There's a school of belief that says gerrymandering doesn't actually have much impact on how members of the House vote. As John Sides, an associate professor at George Washington University and contributor to the Post's Monkey Cage blog wrote in 2013: "The most important influence on how members of Congress vote is not their constituents, but their party."
As proof, he pointed to research showing a lack of correlation between what percentage of a congressman's district voted for a presidential candidate and how its member of Congress voted. There's not so much difference, say, between how a Republican House member votes in a swing district versus one that went 75 percent for Mitt Romney.
The findings suggest even if Congress was made up of more competitive districts in which Democratic and Republican candidates regularly swapped seats every two years, voters might wind up with members who vote nearly as liberal or conservatively as they do now.
The above chart compares DW-NOMINATE scores -- they measure partisanship from -1 (very liberal) to 1 (very conservative) -- to the 2004 presidential vote in all 435 House districts for the 108th Congress. It's true that more conservative Democratic-held districts voted somewhat more conservative, on average, but the difference between the most liberal districts and the most conservative ones is not as sharp as you might think. And the the line is somewhat flat for Republicans too.
There was also a significant gap between how the most conservative Democratic districts vote and how the most liberal Republican districts do -- even as many of these districts overlap greatly on the 2004 presidential vote. The partisan affiliation of a member was a much better indicator of where these members stand on the DW-NOMINATE spectrum than the presidential vote. And this graph, as you can see here, looks very similar for other recent Congresses.
Our partisan Congress is a relatively new thing. In the 1970s, there were a significant number of Democrats and Republicans crossing over. But it's slowly faded, to today, where it doesn't happen much at all. Here's that divide illustrated by Pew, with the House on the right.
The reason for the change, Pew said, was geography. In the 1970s there were moderate Republicans in the Northeast and conservative Democrats in the South who voted according to two things: 1) traditional party ideology and 2) regional issue differences. Today, Southern Democrats and Northeastern Republicans have become an endangered species, and Congress now votes mostly according to traditional party ideology, hence the voting along party lines regardless of a district lies on the political spectrum.
So what if states suddenly adopted redistricting commissions en masse and we got state legislators out of the map-drawing business for good? There would almost definitely be more competitive districts, but perhaps not a ton more. The United States is a country very polarized between rural and urban, after all, and map-drawers' goal is not to create competitive districts, but rather to create compact ones that bring together similar groups of people. In most areas areas of the country, there is simply no prospect of creating new, competitive districts.
And as the numbers above show, even the members who emerged from the increasing number of competitive districts might be about as partisan as the rest of their caucus.
So while the Supreme Court's redistricting decision will be hailed as a sign of progress by good-government types, it's important to note how limited its effect might be on the coming Congresses -- to say nothing of how many states will actually join the few who have already adopted such commissions.