1. Conventional practice1.1 Basic ideas and Model setup (2 time periods*2 groups)
1.2 Extensions
(1) Multiple time periods or multiple groups (for instance, the staggered DiD estimation)
(2) Repeated Cross-Sections
(3) Two-Dimensional Cross-Sections (for instance, Cohort DiD)
(4) Fuzzy DiD
(5) DDD, DDDD, …
1.3 Estimation and inference
(1) Basic model specification
(2) Two-way fixed effects linear regression specification
a. Static TWFE model
b. Dynamic TWFE model (Event study)
2. Recent development2.1 Interpretation of the TWFE estimators (with multiple time periods or multiple groups)
2.2 Problems of the TWFE estimators with heterogeneous treatment effects
2.3 Alternative estimators
(1) Estimators based on some “building blocks”
a. Estimating only the instantaneous treatment effect
b. Estimators based on group-time ATT
c. Estimators based on cohort-specific ATT
(2) Two-stage estimators/imputation estimators