Dear Yvonne,

I am neither a good econometrician nor a Stata specialist, and maybe I misunderstood you posting, but your model setup looks a bit strange too me:
I do not think you can explain any variation between individuals by using only aggregate data, you do need some individual-specific information on the RHS as well. How can you, e.g., explain the differences in personal income between you and your neighbour when you have only information on , let's say, the share of educated people in the neighbourhood ?

Yours sincerely,

Justina Fischer