture-level cities) and a time series dimension (10 years: 2012–2021). Since panel data have a cross-section and a time dimension, utilizing ordinary least squares (OLS) regression to estimate panel data would face problems such as omitted variable bias and cross-sectional heterogeneity (Kim and Wang, 2024). The fixed-effect model adopts a robust standard error structure to correct the heteroscedasticity problem and considers the influence of time dimension, which can better capture the relationship between variables and the long-term trend of panel data (De Chaisemartin and D'Haultfoeuille, 2020), thereby enhancing the precision and consistency of the estimations and increasing the reliability of the regression estimates (Liu L. et al., 2024; Kim and Wang, 2024). Therefore, this research applies a fixed-effect model to examine the association between the PEC and the RLE. The baseline model equation of this research is as follows:

Here, i and t indicate the i^th prefecture-level city and the t^th year, PEC indicates the variable of public environmental concern, and RLE represents the rural living environment. The coefficient β₁ represents the influence of PEC on the RLE. Controls are control variables that may affect the RLE, the coefficient λ represents the influence of control variables on RLE. The two error terms, ϕ_t and ε_jt represent a fixed time effect and model error terms. All the variables in the model are log transformed.

Since data heteroscedasticity can affect the accuracy of the empirical analysis results, this paper has logarithmically treated all the variables in the empirical analysis. Logarithmic processing can make the data distribution closer to the normal distribution and better satisfy the assumption of residual equal variance in linear regression models, thus reducing the influence of heteroscedasticity and further improving the reliability of the empirical results (Silva and Tenreyro, 2006). The study checked correlations between the key variables. Table 3 shows Pearson correlation coefficients for each pair of variables. Considering that the high correlation among independent variables may cause a multicollinearity problem, resulting in distortion of the model estimation results, this paper conducted the variance inflation factor (VIF) test. Table 4 shows that the VIF values of all variables are < 10, with a mean of < 5, indicating that the model does not present serious multicollinearity problems.

To investigate the influence of PEC on RLE, this paper used fixed-effect model for the benchmark regression test, and the results were reported in Table 5. Column (1) included only PEC as the predictor, and the other columns added control variables and lag variables of the PEC. The results of the first and second columns showed that the coefficient of PEC was positive and significant, suggesting that PEC had a soft environmental constraint effect that could enhance the RLE. Research hypothesis 1 was supported. Furthermore, the third and fourth columns considered the lagged effects of PEC on the RLE. The analysis demonstrated that both the first- and second-order lag variables of public environment concerns were positive predictors of RLE. Regarding the coefficients of the control variables, the ln(FDI) was a significant and positive predictor, which supported the “pollution halo” hypothesis of FDI; More foreign investment was associated with a better living environment. The level of industrialization, government intervention, and economic development had negative effects on the RLE.

The analyses so far show that PEC likely influences the RLE through government attention. Mediating effect analysis is an important method of multivariate analysis, which is used to explain how independent variables affect dependent variables through mediating variables, with the purpose of revealing the complex action path and internal mechanism between variables. If an independent variable (X) influences a dependent variable (Y) through a third variable (M), then M is considered a mediator in the relationship between X and Y (Wen et al., 2004). Following the mediation effect analysis method introduced by Wen et al. (2004), this research developed Equation (9) and (8) based on Equation (7) to test whether government environmental attention mediates the association between PEC and the RLE.

Figure 2 illustrates the procedure for testing mediation effects as outlined by Wen et al. (2004). First: Without including the mediator, Equation (7) examines the influence of the PEC on the RLE. If the coefficient β₁ is not significant, it suggests that PEC do not improve the RLE, and the analysis should cease; otherwise, the analysis continues. Following the results shown in Table 3, PEC positively affects the RLE, warranting further analysis. Second: Equation (8) examines the effect of PEC on government attention, yielding the estimated coefficient α₁. Third: Building on Equation (7), government environmental attention is added to form Equation (9) to investigate the association among PEC, government environmental attention, and the RLE, obtaining the estimated coefficients β₃ and β₄. If both coefficients β₄ and α₁ are significantly positive, it implies the existence of a mediating effect of government attention. If coefficient β₃ is not significant, it implies that government attention exerts a complete mediating effect; otherwise, it suggests a partial mediating effect. According to the results of the mechanism examination (see Table 8), both the coefficients of PEC and government attention are significant, and the relationship is positive, showing a partial mediating effect of government attention in the positive association between PEC and the RLE. Hence, Research Hypothesis 2 is validated. Consistent with the theoretical analysis, PEC exerts external influence on the allocation of government environmental attention, enhancing the government's focus, importance, and resource investment in the RLE, thus improving it.