摘要
利用Heckman两阶段法、双重差分模型和"双重差分-分位数回归"法,基于贵州纳雍县两个贫困村跨期十数年的农户追踪调查数据,以受访农户恩格尔系数为标尺探析参与式社区综合发展减缓农户消费贫困的特征化事实及其内在规律。结果表明,参与式社区综合发展对减缓消费贫困有显著的即期效应和滞后效应,且后一种效应更大。不仅如此,参与式社区综合发展即期减贫精度不高、包容性较强,存在"溢出效应",但在一定程度上具有缩小农户间恩格尔系数差距(或消费贫富差距)的功能,且对处于"脱贫率"分位点附近的农户恩格尔系数负向作用最大;及至后期,此减贫方式的"瞄准偏差"极大,仅能降低处于"小康率"分位点的农户恩格尔系数,"漏出效应"突出。
Using the Heckman's two-stage approach,double difference model and 'double difference-quantile regression'method,and based on the tracking survey data for over ten years of two poverty-stricken villages in Nayong County of Guizhou Province,this paper explores the featured facts and inherent law of the participatory comprehensive community development to alleviate consumption poverty adopting farmers' Engel coefficient as scaleplate.The findings suggest that the participatory comprehensive community development has both significantly immediate effect and lag effect to alleviate consumption poverty,which reveals the farmers' Engel coefficient shall be reduced by 7.2 to 9.8 percentage and 12.1 to 22.6 percentage accordingly,and the latter effect is larger.Moreover,the participatory comprehensive community development shows'overflow effect',due to its low immediate poverty-reduction accuracy and strong inclusion.But it shall narrow the gap of Engel coefficient or consumption among different groups of farmers,and the negative effect of Engel coefficient of farmers who are around the'anti-poverty rate'quantile reaches the peak.In addition,as to the late period,the'aiming deviation'of this poverty alleviation mode presents so big that it can just reduce the Engel coefficient of farmers who are around the'well-off rate'quantile,which reveals significant'leakage effect'.
出处
《现代财经(天津财经大学学报)》
CSSCI
北大核心
2017年第11期103-113,共11页
Modern Finance and Economics:Journal of Tianjin University of Finance and Economics
基金
国家社科基金青年项目(15CTY048)
中国农业科学院科技创新工程(ASTIPIAED-2017-03)
关键词
参与式社区综合发展
农户恩格尔系数
消费贫困
Heckman两阶段法
双重差分-分位数回归法
participatory comprehensive community development
farmer's Engel coefficient
consumption poverty
Heckman's two-stage approach
Double Difference-Quantile Regression