All simulators
Group-Time ATTs
Callaway–Sant'Anna cells, event-study aggregation, and the bias of static TWFE with dynamic effects.
Day 71: Heterogeneity-Robust DiD (Callaway–Sant'Anna, Sun–Abraham)ATT(g, t) — cohort rows, period columns
| g \ t | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 |
|---|---|---|---|---|---|---|---|---|---|
| 3 | 0.0 | ref | 2.0 | 2.5 | 3.0 | 3.5 | 4.0 | 4.5 | 5.0 |
| 5 | 0.0 | 0.0 | 0.0 | ref | 2.0 | 2.5 | 3.0 | 3.5 | 4.0 |
| 7 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | ref | 2.0 | 2.5 | 3.0 |
overall ATT (cell average)
3.133
static TWFE
2.320
bias -0.813
event-study θ(e), robusteffect profile of the middle cohort
Three adoption cohorts (periods 3, 5, 7) and never-treated units. Each cell is a clean 2×2 DiD of a cohort against never-treated units relative to its own last pre-period: ATT(g, t). Grey cells are pre-treatment placebos. The event study averages cells with the same time since adoption; the TWFE coefficient (right) uses already-treated cohorts as controls and drifts away when effects grow or differ across cohorts.