If you look at the polls and the forecasts put out by election oracles like Nate Silver, the race to win control of the House of Representatives on Tuesday seems tilted in Democrats’ favor. Michael Li will tell you that if Democrats do take back the House, it will have been against all odds.
In a report released earlier this year, Li and his colleagues at the Brennan Center for Justice found that in order to win a bare majority in the House, Democrats would need to win the national popular vote by 11 points, a margin no party has seen in decades. That’s thanks to gerrymandering, the process of strategically manipulating an election map’s boundaries to preference one political party or group over another. Since 2012, the country’s electoral map has been so heavily skewed toward Republicans that in severely gerrymandered states like Ohio, North Carolina, and Michigan, Democrats could win more than half the vote and still walk away with a minority of congressional seats.
“Gerrymandering will have a major impact on this election.”
Michael Li, the Brennan Center for Justice
It’s little wonder, then that the most competitive places for Democrats this cycle are states like California, Minnesota, New Jersey, and Pennsylvania, where congressional districts weren’t drawn by a single party. In order to win the House, Democrats will need to rack up seats in these places. In the rest of the country, Li says, “gerrymandering will have a major impact on this election.”
Republicans aren’t the only ones who gerrymander. Democrats have done it in states like Maryland where they have control. It just so happens that after the 2010 Census, during the last round of redistricting, Republicans controlled the majority of state legislatures and governorships across the country. And so, Republicans drew the maps to protect their power. They did this by packing Democratic voters into a limited number of districts and then scattering the rest of them across the remaining districts. That ensured that Democrats would always win a small number of districts by a landslide, but would consistently struggle to win even one additional district. Republicans even had a name for this plan. They called it REDMAP.
Li and his colleagues set out to study the effects of this plan in states across the country by measuring something called “responsiveness.” Responsiveness refers to the relationship between the number of votes a party gets and the number of seats they pick up. In responsive maps, those two numbers rise and fall together. In a non-responsive map, a party’s vote share can increase by 10 or even 20 percent and still fail to produce a single additional seat for that party.
For example, witness Ohio. The researchers calculated responsiveness in Ohio and 25 other states by studying Democratic vote share over the last three election cycles and analyzing what vote share would be required to flip each district in that state. In Ohio, they found that Democrats can get four out of 16 seats with about a quarter of the statewide vote. But to get the fifth seat, they’d have to win more than 55 percent of the vote. In other words, more than half of the vote would yield Democrats just a third of the seats.
Michigan’s map tells a similar story. According to the Brennan Center’s projections, Democrats could win about five out of 14 seats with around 38 percent of the votes, but would need nearly 55 percent of the vote to get that sixth seat.
And at the risk of being redundant, here’s North Carolina, where Democrats would need to win nearly 53 percent of the vote just to get a third of the seats.
Maps in all three of these states have faced challenges. This summer, a federal court struck down North Carolina’s map as a partisan gerrymander, but agreed to allow the map to be used for the November midterms. In Michigan, the League of Women Voters has filed a lawsuit against the state, arguing that the Michigan map is also an illegal partisan gerrymander.
Michigan residents can also vote Tuesday on a ballot initiative that aims to eradicate extreme gerrymandering by creating an independent redistricting commission led by citizens. There are four other anti-gerrymandering initiatives on the ballot Utah, Missouri, and Colorado, where there are two: one for congressional redistricting and one for the state legislature. And earlier this year, Ohioans voted overwhelmingly in favor of an initiative that requires 60 percent of state legislators and half of the minority party to approve any proposed map. If they can’t agree on one, map-drawing authority goes to a seven-person commission made up of lawmakers and regulators from both parties.
“If the courts don’t step in and do something about partisan gerrymandering, people seem to be ready to do that themselves, and that’s important,” Li says.
It’s easy to see why voters support these kinds of initiatives. In 2008, California passed Proposition 11, a constitutional amendment that put the map-drawing power in the hands of a 14-member independent commission. Since then, it’s become one of the least gerrymandered states, and therefore, one of the most competitive for Democrats. That’s reflected in the Brennan Center’s projections. Where in Michigan, Ohio, and North Carolina the line comparing vote share to seats plateaus, California’s is almost a straight line. As Democrats gain vote share, they gain seats.
Court-drawn maps also tend to produce more equitable results. In Pennsylvania, a map that was struck down in the courts last year required Democrats to get 56 percent of the vote in order to secure just six out of 18 seats. Here’s a look at the old map’s responsiveness:
The new court-imposed map being used for the 2018 election follows a much straighter curve, yielding Democrats half the seats when they win half of the vote.
There are, of course, exceptions to the Brennan Center’s predictions. In any individual race, there are other factors—like incumbency, candidate quality, and scandals that arise—that can impact a given candidate’s chances of winning. It’s also worth noting that the gerrymandered maps in place in 2018 were drawn to entrench Republicans in 2011. Given how much the party has changed under President Trump, it could be that the map is less reflective of the GOP as it stands today.
Court cases in Michigan, North Carolina, and Ohio stand to change the map slightly, but in most of the country, REDMAP will remain a constant through the 2020 election. “Regardless of whether Democrats manage to overcome it this year,” Li says, “it’s a reality that will be there through the lifecycle of these maps.”
That also means that even if Democrats do win back the House on Tuesday, two years from now they’ll face the unenviable task of defending those same seats all over again. And they’ll be doing it in the midst of a presidential cycle that’s bound to bring out more voters from President Trump’s base.
It’s an unfortunate truth for Democrats in an era of extreme gerrymandering: Until the maps change, no new seat is safe.
More Election Coverage from WIRED