Dem
51
GOP
49
Charlie Cook's PVI (Partisan Voting Index) is widely used to compare House districts to one another and to see which ones are deep blue, deep red, or swingy. Cook regards districts from D+5 to R+5 to be swing districts, since depending on the candidates, either party could potentially win the seat. Back in 1999, there were 164 such swing districts. Now there are 82, exactly half as many. The question is: What happened? David Wasserman has an interesting article analyzing the disappearance of swing seats.
Wasserman starts out by noting that in 1997, there were 26 Democrats in red districts of R+3 or more and there were 14 Republicans in districts D+3 or more. That is another way of saying there were a lot of split tickets, because a Democrat in an R+3 district won even though his or her district voted Republican for president. Also vice-versa. Now there are just five Democrats in R+3 or redder districts and just four Republicans in D+3 or bluer districts. "Hyperswing" districts (D+3 to R+3) have declined from 107 in 1999 to 45 now.
Cook has computed the PVIs for 13 cycles now, so there is a lot of data to be explored. The big question is whether the disappearance of swing districts is due to gerrymandering or to hardening of partisan attitudes. Some districts give a big clue. For example, KY-05, a rural district in eastern Kentucky, has pretty much the same shape now as in 1997. But then it was R+2. Now it is R+32. That is a 30-point shift that can't be attributed to gerrymandering. It is largely the same batch of voters or their children. They simply have become massively more conservative in the past 25 years.
A look at other swing districts that have stopped swinging has shown that 58% ceased to be swing districts without any serious boundary changes. The people there simply became more liberal or more conservative. People moved in, people moved out, or opinions hardened. In the other 42% of the districts, it was changes to the boundaries that pushed them into more partisan territory. In some cases that was due to explicit gerrymandering, but sometimes even when a nonpartisan commission draws the maps, some districts will simply become more partisan because the goal is to divide the entire state up fairly, not to make every district 50-50. That would also have a gerrymandering-like effect and result in weird districts.
Also interesting is that in the nine states where independent commissions draw the maps (Arizona, Colorado, California, Hawaii, Idaho, Michigan, Montana, New Jersey and Washington), the number of swing districts dropped by 39% between 1997 and 2023. This is simply due to people becoming more partisan and not splitting their tickets any more. And across the 19 Republican-controlled districts, swing seats fell by 70%, from 56 to 17, which is consistent with the general observation that Republicans have swung rightward more sharply than Democrats have swung leftward.
So while, gerrymandering certainly has played a role in the loss of swing districts, it is not the only factor out there. And one conclusion is that if one sees swing districts as valuable, more states should opt for having independent commissions draw the maps. Of course, partisans who want to squeeze out every seat, both in red and blue states, oppose that. This is why many of the independent commissions were established by the voters in ballot initiatives. (V)