You’ve probably seen this story all over the media and the internet. In exit polls in the Republican primaries, Donald Trump is getting huge(!) support from Blonde Women. In fact, more Blonde Women have supported Trump than have supported any other candidate.
The pundits are confused. What happened to the “soccer moms” who are looking for decency and comportment in their leaders? What happened to “family values” based on traditional assumptions about gender roles, driving kids to practice in minivans, making wholesome meals, and supporting their husbands?
How is it, the headlines scream, that these women would support a candidate who bragged in his books about famous women he slept with, who talks of wishing he could date his Blonde daughter, and who went to twitter war with the darling of Blonde Womanhood on Fox News, Megyn Kelly? These women are abandoning their core issues in support of a candidate who “tells it like it is” even if the “it” is often offensive. It’s been the most confusing piece of a very unusual campaign season.
Wait, you say you haven’t heard of these stories? You’re telling me that it makes no sense to take a demographic characteristic like hair color and try to make it a predictive variable on voting preferences? That even though they are Blonde Women, their preference for Trump is based on attitudes toward immigration or trade or just plain opposition to Democrats? That maybe some of the Blonde Women really aren’t Blonde at all (not Megyn) but find it useful to color their hair in order to be part of the demographic?
To which I respond: EXACTLY!
If you haven’t guessed, this blog post is really about Evangelicals.
Last Saturday’s results in South Carolina had Trump garnering 31% of the self-identified Evangelical vote. In fact, Evangelicals were exactly as likely to vote for Trump as non-Evangelicals. But the primary determinant of their Evangelical status is dependent upon their telling pollsters that they are. As Sarah Posner said today on twitter:
Pollsters, ahead of elections and at entrance/exit polls, ask voters: “are you born-again/evangelical or not?” I’d bet good money that given binary choice, many voters, particularly in states like South Carolina or Texas or Iowa won’t choose “not.”
For many voters in the Republican electorate, being evangelical is as much a cultural identifier as it is a religious statement. In fact, according to the Pew Religious Landscape Survey, 1 in 8 self-identified evangelicals (12%) rarely or never attend church services. Another 30% attend monthly to a few times a year (I wish Pew broke this category more specifically — the few times a year are more like the seldom than they are like the monthly).
But even if we were to look at the 58% who attend weekly or more, we don’t know that their voting behavior is based on being “values voters”. Back in 2004, when marriage amendments were on the ballot in 16 states, there was a decided attempt (Karl Rove’s brainchild) to get evangelicals to vote in a block. But they aren’t voting on these issues for the most part. They are voting on basic conservative Republican issues.
On NPR tonight, Robert Jeffers of Dallas First Baptist argued that evangelicals have moved on from expecting government to promote spiritual values and instead are willing to support secular leadership who will “solve problems”.
The exit polls from South Carolina didn’t do multivariate analyses of their key factors. Trump won 71% of those who wanted someone to tell it like it is. He won 51% of those who thought immigration was a key issue. He won 36% of those who thought the economy was a key issue.
I’m going out on a limb here, but I’m certain that a significant percentage of the so-called evangelical voters, thought that immigration, the economy, and straight talk were key factors in their decision. Some of them likely have a reasoned perspective connecting their evangelical faith with their position on these issues. But most of them probably haven’t even tried.
Technically speaking, self-identifying as evangelical is orthogonal to these other variables. If we ran a factor analysis, they wouldn’t load on the same vectors at all.
So it’s time to quit freaking out about evangelical voters as a meaningful category. If the pundits need to find things to worry about to fill all those hours of cable news speculation, hair color makes just as much sense.