公共政策评估中的随机控制实验:实施困境与突破策略

Stochastic Control Experiment in Public Policy Evaluation:Implementing Dilemma and Breakthrough Strategy

采用随机控制实验方法来进行政策评估,已成为公共政策或项目评估的“黄金标准”。但是,随机控制实验方法在中国政策评估过程中还运用较少,评估者缺乏对其实施困境与突破策略的深入认识。本研究在深入分析随机控制实验方法及其适用范围的基础上,对随机控制实验的实施困境与突破策略进行了系统的考察与总结。研究发现,采用随机控制实验开展政策评估主要面临着实验方法本身的复杂性、实验方法引发的公共价值分歧以及随机控制实验融入中国政策场域的挑战三个层面的困境。

The Implementation of Randomized Evaluation has become the“gold standard”or“flagship”of public policy or project evaluation.However,due to the lack of understanding of its implementation obstacles and coping strategies by researchers,it is seldom used in actual policy or project evaluation.In this paper,the contents of 100 random impact assessment reports in J-PAL database are systematically investigated,and the coding analysis is carried out around the dilemmas and breakthroughs of the Implementation of Randomized Evaluation in the actual reports.The results show that there are 20 kinds of obstacles in the operation of stochastic impact evaluation,and each obstacle has a certain coping strategy.This study is helpful for evaluators to flexibly use randomized controlled trials to carry out policy evaluation in practice and promote the development of randomized impact evaluation in China.