Hello

I'm interesting to road a panel spatial with n=399 and t=12. In the stata 15 I use to do that the spxtregress command, but this method has the assumption of normality in errors. In the same way, in stata 13 I use xsmle, that addopted the same hypothesy. My problem is that my data dont have such errors distribution. There are an command in stata to road a panel spatial with gmm or IV.

Thanks for the any help