基于倾向得分匹配法的政策参与效应评估

Policy Treatment Effect Evaluation Based on Propensity Score Matching

在一个民主社会中,政府实施的政策项目需要得到准确的评估。但是,政策参与效应的估计需要找到一个"反事实"的比较组,并且有效消除参与的选择性偏差。本文针对倾向得分匹配法要解决的根本问题、基本思路与方法运用、与其他方法的比较三个方面进行评述,揭示了这一方法在评估政策项目效应中的意义。

In a democratic society, policy project implemented by the government needs to be accurately evaluated. However, e- valuation of policy treatment effect needs to find a ＂counterfactual＂ comparison group, and effectively eliminates se- lection bias. This paper reviewed the method of Propensity Score Matching in three aspects： the fundamental problem to be solved, the basic ideas and specific application, the comparison with other methods, and revealed its significance in evaluation of policy project effects.