Extension: A Tale of Two Studies: Revisiting the Unintended Effects of Staggered Legal Changes

Abstract

Post-2003 methodological innovations in causal inference suggest that many older studies investigating the unintended effects of legal changes need reappraisal. In particular, new difference-in-differences estimators have been proposed recently in settings with staggered adoption that allow for treatment effect heterogeneity across groups and dynamic treatment effects that may grow or dissipate over time. These are common to most legal changes occurring at the state level in the United States. We provide a practitioner-oriented overview of those methodological developments, followed by two empirical illustrations. We first revisit a study that finds that abortion legalization led to an increase in sexually transmitted diseases. Second, we revisit an analysis claiming that three-strikes laws caused an increase in homicides. While we find that the conclusions of the second study are robust, we show that some of the conclusions of the first one are sensitive to the new methodological developments.