The Great Midterm Divide

The safest prediction one can make about this year’s congressional elections is that the voters who decide them will look very different from the ones who settled the 2012 presidential contest. The share of minorities and, especially, young people in the electorate will almost certainly decline; the proportion of whites and, especially, seniors will increase.

This shift isn’t new. Midterm elections have long attracted fewer voters than elections in presidential years have, with minorities and young people among the groups most likely to stay home. In the five presidential elections from 1992 through 2008, exit polls conducted for a consortium of media outlets found that voters under 30 cast, on average, 18 percent of the ballots; in the five midterms that immediately followed those elections, young people accounted for just 12 percent of the votes. Voters over 65, by contrast, increased their share of the vote from 15 percent to 19 percent. The decline among minorities hasn’t been as consistent or as severe, but their share of the vote dropped two percentage points from 2004 to 2006, and three from 2008 to 2010, which are big shifts as these things go.

But while the voting falloff between presidential-year and midterm elections has remained constant, its impact has been vastly magnified by a racial and generational realignment that has remade each party’s base of support since the 1980s. In presidential and congressional races alike, Democrats today fare best among minorities, Millennials, and white voters (especially women) who are single or college-educated. Even in a country rapidly growing more diverse, Republicans still rely almost entirely on whites, running best among those who are older, blue-collar, married, rural, and male. In other words, Democrats have become increasingly reliant on precisely the groups most likely to sit out midterms, while Republicans score best among those most likely to show up.

The consequences of these shifts are so profound that political analysts increasingly talk about two American electorates: the one that picks presidents (and has awarded Democrats the popular vote in five of the past six presidential races) and the one that determines midterms (which have usually favored Republicans since 1994). For different reasons, this divergence isn’t particularly good news for either party.

It hasn’t always been this way. In the 1986 midterms, for instance, congressional Democrats won exactly the same share of seniors (55 percent) as they did voters under 30. Nor was that convergence unusual. In the four congressional elections from 1986 through 1992, the biggest gap between Democratic support among voters under 30 and among those over 60 was the party’s two-point advantage among older voters in 1990. In addition, during those years, the differences in voting preference between whites and nonwhites were less dramatic than they are today. Democratic congressional candidates performed better among minority voters than among white voters, but they didn’t face the cavernous deficits with the latter that had been common for the party’s presidential nominees since 1968. In those same four elections from 1986 through 1992, exit polls found that Democratic congressional candidates narrowly carried white voters twice, narrowly lost them once, and split them in the remaining case. All of this dampened the impact of the shift toward an older and whiter electorate in midterm elections.

But starting in the 1990s, and accelerating after 2000, the preferences of old and young, and white and nonwhite, have separated more sharply. The change has come in two stages, starting in 1994. In the backlash against Bill Clinton’s chaotic first two years, white voters backed GOP congressional candidates by a resounding 16-point margin. And in every congressional election since, Republicans have outpolled Democrats among white voters, six times by commanding double-digit margins.

The second important change followed a few years later. After 2000, the political preferences of young and old voters rapidly diverged as the first members of the racially diverse, socially liberal Millennial generation (generally defined as those born after 1980) entered the electorate, and Democratic-leaning Franklin Roosevelt seniors were replaced by the more Republican-leaning Silent Generation and early Baby Boomers. As a result, congressional Democrats have run at least nine points better among young voters than among seniors in each of the past five elections (and at least 16 points better in the past two).

In other words, the racial and generational difference in participation between presidential-year and midterm elections is long-standing; it’s the more recent divergence in preferences that has resulted in the GOP’s midterm advantage. Other factors, of course, also shape the results in these off-year contests. More often than not, the party that won the previous presidential election loses seats in the subsequent midterm. When the incumbent president is unpopular (as Obama is now), his party’s losses are typically greater. And Senate results are always heavily shaped by the map of states on a given year’s docket. But distinct from all these cyclical factors, the electorate’s composition now stands as a structural advantage for the GOP in off-year elections. And in a year like this, when the midterm electorate’s customary whiter and grayer complexion converges with low approval ratings for a Democratic president and a Senate battlefield centered on red states, Democrats understandably feel as if they are caught between colliding storm systems.

Still, the party has some reason for optimism. Democrat Terry McAuliffe won the Virginia governor’s race last year largely because the party mobilized young people and minorities more effectively than in the past, using the same voter-engagement tools that Obama’s team employed in his two presidential campaigns. “Our ability to more directly communicate with people, to find people we were missing … has had an impact,” says Jen O’Malley Dillon, who ran Obama’s field operation in 2012 and is now advising the Democratic Senatorial Campaign Committee on how to boost turnout. The Republican pollster Bill McInturff agrees that it’s wrong to assume Democrats will suffer their traditional slump, particularly in the battleground Senate states both parties are targeting. “There is so much money [going into the races] that we need to be cautious in our old assumptions about composition of the electorate.”

For Democrats, lagging midterm turnout is a key reason the party has controlled the House for only four of the past 14 years it’s held the White House, crimping its ability to implement its agenda. But the turnout differential is also a problem for Republicans, albeit in less obvious ways. The GOP’s historic gains in 2010 discouraged it from making the policy adjustments it needed to appeal to the larger, younger, and more diverse presidential-year electorate of 2012. Instead, the new Republican House majority, believing it had won a national mandate, pursued a confrontational and invariably conservative course that both hurt the party’s overall image and pulled its 2012 presidential contenders to the right on issues from taxes to immigration (remember “self-deportation”?). That tug ultimately diminished nominee Mitt Romney’s appeal to the broader pool of general-election voters.

Something similar could easily happen in 2016 if this year’s older and whiter electorate delivers another big midterm win to the GOP. A Republican majority in both congressional chambers would likely confront Obama aggressively—for example, by voting to repeal any administrative action he takes to provide legal status to millions of undocumented immigrants, or voting to repeal the Affordable Care Act after millions have obtained insurance coverage from it. Those actions would be difficult to sell to the younger and more diverse presidential-year electorate. But conservative momentum following a 2014 breakthrough could create irresistible pressure on the GOP’s next class of White House hopefuls to endorse the congressional agenda, even if doing so complicated efforts to, for instance, woo enough Hispanics to recapture Colorado or Nevada in 2016.

Regardless of what happens next, the turnout gap has already contributed to the whiplash nature of modern politics, with voters careening back and forth between the two parties. This instability has encouraged both sides to treat every legislative choice primarily as an opportunity to score points for the next election. Oscillation, in other words, has encouraged polarization.

But the best news for the Democrats is that, whatever happens this year, eventually demographic change will overwhelm the turnout gap. While Millennials and minorities still participate at lower rates in midterms than in presidential elections, their presence is inexorably growing on both fronts: the minority share of the vote in off-year elections jumped from 14 percent in 1994 to 23 percent in 2010, and this year will likely come in somewhere between that figure and the 28 percent from 2012. If Republicans can’t attract more votes from the growing numbers of minorities, Millennials, and white-collar white women who have powered the Democrats’ success in recent presidential elections, demographics will ultimately threaten the GOP’s hold on the House, too. “Obviously the Democratic presidential coalition continues to expand,” notes Ruy Teixeira, a leading liberal analyst of voting patterns. “Eventually you reach the point where even turnout differentials aren’t enough to derail it.” That’s an encouraging long-term prospect for Democrats—but it may be cold comfort if lagging turnout among their best voters contributes to another brutal midterm this year.