Arellano-Bond dynamic panel-data estimation

[Nigel Taylor]

I am making a big guess here, I am wondering if is it possible to run a GMM model with fixed effects? I think it is not possible as I am mostly seeing fe being run with the xtreg command.

[Sebastian Kripfganz]

help xtabond
help xtdpdsys
help xtdpd
ssc describe xtabond2
ssc describe xtseqreg

[Nigel Taylor]

Thanks Sebastion but none of those commands indicate that fe can be run together. Hence, I am assuming it can't be done in a GMM model. I was just wondering if anyone had look into it.

[Sebastian Kripfganz]

I am not quite sure what you have in mind. All of these commands assume (or allow for) an error-component structure with unobserved unit-specific effects ("fixed effects").

[Dario Maimone Ansaldo Patti]

Nigel a look at Roodman's paper "How to do xtabond2", p. 99 could be helpful for your needs. This is the link: http://www.stata-journal.com/sjpdf.h...iclenum=st0159