Bayesian Regression Tree Models for Causal Inference: Regularization, Confounding, and Heterogeneous Effects

31 Pages Posted: 6 Oct 2017 Last revised: 22 Jul 2018

See all articles by P. Richard Hahn

P. Richard Hahn

Arizona State University (ASU) - School of Mathematical and Statistical Sciences

Jared Murray

University of Texas at Austin - Red McCombs School of Business

Carlos M. Carvalho

University of Texas at Austin - Red McCombs School of Business

Date Written: October 5, 2017

Abstract

This paper develops a semi-parametric Bayesian regression model for estimating heterogeneous treatment effects from observational data. Standard nonlinear regression models, which may work quite well for prediction, can yield badly biased estimates of treatment effects when fit to data with strong confounding. Our Bayesian causal forests model avoids this problem by directly incorporating an estimate of the propensity function in the specification of the response model, implicitly inducing a covariate-dependent prior on the regression function. This new parametrization also allows treatment heterogeneity to be regularized separately from the prognostic effect of control variables, making it possible to informatively “shrink to homogeneity”, in contrast to existing Bayesian non- and semi-parametric approaches.

Keywords: causal inference, treatment effects, heterogeneous effects, subgroup effects, tree methods, Bayesian, nonparametric regression, nonlinear regression

JEL Classification: C11, C14, C30, C45

Suggested Citation

Hahn, P. Richard and Murray, Jared and Carvalho, Carlos M., Bayesian Regression Tree Models for Causal Inference: Regularization, Confounding, and Heterogeneous Effects (October 5, 2017). Available at SSRN: https://ssrn.com/abstract=3048177 or http://dx.doi.org/10.2139/ssrn.3048177

P. Richard Hahn (Contact Author)

Arizona State University (ASU) - School of Mathematical and Statistical Sciences ( email )

Tempe, AZ 85287-1804
United States

Jared Murray

University of Texas at Austin - Red McCombs School of Business ( email )

Austin, TX 78712
United States

Carlos M. Carvalho

University of Texas at Austin - Red McCombs School of Business ( email )

Austin, TX 78712
United States

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
175
Abstract Views
868
rank
199,863
PlumX Metrics