By Jay Barth
We are now at a point where we have solid numbers from across the state from Tuesday’s election and can begin to understand how Donald Trump won a surprisingly strong victory in Arkansas and, in the only close statewide vote of the day, how the state became the first Bible Belt state to create a medical marijuana program at the ballot box. Here’s a little of what we know from those numbers.
Unsurprisingly, Trump’s vote across the counties of the state was driven by four demographic forces according to a simple regression analysis of Tuesday’s results at the county level: education levels, the percentage of African-Americans, the percentage of Latinos, and income levels. As the percentage of a county’s residents with a bachelor’s degree or more goes up, Trump support went down controlling for other factors. In terms of race and ethnicity, as the percentage of a county’s residents who are African-Americans or Latinos increases, Trump support also decreases. On the other hand, all things being equal, the median household income in a county is positively connected with a county’s vote for Trump. The percentage of a county’s population that is a member of an evangelical denomination also approaches statistical significant with the percentage of evangelicals positively correlating with support for Trump. Thus, the Trump coalition in Arkansas looks much like the traditional GOP coalition of wealthier, whiter, and more religious voters. But, added to it are less-educated voters. All told, a powerful coalition.
While demography explains the Trump vote, it tells us much less about the patterns on Issue 6 — the medical marijuana amendment. Two demographic forces —the percentage of the county’s citizens with a bachelor’s degree or advanced education and the percentage of the county’s populations who are African-American — are both positively related to support for the measure at the ballot box, according to a regression analysis of Tuesday’s results.
Also significant was the evangelical vote in a given county; the more adherents to evangelical churches in a county, the lower the percentage of the vote for medical marijuana. However, these variables — along with the median household income in the county and the percentage of Latino citizens in the county, neither of which are statistically significant—explain less than a quarter of the variation vote on Issue 6 across the state. This suggests other key forces beyond demographics were at work on the vote on the issue.