Redistricting is one of my pet peeves. I’m a fan of proportional representation in the first place, which would remove the problem of redistricting completely, but if we’re going to do it this way let’s do it right. But we’re not: as people on twitter were noticing,
Pennsylvania: 2.71 million votes for D house candidates. 2.64 million votes for R house candidates. House members elected: 13 GOP, 5 Dems.
— sethdmichaels (@sethdmichaels) November 7, 2012
The way this works is pretty simple. Imagine a state with 100 voters, 50 Democrats and 50 Republicans. It has 10 congressional districts with ten people each. The GOP in our hypothetical state controls the state house, and wants to maximize the congressional seats won by Republicans. So what they’d do is create three districts that, combined, have 23 Democrats and only 7 Republicans. The Democrats easily win those three districts, but the other seven districts have, combined, 43 Republicans and only 27 Democrats. Boom: three Democratic seats, seven Republican seats.
This is exactly what the GOP in Pennsylvania did – they put a whole lot of Democrats in those five districts. I did some quick and dirty math based on the AP figures published by the Chicago Tribune. The average winning Democrat in PA took 76.2% of the vote; the average winning Republican took only 58.7%. The strongest-performing winning Republican had a lower winning percentage than all but one of the winning Democrats. Since every vote past 50% is (in effect) “wasted,” the GOP forced the Dems to waste a whole lot of votes in a small number of districts to ensure themselves victory in the rest of the state.
District | Winning Party | Winning Party % |
2 | Democrats | 89.4% |
1 | Democrats | 85.0% |
14 | Democrats | 76.9% |
13 | Democrats | 68.9% |
10 | Republicans | 65.9% |
18 | Republicans | 63.8% |
5 | Republicans | 62.9% |
9 | Republicans | 61.6% |
17 | Democrats | 60.5% |
4 | Republicans | 59.7% |
7 | Republicans | 59.5% |
11 | Republicans | 58.5% |
6 | Republicans | 57.2% |
8 | Republicans | 56.7% |
15 | Republicans | 56.6% |
16 | Republicans | 55.0% |
3 | Republicans | 54.7% |
12 | Republicans | 51.5% |
This is actually contrary to the way redistricting has worked in the past. Usually the party in power tries to create a bunch of safe seats for its incumbents, and forces the opponents incumbents to fight in evenly matched districts in the hopes that they could be knocked off. This works well for the individual politicians, but is less effective for the party at large. The PA GOP put party ahead of individual interests, and it paid off. Meanwhile, the Illinois Democratic Party stuck to the old formula; they protected a bunch of incumbents, and though the state voted 53.8%-44% Democratic, they “only” won 12 of 18 seats.
District | Winning Party | Winning Party % |
7 | Democrats | 84.7% |
4 | Democrats | 83.2% |
18 | Republicans | 74.2% |
1 | Democrats | 73.9% |
15 | Republicans | 68.9% |
3 | Democrats | 68.8% |
9 | Democrats | 66.1% |
5 | Democrats | 65.6% |
2 | Democrats | 63.0% |
16 | Republicans | 61.9% |
6 | Republicans | 59.2% |
14 | Republicans | 58.8% |
11 | Democrats | 57.7% |
8 | Democrats | 54.7% |
17 | Democrats | 53.3% |
12 | Democrats | 51.5% |
10 | Democrats | 50.5% |
13 | Republicans | 46.6% |
This problem is bigger than the Pennsylvania GOP’s tactical brilliance and the Illinois Democrats’ relative lack thereof. My quick and dirty calculations nationally (not all precincts have reported, and AP doesn’t report uncontested races): 53,627,450 people voted for Democratic candidates for the House, or 50.6% of the two-party vote; 52,162,619 people voted for Republican candidates for the House, or 49.3% of the two-party vote. That’s pretty darn close to the presidential vote. But the GOP controls 27 state houses outright, compared to only 15 state houses controlled by Democrats (the rest are split or nonpartisan). That’s a lot of redistricting done by Republicans. The Republicans control the House not because more people voted for Republicans, but because either through GOP redistricting or chance, the Democrats had more “wasted” votes.
Only the losers complain about the refs, and lord knows the Dems won a whole lot last night. Until we change the way house districts are drawn, the rules of the game are what they are and the GOP played the game better than the Democrats. But I’m already hearing Boehner talk about how the House results are a “mandate” for Republicans. It’s important to keep in mind that the GOP House majority is structural, not democratic, and more people voted for a blue House than a red House.
[ Update: Dave Weigel at Slate notices the phenomenon in a couple posts; Ian Millhiser at Think Progress ran the same numbers I did for the national House vote; and Dylan Matthews at Wonkblog looks at the seat share/presidential vote disparity in the states where Republicans ran redistricting. Matthews’ final sentence:
This suggests that it’s going to be tough for Democrats to make big gains in the House until 2022, when the districts are drawn again following the Census. And for that to happen, they’d have to do quite well in the 2020 state legislature elections.
That’s right on. But the problems that keep the Dems from making gains in the House until 2022 will also make it really hard for them to do well in the 2020 state legislature elections.]
Pingback: More on Redistricting and the 2012 Congressional Elections | Musings of a Failed Political Scientist