Something stray
(I wrote this a couple weeks ago, and never posted, but since the WaPo did a piece this AM, I thought I'd throw it up.).
In the census, prison inmates are counted as residents wherever they are in prison. With most prisons located in rural areas, that means primarily urban people are relocated by the census to be citizens of rural areas.
Now, this is where it gets weird. Although these are non-voting citizens, they are counted for the purposes of assigning Congressional districts which means the Congressional districts with prisons inside them tend to slightly overrepresent the votes of the legitimate voters as these districts have so many forced non-voters.
It's a relatively slight distortion, but it's there, overrepresenting the rural.
In the census, prison inmates are counted as residents wherever they are in prison. With most prisons located in rural areas, that means primarily urban people are relocated by the census to be citizens of rural areas.
Now, this is where it gets weird. Although these are non-voting citizens, they are counted for the purposes of assigning Congressional districts which means the Congressional districts with prisons inside them tend to slightly overrepresent the votes of the legitimate voters as these districts have so many forced non-voters.
It's a relatively slight distortion, but it's there, overrepresenting the rural.
posted by mikevotes at
7:02 AM
2 Comments:
I have no idea, nor the time to research, if that non-voter count has any mathematical effect on the Congressional vote balance.
Since the Census numbers are used to allocate federal funds (including funds for prisons) I think the prisoners must be counted.
By
Unknown, at 11:06 AM
It's a relatively minimal effect. A congressional district in 2000 was about 625,000, and you gotta figure a big prison is 20-25,000, so small per district.
But collectively, you're talking about transferring 1.2 -1.4 million people from urban to rural.
By
mikevotes, at 12:38 PM
Post a Comment
<< Home