Estimation in high-dimensional linear regression: Post-Double-Autometrics as an alternative to Post-Double-Lasso

TL;DR

This paper introduces Post-Double-Autometrics, a new method based on Autometrics, which outperforms Post-Double-Lasso in high-dimensional linear regression, especially in finite samples, and provides new insights into economic growth convergence.

Contribution

The paper proposes Post-Double-Autometrics as a superior alternative to Post-Double-Lasso for high-dimensional regression estimation.

Findings

Post-Double-Autometrics reduces omitted variable bias in finite samples.

The new method outperforms Post-Double-Lasso in empirical tests.

Application to economic growth offers new evidence on convergence.

Abstract

Post-Double-Lasso is becoming the most popular method for estimating linear regression models with many covariates when the purpose is to obtain an accurate estimate of a parameter of interest, such as an average treatment effect. However, this method can suffer from substantial omitted variable bias in finite sample. We propose a new method called Post-Double-Autometrics, which is based on Autometrics, and show that this method outperforms Post-Double-Lasso. Its use in a standard application of economic growth sheds new light on the hypothesis of convergence from poor to rich economies.