In its final stages, the tumultuous 2018 midterm election appears to be moving in contradictory directions, with Democrats and Republicans alike finding legitimate reasons for optimism amid the daily flurry of new polls.
But these seemingly disparate signals actually represent a mirror image of the same powerful force: a tightening connection between voters’ attitudes about President Trump and their preferences in House and Senate races next month.
That dynamic is lifting Republican hopes in the red-state Senate contests central to the battle for the upper chamber. It’s also dimming Democrats’ prospects in the exurban and small-town districts mostly on the periphery of their target list in the House. But it is simultaneously solidifying the Democratic advantage in many of the white-collar suburban seats, especially those that preferred Hillary Clinton over Trump in 2016, key to their hopes of recapturing a House majority.
“We are on two different tracks, moving in two different directions,” says Tom Davis, the former chair of the National Republican Congressional Committee.
Both sides of this equation capture the continued transformation of congressional elections into quasi-parliamentary contests that are turning less on assessments of individual candidates and more on broader attitudes about whether voters want Congress to enable the president or check him. “In the last few elections, we’ve moved more toward a parliamentary system where you are voting based on the leader and not necessarily on your local candidate,” says Glen Bolger, a longtime Republican pollster.
This long-term shift toward nationalized congressional elections decided largely by attitudes about the president is affecting contests both in the places where Trump is popular and in those where he’s disliked. That means the party most satisfied with next month’s results may be the one with the most candidates who succeed at swimming against that current: Democrats who find ways to win in Trump country and Republicans who hold on in the major metropolitan areas moving away from him. At this point, pollsters in both parties don’t expect to find many examples on either side. “It’s going to be tough for both,” says Mark Mellman, a Democratic pollster. Bolger concurs: “It’s possible that both are disappointed.”
The correlation between attitudes about the president and voting choices in congressional elections has been increasing over the past generation. In a recent paper examining long-term survey data, Gary Jacobson, a University of California at San Diego political scientist who specializes in Congress, found that the congruence between voters’ assessments of the president and whether they supported his party’s candidate in House and Senate contests ticked up in the 1980s and 1990s before rising more rapidly in midterm elections during this century.
Exit polls now track a powerful connection between attitudes toward the president and votes in both House and Senate elections. In the past three midterm elections—2006, 2010, and 2014—exit polls found that 84 to 87 percent of voters who approved of the president’s performance voted for his party’s candidates in House elections. Simultaneously, 82 to 84 percent of those who disapproved of the president’s performance voted against his party’s House candidates. Since most voters in each election disapproved of the president’s performance, his party suffered substantial House losses each time.
The general assumption among political operatives is that because senators are better known than House members and raise more money to create an independent identity through television advertising, it’s easier for them to resist this trend. And in almost every election cycle, there are senators who conspicuously notch victories in states that lean the other way in their presidential preference. For instance, two of the most threatened Democratic senators in this election—Claire McCaskill in Missouri and Joe Donnelly in Indiana—won in 2012 while their states were voting for Mitt Romney over Barack Obama, largely because each benefited from a deeply flawed opponent.
The general trend, though, is for Senate races also to increasingly align with attitudes toward the president. In 2006, Republicans won six of 10 Senate contests in states where exit polls put George W. Bush’s approval rating at 46 percent or higher, but lost 19 of 20 in states where he stood at 45 percent or less. In 2010, Democrats won nine of the 10 Senate races in states where Obama stood at 48 percent or higher, but only two of the 15 where he had fallen to 47 percent or less. (Joe Manchin, facing a tough reelection fight again this year in West Virginia, was one of those two.) In 2014, Democrats lost Senate races in 14 of the 15 states where Obama’s approval rating slipped to 42 percent or less. In the 2016 election, every Senate race was decided the same way as the presidential contest in that state. “In the ’80s and ’90s, you could point to so many individual senators who bucked the partisan trends in their states to win, and you can’t do that as much anymore,” says Bolger.
Many experts see indications that this long-term trend may reach a new peak this year. As Jacobson notes, the gap between Trump’s job-approval rating among voters in his party and the other is the widest ever recorded for a president during a midterm. Voters who strongly approve or disapprove of the president are also especially likely to vote for or against his party’s candidates, and the share of voters expressing such strong emotion about Trump—especially strong disapproval—is unusually high. All of this has manifested in the latest polls: The recent Washington Post/Schar School survey of 69 battleground House districts, for instance, found that 91 percent of those who approved of Trump’s job performance said they intended to vote Republican, while 88 percent of disapprovers said they would support Democrats. The latest national ABC/Washington Post survey reported similarly stratospheric numbers.
The emotional fight over Supreme Court Justice Brett Kavanaugh’s disputed confirmation—which inflamed gender, cultural, and partisan divides—is reinforcing this separation. GOP strategists see two benefits from the confrontation in red-leaning states and House seats: more Republicans saying they intend to vote and more voters focusing on which party they want to control Congress. But in the swing, mostly white-collar, seats that will likely determine control of the House, Democrats see their advantage hardening among college-educated white women and, to a somewhat lesser extent, white-collar white men, already recoiling from Trump. “Closing with these cultural arguments, meaning the Court fight, continues to move things further apart,” says Davis.
The bottom line for November 6 is that these developments strengthen the Republican chances in the Senate, where Democrats are defending 10 seats in states that voted for Trump; of the four Republican-held Senate seats Democrats are targeting—Arizona, Nevada, Texas, and Tennessee—all but Nevada are also states Trump won. Relative to his vote in 2016, Trump’s support has slipped at least somewhat in almost all of those states. Except for a concentration of four key Rust Belt states, Trump’s approval rating in most of them remains at about 50 percent or above in at least some surveys, which boosts Republicans. Not coincidentally, Democratic senators are in their best position in those Rust Belt exceptions: Ohio, Michigan, Wisconsin, and Pennsylvania.
In the House, these late-season trends may be simultaneously reducing the odds that Democrats win 45 seats but somewhat increasing their chances of winning at least the 23 seats they need to recapture the majority. Democratic hopes look to be fading in many of the strongly Republican-leaning exurban and small-town seats that made the outer edge of their target list earlier this summer. Democratic chances in those places always turned on low GOP turnout, and that now appears much less likely.
But, especially after the Kavanaugh confrontation, Republican strategists are growing glummer about their chances of holding well-educated and often diverse seats in suburbs around Denver, Kansas City, Chicago, Detroit, Philadelphia, Northern Virginia, parts of New Jersey, and Orange County, California. Distaste for Trump is “precisely why those OC seats are so troublesome,” says John Thomas, a California-based GOP strategist. “The attempts to localize the race were not effective. In these congressional races, people are coming to vote their stamp on Trump.”
Still, even in the House, the increased nationalizing of the election isn’t without risk for Democrats. They retain opportunities in enough suburban seats to win a majority, but with some of their longer-shot blue-collar and non-metro targets falling off the board, they can’t afford to squander many of their white-collar chances. That will be especially true if Republicans can swipe even a few of the blue-collar, Democratic-held districts that voted for Trump in 2016 (such as two in Minnesota). And while Democrats retain opportunities in a significant number of districts that narrowly preferred the president in 2016, it’s always important to remember that they can’t recapture the majority without making some incursions into Trump country, because he won more House districts (228) than Clinton did (207).
Whichever party controls the House or Senate after the election, the safest prediction about the trend toward increasingly parliamentary elections is that it will deepen the trench between the two Americas now contending for control of the nation’s direction: diverse major metropolitan areas in most states consolidating even more firmly around Democrats, and preponderantly white non-metro areas where Republicans remain dominant. It may be worth filing a screenshot of that contrast when the television networks feature their red and blue maps on Election Night—because it could be only a preview of the chasm ahead when Trump is on the ballot himself in 2020.