The paper proposes a supervised machine learning algorithm to uncover treatment effect heterogeneity in classical regression discontinuity (RD) designs. Extending Athey and Imbens (2016), I develop a criterion for building an honest "regression discontinuity tree", where each leaf of the tree contains the RD estimate of a treatment (assigned by a common cutoff rule) conditional on the values of some pre-treatment covariates. It is a priori unknown which covariates are relevant for capturing treatment effect heterogeneity, and it is the task of the algorithm to discover them, without invalidating inference. I study the performance of the method through Monte Carlo simulations and apply it to the data set compiled by Pop-Eleches and Urquiola (2013) to uncover various sources of heterogeneity in the impact of attending a better secondary school in Romania.
翻译:本文提出了一种监督的机器学习算法,以发现传统回归不连续(RD)设计中的治疗效果异同。 扩展AYES和Imbens(Imbens)(_),我为建立一个诚实的“递减不连续树”制定了标准。 树叶的每一片树叶都含有RD对治疗(由共同断开规则所指定)的估算值,以某些预处理共变体的值为条件。先验性未知的是哪些共变对于捕捉治疗效果异异异异,而且算法的任务是在不否定推理无效的情况下发现它们。 我通过蒙特卡洛模拟研究这种方法的性能,并将其应用到由Pop-Eleches和Urquiola(2013年)汇编的数据集中,以发现罗马尼亚上更好的中学所产生的各种异异同源。