Class 25: Difference-in-Differences, Part II

Published

April 14, 2026

The goal for today is to apply DD to real data on hospital uncompensated care and Medicaid expansion, and to discuss current issues with DD designs in practice.

Objectives

  • Understand and explain some of the difficulties when interpreting the results of regression-based DD estimates (TWFE with staggered adoption)
  • Be aware of alternative estimators that may be applicable in some settings

Activities

  • Estimate an event study with constant treatment timing
  • Estimate an event study with differential treatment timing
  • Discuss the TWFE decomposition and modern DD estimators