clustered errors with general weighting
Posted: Fri Mar 24, 2017 7:30 pm
I have set up a repeated cross-section as an unbalanced panel and clustered using the state as cross-section identifier. It clusters correctly. However now I need to add observation specific weights and there seems to be no option for variance weights, as in weighted least squares. Transforming all the variables would be a huge chore because it is a very large model. Any suggestions?
The stata command I am trying to replicate is something like regress y x1 x2 [pweight=perwt] , vce(cluster state)
The stata command I am trying to replicate is something like regress y x1 x2 [pweight=perwt] , vce(cluster state)