Social trust, social capital, and subjective well-being of rural residents: micro-empirical evidence based on the Chinese General Social Survey (CGSS)

The study examines how social trust influences subjective well-being (SWB) among rural Chinese residents and whether social capital mediates this relationship. Motivated by evidence that GDP growth alone does not ensure happiness, the paper emphasizes non-market determinants of life quality such as trust and social capital. China’s pronounced rural–urban divide and deficiencies in rural education, healthcare, and social security may shape trust, cooperation, and happiness. Rural China provides a pertinent context due to dense, family-based networks and the prominent role of informal institutions. The research focuses on rural residents, a comparatively understudied group, and asks: Does social trust increase rural residents’ happiness, and does social capital mediate this effect? The work aims to inform policy for rural, developing, and non-Western contexts.

The paper integrates theory and prior evidence on trust, social capital, and SWB. Social trust—reliance on most people in society—reduces social anxiety, fosters reciprocity, and lowers transaction costs, potentially enhancing well-being (Putnam, Fukuyama). Empirical findings are mixed across contexts: positive associations in many settings (e.g., Asian countries, Japan, workplaces), with some null results in OECD panels. In rural China, higher trust in strangers and trust between villagers and cadres have been linked to greater SWB, and political trust influences subjective evaluations. Hypothesis 1: Social trust increases the happiness of Chinese rural residents. Social capital, conceptualized by Bourdieu, Coleman, and Putnam as resources embedded in social networks and civic/political participation, is associated with better health, coping, and SWB, with heterogeneous effects by gender, age, and urban-rural status. Chinese studies generally find positive links between social capital and rural happiness. Hypothesis 2: Social capital increases the happiness of Chinese rural residents. As a mediator, social capital may transmit trust’s effects on SWB through improved mental health (social support, self-esteem) and economic benefits (employment, income, reduced poverty and credit constraints). Trust fosters cohesive networks and civic engagement, supporting social capital formation. Hypothesis 3: Trust has a significant positive impact on the happiness of rural residents through social capital. Research gaps include limited focus on rural residents, overemphasis on informal ties, and limited incorporation of social capital as a mediator. This study uses nationally representative data, models mediation, and emphasizes social networks alongside formal organizational ties.

Data: The study uses nationally representative CGSS data from 2012, 2013, and 2015, employing stratified four-stage unequal probability sampling across 32 provincial-level regions. Urban samples and responses of “don’t know”, “inapplicable”, or “refused” were excluded, yielding 10,014 valid rural observations. Measures: Happiness (SWB) follows the WVS item “Taking all things together, would you say you are happy?”, coded on a 5-point scale (very unhappy = 1 to very happy = 5). Trust is constructed by summing responses to two questions: generalized trust (“Most people can be trusted”) and a recoded item on opportunism (“Others will take advantage of you if you are not careful,” recoded so higher values indicate less agreement with opportunism). Social capital is measured using social and political relations. Social relations: summed frequencies of friends gatherings, relatives gatherings, watching sports games, and watching concerts. Political relations: summed indicators for Communist Party membership (1/0), trade union membership (1/0), religious belief (1/0), and participation in village elections (1/0). An entropy evaluation method (EEM) is used to weight and aggregate social and political relations into a social capital index. Controls include age, gender (1=male), ethnicity (1=Han), log income (last year), marital status (1=married/cohabiting), employment (1=employed), social insurance participation, self-rated health (1–5), education (13-category scale), family assets (housing count, car ownership, investment), and regional dummies (East, Middle, North-east; West as reference). Empirical strategy: A mediation framework (Baron and Kenny; Wen et al.) is estimated via ordered probit for the ordinal happiness outcome, with three equations: (1) happiness on trust and controls (total effect β1), (2) social capital on trust and controls (γ1), and (3) happiness on trust, social capital, and controls (direct effect δ1 and mediator effect δ2). Mediation is assessed by significance of paths, Sobel tests, and bootstrap confidence intervals for indirect effects (γ1*δ2). Heterogeneity analyses split samples by age group (<30; 30–59; ≥60) and by provincial marketization (Fan et al., 2017) into high versus low market competition regions. Robustness checks replace trust with “trust in strangers” and the outcome with perceived social fairness (both 1–5 scales); due to differing item formats across years, robustness uses CGSS 2013 and 2015.

Descriptives: About 14.73% reported very happy, 60.62% somewhat happy, 15.83% no opinion, 7.38% somewhat unhappy, and 1.448% very unhappy; mean happiness = 3.798. Empirical results (Table 3): - Trust positively predicts happiness (Model 1: β = 0.186, p < 0.01). - Trust positively predicts social capital (Model 2: γ1 = 0.022, p < 0.05). - In the full model, both trust (δ1 ≈ 0.185, p < 0.01) and social capital (δ2 = 0.055, p < 0.01) positively predict happiness, indicating partial mediation. - Sobel test for mediation is significant (0.0011, p < 0.05). Bootstrap confirms a small but significant indirect effect: r(ind_eff) = 0.00113532 (95% CI percentile: 0.0000587, 0.0023684), with a direct effect r(dir_eff) = 0.15116851 (95% CI percentile: 0.1311723, 0.1703598). Controls: Higher age, income, health, being married, owning a car, more social insurance, and more houses are associated with higher happiness; being male and being employed are associated with lower happiness. Regional differences show western regions relatively happier; East and North-east dummies are generally negative relative to West. Heterogeneity (Table 5): - By age: Trust is positively associated with happiness across all age groups. Social capital significantly predicts happiness for ages 30–59 and ≥60, but not for <30; mediation is significant only for ages 30–59 (partial mediation). - By marketization: Trust and social capital are positive in both high and low competition regions, with significant mediation in high-market competition areas only (Sobel p < 0.1). Robustness (Table 6): Using 2013/2015 data, “trust in strangers” and social capital positively predict perceived social fairness (p < 0.01). Social capital retains a significant partial mediating role (social capital coefficient ≈ 0.418; Sobel test 0.00338***). Overall, results support H1 (trust→happiness), H2 (social capital→happiness), and H3 (trust→happiness via social capital), with stronger mediation among adults ≥30 and in more marketized regions.

Findings indicate that social trust enhances rural residents’ SWB both directly and indirectly through social capital. Trust facilitates cooperative behavior, lowers perceived social risks, and reinforces social cohesion, which fosters denser networks and civic participation. Social capital then supports well-being via two channels: (1) psychosocial support and improved mental health (emotional communication, positive affect, self-esteem) and (2) economic benefits (access to jobs, income growth, reduced poverty and credit constraints), thereby mitigating stress. Patterns in control variables align with global happiness research, reinforcing the credibility of results (e.g., positive effects of health, marriage, income). Regional differences may reflect government investments and varying baselines of development; western regions show higher life satisfaction, potentially due to improvements from a lower baseline and policy focus. The heterogeneity analyses suggest social capital becomes more functional as individuals age and as regions become more marketized, where networks can be leveraged more effectively for economic and social opportunities. Robustness checks using perceived fairness corroborate the central mechanisms, indicating that generalized trust and social capital also bolster perceptions of social equity, a correlate of well-being. The study discusses broader relevance for developing countries with similar rural social structures, while noting that China’s unique institutional and infrastructural context may condition generalizability.

The study demonstrates that social trust significantly increases the happiness of rural Chinese residents and that social capital partially mediates this relationship. Using large, nationally representative CGSS samples (2012, 2013, 2015), ordered probit estimations, and mediation tests (Sobel and bootstrap), the findings consistently support three hypotheses: trust→happiness, social capital→happiness, and trust’s indirect effect via social capital. Mediation is pronounced for adults aged 30 and older and in provinces with higher market competition. Policy recommendations emphasize building environments that strengthen social networks and facilitate social interactions: - Prioritize support for disadvantaged groups (low income, low education, elderly, women, children) through regular outreach by village officials to encourage participation in social activities and provide material and emotional support. - Organize meaningful village council and community activities (e.g., information sharing on jobs by migrant workers, technical exchanges by large-scale producers, dissemination and scrutiny of welfare policies), and encourage neighborhood exchanges and community events (e.g., group dining, Rural Basketball Association). Such initiatives can enhance social capital, thereby amplifying the positive impact of trust on well-being. Future research can further explore causal identification and the roles of different forms of social capital across contexts.

Generalizability may be context-dependent: the paper notes that China’s distinctive institutional and infrastructural conditions mean the findings are most applicable to developing countries with similar stability and infrastructure. Robustness tests relied on alternative indicators available only in 2013 and 2015 due to measurement differences (e.g., trust-in-strangers scale), resulting in different sample sizes for those analyses.