Don’t forget that I am offering a discount on subscriptions through December 25th - $40 for the year. You can take advantage through this link.
There’s this book that came out about twenty years ago that really shaped my time in graduate school - it was called What’s the Matter with Kansas? by Thomas Frank. The author, who grew up in Kansas, describes how it used to be a progressive place. He goes to college, returns home, and finds that his once-progressive state has turned deep red. The book is his attempt to puzzle through why that happened.
Frank’s thesis is straightforward: the Republican Party managed to convince a large number of evangelicals to prioritize issues like abortion and same-sex marriage at the ballot box instead of economic concerns. He portrays evangelicals as predominantly lower-middle-class, poorly educated people who were duped into voting against their economic self-interest by not supporting Democrats, who, in theory, would have offered more in social service programs.
The argument sounds so appealing to folks in the media ecosystem that it sold a ton of copies and got Thomas Frank on The Daily Show. However, there’s a serious problem with the book - the entire thesis was dead wrong at the time (PDF). And, if you think about just for a second, you realize that Frank’s entire argument rests on the foundation that evangelicals are at the bottom end of the socioeconomic spectrum and would benefit from leftist tax policy.
But what if that assumption is fundamentally false? Let me explain.
Here is the educational attainment of evangelicals in the Cooperative Election Study (CES) over the last fifteen years:
In 2008, half of all evangelicals had no more than a high school diploma, and fewer than 20% had earned a bachelor’s degree. That’s a significant education gap. However, over time, the gap began to close. Looking at data from 2020–2023, the share of evangelicals with only a high school diploma dropped to 43%, while the share with a bachelor’s degree rose to 30%—a 50% increase from 2008. Additionally, 40% of evangelicals had earned at least a two-year college degree, up from 27% in 2008.
What about the broader baseline? These gains in education are significant, but they need context. Let’s compare evangelicals to the rest of the CES sample regarding college attainment over the last 15 years.
In 2008, about a quarter of the sample had earned a four year college degree. As previously mentioned, about 19% of evangelicals could say the same thing. Those who claimed that they were not born-again had a much higher level of educational attainment, though, at 28%. The gains in education for each of the three samples was fairly modest between 2008 through 2014 or so. The lines are relatively flat. But then, right around the time that the country elected Donald Trump, there was a significant upturn.
By 2023, 36% of non-born-again respondents had a bachelor’s degree—two points higher than the overall sample average of 34%. Evangelicals also saw notable gains, with 30% holding a four-year degree, an increase of 11 points since 2008. Evangelicals are clearly closing the education gap.
Let’s go another step here and break this down by race, too. This is the share of evangelicals who have earned a four year college degree by race in 2008 vs 2023.
Breaking this down by race offers even more insight. Here’s the share of evangelicals with a bachelor’s degree in 2008 vs. 2023:
Hispanics: Increased from 11% to 16%, well below the evangelical average.
Whites: Increased from 20% in 2008 to 33% in 2023, now comparable to the national average.
The claim that white evangelicals are poorly educated is empirically false. Their level of college attainment is now on par with the average American, undermining a key pillar of Frank’s argument.
Now, let’s address the political side. Here’s the partisan composition of college-educated evangelicals, broken down by race:
Okay, this is worth ruminating on just a bit. Among college educated white evangelicals, there’s a clear trend away from a Republican affiliation. About 80% of those folks aligned with the GOP back in 2008. That has eroded and now the share of white evangelicals with a college degree who say they are Republicans has dropped to 60%. Meanwhile, the Democrats have made some inroads in the last couple of years.
The Hispanic lines are interesting, too. Among college educated Hispanic evangelicals, the GOP had a big advantage during the early Obama years. About 58% of this group said that they were Republicans while the Democrats were about 20 points lower. But right around the time of Trump’s election, those lines had collided and now in the last 8 years, the Democrats and Republicans make up equal portions.
But let me complicate this a little while also giving you a more complete picture of what is happening in the word of partisanship and ideology. I took all five racial groups and broke them into three levels of educational attainment. Then I calculated the mean partisanship score and ideology score in 2008 vs 2023. This may give us a sharper view of how things have actually shifted.
The first thing that jumps out is that graph in the top right - it’s white evangelicals with a college degree. You can clearly see that they have moderated on both dimensions since 2008. That’s not the case for white evangelicals with lower levels of educational attainment. Those groups just haven’t moved that much at all.
What else jumps out to me? You can see that Black evangelicals have moved toward the middle in these graphs. That’s especially the case among those who have not earned a college degree. The exodus of Black Protestants from the Democratic party is pretty evident in these results.
The Hispanic graphs are a head scratcher, though. Among Hispanic evangelicals with a high school diploma, they have moved to the right. But those with more education have moved in the opposite direction. I am going to be very interested to see what was happening there in 2024, for what it’s worth. I bet there’s some really big movement.
But all this is just tilting at windmills until we see the actual data about vote choice. So, that’s what I did below - it’s the two-party vote share for the Republican in presidential elections from 2008 through 2020.
Yeah, all that talk about educated white evangelicals abandoning the GOP is just non-existent. They voted for Trump at the same rate as they supported McCain back in 2008. They may change their affiliation, but their behavior at the ballot box is rock steady. But you can see how Trump did really well among white evangelicals with a low level of education. McCain only got 75% of their votes. Trump did 11 points better than that in 2020.
What else jumps out to me? I think you can see how Trump enjoyed some gains among Black evangelicals in 2020 compared to 2016. That was true at every level of education. The Hispanic and Asian numbers are a real mixed bag, honestly. I stared at them for a while but no clear conclusions emerged.
Let me distill my thoughts about the Frank thesis in a couple of bullet points.
Frank’s thesis mischaracterized evangelicals as poorly educated and economically disadvantaged. In reality, many were climbing the socioeconomic ladder, seeking lower taxes and opposing same-sex marriage, aligning with the GOP's platform.
The stuff about abortion is just way off, though. In 2022, the state of Kansas voted on a a proposed state constitutional amendment that would have said there was no right to an abortion. It failed spectacularly - only 41% of voters were in favor. To say that abortion is this huge motivating issue, even for white evangelicals is nonsense at this point and I am increasingly losing my patience with anyone who says that it’s still a defining political issue.
Maybe Frank’s entire argument was just 20 years too early. The initial read on the election of 2024 was that Trump made huge gains among voters at the bottom end of the income spectrum. The GOP wasn’t running up the score with poor folks in 2000. Now, they truly are. So, yes, poor white evangelicals do support Republican and I don’t think it’s as simple as saying it’s about Culture War issues. It’s also about inflation and immigration, too.
White evangelicals vote for Republicans because they are Republicans. An increasing number of them have a strong educational background and make a good wage. They want low taxes and minimal regulation. They also orient their lives around their understanding of “family values.” We can talk about the hypocrisy of that all day long, but they generally aren’t super chill with books in the public schools about sexual behavior, nor do they encourage their children to explore their gender identity. They want their kids to be straight and cisgender and marry someone of the opposite sex and have a couple of kids and live close to them. The average Kansan believes that Trump will get them closer to those goals than the Democratic party.
Code for this post can be found here.
I feel like abortion is still a political issue in the places where it remains illegal; it's just not translating to the individual politicians that people vote for. Missouri had a similar vote this year to lessen reproductive healthcare regulations but they voted in the same people that enacted those regulations and pledge to do it again.
Frank’s big mistake in his thesis here is adding in wealth and educational attainment to the equation. I do think what the Republican Party is doing right now is trying to drive a culture war to distract from issues of class. Abortion has been the driving force for a while, but Ryan is right: it seems to be losing its role as the defining political issue. That is why conservatives have pivoted to transgender rights as they realize abortion is not a winning issue.