How does social capital promote consumer participation in food safety governance? Evidence from online food consumers in China

The study investigates how social capital promotes consumer participation in food safety governance and through which mechanisms, addressing the collective action dilemma—particularly free-riding—that discourages individual engagement. In China’s rapidly evolving online food marketplace, consumers often face information asymmetry and coordination challenges that hinder reporting, monitoring, and safeguarding behaviors. Grounded in social capital theory, the paper examines whether and how dimensions of social capital (trust, norms, networks, pro-social behaviors, and community cohesion) can lower participation costs, curb free-riding, and increase the likelihood and intensity of consumer participation. The authors hypothesize (Hypothesis 1) that social capital both offsets the negative effect of free-riding and directly promotes consumer participation, shifting engagement from traditional complaint/reporting toward broader, more integrated actions such as information sharing and direct supply of safe food.

The paper builds on the collective action literature that highlights free-riding as a core barrier to public participation in governance and commons management (e.g., Hardin, 1968; Ostrom, 1990). It references prior work (Su et al., 2022) identifying free-riding as a key inhibitor of consumer involvement in food safety governance and advances the argument that social capital can reduce participation costs and overcome such dilemmas. It also invokes the perspective that, unlike traditional commons that can be depleted, social capital may grow with use (e.g., Brondizio et al., 2009), offering a theoretical rationale for its role in sustaining collective action in food safety governance.

Design and data collection: Cross-sectional online survey of Chinese online food consumers conducted in August 2016 over one month. Initial N=1,286; after rigorous data quality screening, valid N=1,229 (effective rate 95.57%). The survey link was randomly distributed to online shoppers nationwide via an accessible e-commerce platform, capturing diverse regions (rural, fringe, second-tier cities, provincial capitals, first-tier cities), ages, education levels, and incomes. Quality control and exclusions: Questionnaires were excluded for (1) illogical subjective answers; (2) inconsistencies across items; (3) patterned responding; (4) unrealistically short completion times (<10 minutes, given average 27.3 minutes); (5) incorrect answers to trap questions. Sample profile highlights: 53.86% female; 69.57% married; education skewed high (61.76% bachelor’s, 8.46% master’s+); age concentrated 20–39 (80.56%); broad geographic spread; income distribution centered on middle-income brackets. Measures (Table 2): Dependent variable PR (participation rate) constructed by summing three 5-point Likert items: making complaints/reporting calls, claiming compensation, exposing unsafe food issues on internet platforms (Weibo/WeChat). Higher scores indicate greater participation (mean 8.075, SD 3.127). Core independents: FR (free-riding inclination: 1 very unfavorable to 5 very favorable; mean 2.548, SD 1.144); PC (policy context: 1 area with policies reducing participation cost; mean 0.156, SD 0.363). Social capital dimensions and community cohesion: TRUST (binary ability to borrow money from others in community), NETWORKS (acquaintance with community members, 1–5), NORMS (perceived honesty norm, 1–5), SOB (self-reported pro-social actions to improve community, 1–5), CE (cooperative activities under shared values, binary). Controls: gender, age (years), marital status, education (ordered scale), household income (million yuan), family size, sleep sufficiency (1–5), alcohol drinking frequency (1–5), rural residence. Analytical strategy: Ordered regression (ordered probit) to estimate determinants of participation intensity (PR). Multiple models incrementally included social capital dimensions to assess their direct effects and their capacity to offset free-riding’s negative effect. To analyze participation pathways, the study identified four modes—traditional complaint/reporting, information sharing via social media, direct supply of safe food (e.g., participation in planting/production for own supply), and establishing relationships (buying from fixed stalls/sellers)—and enumerated 16 possible combinations. A multinomial logit (MNL) model was employed to model selection among these participation schemes; IIA considerations are discussed alongside comparison to MVP models, with MNL used in reported results.

- Social capital promotes consumer participation in food safety governance and mitigates the negative effect of free-riding, supporting Hypothesis 1. Ordered probit estimates show significant positive associations for several social capital dimensions with participation intensity. - Trust, norms, networks, and pro-social behavior (SOB) significantly increase participation; community cohesion (CE) shows no robust significant effect in the main models. Example coefficients (Model 3/8/9/10): TRUST ≈ 0.29–0.32 (p<0.01), NORMS ≈ 0.10–0.23 (p<0.05 or better), NETWORKS ≈ 0.12–0.22 (p<0.05 or better), SOB ≈ 0.14–0.18 (p<0.01). CE is small and non-significant. - A single dimension (e.g., trust alone) is insufficient to fully eliminate free-riding’s adverse effect; multiple social capital dimensions operating together are needed to neutralize the negative association of FR with participation. - Mechanisms: Social capital enhances participation via two primary pathways—(1) promoting the sharing of food safety information online and (2) promoting consumers’ direct supply of safe food (e.g., participation in production for self-supply). Empirically, participation has shifted from purely traditional complaint/reporting toward integrated modes centered on these two paths. - Controls: Marriage, education, family size, sufficient sleep, and drinking frequency are positively associated with participation; age is slightly negatively associated. Policy context (PC) and rural residence are generally not significant. Sample size across models is N=1,229; model fit measures (pseudo R2) are modest but jointly significant (p<0.001).

Findings confirm that social capital helps overcome the collective action dilemma in food safety governance by reducing participation costs and counteracting free-riding. Rather than merely increasing stated willingness, social capital is linked to concrete actions—complaints, information sharing, and direct safe food supply—thereby closing the gap between attitudes and behaviors. The multidimensional nature of social capital is critical: trust must be complemented by shared norms, dense networks, and pro-social engagement to neutralize free-riding’s dampening effect. Participation patterns are evolving. While traditional complaint/reporting remains relevant, consumers increasingly engage through information sharing and direct supply strategies, often combining modes. This integrated approach reflects the role of social capital in enabling credible information flows (reducing verification costs, mitigating misinformation/information overload) and in supporting cooperative arrangements that secure safer food supplies. These dynamics have policy relevance in China and other developing contexts where consumer engagement is needed to bolster food safety governance.

Social capital accumulation promotes consumer participation in food safety governance by inhibiting free-riding and lowering participation costs. Empirical evidence from 1,229 Chinese online food consumers shows that multiple dimensions of social capital—trust, norms, networks, and pro-social behavior—significantly increase participation, while community cohesion alone is not a reliable driver. Social capital channels participation through two key paths: information sharing and direct supply of safe food, marking a shift from a sole emphasis on complaint/reporting toward more comprehensive, integrated modes of engagement. Contributions include extending prior research on the collective action dilemma in food safety governance by demonstrating the mechanisms through which social capital fosters participation. Conceptually, treating social capital as a ‘commons’ that grows with use offers an explanation for its capacity to sustain collective action. Future research could further unpack organizational and institutional mechanisms that translate social capital into specific participatory behaviors across diverse contexts, and examine how digital platforms and community structures amplify or constrain these effects.