Table 2 linear DID estimation process with two-way fixed effects models
| Â | Cohort A (no change in WLL) | Cohort B (WLL increased by 1 unit) |
|---|---|---|
t | β1WLLAt + β2ZAt + αA + γt + ϵAt | β1WLLBt + β2ZBt + αB + γt + ϵBt |
t + 1 | β1WLLAt + β2ZA, t + 1 + αA + γt + 1 + ϵA, t + 1 | β1(WLLBt + 1) + β2ZB, t + 1 + αB + γt + 1 + ϵB, t + 1 |
E(Difference) | β2(ZA, t + 1 − ZAt) + (γt + 1 − γt) | β1 + β2(ZB, t + 1 − ZBt) + (γt + 1 − γt) |
DID | β1 + β2[(ZB, t + 1 − ZBt) − (ZA. t + 1 − ZAt)] | |