The 2016 American National Election Study (ANES) Pilot Study asked respondents who they would vote for in a series of randomized general election matchups between Hillary Clinton and a series of Republican candidates (we focus on Donald Trump, Ted Cruz, and Marco Rubio). Bernie Sanders and John Kasich were not included in these matchups.

We model respondents’ choices between Hillary Clinton and the Republican candidate across the treatments of Trump, Cruz, or Rubio as the nominee. We include party ID, ideology, education, income, gender, race, and trade attitudes as predictor variables, also estimating interaction effects between GOP candidate, party ID, and education/trade attitudes/gender in a bivariate probit model.

Below we plot the interaction effects on predicted probabilities of supporting the Republican candidate. The main finding is that Trump divides the electorate by gender, education, and trade attitudes like no other. Among independents (coded “4” on the party ID scale), for instance, there are huge gaps in predicted Trump support on each of these variables. Across party identification, women, those with college/postgrad degrees, and voters supportive of free trade agreements are much less likely to support Trump than men, those with high school degrees or less, and voters opposed to free trade agreements.

These findings, of course, are hardly surprising, although the magnitude of the effects took us aback. If Trump is the GOP nominee, his path to victory would rely on the bloc of the electorate least likely to vote. Based on weighted values from the 2016 ANES Pilot Study, only 38% of independents with a high school diploma or less said they voted in the 2012 election (which, as a self-report measure, is itself an overestimate). This is a meaningful slice of the electorate—comprising about 6% of all voters, again with weights applied—but one whose turnout has historically lagged. Hence, the real uncertainty surrounding a Trump nomination.

Click images to enlarge





NOTE: “1” indicates less than high school education, “6” indicates postgrad degree.



NOTE: “1” indicates strong opposition to free trade agreements, “7” indicates strong support.