Heterogeneity in terms of hte lenght of the cases pre metoo. i would imagine that there is a relationship between win probability and length of case. even if metoo doesn't take hte length that a case takes. so the cases taht overlap are a different amount
control for duration on the RHS.
reg y i.overlap duration, r
overlap = 1 if SH case resolved before Metoo
= 0 if started before but resolved after, and you're SH
Heterogeneity in terms of hte lenght of the cases pre metoo. i would imagine that there is a relationship between win probability and length of case. even if metoo doesn't take hte length that a case takes. so the cases taht overlap are a different amount
control for duration on the RHS.
reg y i.overlap duration, r overlap = 1 if SH case resolved before Metoo = 0 if started before but resolved after, and you're SH