Hi there,

I want to compare two ivregress models with two independent samples.
I am trying to see if there is any difference between pre-policy change period and post-policy change period. The sample for each period (pre and post) is independent.

ivregress 2sls Y X1 X2 X3 (X4=Z) if P_Change==0
ivregress 2sls Y X1 X2 X3 (X4=Z) if P_Change==1

where P_Change is a dummy variable for policy change (0 = pre-change; 1 = post-change).
I have been looking for something like suest for ivregress, but nothing seems to be working.
I have reviewed the previous posts related to this topic and have tried gmm method or stacked model using ivregress, but nothing seems to be giving the right estimates (the same estimate for each period as the one from each ivregress model).

Any suggestions or comments would be appreciated.

Thanks,
Nami