摘要
文章使用2013年中国家庭收入调查(CHIP)农村住户数据,构建了基于OLS回归模型和分位数回归模型的代理家计调查瞄准框架。基于四类测量方法系统评估和比较了农村低保政策的实际瞄准效果与代理家计调查下的瞄准效果。研究结果显示,代理家计调查框架拟合优度较高,代理指标与家庭人均收入水平关系显著;采用代理家计调查瞄准方法时,农村低保的瞄准效果明显优于实际瞄准效果;在改善农村低保瞄准方面,基于OLS分地区回归和分位数回归的代理家计调查框架比基于OLS总样本回归的代理家计调查框架的效果稍好;代理家计调查方法在降低西部地区漏保率方面效果更好,而在降低东部和中部地区错保率方面效果更为显著。因此,在农村低保政策执行中引入代理家计调查方法有助于提升其瞄准效果。
Using the rural sample of 2013 CHIP data, the paper proposes a proxy means tests(PMT) method based on the OLS and quantile regression models. Using a rich set of targeting measures, we systematically evaluated and compared actual and PMT targeting performance of the Rural Subsistence Allowance(RSA). We found that, first, the regressions based on the PMT method all have a high level of goodness of fit. Second, the proxy indicators are all significant predictors of household per capita income. Second, RSA's PMT targeting was more accurate than its actual targeting. Third, RSA's PMT targeting performance is better when estimated for the respective eastern, central, and western samples than that for the national sample, and also better when estimated based on quantile regression model than based on OLS model for the national sample. Fourth, the PMT targeting is more effective in reducing the exclusion errors in western than that in eastern and central regions, but more effective in reducing the inclusion errors in eastern and central than that in western region. These results suggest that applying the PMT method in China's RSA contributes to the improvement of its targeting performance.
出处
《中国人口科学》
CSSCI
北大核心
2018年第3期73-84,共12页
Chinese Journal of Population Science
基金
国家自然科学基金青年项目"中国农村低保救助的瞄准
减贫效应和行为激励研究"(编号:71703008)的阶段性成果
