The "Missing White Voters," Revisited
In the aftermath of the 2012 elections, I was perplexed. Most analysts had believed that if turnout dropped, Mitt Romney would be president. Yet turnout was clearly down by a substantial degree, yet President Obama had a surprisingly strong showing.
After a quick look at the data and some math, it became apparent to me that some white voters simply went missing. That is, they cast ballots in 2004 and 2008, but did not do so in 2012. If the GOP could persuade them to turn out, the party could help close the gap with Democrats. This became known as my “Missing Whites” hypothesis.
Who were these missing whites? If you look on the second page of the article, you can see a map I created of Ohio, comparing turnout in 2008 to 2012. It was fairly obvious that turnout was strong in urban and suburban areas, but weak in rural areas. I reasoned that these were probably people who liked George W. Bush and perhaps John McCain, but were turned off by Mitt Romney’s wealth and patrician air. If Republicans nominated someone with more working-class appeal, I reasoned, these people could be motivated to vote.
In 2013, after all the votes had been counted, I was able to perform a more thorough analysis. It became apparent to me that these voters were blue-collar, rural and lived in areas with strong showings for Ross Perot. This reinforced my view that if Republicans softened their economic libertarianism and nominated the right candidate, this bloc might turn out and help the party win with only modest improvements among nonwhites.
This provoked great controversy. To this day I am a bit perplexed, as I was effectively saying to Democrats, “You were correct, the GOP has to make some changes to its approach in order to win.” Part of the backlash was plainly inspired by people who did not read the article. More importantly, my observation was expressly limited to the idea that missing whites could help the GOP win. People interpreted this as advocating for a “whites-only” GOP, which I expressly disclaimed, both in that article and in a subsequent piece.
It is far too early to say for certain, but what happened Tuesday certainly looks consistent with the missing-whites thesis. First, there’s very little doubt that Trump’s vote share soared when compared to Mitt Romney’s in northern rural areas. But turnout seems to have been robust as well. We once again will have to wait until all the votes are counted to conduct a more robust analysis, but consider Ohio once again:
For this iteration, the darkest red counties are ones where turnout was up or the same as 2012 levels, while lighter shades suggest turnout was down. As you can see, the Columbus and Cincinnati suburban counties – Delaware and Union, Clermont and Warren – were dark red. This is expected, given the population growth there. Notice, however, that the urban counties – Hamilton, Franklin, Cuyahoga, Stark, Summit, Lucas and Montgomery – are light. Turnout was down substantially there.
But look at the rest of the state, especially the southeast. There is a lot of red there, which is all the more surprising given population declines. It isn’t yet clear that whites there turned out in increased numbers, but they clearly avoided the decreases we saw in other areas of the state, and nationwide. Again, we need more data, but for now the best indications are that these voters were, in fact, inspired by a Republican candidate with more blue-collar appeal. Donald Trump did do better with nonwhites than Mitt Romney, which played a significant role in his victory. But there’s little doubt that a strong showing with these rural whites, who are disconnected from the global economy that increasingly defines urban and suburban environs, played a major role in his win.