Specification Tests for Generalized Propensity Scores Using Double Projections

36 Pages Posted: 21 May 2020

See all articles by Pedro H. C. Sant'Anna

Pedro H. C. Sant'Anna

Vanderbilt University - College of Arts and Science - Department of Economics

Xiaojun Song

Peking University - Guanghua School of Management

Date Written: March 30, 2020

Abstract

This paper proposes a new class of nonparametric tests for the correct specification of generalized propensity score models. The test procedure is based on two different projection arguments, which lead to test statistics with several appealing properties. They accommodate high-dimensional covariates; are asymptotically invariant to the estimation method used to estimate the nuisance parameters and do not requite estimators to be root-n asymptotically linear; are fully data-driven and do not require tuning parameters, can be written in closed-form, facilitating the implementation of an easy-to-use multiplier bootstrap procedure. We show that our proposed tests are able to detect a broad class of local alternatives converging to the null at the parametric rate. Monte Carlo simulation studies indicate that our double projected tests have much higher power than other tests available in the literature, highlighting their practical appeal.

Keywords: Double projections; generalized propensity scores; multiplier bootstrap; multi-valued treatment; specification tests

JEL Classification: C14, C21, C22, C25

Suggested Citation

Sant'Anna, Pedro H. C. and Song, Xiaojun, Specification Tests for Generalized Propensity Scores Using Double Projections (March 30, 2020). Available at SSRN: https://ssrn.com/abstract=3564679 or http://dx.doi.org/10.2139/ssrn.3564679

Pedro H. C. Sant'Anna (Contact Author)

Vanderbilt University - College of Arts and Science - Department of Economics ( email )

Box 1819 Station B
Nashville, TN 37235
United States

HOME PAGE: http://https://sites.google.com/site/pedrohcsantanna/

Xiaojun Song

Peking University - Guanghua School of Management ( email )

Peking University
Beijing, Beijing 100871
China

Here is the Coronavirus
related research on SSRN

Paper statistics

Downloads
8
Abstract Views
97
PlumX Metrics