Doubly robust local projections difference-in-differences

TL;DR

This paper introduces a doubly robust extension of local projections difference-in-differences (LP-DiD) for staggered treatments, ensuring consistency under model misspecification and providing advanced inference methods.

Contribution

It develops DRLPDID, a new estimator that combines local projections with doubly robust techniques for improved causal inference in staggered treatment settings.

Findings

DRLPDID matches regression-adjusted LP-DiD under outcome-model alignment.

It outperforms IPT-only variants when the propensity score model is misspecified.

In an application, DRLPDID provides more robust and less negative estimates than unadjusted LP-DiD.

Abstract

This paper develops a doubly robust extension of local-projections difference-in-differences (LP-DiD) for staggered absorbing treatments. The resulting estimator, DRLPDID, preserves the LP-DiD local-stack ATT target and is consistent when either the local untreated-outcome regression or the local treatment-probability model is correctly specified. It also delivers influence-function-based inference for post-treatment summaries and multiplier-bootstrap bands for dynamic paths. In Monte Carlo designs with covariate-driven selection, DRLPDID matches regression-adjusted LP-DiD under outcome-model alignment and clearly outperforms the IPT-only variant under propensity-score misspecification. In the no-fault-divorce application, DRLPDID tracks robust staggered-adoption estimators and is less negative than unadjusted LP-DiD.