2018/06/12 by Denis Nekipelov, Vira Semenova, Nekipelov, Denis +4 · 1 citation
Economics, Econometrics and Finance · Mathematics · Social Sciences · #Advanced Causal Inference Techniques #Econometrics (econ.EM) #Economic Policies and Impacts #FOS: Computer and information sciences #FOS: Economics and business #FOS: Mathematics #Gender, Labor, and Family Dynamics #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Statistics Theory (math.ST)
paper · pdf · doi:10.48550/arxiv.1806.04823
openalex publication_date 2018/06/12 · openalex created_date 2022/08/27 · openalex updated_date 2026/07/28
This paper proposes a Lasso-type estimator for a high-dimensional sparse\nparameter identified by a single index conditional moment restriction (CMR). In\naddition to this parameter, the moment function can also depend on a nuisance\nfunction, such as the propensity score or the conditional choice probability,\nwhich we estimate by modern machine learning tools. We first adjust the moment\nfunction so that the gradient of the future loss function is insensitive\n(formally, Neyman-orthogonal) with respect to the first-stage regularization\nbias, preserving the single index property. We then take the loss function to\nbe an indefinite integral of the adjusted moment function with respect to the\nsingle index. The proposed Lasso estimator converges at the oracle rate, where\nthe oracle knows the nuisance function and solves only the parametric problem.\nWe demonstrate our method by estimating the short-term heterogeneous impact of\nConnecticut's Jobs First welfare reform experiment on women's welfare\nparticipation decision.\n