> The policy of counting them as living in the place where they are imprisoned diverts federal and state monies away from the places that they came from.

I think, viewed more carefully, one would be forced to recognize that the effect described results instead from the policy of allocating money for programs by total Census counts without considering the fitness of those counts to the purpose of the particular programs and making suitable adjustments (and that that policy is going to have some infelicities, if not the exact same ones, no matter what general policies are adopted for Census counts.)