Regularization and confounding in linear regression for treatment effect estimation

By P. Richard Hahn, Carlos M. Carvalho, David Puelz, Jingyu He. 2016

97 citations recorded by the source corpus.

This paper investigates the use of regularization priors in the context of treatment effect estimation using observational data where the number of control variables is large relative to the number of observations. First, the phenomenon of regularization-induced confounding is introduced, which refers to the tendency of regularization priors to adversely bias treatment effect estimates by over-shrinking control variable regression coefficients. Then, a simultaneous regression model is presented which permits regularization priors to be specified in a way that avoids this unintentional re-confounding. The new model is illustrated on synthetic and empirical data.

Subfields: Economics and Econometrics, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/e1fe5b2c3e2b64bb029b884f5da2d552f0b588f2

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Economics and Econometrics

1. Jundong Li
2. Huan Liu

In Statistics and Probability

1. Lu Cheng
2. P. Richard Hahn
3. Jundong Li
4. Ruocheng Guo
5. Huan Liu

Citations in the corpus

• Bayesian Regression Tree Models for Causal Inference: Regularization, Confounding, and Heterogeneous Effects (with Discussion) (2020, unknown): “We provide some calculations showing this in our paper, and extensive evidence in Hahn et al. (2016b).”
• A Survey of Learning Causality with Data (2018, background): “[59] propose to handle the problem regularization-induced confounding (RIC) with BART [58].”
• A Survey of Learning Causality with Data: Problems and Methods (2018, background): “RIC happens when E[Y |D,X] heavily depends on the features X rather than the treatment D.”
• Bayesian regression tree models for causal inference: regularization, confounding, and heterogeneous effects (2017, unknown): “induced confounding” of Hahn et al. (2016) in the context of nonparametric Bayesian models as utilized by Hill (2011).”