I am working with duration data. My analysis uses as dependent variable a binary variable indicating whether the individual has divorced yet in time t (Y=1) or not (Y=0).
However, the literature on divorce uses mostly the duration (Y=1,2,3,...J) as the dependent variable, and follows on using a Fixed Effects estimator.
I acknowledge that the two approaches estimate different things, but I also see that the are slightly different perspectives over the same thing, aka how a policy affects duration. Does anyone have insight over which approach is better or what are the limitations of each?