GIVEN THE unpredictability of American politics during Donald Trump’s presidency, the contest for control of the House of Representatives has been remarkably stable. In the summer of 2017 Democrats led “generic-ballot” polls, which ask respondents which party they plan to support for Congress, by around seven percentage points. The opposition’s lead now sits just above eight points. Democratic candidates’ performances in special elections for vacant legislative seats, and Mr Trump’s approval rating of just over 40%, are also consistent with a wave election roughly of the magnitude seen in 2006, the last time the party re-took the lower chamber in a mid-term election.
That precedent is not as auspicious for Democrats as it might appear. Thanks to the Republicans’ advantage in opportunities to gerrymander districts after their mid-term victories in 2010, as well as Democratic voters’ increasing tendency to cluster in electorally inefficient cities, even a 2006-level share of the popular vote, at 53.5% (after accounting for uncontested races), would be consistent with a close race for the House. To feel more confident, Democrats would need one of two sources of comfort. They would either have to believe they were likely to enjoy a 2008-style wave—when they won nearly 55% of the vote—or need evidence that districts Republicans designed to be impregnable seven years ago are now ripe for the picking.
Two relatively new sources of data provide support for both these hypotheses. The first is the Democrats’ third-quarter fundraising haul. In districts where both the Democratic and Republican nominees have filed reports, Democrats collected nearly twice as much money from individual donors as Republicans have. The most obvious benefit this discrepancy offers the Democrats is the ability to out-spend their opponents. However, donations from the public, as opposed to those from political committees or parties, also serve as a helpful proxy for two hard-to-measure factors: candidate quality and voter enthusiasm. If Democrats are out-raising Republicans by such a margin, that suggests both stronger candidates and a base that is more revved-up than normal.
The other favourable piece of evidence for the Democrats is district-by-district polling. In a wide range of seats where voting histories suggested that the opposition party would face a steep uphill climb, multiple surveys have shown Democratic candidates in surprisingly close races. In West Virginia’s third district, which voted for Mr Trump in 2016 by a 76-24 margin, Richard Ojeda, an army veteran and Democratic state senator, who himself plumped for Mr Trump in 2016, is within a few points of the lead. Even more striking are districts expected to be competitive that Democrats appear to have already put away, such as Iowa’s first or Colorado’s sixth. District-level polling has offered some bright spots for Republicans as well, but far fewer.
Individual district polls are of course prone to large errors. However, they are just as likely to underestimate the Democrats’ performance as the Republicans’. Although Democrats do not hold clear leads in enough seats to take control, they are in the hunt in so many different contests that the Republicans would probably need a systematic national polling error in their favour to hold on to the House.
Such misfires do happen, with some frequency. In both 2012 and 2016, district polls lowballed one party’s House candidates by nearly three percentage points. A repeat of the 2016 miss could mean a long election night—and perhaps weeks of uncertainty, since California continues to count votes into December.
The Democrats’ fundraising makes life hard for election handicappers. Since no party has ever dominated the money race to this degree, there are no relevant historical examples on which to train a statistical model. Extrapolating the impact of campaign contributions in the past, The Economist’s forecast would give the Democrats a 92% chance to take the House. However, the party may well have approached or surpassed a yet-unseen upper bound on the value of fundraising. Removing this variable from the model reduces the Democrats’ chances to 79%. Either way, they are clear favourites but far from a sure thing. It was only two years ago that an unlikely but similarly plausible election result put Donald Trump in the White House.