Time Fixed Effects with areg?

[Thomas Wake]

Hi,

I am running a DID regression with panel data. I have just used cem command to match my data, and I wish to use this in my analysis.
However, I can only seem to use this for areg, and not xtreg (as this does not allow for weights). Would somebody please be able to tell me which is more appropriate - areg with matching, or xtreg with year fixed effects? Or would it be best to include both? If somebody could please explain the implications of this to me, it would be hugely appreciated.

Many thanks

[Maxence Morlet]

I would recommend Sergio Correia's reghdfe, for multi-dimensional fixed effects.

As an aside, xtreg does allow weights, they just can't vary within your unit variablle. Therefore, you may have to average out the weights within each unit, and employ the average.