Climate risks and rural labor outflow are jointly challenging the sustainability of farm households in disaster-prone agricultural regions. Agricultural insurance, as an important risk management tool, may influence not only farm income but also farmers’ production and labor decisions. While previous studies on agricultural insurance have primarily focused on staple crops and income stabilization, the effects of labor allocation in facility-based intensive agriculture remain underexplored. This study addresses this gap by examining how participation in agricultural insurance affects farm household labor allocation in the context of greenhouse vegetable production. Using survey data collected from July to August 2025 from 247 greenhouse vegetable farmers in ShouGuang, China’s largest vegetable production base, we employ Probit and Two-Stage Least Squares (2SLS) models to address endogeneity and mediation analysis to explore the underlying mechanisms. The findings show that: (1) Agricultural insurance participation significantly increases the probability of farm households increasing agricultural labor input in the next planting season. (2) Agricultural insurance reduces farmers’ likelihood of exiting agricultural production through increased investment in agricultural technology, but this effect is constrained by the availability of investment capital. (3) The impact of agricultural insurance on labor allocation varies across different farmer groups, and is more pronounced among middle-aged and younger farmers with relatively small farm sizes, lower disaster risk, and higher levels of education. Several insurance-related policy recommendations are proposed to stabilize the agricultural labor force and mitigate rural labor decline—key dimensions of sustainable rural development. Given the single-region focus and sample size, the generalizability of these findings requires further validation in other agricultural contexts.
As agricultural marketization deepens, market risks have become a key factor affecting rural household livelihoods, while the trend toward differentiation among rural households has further exacerbated heterogeneity in resource allocation strategies and risk-coping capabilities across different groups. Based on field survey data from 763 rural households in the border areas of Yunnan Province, this study employs a combination of the Target MOTAD model and OLS regression to analyze the optimal production mix, resource utilization efficiency, and determinants of agricultural income for different types of farming households. Empirical results indicate that capital shortages are the primary factor constraining the full utilization of agricultural resources and farmers’ income growth; non-agricultural households, through proactive risk management, demonstrate significantly stronger risk resilience than agricultural households; and differences in the income structures of these two types of households drive them to adopt differentiated risk response strategies. Based on these findings, this study proposes the following optimization pathways: First, enhance rural households’ ability to accumulate capital to overcome the bottlenecks of resource idleness and income growth; second, promote moderate-scale agricultural operations to improve resource utilization efficiency; third, implement differentiated risk response strategies based on the heterogeneous characteristics of farming households. This research provides theoretical references and practical foundations for optimizing resource allocation among rural households in Yunnan’s border regions and enhancing agricultural risk-response capabilities.