How can rural e-commerce promote green production on family farms in China? Evidence from Hunan Province

Family farms are crucial for promoting the green transformation of China's agriculture, but their green production (GP) level remains low under the traditional sales model. Rural e-commerce (RE), facilitated by the “Internet +”, has the potential to overcome this challenge; however, its theoretical and empirical efficacy remains unproven. This article aims to verify the impact of RE on family farms' GP. To do so, we establish a theoretical framework, and use endogenous switching regression (ESR) to investigate the impact based on survey data collected from 836 farms in Hunan province, China. We find that participation in RE can significantly promote GP on family farms, with an average treatment effect of 1.175, indicating a 24.05 % improvement in GP. The finding remains robust after a series of tests. The analysis of heterogeneity shows RE is more effective in improving the GP of family farms with a social e-commerce model, large-scale operation, growing food crops, and cooperative membership. It is also more conducive to promoting GP in family farms without product certification, atypical demonstration, and low land tenure security. Further mechanism analysis results show the impact is mainly driven through five mechanisms: expected price incentive, transaction cost saving, mitigation of credit constraint, convenient information access, and quality reputation guarantee. These results suggest that the government enhances investment in rural e-commerce infrastructure, actively promote social e-commerce, develop differentiated support policies, and facilitate practical channels for e-commerce to drive green transformation.