The first two months of 2018 have given both parties reasons for optimism. Republicans have cut Democrats’ lead in the FiveThirtyEight average of generic congressional ballot polls in half, from roughly 12 points in December to about six points today. As debates over unpopular health care and tax bills have subsided, news of strong economic data and record-high stocks (at least until late last week) has likely aided the GOP.
However, most new district-by-district fundraising and polling numbers are downright terrible for Republicans, even in seats previously thought to be safe. In the fourth quarter of 2017, 39 Republican House incumbents were outraised by at least one Democratic challenger, and private polls and special election results suggest Democrats are highly competitive even in some districts President Trump won by wide margins.
At first glance, these two data trends might seem at odds with each other. How could Democrats’ lead in national polls be shrinking while their odds in individual districts improve? The answer: the “macro” outlook for the House (national polls) and the “micro” view (district-by-district) aren’t diverging; they’re coming into alignment.
Democrats probably need at least a six or seven point lead on the generic ballot to win the majority, thanks to the GOP’s redistricting edge and Democratic voters’ tendency to cluster and waste votes in safe districts. Democrats have been above that threshold most of the past year. However, only gradually have the GOP’s district-level problems come into view, as more Democrats announce candidacies and fundraising totals.
Republican leaders believe they can save their majority with a four-pronged approach: emphasize strong economic fundamentals, a muscular national security posture, opposition research against untested first-time Democratic candidates and the possible return of Nancy Pelosi as speaker. But historically, it’s been difficult to frame midterms as anything other than referenda on the president and party in charge.
This week, we’re shifting our ratings in 21 races towards Democrats. If anything, that still understates Democrats’ potential in individual races. If Democrats win the national House vote by six points (as today’s polls indicate), House control would be a coin flip. But according to our new ratings, if each party were to win an even number of Toss Up races, Democrats would only win 13 or 14 seats — ten shy of the 24 they need.