In 2012, the Democrats infamously won a plurality of votes cast in House of Representatives races nationwide, but didn't end up winning back the majority of seats. This year, that's not the case — the votes are still being tallied, but right now the national House vote is 52 percent for Republicans and 45 percent for Democrats, according to a count kept by Dave Wasserman, Loren Fulton, and Ashton Barry of the Cook Political Report.
However, the GOP is certain to end up with more than 52 percent of the seats in the new House. Depending on the outcome of a few uncalled races, it looks like they'll win around 250 seats overall, which would mean they've won 57 percent of seats in the chamber. To help explain the discrepancy, here's a list of the states where the percentage of seats won by a party most differs from the party's actual total vote share in House races:
:format(webp):no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/2437830/pop_vote_vs_seats_corrected2.0.png)
Overall, 86 of these seats went for Republicans, and 48 for Democrats. Now, this isn't a simple metric of gerrymandering, for several reasons.
- First, many of these are small states — because in states with only a few House seats to go around, it's intrinsically more difficult to allot those seats proportionately to the vote. For instance, Idaho only has two seats to allot, so one party will have to end up with either 100 percent of 50 percent of them, even though the actual votes were split 64/36.
- Second, even in larger states there's a natural "winner's bonus" inherent to our majoritarian, single-member-district system. This is because the losing voters in each district don't affect the House delegation at all. When one simulates various districts and election outcomes under such a system, the party that wins more votes naturally ends up with some advantage in the overall seat share.
- Third, there were some House races where one party or the other didn't put up a candidate. A whopping 6 out of 9 races in Massachusetts were uncontested, and there was no actual voting in those races. So our party vote share totals for Massachusetts just uses the three races that were contested, which isn't ideal.
- And fourth, people choose where to live. Many Democrats live close to each other in urban areas. Because they're naturally "packed" in, it's unnatural to draw a map dispersing those voters across many districts. Maps that favor Republicans are just easier to draw — geography has a built-in pro-GOP bias. Jowei Chen and Jonathan Rodden wrote that a map accurately reflecting the partisan split in most states is "unlikely to emerge by chance from a nonpartisan process."
If one looks at the bigger states, however, we can see what look to be substantial effects of gerrymandering. In North Carolina, the GOP won 53 percent of the votes and 77 percent of the seats (10 out of 13). In Pennsylvania, they won 53 percent of the votes and 72 percent of the seats (13 out of 18). In Ohio, they won 56 percent of the vote and 75 percent of seats.
However, some of those GOP advantages were countered in states like Maryland, where Democrats won about 57 percent of the vote but carried 88 percent of the seats (7 out of 8), and the aforementioned Massachusetts where just 3 of 9 races were contested, and all were won by Democrats. So overall, it looks like the net effect of gerrymandering this year might have padded the GOP's margin a bit — but not a ton.
