How Trump Picked the Democratic Lock and Won the Presidency

How Trump Picked the Democratic Lock and Won the Presidency
AP Photo/Steve Helber, File
Story Stream
recent articles

The following is an excerpt from "Demographic Coalitions: How Trump Picked the Democratic Lock and Won the Presidency" – David Byler’s chapter in the newly published "Trumped: The 2016 Election That Broke All the Rules" (from Rowman & Littlefield). Ellipses indicate skipped parts of the text. Citations can be found in the book.

“Donald Trump Can’t Win the White House Because [fill in the blank].”

Throughout this cycle, journalists, academics, bloggers, and analysts wrote countless pieces with titles like that one. During the GOP primary, Trump attacked Republican elites rather than courting them and seemed unable to widen his appeal beyond 30 to 40 percent of the electorate, so many observers concluded he couldn’t win. During the general election, Trump constantly generated controversies, lacked a traditional ground game, and had historically low favorability ratings—which again led many to argue that he would inevitably lose.

Perhaps the most popular argument in this genre was that demographics would keep Trump from winning the presidency. The basic idea was that Trump was toxic with too many groups—college graduates, Hispanics, African-Americans, women, millennials, and more—and he had no way to put enough votes together to win. This argument isn’t new. Since before the beginning of the Obama era, a number of political data wonks have argued that as America becomes more racially diverse and well-educated, Republicans (particularly those who take a hard line on illegal immigration) will find it difficult, if not impossible, to win the presidency.

Had Trump lost, analysts would still be arguing over this issue. Many liberals would have taken a Hillary Clinton victory as a sign that the Obama coalition was unstoppable, and many conservatives would have argued that a different presidential candidate would have been able to persuade Democratic voters or energize new voters enough to win back the White House.

But (despite winning the popular vote) Clinton lost. If Trump or some other Republican can improve on his showing in some future election, he or she could win the popular vote as well as the Electoral College. So it’s worth asking: What happened? How did Donald J. Trump build a winning coalition in the face of demographic headwinds? How did the Democratic coalition change in this election? What do changes in both coalitions mean for the future of American politics? I used a data-driven approach to tackle these questions.

Essentially, I found that Trump used a two-step process—consolidating the Republican base and then earning massive levels of support from whites without a college degree—and that the parts of the Obama coalition that were supposed to doom Trump (African-Americans, Hispanics, and college-educated whites) didn’t show up in great enough number to beat him in the Electoral College. I’ll go into detail on both of those points and then discuss what that means for politics in the coming years.


Trump’s process for winning the Electoral College can basically be divided into two steps. First, he managed to keep the Republican base mostly in line. According to exit polls (note that exit polls are an imperfect measure, but they’re useful for broad descriptions), Trump won 88 percent of self-identified Republicans. He also won 80 percent of evangelical Protestants and 81 percent of self-described conservatives. Obviously there’s significant overlap between these groups, but these numbers illustrate that Trump was able to keep the ideologically conservative parts of the GOP base behind him. Trump also did very well with white voters. He equaled 2012 Republican nominee Mitt Romney’s 20-point margin among whites while winning both college-educated and noncollege whites. …

These data suggest (but don’t by themselves prove) that Clinton and Trump won many of the same voters, respectively, that Obama and Romney did in 2012. Clinton and Trump were both able to lay down familiar foundations before attempting to expand their coalitions.

This leads us to the second step of Trump’s strategy—run up the score with non-college-educated whites. According to national exit polls, Trump won whites without a college degree, 66 percent to 29 percent. That’s a 12-point shift in the Republicans’ favor from 2012. But topline numbers like these are only so informative. In order to really understand Trump’s coalition, we need to know where he gained, where he didn’t, and why he gained these votes from this group specifically.

We’ll tackle geography first—Trump’s gains weren’t evenly distributed across the country, as figure 2.3 shows.

The geographic pattern here is relatively clear. Trump made significant gains in the broad Midwestern/mid-Atlantic region. He gained in large, less populous swaths of Michigan, Wisconsin, Ohio, Iowa, Minnesota, the Dakotas, New York, Pennsylvania, and Maine. In other words, Trump made significant strides in the Central Northern region, and the Northeast outside the D.C.-Philadelphia-New York-Boston “Acela” corridor.

In most of those states, Trump seemed to gain because he won over areas with large concentrations of whites without a college degree. ….  In many of the most electorally important heartland states (e.g., Ohio, Pennsylvania, Iowa, Michigan, and Wisconsin), Trump improved on Romney’s showing in areas with high concentrations of non-college-educated whites. Trump also appeared to make gains with these voters in Minnesota and Maine—normally safe blue states that Clinton won by a relatively small margin. Linear regression on county-level results, like every other part of statistics, has its limitations (e.g., this regression looks at county-level data, so it doesn’t prove anything about individual behavior in the strictest sense), but it does strongly suggest that in key areas, Trump’s increased strength came from newfound support among blue-collar whites. This strong statistical relationship didn’t hold everywhere, however. In some states, longer-term trends complicated the picture.

For example, rightward trends in coal country threw off the predictive power of college education in Kentucky and West Virginia. Those two are perfect states for Trump—West Virginia has the lowest rate of college education among whites in the nation, and Kentucky has the second lowest. Both states have seen tough economic times, and neither has a large concentration of racial minorities that would push back on a heavily Republican white vote. Trump gained in both states, but education level among whites was not a particularly good predictor of where he gained. A quick look at the map in Figure 2.4 shows why.

Voters in Kentucky coal region (the southeastern portion of the state) didn’t shift to Trump as much as the rest of the commonwealth, and the vote in coal-rich, economically damaged southwestern West Virginia shifted less than one might expect based on educational attainment there. That’s partially because long-term trends allowed the GOP to pick up those regions before Trump even registered as a Republican.

The Democratic coalition has become more urban and cosmopolitan in recent decades (consider the contrast between  Bill Clinton’s and Barack Obama’s style and platform), causing Appalachia to move toward the Republicans. As Democrats embraced more environment-focused energy policies, voters in coal counties have come to view Republicans as a more reliable ally. In other words, some of the most economically downtrodden white areas of the country were moving away from Democrats before Trump—in some cases so much that he had limited upside there.

A similar pattern explains why education wasn’t as predictive in the South as in other regions. In Mississippi, Alabama, Georgia, Louisiana, Texas, and Arkansas, the predictive power of education was much lower than it was in many Midwestern and mid-Atlantic states. That’s partially because Republicans have been increasing their hold on white working-class voters in the South for decades. The South’s transition from a top-to-bottom Democratic stronghold to a supermajority Republican region is a complicated story that involves economics, culture, and race. It’s sufficient to say that voting in the South became racially polarized before Trump came on the scene—that is, Republicans often win supermajorities of white voters, Democrats win supermajorities of African-Americans, and that gives the GOP a substantial edge in almost all statewide and national elections across the region (apart from swing states Virginia, North Carolina, and Florida). In other words, non-college-educated whites in the South were already voting heavily for the GOP, so Trump could only gain so much with them. …

Excerpted from “Trumped: The Election That Broke All the Rules” (Rowman & Littlefield, 2017). All rights reserved. No part of this excerpt may be reproduced or printed without further permission in writing from the publisher.

David Byler is an elections analyst for RealClearPolitics. He can be reached at Follow him on Twitter @davidbylerRCP.

Show commentsHide Comments