Weighted regression with unbalanced Panel Data

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
Report Content as Inappropriate

Weighted regression with unbalanced Panel Data

This post has NOT been accepted by the mailing list yet.

I have unbalanced panel data and am trying to do regressions with and without fixed effects.
I am trying to identify a treatment effect and therefore each observation is assigend the treatment varibale after [0;1].
Treatment occurs in different periods and therefore some individuals have e.g. 2 observations before and 8 after, others the other way around and so on.

I would like to apply weighting to account for these differences. So I assigned a weight to each observation according to the number of times there are observations for each individual in the before and after periods.

Of course, afterwards weights vary within individuals for the before and after periods.

If I now try to use xtreg with fixed effects , Stata tells me "weight must be constant within id".

Is there a way to get around that.

Thanks for helping me!