We’ve all had several months to make up our minds about whether Trump voters are racist or not, and one might think that there’s nothing more to be said on the topic. But the American National Election Studies (ANES) group recently released its 2016 data file – the gold standard of election surveys in the United States, with data on thousands of subjects who answered hundreds of questions about political topics, including voting behavior.
I’m especially interested in this data set because it asks not only whom the respondents voted for in the 2016 election, but also who they voted for in the 2012 election. Most “Trump voters” are simply loyal Republicans, which means that analyzing them tells us a whole lot about Republicans and very little about what’s distinctive about “Trumpism.”
But if I limit my analysis to former Obama voters, I can learn two interesting things:
- What’s so special about Trump, that convinced an unusually large number of former Democrats to vote for him?
- What sorts of things matter to a subgroup of Trump voters that might be unusually easy for the Democrats to win back, given that they have proven themselves willing to vote for a (black) Democrat in the past?
On to my first analysis:
Who were the Obama/Trump voters?
(This is my first WordPress post after a long stretch of writing sloppily on Facebook, so I’ve yet to settle on a format. I’m not planning on including many charts or graphs in this series of posts, but if readers have specific requests or questions I will try to address them.)
The first thing I did was to run some simple descriptive numbers on three groups of survey respondents who said they voted in both elections: Obama/Clinton voters, Romney/Trump voters, and Obama/Trump voters. I also looked at Romney/Clinton voters, but there were fewer of them and they were not the main focus of my investigation.
First, I found that Obama/Trump voters were slightly more likely to be white than Obama/Clinton voters; much less likely to be black, but only slightly less likely to be Hispanic or Asian.
Based on this finding, the specific question is “which white (or at least, non-black) Obama voters voted for Trump?” In order to focus more narrowly on that question, I limited all further analyses to white voters only.
White Obama/Trump voters:
- Were much less likely to have a college degree than either Obama/Clinton voters or Romney/Trump voters.
- Had slightly lower incomes than Obama/Clinton voters and much lower incomes than Romney/Trump voters. This is the kind of thing you miss if you look only at Trump voters as a whole.
- Were somewhat older than Obama/Clinton voters and even slightly older than Romney/Trump voters.
- Were more likely to live in the Upper Midwest than Obama/Clinton voters, but also more likely to live in the broader Midwest and the Northeast. White Obama voters on the West Coast or in the Southwest were very unlikely to vote for Trump; so few whites in the South voted for Obama in the first place that it’s hard to say much about them.
- Were only slightly more likely to be male than Obama/Clinton voters.
Overall, these data paint a picture quite similar to the dominant media portrait of an Obama/Trump voter – older, working class, low income, and white; one might also guess “rural” but none of the questions in the survey seemed to address the urban/rural divide.
In part 2 of this series, I will describe the results of multiple regression analyses of white Obama voters, uncovering which survey answers predicted voting for Trump.