Hi!
I am trying to measure the effects of having children on sick-leave using xtlogit fixed effects to calculate the probabillity of having a sick-spell. My problem is that in the data is norwegian, where there is paid maternity leave. In my data women om maternity-leave are registed as working, but then with no sick-leave, which messes up the regression. The panel data is sorted by an idnr, and has observations over 2-8 years.
I wish to somehow omit the year following an increase in children in the regression, so that the paternity leave does not interfere with the sick-leave variable. Does anybody have an idea on how this is possible?
I am trying to measure the effects of having children on sick-leave using xtlogit fixed effects to calculate the probabillity of having a sick-spell. My problem is that in the data is norwegian, where there is paid maternity leave. In my data women om maternity-leave are registed as working, but then with no sick-leave, which messes up the regression. The panel data is sorted by an idnr, and has observations over 2-8 years.
I wish to somehow omit the year following an increase in children in the regression, so that the paternity leave does not interfere with the sick-leave variable. Does anybody have an idea on how this is possible?
Comment