Thomas Wood studies political campaigns and their effects on voters’ attitudes and behavior. Wood recently worked as director of experiments and modeling for the Jeb Bush presidential campaign. He teaches "Voters and Elections." He discusses his experience during the Republican presidential primary and assesses student interest elections.
Tell us about your recent experience on the 2016 presidential campaign trail.
I was the director of experiments and modeling for the Jeb Bush campaign, which meant I worked day to day in the field. About every two weeks we would do one to five thousand surveys, and then we would take those answers, append them to voter and consumer data, and run models about the way that socioeconomic markers impacted primary ballot choices in the last 15 days.
When you think about primary voter data, you’re asking people to make a choice between 15 candidates from the same party, so you have to throw away most of the data that campaigns rely on. You are choosing between seven different types of vanilla ice cream. We observed quite quickly that class was going to be the most important cleavage between the Republican cohort, so from there, we had to spend time figuring out the best ways to measure socioeconomic indicators from survey data.
For every voter, we had 14,000 characteristics that boiled down to the magazines you read, whether you had ever bought mail-order steaks, the number of credit cards you have. The mail-order steaks piece of information is a big one. If you sent away for frozen steaks in the mail, you tended to fall into a higher socio-economic status. These funny little patterns tend to matter a lot in a primary race.
What we found worked well was county-level data indicators of real estate. For example, home heating type — whether or not you had electric heating or wood fire heaters — was a hugely predictive indicator of who you were going to vote for.
Thomas Wood, @osupolisci prof, shares his #election2016 experience and insights w/ @ASCatOSU #votertrends #ASCDaily
How are student attitudes towards this election different than the general population?
It doesn’t compare. The students at Ohio State are way more engaged in the political process in a way that is completely anomalous to the way the general public engages in politics. College is the weirdest time for examining mass attitudes because the environment is so much more diverse and inclusive than most American communities and workplaces.
I study unorthodox, heterodox mass attitudes, and there is a group of people who think all public opinion reporting is a shell game. It’s interesting to teach a class on mass attitudes when the perception is that the data you get from survey research is a con.
Students are very sophisticated, but it’s been difficult to discuss the election in class from a policy perspective. They are really on edge right now. What I’ve seen, though, is that both young Democrats and young Republicans are dissatisfied with modern society.
How did the experience of working in the field influence your teaching?
One thing that has changed is the way I talk to young undergrads. If a student has a capacity for statistics, I tell them to take all the computing classes they possibly can. There is such a gaping moor of talent and so much appetite for students who can take a voter file, score it, and hand it back to a campaign that they will never be out of a job.
The number of people who get politics and also understand computational statistics is so small; there is a massive career opportunity here, because after you have campaign experience, you can take these skills to businesses as well because they do similar types of research for consumer data.