Subjective well-being (SWB), a measure of individual happiness, is a key indicator of government policy effectiveness. While numerous studies explore factors influencing SWB, including socio-demographic and contextual factors, research on the impact of environmental governance programs, particularly in developing and non-democratic societies, is limited. This study addresses this gap by examining the effects of China's RLEI program, the largest rural environmental governance program in history, on the SWB of its 600 million rural residents. The research questions are: Does RLEI improve rural residents' SWB? What is its monetary value? Are the effects heterogeneous? What are the mechanisms? The study contributes to the literature by providing the first rigorous quantitative estimation of rural environmental governance programs' impact on SWB in China, extending the understanding of environmental governance in non-democratic settings, exploring the mechanisms through which RLEI influences SWB, and estimating the monetary value of RLEI using the life satisfaction approach.

Existing literature confirms that SWB is influenced by socio-demographic factors (income, education, age, health, employment, marital status) and macro contextual factors (social security, social media use, ecosystem services, species diversity, city size, income gap). Studies have shown a negative association between environmental pollution and SWB. However, research on the positive impact of environmental governance programs on SWB, especially in rural areas of developing countries like China, is scarce. This study aims to fill this gap, especially considering the significant contribution of rural pollution to China's overall water pollution and its disproportionate impact on the welfare of low-income rural residents.

This study utilizes data from the China Labor-force Dynamics Survey (CLDS) from 2016 and 2018, encompassing 3747 rural residents across 29 provinces. The dependent variable is rural residents' SWB, measured by a self-reported happiness scale. The independent variable is village-level RLEI, constructed from CLDS questions on sewage treatment, waste disposal, and livestock manure treatment. Control variables include gender, party membership, marital status, household income, village transportation, and village expenditure on public goods. The Ordered Probit model is used for the baseline analysis to account for the ordinal nature of the SWB variable. To address potential selection bias in village participation in the RLEI program, the study employs propensity score matching (PSM) using logit regression to estimate propensity scores and then matching methods (nearest-neighbor, radius, and kernel matching) to calculate the average treatment effect on the treated (ATT). Further, to address potential endogeneity issues, an instrumental variable (IV) approach using the average RLEI score of other villages in the same province as an instrument is employed. The life satisfaction approach (LSA) is used to estimate the monetary value of RLEI. Heterogeneity analysis is conducted across regions (East, Midwest, Northeast), age groups (junior/old vs. young/middle-aged), education levels (low vs. high), and working status (working vs. non-working). Robustness checks are performed using alternative regression methods (Ordered Logit, OLS), different SWB scoring methods, life satisfaction as the dependent variable, and exclusion of migrant households.

The baseline Ordered Probit regression shows a significantly positive effect of RLEI on rural residents' SWB. The marginal effects indicate that a one standard deviation increase in RLEI increases the probability of higher SWB scores and decreases the probability of lower SWB scores. Indirect effects analysis reveals that RLEI's positive impact on SWB operates through improvements in income (both agricultural and non-agricultural), consumption expenditure (especially gifts and gratuities), and health (both physical and mental). Among the different RLEI components, the rural waste program has the largest positive impact on SWB. The LSA analysis estimates that the improvement in SWB due to RLEI is almost equivalent to the effect of household income, with the rural waste program exhibiting an even larger monetary value (3.67 times the effect of household income). PSM analysis, after addressing selection bias, consistently shows a significant positive impact of RLEI on SWB using different matching methods. The IV approach, addressing potential endogeneity, confirms the significant positive impact of RLEI on SWB. Robustness checks using alternative regression models, different SWB value assignments, life satisfaction as the dependent variable, and exclusion of migrant households all support the baseline finding. Heterogeneity analysis reveals that RLEI has a greater positive impact on SWB for junior and old-aged, low-educated, Midwestern and Northeastern, and working rural residents. In contrast, the effect is insignificant in the Eastern region, young and middle-aged group, high-educated group, and for non-working rural residents.

The findings strongly suggest that RLEI significantly contributes to improving the SWB of rural residents in China. The study highlights the importance of considering the multifaceted indirect effects of environmental governance programs on well-being, going beyond simply addressing environmental pollution. The substantial monetary value of RLEI, particularly the rural waste program, underscores the economic benefits of these investments. The heterogeneous effects across different groups highlight the need for targeted policy interventions that consider factors like age, education, region, and employment status. The study's findings have significant implications for policymakers and researchers, emphasizing the importance of investing in rural environmental improvements not just for environmental sustainability but also for enhancing the well-being of rural populations.

This study demonstrates a significant positive impact of China's RLEI program on rural residents' SWB, primarily through improved income, consumption, and health. The findings highlight the economic and social benefits of such investments and the need for targeted interventions based on regional and demographic factors. Future research could utilize a difference-in-differences approach with longer panel data or randomized controlled trials for more robust causal inference.

The study's use of two waves of CLDS data limits the establishment of a definitive causal relationship between RLEI and SWB. While PSM and IV methods mitigate selection bias and endogeneity, perfect causal identification remains challenging. The self-reported nature of SWB introduces potential measurement error. Future research should address these limitations by employing longer panel data, more sophisticated econometric techniques, or experimental approaches.