Yes, midterm elections are almost always referenda on the incumbent president, but the real driving force is turnout. With about a third fewer people casting ballots in midterm elections compared with presidential-year elections, which side’s supporters actually turn out is critical. All things being equal, older people tend to vote in higher numbers than those who are younger, whites tend to vote in higher proportions than minorities, and both factors tend to help Republicans. But all things are rarely quite equal. In midterm elections, one party’s voters almost always, as my mom used to say, “have their noses out of joint” more than the other’s.

In 2006, with the war in Iraq going badly and President George W. Bush unpopular, it was Democrats and liberals whose noses were out of joint, while Republicans were unmotivated. The GOP suffered a disastrous electoral defeat, losing 30 House seats, six Senate seats, and control of both chambers.

In the midterm elections of 2010 and 2014 during President Obama’s tenure, it was conservatives and Republicans, particularly those of the tea-party variety, that were all torqued up and voted in disproportionate numbers while complacent Democrats stayed home. In 2010, Democrats lost 63 seats and control of the House, as well as six Senate seats—though they held their majority in that chamber. In 2014, Democrats lost nine Senate seats and their majority there along with 13 House seats, deepening their deficit in that chamber.

Bloomberg’s Josh Green reported an important story this week of a national survey of 800 registered voters, conducted Aug. 29-Sept. 2 by Public Opinion Strategies for the Republican National Committee, that showed almost half of self-identified Republicans do not believe that Democrats will win back control of the House. Of that group, 57 percent of those that consider themselves strong supporters of President Trump don't believe the GOP's control of the House is in danger, while only 37 percent of them think the Democrats can capture a majority.

The point of this is not to question the political prognostication prowess of Republicans or Trump supporters; it’s that they simply don’t feel threatened. Green quotes the study as saying to Republican Party leaders, “We need to make real the threat that Democrats have a good shot of winning control of Congress.” It would seem that the Trump base is hearing the president talking about a red wave and, as a result, is overconfident about the GOP’s standing even as the House is teetering on the edge and the Senate is split 51-49. And with three-quarters of the nation’s governorships and four-fifths of its state legislative seats on the ballot this year, the huge gains that the GOP scored on the state level during those Obama years are in real jeopardy.

Neil Newhouse, the POS partner who oversaw the study, was quoted in the report saying, “While a significant part of that lack of intensity is undoubtedly due to these voters’ sentiments toward the president, it may also be partly because they don’t believe there is anything at stake in this election.” Newhouse concludes: “Put simply, they don’t believe that Democrats will win the House.” Those with the strongest support for Trump were those who least perceived that the GOP majority was in danger.

The report went on to say, “Those voters who ‘somewhat approve’ of Trump and those who support the president’s policies but not his leadership style are the ones posing a challenge to the party. They are about a quarter of the electorate and are somewhat less supportive of GOP candidates and less interested in the upcoming election. While a significant part of that lack of intensity is undoubtedly due to these voters’ sentiment toward the president, it may also be partly because they don’t believe there is anything at stake in this election.”

The survey gave Democrats a 9-point advantage on the generic-congressional-ballot test with substantial gender and generational gaps; Democrats led among women of all ages but among men between 18 and 44 years of age as well. The 9-point Democratic advantage is consistent with the 8.4- and 8.9-point Democratic leads, respectively, in the RealClearPolitics and Nate Silver’s FiveThirtyEight averages of public polls.

To put the situation most simply, independent and swing voters are angry with Trump and disappointed in the Republican majority, and that portends a bad night for the GOP. But the difference between a bad night for Republicans and a horrific one is if those Republican voters who do support Trump strongly are overconfident and don’t vote in big numbers, while those who more tepidly back Trump and the party are unmotivated and stay home as well. With Democratic and liberal voters on the warpath, this is how catastrophic outcomes occur.

Interestingly though, it continues to look like the House, gubernatorial, and state legislative elections are being contested in one world while the Senate races, mostly being fought in states that Trump carried in 2016, are in another. RealClearPolitics’ “Senate No Toss Ups” chart, reflecting the averages of the state polling in each Senate seat, with each state pushed in the direction of the poll average, suggests no net change in the Senate at this point. If you just go by the public-poll averages, Democrats would edge out Republicans in three seats held by the GOP, beating incumbent Dean Heller in Nevada and winning open seats in Arizona and Tennessee, while Republicans would be unseating three Democrats, Bill Nelson in Florida, Claire McCaskill in Missouri, and Heidi Heitkamp in North Dakota.

Now there are real limitations to this method—the quality of state-level public polling varies enormously and is nothing like the sophistication of some of the best campaign pollsters on both sides—but it is an interesting exercise and one worth keeping an eye on.

This story was originally published on on September 21, 2018

More from the Cook Political Report

CPR 2024 National Polling Average
Cook Politcal Logo