精准扶贫政策对城乡居民收入差距的影响研究——以山西省为例

IMPACT OF TARGETED POVERTY ALLEVIATION ON THE URBAN RURAL HOUSHEOLD INCOME GAP——A CASE STUDY OF SHANXI PROVINCE

关键词:  精准扶贫扶贫效果城乡居民收入双重差分模型山西省

Key words:  targeted poverty alleviation  poverty alleviation effect  urban rural income gap  difference in difference method  Shanxi province

摘要:

［目的］实施精准扶贫政策以来，我国共减少贫困人口6 853万人。在脱贫攻坚取得重大进展的背景下，对比政策实施前后城乡居民收入的变化，对于揭示精准扶贫政策的作用具有重要意义。［方法］以贫困范围广、贫困程度深的山西省为对象，基于2010—2017年精准扶贫政策实施前后的统计数据，采用双重差分模型研究不同贫困程度县（区）城乡居民收入差距变化。［结果］精准扶贫政策的效果主要为：贫困县的城乡居民收入差距缩小幅度约为非贫困县的3倍，而其中国家级贫困县的收入差距缩小幅度约为省级贫困县的2倍； 共同趋势检验和安慰剂检验的结果验证了模型的稳健性。对比采用最小二乘法（OLS）得到的精准扶贫政策效果，揭示出如果忽略城乡居民收入差距随时间的变化，则会高估精准扶贫政策效果约15~4倍，再次证明模型的有效性。通过引入控制变量探讨政策以外的其他影响因素，发现城镇化水平、经济发展水平、第三产业占比、人均农业产值有助于缩小城乡居民收入差距，人均土地状况会扩大城乡居民收入差距，固定资产投资和消费水平的影响不显著。［结论］精准扶贫政策对于缩小城乡居民收入差距具有显著效果，且对深度贫困地区的影响更明显。

Abstract:

Since the implementation of targeted poverty alleviation policy, China has lifted 68.53 million people from poverty. Under the background of the massive poverty alleviation campaign, it is crucial to evaluate the effect of the targeted poverty alleviation policy by investigating the changes of household income gap between urban and rural residents before and after policy implementation. This paper took Shanxi province as the study area because it had been suffering from chronic poverty. The study period was from 2010 to 2017, as the policy was implemented in 2013. The difference in difference (DID) method was used to reveal the changes of income gap between urban and rural residents in counties of different poverty degrees. The results showed that the reduction of income gap between urban and rural residents in impoverished counties was three times more than that in non impoverished counties since the implementation of the policy, and the reduction of income gap in national level counties was twice more than that in provincial level impoverished counties. The result robustness was validated by the common trend test and the placebo test. When comparing to the results of ordinary least squares (OLS) method, the results showed that the policy effect would be overestimated for 1.5 to 4 times if we neglected the variation of income gap over time, which further proved the validity of the model. When control variables were included in the model, the factors that may positively affect the income gap included urbanization level, economic development level, proportion of tertiary industry and agricultural output per capita, while land holding per capita had negative effect, and fixed asset investment and consumption level had no significant impact. It concludes that the targeted poverty alleviation policy has a significant impact on reducing the income gap, and the effect is more remarkable in poorer areas.