The party in power usually takes a beating in midterm congressional elections. But despite Donald Trump’s ostentatious incompetence and plummeting popularity, it looks like the Democrats could well fail to take back a House majority. And that’s at least in part because while Democrats have a national popular vote majority, they are not winning voters in the right places. In Vox, Matt Yglesias lambasts this apparent unfairness:
If elections for the US House of Representatives were held today, polling averages suggest Democrats would get a little bit over 54 percent of the vote. […]
But here’s the thing. According to Elliott Morris’s model for Decision Desk HQ, 54 percent of the vote won’t deliver Democrats a landslide House majority. […]
There is a somewhat tedious debate involving political scientists, journalists, and election analysts as to whether we should characterize this situation — in which 54 percent of the vote wins Democrats 47 percent of the seats — as the result of “gerrymandering” or just “clustering” into an inefficient geographical pattern.
But whatever you call it, it’s an ugly number.
To the extent that this is the product of Republican gerrymandering, it is indeed a serious problem. Districts should be drawn by non-partisan commissions, not state legislatures. But according to one recent estimate, redistricting has only produced about 12 extra Republican seats, on net. That means that even if all recent redistricting were rolled back, the Democrats would still struggle to win the House with a lopsided 54 percent popular vote majority in individual House races. Why? Because the GOP draws on a geographically dispersed electoral coalition, while Democrats depend on super-majorities in districts around metropolitan areas, especially on the coasts. It doesn’t matter whether Nancy Pelosi wins 75 percent or 85 percent of the vote in her San Francisco district; the seat is still safely blue.
But perhaps the U.S. system of territorial representation itself is unfair. Perhaps, as Dylan Matthews said on Twitter, “if you get 54 percent of the vote, you should get 54 percent of the seats, regardless of where your voters live.” That sounds intuitively compelling. But people don’t vote for a generic, depersonalized party for Congress; they vote for a specific individual. So the claim that “Democrats” will get 54 percent of the vote is a shorthand that doesn’t capture the distinctiveness of individual elections, where some voters cast ballots based on a candidate’s personality and charisma and not just his party affiliation.
But let’s grant that the Democrats are systematically disfavored, as a party, because of their super-strength along the coasts. There is a long-term and constitutional logic to this even if it sometimes seems to create undesirable political outcomes. One of the purposes of proportional representation is to discourage regional factionalism. The American electoral system punishes parties for relying too heavily on a certain region of the country for their political strength. This can be a good thing—it’s not conducive to productive politics if the parties are geographically walled off from one another. In extreme cases, if a certain region of the country drifts too far away from another region politically, and the minority region is out of power at the federal level, that could set the stage for secession or Civil War. At a time when political violence in the United States seems to be on the upswing, it’s especially easy to see the perils of especially balkanized political coalitions.
The U.S. constitutional system is currently all but screaming at Democratic Party to broaden its geographic base, and threatening it with continuing powerlessness if it fails to do so. A more centrist Democratic Party of the Bill Clinton variety would peel off GOP support in the heartland, and perhaps prompt the Republicans to move away from white working class populism and try to compete in some districts closer to city centers. It may be unfair that Republicans have no incentive to move towards the center first. But such a party system would almost surely be more stable. It remains to be seen whether the framers’ design will actually pull us back in that direction, or whether it will merely continue to be a vehicle for ever-more-vituperative polarization and distrust.