The impact of rural living environment improvement programs on the subjective well-being of rural residents in China

Subjective well-being (SWB) is widely used to assess policy and public goods performance and is central to the UN SDG goal on well-being. Despite extensive research on determinants of SWB (income, education, age, health, employment, marital status, social security, social media, ecosystem services, biodiversity, city size, inequality), fewer studies analyze how environmental governance programs influence SWB, particularly in rural contexts of developing and non-democratic countries. China faces severe rural environmental pollution contributing substantially to national water pollution loads. In response, the government launched the Rural Living Environment Improvement (RLEI) program—an unprecedented, large-scale initiative focusing on waste treatment, wastewater disposal, toilet upgrades, and village appearance, with sustained policy support and investment. This study asks: (1) Does RLEI improve rural residents' SWB? (2) What is its monetary value? (3) Are effects heterogeneous across contexts and populations? (4) Through what mechanisms does RLEI affect SWB? The paper posits Hypothesis 1: RLEI positively affects SWB; Hypothesis 2: effects are mediated by income, consumption expenditure, and health; Hypothesis 3: rural waste programs yield larger SWB gains than sewage and livestock manure programs. The study contributes by providing rigorous quantitative evidence from China (a large developing, non-democratic context), elucidating mechanisms, and valuing RLEI via the life satisfaction approach.

Prior literature shows SWB depends on individual factors (income, education, age, health, employment, marital status) and contextual factors (social security, social media, ecosystem services, biodiversity, city size, inequality). Environmental pollution is consistently associated with lower SWB (e.g., air pollution reduces happiness and life satisfaction). Yet, empirical evidence on environmental governance programs’ impacts on SWB—especially in rural settings—is limited. Given market failures and policy centrality in developing/non-democratic settings, governance programs may be pivotal for SWB. China’s RLEI policy suite since 2014 targets rural sewage, waste, and livestock manure with rising investments, which theory and evidence suggest should enhance SWB through improved environmental quality (health and psychological stress relief), satisfaction from participation and reduced inequality perceptions, and potential economic and social benefits. This study fills the gap by empirically testing RLEI’s effects on SWB and comparing different program types.

Data: China Labor-Force Dynamics Survey (CLDS) 2016 and 2018 waves (nationally representative across 29 provincial-level regions). After cleaning and integrating, the study analyzes an unbalanced panel of 3747 rural residents across 20 provincial-level regions. Variables: - Dependent variable (SWB): Response to “Overall, do you think you are happy with your life?” on a 1–5 Likert scale (1 very unhappy to 5 very happy). - Independent variable (RLEI): Village-level index reflecting implementation of three components: sewage treatment, waste disposal, and livestock manure treatment. Each is coded 0/1 based on CLDS village questions, summed to 0–3; to mitigate extremes, add 1 and take natural log (LnRLEI). Component-specific logs (LnRural sewage, LnRural waste, LnLivestock manure) are similarly defined. - Controls: gender, party membership, marital status, household income (log), village transportation (hardened road availability), and village expenditure on public goods (log). Year and village fixed effects are included. - Mediators: income (agricultural and non-agricultural income, logs), consumption expenditure (household travel; gifts and gratuities, logs), health (physical: hospitalization since last July, coded 1=No; mental: frequency of sadness, higher=better, reverse-coded per description). Models: 1) Direct effect: Ordered Probit model regressing SWB on LnRLEI, controls, village and year dummies. Marginal effects computed for SWB categories. 2) Mediation: Interaction model includes LnRLEI×Mediator and mediator main effects in Ordered Probit to test whether income, consumption, and health mediate RLEI’s impact. 3) Program-type effects: Separate Ordered Probit models for sewage, waste, and livestock manure program measures (and their mediation interactions). 4) Monetary valuation (Life Satisfaction Approach, LSA): Compute marginal effects of RLEI and income on SWB at SWB=4 and derive money-metric valuation as the ratio of marginal effects. 5) Selection bias correction (PSM): Estimate propensity scores (logit), verify overlap, perform balance tests, and compute ATT using nearest-neighbor, radius, and kernel matching. 6) Endogeneity (IV-Oprobit): Use average RLEI of other villages in the same province (excluding own village) as instrument (lnAvg RLEI). First stage regresses LnRLEI on instrument and controls; second stage Ordered Probit for SWB on fitted RLEI. Report first-stage F and test of endogeneity. Robustness checks: Alternative estimators (Ordered Logit, OLS), recoding SWB to binary (under/over-reporting adjustments), replacing SWB with life satisfaction measure, and excluding migrant households (planned/already migrated). Heterogeneity analyses: Split samples by region (Eastern, Midwestern, Northeastern), age (junior and old-aged: <18 or >60 vs. 18–59), education (low vs. high), and working status (working vs. non-working).

- RLEI significantly increases rural residents’ SWB. Ordered Probit coefficient on LnRLEI: 0.154 (p<0.05). Marginal effects (per one SD increase in LnRLEI=0.315 from mean 1.068): probability changes are −0.189% (very unhappy), −0.378% (relatively unhappy), −0.976% (so-so), +0.409% (relatively happy), +1.103% (very happy). - Mechanisms: Interaction terms indicate RLEI enhances SWB via increases in income (agricultural and non-agricultural), higher consumption expenditure (travel; gifts/gratuities), and better health (physical and mental). Reported mediation coefficients are positive and significant across these channels (e.g., Table 6 shows significance across mediators). - Program-type comparison: Rural waste program exhibits the strongest positive effect on SWB relative to sewage and livestock manure programs (Table 7). Mediation paths are also stronger for rural waste (Table 8). - Monetary valuation (LSA): At SWB=4, marginal effect of RLEI on SWB is 0.013; household income marginal effect is 0.010. The paper concludes a 1% increase in RLEI is equivalent to 130% of the effect of household income on SWB. By component: rural sewage ~1.20×, rural waste ~3.67×, livestock manure ~1.10× the effect of household income on SWB. - Selection and endogeneity: PSM ATT estimates show significant positive impacts of RLEI on SWB across matching methods (ATT ≈ 0.632–0.671, p<0.01). IV-Oprobit: instrument strongly predicts RLEI (first-stage F=213.93); second-stage shows RLEI positively affects SWB (e.g., LnRLEI coefficient ≈0.471, p<0.01), supporting a causal interpretation. - Robustness: Results hold under Ordered Logit and OLS, with binary recoding of SWB, substituting life satisfaction as outcome, and excluding migrant households. - Heterogeneity: Effects are larger and significant in Midwestern and Northeastern regions; insignificant in Eastern region. Stronger for junior and old-aged vs. young/middle-aged; significant for low-educated but not high-educated; positive for both working and non-working, with a larger effect among working residents.

The findings directly address the research questions: RLEI improves SWB among rural Chinese residents and does so primarily through higher incomes (agricultural and non-agricultural), increased consumption (travel and social expenditures), and improved physical and mental health. The rural waste program’s broader daily labor needs, recycling benefits, and multiple health co-benefits likely explain its larger impact compared with sewage and livestock manure programs. The monetary valuation indicates substantial welfare gains from RLEI relative to household income effects, underscoring the high social value of environmental governance in rural settings. These results are significant for the literature by demonstrating that large-scale environmental governance programs in a developing, non-democratic context can materially enhance SWB, complementing evidence from developed democracies. Policy-wise, they support continued and targeted investment in RLEI, particularly in regions and populations where gains are greatest, and suggest prioritizing interventions with broader employment and health spillovers (e.g., rural waste management).

This paper provides rigorous quantitative evidence that China’s Rural Living Environment Improvement (RLEI) program significantly enhances rural residents’ subjective well-being. Mechanistically, gains operate through increased income, higher consumption, and improved health, with the rural waste program yielding the strongest impacts. Monetary valuation via the life satisfaction approach shows that RLEI’s SWB benefits are comparable to or exceed those associated with household income, with component-specific values of approximately 1.2× (sewage), 3.67× (waste), and 1.1× (livestock manure) relative to income effects. Heterogeneity analyses indicate larger benefits in Midwestern and Northeastern regions, and among junior/old-aged, low-educated, and working residents. Policy recommendations include sustained and expanded RLEI implementation, especially in under-served regions; strengthening weaker components (sewage and livestock manure) through infrastructure and mechanization; and tailoring efforts to populations most affected by local environmental quality. Future research should pursue stronger causal identification using longer panel data (e.g., difference-in-differences) or randomized interventions to validate and refine these findings.

- Data limitations: Only two CLDS waves (2016, 2018) include the requisite RLEI variables, limiting causal identification over time. - Methodological constraints: Despite PSM and IV-Oprobit addressing selection and endogeneity, residual confounding and reverse causality cannot be completely ruled out. - Measurement: SWB is self-reported and subject to reporting biases; village-level RLEI measures may mask within-village heterogeneity. - External validity: Results pertain to Chinese rural contexts and program designs; generalization to other countries or governance regimes should be cautious. Future work could leverage longer longitudinal data with difference-in-differences designs or field experiments (randomized interventions) to establish causality more firmly and to explore additional mediators and spillovers.