Dec 15, 2021
Moving Beyond the Classic Difference-in-Differences Model
Published Jun 24, 2021
Motivated by the context of opioid policy research, this workshop provides an overview of key methodological challenges faced when evaluating the effectiveness of state-level policy using annual state-level data, as well as potential solutions and practice guidelines. We begin by discussing the commonly used linear difference-in-differences (DID) model and then cover alternative approaches involving autoregressive models that have been shown to be more promising. We continue the webinar by discussing confounding and heterogeneous policy effects, effects of concurrent policies, and additional design-based approaches (e.g., event studies and stacked comparative interrupted time series).