This isn’t some immutable law of nature or anything. In the past I calculated the difference of lean between the mean district and the median district for a bunch of states and several states managed to overcome this. If I have time, I might do it again.
As a quick and dirty substitute I can do something similar using Cook Partisan Voting Index.
Down the page there’s a collapsable table of the Cook PVI for all 435 congressional districts in the country. It doesn’t take much work to pull that table into a spreadsheet and calculate:
- The mean district is D+0.41.
- The median district is R+1.
So there’s a skew of about R+1.41. If there is some built in structural advantage toward low population density areas and a detriment to high population areas, this is size of that effect across the entire country.
Wisconsin has eight congressional districts so we can also calculate:
- The mean Wisconsin district is R+1.13.
- The median Wisconsin district is R+7.
This results in an R+5.87 skew which means Wisconsin congressional districts are skewed toward Republicans 4.46 points more than the nation as a whole.
New Jersey on the other hand, with 12 CDs, has a mean of D+7.08 and a median of D+7 which is almost has no skew at all.