通过多变量Prob it模型考察了我国农村贫困在两期之间的状态依赖性,同时分析了决定我国农村居民当期免于贫困的因素,研究发现:(1)人力资本投资中的教育投资和健康投资的反贫困效果最好;(2)从贫困的状态依赖性来看,我国农村低收入者没有陷入明显的持续贫困,收入流动性较强;(3)农村女性收入状况逐年改善,婚姻状况有利于缓解贫困,国家公务员或者企业经理,政府机关、国有企事业单位性质的员工陷入贫困的概率更低,东部地区居民容易免于贫困;(4)改革开放的制度变迁给农村居民带来收入变动的机遇,同时也增加了他们收入的不稳定性;(5)回归结果对低收入门槛的定义比较敏感,有必要在收入分布的"绝对贫困"区域上方再划定一个"基本贫困"区域。
This paper discusses poverty state dependence in Chinese rural area with multivariate Probit model and analyzes the current antipoverty determinants.It finds that:(1) education investment and health investment have the best effect;(2) seeing from the poverty state dependence,low-earning rural people do not fall into obviously continuous poverty,and the income mobility is strong;(3) the income of the rural female has improved in recent years and their marriage is helpful to eliminating poverty.Civil servants,business managers and employees working for government,state-owned business are probably not getting into poverty.And residents in Chinese east area are more likely shake off poverty;(4) institutional change in Chinese reform and open policy not only increases the income of rural residents,but also enhances their income instability;(5) the regressing outcome is sensitive to low incoming threshold,so it is necessary to demarcate a "basic poverty" area on the top of "absolute poverty" area in income distribution.