I have a panel data set where one of the variables is employment status is where 1= employed and 0 = not employed. The other variables are for those who have employ=1, such as place of employment, sector, number of hours worked etc.

I would like to run a fixed-effects model where I have the variables above as independent variables. However, those who have employ=0 have missing values for the workplace, sector , number of hours etc.

Under crossectional data, I would have made a subset of employed individuals and just analysed that but I don't know how to deal with this issue in panel data.

  • $\begingroup$ @kjetilbhalvorsen Thank you! It did give me insight on how to deal with the problem ! $\endgroup$ – Lorato May 18 '20 at 14:58