Collective action improves elite-driven governance in rural development within China

The study addresses how collective action mediates the relationship between rural elites and rural development, and under what conditions it can mitigate elite capture. Despite improvements in rural living standards, significant inequalities persist, making rural development key to achieving SDGs. Elites—actors with disproportionate governing or economic power—often drive local decision-making. While elite leadership can spur development, it can also lead to elite capture, undermining equitable distribution of public resources. Collective action is posited as a mechanism that can influence elite-driven governance by engaging both elites and non-elites in decision-making and service provision. In China, two prominent collective action forms—village collectives (public affairs) and farmer cooperatives (private agricultural business)—offer contrasting institutional logics and participation structures. The study distinguishes between governing elites (political authority) and economic elites (wealth/influence independent of elections) and sets two goals: to clarify how village collectives and farmer cooperatives mediate links between governing/economic elites and rural development, and to identify conditions under which collective action can control elite capture.

Prior research shows mixed effects of elite-driven governance: some elite leaders catalyze economic growth and service provision, while others engage in elite capture, corrupting planning and resource allocation. Collective action has been shown to shape participation and accountability, potentially limiting elite capture depending on project design, community context, and broader political-economic conditions. In China, village collectives manage public goods and collectively owned property, whereas farmer cooperatives address market failures for smallholders by improving access to information, finance, and buyers. Theoretical debates on collective action (Olson’s rationale for authority intervention vs. Ostrom’s self-governance) frame whether and how authority supports or undermines collective management. This study extends these debates by empirically testing the mediating role of collective action between distinct elite types and development outcomes, and by exploring mechanisms—trust, reciprocity, reputation, and transparency—that may constrain elite capture.

Study area and data: The study focuses on Jiangsu Province, China, a developed region with marked north–south heterogeneity in rural collective economies. Quantitative data come from the 2017 village-level statistical system (“village card”) for 604 administrative villages across seven counties (Changshu, Hai’an, Jingjiang, Shuyang, Sihong, Tongshan, Yixing), covering over 20 indicators on geography, population, agriculture, infrastructure, elites, and collective action. Qualitative data: In 2021, the authors conducted in-depth interviews and questionnaires in eight stratified-randomly selected villages from three counties: Jiangxiang and Xiajia (Changshu, south), Wangzhuang and Dajue (Jingjiang, central), and Zhangdagou, Xingyi, Xiezhuang, and Dongwangmiao (Shuyang, north). Respondents included 8 governing elites (standing Communist Party committee members), 29 economic elites (top ~5% income, respected), and 197 non-elite farmers. Supplementary materials from government sites and village archives documented socioeconomics, governance rules, and histories. Variables: Dependent variable—rural development measured by farmers’ per capita net income (noted limitations as incomplete reflection of social conditions; qualitative data complement interpretation). Independent variables—economic elites (proxied by the ratio of households to number of new professional farmers and self-employed entrepreneurs) and governing elites (proxied by the salary of the standing member of the local Communist Party or alternatively by the head of the village leadership committee, as performance-linked pay reflects leadership). Mediators—collective action capacity measured as: (1) village collectives by per capita business income of village collective (outcome-based measure due to universal membership), and (2) farmer cooperatives by participation rate (households participating/total households). Controls—11 indicators capturing community attributes and agriculture: township government residence, landform/topography, total area, resident population, population density, proportion permanent-to-registered population, construction land for village collective, arable land, land transfer, facility agriculture, and drainage/irrigation stations. Statistical analysis: Continuous variables were min–max normalized to [0,1]. Multiple OLS regressions (four models) evaluated associations between rural development and (i) controls; (ii) collective action; (iii) elites; and (iv) both collective action and elites with controls and regional fixed effects (north/central/south Jiangsu). Path analysis modeled direct and indirect (mediated by village collectives and farmer cooperatives) effects for two elite types, estimating paths a (elite→mediator), b (mediator→outcome), c' (direct), with significance testing. Multiple mediation effects were tested with 5,000 bootstrap samples (percentile, BC, BCa CIs). Robustness checks used the alternative governing elite indicator (head of village leadership committee) in OLS, mediation, and path models. Qualitative analysis triangulated mechanisms from interviews, surveys, and documents, focusing on resource reallocation/use efficiency, participation, transparency, and institutional innovations (e.g., points-based reward systems, public sponsorship postings).

• Collective action as mediator: Both village collectives and farmer cooperatives are positively associated with rural development, with village collectives having a larger standardized effect than farmer cooperatives (OLS model 4: village collectives coefficient 0.3834, p<0.01; farmer cooperatives 0.1539, p<0.01).
• Elite effects: Governing elites have a significant positive association with rural development (OLS model 4: 0.2891, p<0.01). Economic elites show no significant direct association in OLS.
• Path analysis (economic elites): No significant direct effect on rural development, but significant indirect effects via mediators: EE→Village collectives a=0.0176 (p<0.05), VC→RD b=0.4942 (p<0.01), product 0.0087; EE→Farmer cooperatives a=0.0933 (p<0.05), FC→RD b=0.1662 (p<0.01), product 0.0155. Total indirect effect 0.0242 (bootstrap SE 0.0092, Z=2.63; CIs exclude zero).
• Path analysis (governing elites, standing Party member): Significant direct effect c'=0.2893 (p<0.01). Indirect effects via VC: a=0.1030 (p<0.01), b=0.3859 (p<0.01), product 0.0397; via FC: a=0.4932 (p<0.01), b=0.1542 (p<0.01), product 0.0761. Total effect 0.4051 (sum of direct and indirect). Bootstrap mediation confirms significance (TOTAL indirect 0.1158, SE 0.0311, Z=3.72).
• Robustness: Using the head of village leadership committee as the governing-elite indicator yields consistent OLS and mediation/path results (e.g., OLS coefficients ~0.2608–0.3628 for governing elites; total indirect 0.1045 with both mediators significant).
• Mechanisms from case studies: Collective action increases resource reallocation and use efficiency (e.g., mobilizing cleanup and reclaiming idle land for leasing; enterprise attraction; job creation). Points-based reward systems and publicly posted sponsorships leverage reputation effects, fostering reciprocity and participation while discouraging elite capture. Survey evidence shows majorities perceive income gains and enhanced social capital (e.g., 28.83–62.79% report stronger trust/ties across action types; nearly 30% report more equitable society). Higher perceived participation and influence correlate with higher village collective income (e.g., Jiangxiang and Xiajia).

Findings address the research questions by showing that collective action mediates the relationship between elites and rural development and that its effectiveness depends on institutional design, participation, and transparency. Governing elites exert strong direct and indirect effects, catalyzing and sustaining both public (village collectives) and private (farmer cooperatives) forms of collective action, thereby justifying the constructive role of authority. Economic elites mainly contribute indirectly by enhancing knowledge diffusion, integrating smallholders into commercial networks, and providing financial support for public goods. Designing collective action to leverage reputation, trust, and reciprocity norms—through public posting of contributions and points-based reward systems—promotes elite and non-elite participation and constrains elite capture. Democratization and transparency strengthen social capital and accountability, improving governance outcomes. The results integrate Olson’s argument for authority intervention with Ostrom’s emphasis on self-governance by demonstrating that authoritative leadership combined with participatory, transparent institutions yields better rural development and mitigates elite capture.

Collective action is an effective governance tool that mediates the influence of elites on rural development. Village collectives and farmer cooperatives, when designed to foster participation, transparency, and reputation-driven reciprocity, mobilize social and natural resources, improve smallholder livelihoods, and enhance agricultural productivity. Governing elites play a more decisive role than economic elites in organizing, facilitating, and contributing to collective action. Innovative, low-cost institutional designs—such as points-based rewards and public recognition of contributions—promote participation and deter elite capture. Policy should enable authority intervention that supports bottom-up collective action while ensuring democratic participation and transparency. Future research should examine how governing and economic elites interact across different network structures and how collective action scales across regions and levels of governance.

• Measurement constraints: At fine spatial scales, comparable indicators are limited. Village collective capacity is proxied by per capita business income rather than participation (universal membership), and farmer cooperative capacity by participation rate. Economic elite leadership is proxied by prevalence rather than direct leadership quality due to data constraints.
• Outcome proxy: Per capita net income captures economic development but not broader social outcomes; qualitative data partially mitigate this limitation.
• Causal inference: Observational design with OLS and path analysis implies associations rather than definitive causality; unobserved confounding may remain despite controls and regional fixed effects.
• Generalizability: Findings are from Jiangsu Province, which is relatively developed and heterogeneous; results may not generalize to other provinces or countries without adaptation.
• Scope: The study does not analyze cases lacking elites, limiting conclusions about contexts without elite presence.