Testing for Rank Invariance or Similarity in Program Evaluation
Testing for Rank Invariance or Similarity in Program Evaluation
复制标题
测试程序评估中的排名不变性或相似性
DOI：
--
复制
发表时间：
2017
影响因子：
8
通讯作者：
Shu Shen
中科院分区：
文献类型：
--
作者：
Yingying Dong;Shu Shen
摘要本文讨论了等级不变性或等级相似性的可测性含义，这是程序评估和分位数处理效应(QTE)文献中常见的假设。我们非参数地识别、估计和检验潜在等级的反事实分布，或分布的特征。拟议的测试允许内源性治疗，外源性治疗是一个特例。这些测试基本上不需要任何额外的假设，除了那些识别和估计QTE的假设。我们应用拟议的测试来调查《就业培训伙伴法》培训是否会导致受训者系统性地改变他们在收入分配中的排名。
Abstract This paper discusses testable implications of rank invariance or rank similarity, assumptions that are common in program evaluation and in the quantile treatment effect (QTE) literature. We nonparametrically identify, estimate, and test the counterfactual distribution of potential ranks, or features of the distribution. The proposed tests allow treatment to be endogenous, with exogenous treatment following as a special case. The tests essentially do not require any additional assumptions other than those to identify and estimate QTEs. We apply the proposed tests to investigate whether the Job Training Partnership Act training causes trainees to systematically change their ranks in the earnings distribution.