How livelihood capital shapes poverty return risk? Evidence from rural monitoring households in Western Hunan, China

Introduction Against the background of post-poverty alleviation governance, livelihood capital acts as the core endogenous resource determining rural monitored households’ capacity to avert poverty relapse risks. Existing literature fails to fully differentiate three official categories of at-risk rural households, and lacks systematic evidence on heterogeneous mechanisms of multi-dimensional livelihood capital affecting poverty relapse prevention. This study fills such research gaps and provides empirical basis for classified targeted anti-poverty governance. Methods Taking 19,326 monitored households across eight contiguous poverty-alleviation counties in Western Hunan as samples, including 5,896 poverty-relieved unstable households, 8,321 marginal vulnerable households, and 5,109 households with sudden severe hardships, this paper establishes a five-dimensional livelihood capital evaluation system covering natural, physical, human, social and financial capital. The entropy-weighted TOPSIS model is adopted to quantify livelihood capital endowment gaps, and four groups of binary Logistic regression models (full sample plus three sub-samples) are constructed to identify differentiated influencing factors of poverty relapse risk elimination. Results The results show that 81.77% of all sampled households have eliminated relapse risks; marginal vulnerable households reach the highest risk elimination rate of 97.03%, while households with sudden severe hardships record the lowest rate at 69.90%. Households free of poverty-returning risks possess higher comprehensive livelihood capital scores, with financial capital as their dominant advantage. Livelihood capital exerts obvious heterogeneous effects across groups: financial capital dominates risk elimination in the full sample; human capital serves as the core driver for poverty-relieved unstable households; financial capital determines risk resistance for households with sudden severe hardships; social capital generates significantly positive effects on marginal vulnerable groups. Theoretically, this paper subdivides monitored households into three official types and enriches the empirical evidence of livelihood capital’s poverty reduction effects, expanding the application boundary of the sustainable livelihood framework in mountainous anti-poverty research. Practically, it provides a data-driven risk identification tool for regional dynamic poverty monitoring. Conclusion Local governments should formulate differentiated assistance policies matching livelihood capital disadvantages of different household groups. Activating multi-dimensional livelihood capital endowments can effectively strengthen rural households’ risk resistance, consolidate long-term poverty alleviation achievements, and fundamentally curb repeated poverty among rural monitored households.