关键词: 多维特别关注指数, 易返贫致贫程度, 影响因素, 中介效应

Abstract: This paper takes the requirement of the monitoring and supporting mechanism of preventing re-poverty as the starting point, and uses the micro-data of a province in the southwest of China as the sample to construct a Multidimensional Special Concern Index, which provides a scientific theoretical method for the effective identification of marginal households prone to poverty and unstable households out of poverty. After identifying the monitored persons, we analyze the potential poverty-causing factors with the Mlogit model, and analyze the mechanism of each influencing factor through mediating effect test. The empirical results shows that: Firstly, multidimensional special concern index can effectively meet the multi-dimensional needs of re-poverty prevention monitoring, and emphasize the importance of income dimension. Secondly, income deproiation, education deproiation and health deprivation are the keywork in the monitoring and supporting mechanism of preventing re-poverty, which should be paid more attention to. Thirdly, policy support for marginal households prone to poverty should focus on developing industries, promoting employment and other methods that are likely to produce long-term results in increasing income. Support for unstable households out of poverty should focus on strengthening the supply of bottom-line security policies. Fourthly, household head characteristics, family characteristics and their located areas can influence the degree of re-poverty through employment channel, industrial channel and social security channel. This paper provides a new perspective and empirical reference for the establishment of monitoring and supporting mechanism of preventing re-poverty.

Key words: Multidimensional Special Concern Index, Re-poverty Degree, Influence Factors, Mediating Effect