Sunday, April 24, 2005 :::
I posted estimates last week for the reapportionment for the 2010 census, based on a very simple linear extrapolation of population change. Later in the week, Taranto happened to point to new estimates by the census bureau, which I assume uses projection techniques that are somewhat more sophisticated than linear extrapolation (if they don't, I don't want to know about it).
Using their numbers moves one congressman from New York to Louisiana. That's the only change; as I had suspected, for a date as close as 2010, linear extrapolation isn't all that much worse than the state of the art. They also, though, have estimates for 2020 and 2030. I don't have much confidence in any predictions that far out, but that didn't stop me from reapportioning based on them. After all, I'd already written the program.
2003:
Iowa -1
Ohio -1
Texas +1
Utah +1
2010:
Alabama -1 (next in line)
Arizona +1
Calfornia +1
Florida +2
Georgia +1 (got the 435th)
Illinois -1
Iowa -1
Massachusetts -1
Missouri -1
Nevada +1
New York -2
Ohio -2
Pennsylvania -1
Texas +3
Utah +1
2020 (relative to 2000):
Alabama -1
Arizona +3
California +1
Florida +5
Georgia +1
Illinois -2
Iowa -1
Louisiana -1
Massachusetts -1
Michigan -1
Missouri -1
Nebraska -1
Nevada +1
New Jersey -1
New York -4
North Carolina +1
Ohio -3
Oregon +1
Pennsylvania -3 (next in line to retain one)
Texas +5
Utah +1
Virginia +1 (got the 435th)
Washington +1
West Virginia -1
2030 (relative to 2000):
Alabama -1
Arizona +5
California +2 (next to gain another)
Connecticut -1
Florida +9
Georgia +1
Illinois -3
Indiana -1
Iowa -1
Kentucky -1
Louisiana -1
Massachusetts -2
Michigan -2
Minnesota no change, but got the 435th
Missouri -1
Nebraska -1
Nevada +2
New Jersey -1
New York -6
North Carolina +2
Ohio -4
Oregon +1
Pennsylvania -4
Rhode Island -1
Texas +8
Utah +1
Virginia +1
Washington +1
West Virginia -1
Wisconsin -1
UPDATE
::: posted by Steven at 10:18 PM