Rural household waste sorting and its social, environmental and economic impacts: implications for sustainable governance

Purpose This study addresses the “last-mile” governance dilemma of rural household waste sorting in developing countries, where participation is high but effectiveness is low. Guided by public-goods, externality, and collective-action theories, it aims to investigate two main questions: (1) what are the multidimensional (environmental, health, economic and social) welfare effects of rural waste sorting and (2) how do these effects vary across public and private domains, challenging the assumption that the benefits of waste sorting are homogeneous. Design/methodology/approach The study utilizes survey data collected from 939 rural households across two distinct regions in China: Jiangsu and Gansu. To rigorous analyze the data while accounting for potential self-selection bias and unobserved heterogeneity among participants, the authors employ a sophisticated econometric approach. Specifically, they combine Propensity Score Matching (PSM) with an endogenous switching regression (ESR) model. This methodology ensures robust causal inference regarding the multidimensional welfare effects and allows for a detailed comparison of outcomes between public and private behavioral domains. Findings Waste sorting yields positive but uneven welfare effects, following a gradient of social > environmental > health > economic benefits. This indicates that current participation is primarily driven by social norms rather than economic or health returns. Crucially, situational context matters: informal normative pressure in the public domain inflates perceived environmental benefits. Robustness checks reveal a “numerical illusion,” showing that the environmental effects realized through externally mobilized participation may be overestimated by up to 25.4%, highlighting a gap between reported and actual governance success. Originality/value This study advances behavioral and environmental governance research by dismantling the assumption of homogeneous welfare effects in waste sorting. Academically, it is among the first to quantify domain-specific, multidimensional outcomes and expose a significant measurement bias (“numerical illusion”) in public interventions. For policymakers and managers, it provides a crucial diagnostic framework to shift rural governance from arbitrary, campaign-driven participation targets toward outcome-based, scenario-adapted incentives that tie behavioral outcomes to tangible economic and health returns.