The Four Groups That Will Decide the Presidential Race

The outcome of the election hinges on how pronounced a handful of demographic trends turn out to be.

Early voting in North Carolina (Jonathan Drake / Reuters)

For all the turmoil, turbulence, and sheer reality-show melodrama of the 2016 presidential campaign, the actual results appear more likely to deepen long-standing trends in the electorate than to shatter them.

That’s been one of the paradoxes of this extraordinary election year. With both Donald Trump and Hillary Clinton facing unfavorable opinions from a majority of the voters, this has been a demolition derby of a campaign that has left both sides sputtering toward the finish line with dented fenders and cracked windshields. Yet a race that has unfolded like no other still appears on track to reinforce and intensify the trends that have defined the competition for the White House over the past quarter-century.

The polarizing nature of Trump’s candidacy in particular is pushing many of the dynamics that have shaped the electoral competition since the 1990s to new heights. With his brusque message of defensive nationalism, he is well positioned to extend the GOP advantage in some places where it is already strong, both demographically (working-class whites and evangelical Christians) and geographically (non-metropolitan areas, Appalachian and Interior Plains states). But he appears certain to compound the party’s problems among voters (college-educated and secular whites, minorities, Millennials) and in places (the nation’s largest urban centers, coastal states) where the GOP was already facing crippling deficits. In particular, the distance between blue-collar white voters drawn to Trump with passionate intensity and both the college whites and minorities resisting him may reach record heights.

The cumulative effect may leave Republicans relying even more heavily on the voters and regions most uneasy about the United States’ cultural and demographic change, or what I’ve called the “coalition of restoration.” Conversely, the election could substantially expand the Democratic advantage among the groups and regions most comfortable with those social changes, what I’ve called the “coalition of transformation.” Whoever wins, the safest prediction is that this election will widen every divide that fractures American politics—along lines of race, education, generation, and geography.

A Clinton victory would mean that Democrats have won the popular vote in six of the past seven presidential elections, or since 1992. That would be unprecedented: No party has won the popular vote six times in seven tries since the formation of the modern party system in 1828. Conversely, a Trump win would measure just how much separates a major portion of the electorate from the leadership class in virtually every American institution, ranging from business to national security to media, in the form of newspaper editorial boards—all of which have coalesced in virtually unprecedented fashion against the tumultuous GOP nominee.

A series of demographic and geographic factors will determine the result tomorrow, while also sending critical signals about the future direction of American politics. Together, those underlying elements amount to the tectonic plates of the 2016 election. Today we explore the key demographic plates that will shape the election’s outcome; tomorrow we will look at the principal geographic dynamics.

Does the class inversion deepen?

One of the defining characteristics of American politics over the past generation has been the class inversion: the reversal of political allegiance among blue- and white-collar white voters. Through the first decades after World War II, every Democratic nominee from Adlai Stevenson through Jimmy Carter consistently ran better among white voters without a college education than whites who held advanced degrees.

But Republican gains, starting in the late 1960s, among whites without degrees and Democratic advances among college-educated whites, which accelerated in the 1990s, have reversed that pattern. Starting with Al Gore in 2000, every Democratic nominee has won a higher share of the vote among whites with a college degree than whites without advanced education. In his first victory in 2008, President Obama ran seven points better among college-educated than non-college-educated whites, the widest such gap ever for a Democratic nominee.

This campaign seems poised to shatter that record and accelerate the class realignment of the two parties’ coalitions. From the outset of his candidacy, Trump has established a visceral connection with many non-college-educated white voters, especially men. Competing in a 17-person field in the GOP primary, Trump still won nearly half of all non-college-educated Republicans; he has led Clinton, usually by gaping margins, among that group in virtually every general-election national survey.

Republicans almost always now win working-class whites. The only Democrat to carry them since 1980 was Bill Clinton in 1992 and 1996, and he never attracted more than 44 percent of them. (That was enough for a slim plurality, though, because many working-class whites supported Ross Perot’s third-party candidacies.) The real question for Trump is how far he can push the margin among these voters. In each election since 2000, exit polls show that the GOP nominee has carried non-college-educated white voters by at least 17 percentage points; Mitt Romney won them by 26 points, drawing 62 percent of them in 2012. Pre-election polls show that Trump has a strong chance to extend that advantage. The ultimate mark of success for him would be to rival Ronald Reagan’s dominant performance in 1984, when he won 66 percent of non-college-educated whites, beating Walter Mondale among those voters by 32 percentage points. (Whether Trump can reach that high will likely depend largely on his performance among blue-collar white women, as discussed below.)

But while Trump has made these inroads among working-class whites, he faces unprecedented resistance among whites holding at least a four-year college degree. Though Democratic nominees now routinely run better among whites with a degree than those without one, none of them have run well enough to actually win most college-educated whites. In fact, no Democratic nominee in the history of modern polling, dating back to 1952, has ever won most whites holding a four-year degree or more, according to exit polls and the American National Election Studies. Hillary Clinton appears poised to break that record: Almost all pre-election polls have shown her leading among college-educated whites. She seems virtually certain to at least double, if not triple, the widest Democratic advantage ever among college-educated white women, which was Gore’s 8 percentage points in 2000. As of Saturday, the ABC News/Washington Post tracking poll gave her a thumping 25-percentage-point lead with these women.

Less clear is how college-educated white men will vote. Many of these ordinarily Republican-leaning voters—the GOP nominee has carried them by double digits in all but three elections since 1980—express skepticism toward both candidates, and polls have varied widely on their preferences. Some late surveys show more of them drifting back toward their usual Republican inclinations, though the ABC News/Washington Post tracking poll as of Saturday showed them breaking about evenly between the two rivals. At the least, Trump appears likely to fall well short of margins of 20 percentage points or more that these men have given the GOP nominee in three of the past four elections.

These contrasting trends among college-educated and non-college-educated white voters seem certain to produce a much wider gulf between the two groups than the record seven-point gap President Obama saw in 2008. In the final NBC News/Wall Street Journal national poll released Sunday, Clinton ran fully 22 points better among college-educated whites—leading among them and drawing 47 percent support—than non-college-educated whites, where she drew just 25 percent support. That could point to a lasting new order in American elections.

Two other trends in the class breakdown among whites are worth watching. One is regional variation. Democrats since 1992 have dominated the five key swing states in the Rustbelt—Ohio, Iowa, Michigan, Pennsylvania, and Wisconsin—largely because they have performed slightly better among these states’ large populations of blue-collar whites than Democrats have among those voters nationally. Trump’s hopes may turn on whether he can end that advantage in enough states.

Conversely, one reason Democrats have struggled to penetrate the GOP hold on the South is that they have underperformed their showings elsewhere among college-educated whites, many of whom in Dixie are more culturally conservative than their counterparts in other regions. Hillary Clinton’s performance among college-educated whites in the suburbs of Charlotte and Raleigh will likely determine whether she recaptures North Carolina, which tilted to Romney in 2012 after Obama carried it four years earlier. Even if Hillary Clinton does not win Georgia or South Carolina, not to mention Texas, Democrats will be closely watching for signs of inroads in white-collar suburbs that would be indispensable to any future victories there. (The same is true of Arizona in the southwest.)

The last factor worth watching for is the turnout of white-collar and blue-collar white voters. The two main data sources on the electorate’s composition differ about the relative presence of the two groups, with the Census Bureau showing non-college-educated whites to be a larger share of the total vote than the exit polls conducted by a consortium of media organizations do. Yet, as America diversifies and education levels rise, both data sources have shown non-college-educated whites declining as a share of the electorate by about three percentage points every four years, with minorities steadily growing—as discussed below—and college whites either holding steady or slightly increasing.

Since 2012, a loud faction of GOP thinkers has argued that the key to the party’s revival was not so much broadening its reach to minority, Millennial, and socially liberal voters who have largely rejected the party, but increasing turnout among blue-collar, religiously conservative, and non-urban whites. No candidate seems better designed to execute that mission than Trump, who has built his campaign around both the priorities and resentments of working-class white America. Trump is generating huge margins with those voters, but he has also clearly demonstrated one flaw in that strategy: Any message polarizing enough to mobilize millions of “missing” blue-collar white voters also risks alienating not only minorities but many white-collar whites. The results will measure whether any votes Trump gained among blue-collar whites was equaled or exceeded by losses among minorities and white-collar whites. A Trump victory would provide validation for those who want the GOP to prioritize mobilizing conservative whites over outreach. But if the blowtorch intensity of Trump’s courtship can’t reverse the steady decline in non-college-educated whites’ share of the vote, the missing-white-voter theory will look like a dead end.

Does the minority presence continue growing?

Beyond the class inversion, the other key reason for the Democrats’ revival in presidential politics since 1992 is the growth of the minority population. Although change hasn’t come to the electorate as fast as it has to the overall society, the minority share of the vote has roughly doubled since 1992, whether one measures it from the exit poll or Census data. In 2012, the Census put the minority vote share, 26 percent, slightly below the level in the exit polls, 28 percent. But both show non-white voters rising by about two percentage points every four years as a portion of the total votes cast.

One reason for that growth in the total minority vote under President Obama was extraordinarily high levels of African American turnout; black voters actually turned out at slightly higher rates than whites in 2012. The Hillary Clinton campaign doesn’t expect black voters to reach that peak again; the initial trends in early voting among African Americans have prompted concern among Democrats that the slippage could be greater than they expected, though those fears eased somewhat by the weekend.

Yet, even if black turnout dips, the total minority vote share could rise anyway because non-whites represent a larger share of the eligible voter pool than they did four years ago. (A smaller percentage of a larger pool could still generate an increase in voters.) In particular, fully four million more Latinos are eligible to vote this year than in 2012, according to the Pew Research Center. In 2012, only a little less than half of eligible Latinos voted. But early voting among Latinos, driven by antipathy to Trump, has soared in key states. Long lines stretched deep into the Friday-night deadline for early voting in Las Vegas; in Florida, the Clinton campaign on Friday said more Latinos had already voted than in the entire early-voting period in 2012. Texas and Arizona have reported big spikes in participation, too.

Latino Decisions, a Democratic polling firm that focuses on Hispanic voters, initially projected that between 13.1 million and 14.7 million Latinos might vote in 2016, up from just over 11 million in 2012 and just under 10 million in 2008. But the surge in early voting has convinced them that those projections might be too low: Now they say it’s possible that Latinos reach 15 million. Population trends could also slightly bump up the combined share of the vote cast by Asian Americans and mixed-race voters from its five percent in 2012.

If the minority vote share continues rising along the roughly two-points-per-four-years trajectory of the past quarter-century, Trump’s hill will get steeper—and the missing-white-voter theory will grow more implausible. Margins matter, too. Before Obama, Republicans often won 10 percent to 12 percent of African Americans; if Trump can regain that modest foothold, it will boost him in states like Ohio, Pennsylvania, and North Carolina, and help him repel Hillary Clinton’s unexpected advance in Georgia. Trump made a steady push for black support during the general election, but is laboring under the stain of his leadership in the “birther” claim that Obama was not born in America, as well as his embrace of barbed law-and-order rhetoric in this campaign. Democrats are hoping that Trump’s sharp language on immigration will produce record margins for Hillary Clinton not only among Latinos, but also among Asian Americans. In 2012, exit polls showed Obama carrying Latinos by 44 percentage points and Asians by 47 points.

How far does the gender gap extend?

In January, one Republican strategist told me the gender gap in a Trump-Clinton election could resemble the Grand Canyon. That prediction looks prescient. The key to the gender gap’s magnitude may be how far into traditionally Republican groups Trump’s troubles among women extend. He’s facing cavernous deficits among women of color, but that’s not particularly unusual for a GOP nominee. As already noted, he also seems virtually guaranteed to lose college-educated white women by the biggest margin for a Republican ever. But those women usually tilt Democratic, too, if not by such decisive margins: Although Obama lost them by six points in 2012, the Democratic nominee had carried them in four of the previous five elections.

Trump would face even greater difficulties if he also cedes substantial ground among white women without a college education. These women, once described as “waitress moms,” are often economically strained and more culturally conservative than their white-collar counterparts, and they have typically leaned Republican. Bill Clinton in 1996 is the only Democratic presidential nominee to win them since 1980, and in each of the past three elections they have preferred the GOP nominee by at least 17 percentage points. Like the ordinarily Republican-leaning college-educated white men, these women have appeared torn between their partisan leanings and their ambivalence, if not hostility, toward Trump; both these groups have proven the most volatile in polling leading into the election. As of Saturday morning, the ABC News/Washington Post tracking poll showed Trump leading among these women by a resounding 34 percentage points, even greater than Reagan’s advantage in 1984; but an NBC News/SurveyMonkey poll released a few days earlier showed him holding a much more modest 12-percentage-point edge.

The marriage gap will be worth watching, too. Polls have shown Trump facing heavy resistance from single white women, who have voted Democratic in each election since 1992—although Obama’s margin among them sagged to just six percentage points in 2012, by far the smallest Democratic advantage over that period. Even more worrisome for Trump would be erosion among married white women, who have preferred Republicans in every election since 1984 and gave Romney a 25-percentage-point margin last time.

Trump is guaranteed to win white men, though his difficulties among those with college degrees mean he may struggle to match Romney’s 27-point advantage among all white men in 2012, which was the highest for Republicans since 1988. Since 1980, Bill Clinton in 1996 is the only Democrat to carry even a plurality of white women; Hillary Clinton has a real chance, though no guarantee, of becoming the second to do so.

Can Democrats retain their Millennial advantage?

Since the first Millennials cast presidential ballots in 2000, Democrats have built a widening advantage among young people. In 2000, when only the oldest Millennials voted, voters under 30 split about evenly between Gore, at 48 percent, and George W. Bush, at 46 percent. As more Millennials entered the electorate, John Kerry pushed that number to 54 percent in 2004, before Obama won two-thirds of younger voters in 2008. Obama suffered some erosion in 2012—he lost white Millennials after winning them four years earlier—but overall still captured 60 percent of voters under 30 in his reelection.

The stakes have increased in the battle for Millennial allegiance as the generation has swelled its presence in the electorate. This year, for the first time, Millennials will equal baby boomers as a share of eligible voters, though they won’t match the boomers as a portion of actual voters because they vote at lower rates. But by 2020, Millennials will be a larger share of eligible voters than any other generation, and they will almost certainly be the largest generation among actual voters, too. And for the first time in 2020, the post-Millennial generation will enter the electorate. The non-partisan States of Change project forecasts that by 2020, Millennials will represent 34 percent of eligible voters and post-Millennials another 3 percent. Those groups will race past the 28 percent of voters from the baby boom. (And by as soon as 2024, Millennials and post-Millennials will approach 45 percent of the electorate, while boomers will shrink to about one-fourth.)

This generational transition will dramatically accelerate the electorate’s racial reconfiguring. About 80 percent of baby boomers are white—largely because the United States virtually shut off immigration between 1924 and 1965. But 44 percent of Millennials, and nearly 49 percent of the post-Millennials, are non-white, according to calculations by Bill Frey, a Brookings Institution demographer. The contrast between an increasingly diverse youth population, and a preponderantly white senior population—what I’ve called “the brown and the gray”—seems likely to dominate American politics for years to come.

Hillary Clinton has never connected easily with Millennials: Obama carried nearly 60 percent of them against her in their 2008 primary duel, and she lost 71 percent of them to Bernie Sanders during this year’s Democratic primary, faring about as poorly among Millennial women as men. But polls have consistently shown that Trump is deeply unpopular with this group, with large proportions describing him as racist, disrespectful to women, and unqualified for the presidency. At a time when Millennials recognize they embody the United States’ increasingly diverse future, Trump has stamped the GOP to many of them as a party determined to restore a past dominated by white men.

Despite Millennial ambivalence about Hillary Clinton, late surveys showed them moving toward her in much greater numbers than earlier in the campaign, when many expressed support for Gary Johnson or Jill Stein. Republicans face the very real risk of a third consecutive wipeout with Millennial voters—and the possibility that their nominee may attract less than 30 percent of them just as they are on the verge of becoming the electorate’s largest generation.

Hillary Clinton hasn’t particularly inspired Millennials, but even more than Obama she has aligned the Democrats’ agenda with those voters’ priorities on criminal-justice reform, immigration, gay rights, gun control, and climate change, to name a few subjects. If she wins, she will have the opportunity to cement the Democrats’ hold on them. For Republicans, further alienating the Millennial generation—and the even more diverse young people who will begin filing into voting booths behind them in 2020—may be the heaviest price that Donald Trump’s candidacy imposes on the party, even if he overtakes Hillary Clinton by mobilizing his impassioned coalition centered on older and blue-collar whites.

Assistant editor Leah Askarinam contributed.