By Paul A. Djupe, Ryan P. Burge, and Andrew R. Lewis

The reckoning of Donald Trump’s election and Roy Moore’s senate campaign in Alabama has led a few evangelical leaders to question the usefulness of the label. Lately, this critique has taken a misguided turn – criticizing the veracity of the polling on evangelicals.

Thomas S. Kidd, a historian at Baylor University, has been one of the most outspoken. He recently published a post on Vox.com arguing, especially, that modern survey measurements of an “evangelical” sow confusion and may even mislead the American public. He accuses the media of pouncing on instances of evangelical hypocrisy, which, in his view, are manufactured from biased polling. In Kidd’s view, there are true evangelicals who, properly measured, are not the moral hypocrites equivocating about whether a candidate who targeted teens should be supported. Tuesday night, as Roy Moore cost Republicans a Senate seat, Kidd doubled down on Twitter:

There is no doubt that surveying has become more difficult in recent years, as alluded to in Kidd’s Vox post. There are also significant disagreements about how to measure evangelicals (see here and here). However, there are problems, both logical and especially empirical, with Kidd’s assertions. Most important, it is simply false that we have “no idea who these people are.” Prior data gives us a great deal of insight into who self-identified (or affiliated) white evangelicals are, and Kidd’s skepticism is unwarranted.

We’ll briefly take up some of his more important points.

1. “For instance, if you ask more probing questions, it turns out that significant numbers of these “evangelicals” do not go to church.”

First, Kidd suggests that evangelicals do not attend church much. The figure below examines the proportion of evangelicals (by denominational affiliation) that never attends church compared to those in other religious traditions. It is just not true that significant numbers do not go to church – only about 10 percent of these evangelicals never attend. On the flip side, 54.7 of (self-id) white evangelicals attend weekly or more often; the figure is 48.8% of white evangelicals affiliates. These are very high rates of church attendance.

Perhaps, as Kidd seems to suggest, the difference is over how evangelicals are classified in many polls. Do self-identified evangelicals attend church less than evangelicals categorized based on denominational affiliation? The figure below clearly shows that the two measurement schemes return nearly identical estimates of church attendance. And the measure that Kidd decries the most, that based merely on identification with the term (“born again or evangelical” in the CCES), shows a slightly higher level of church attendance than those who answer religious affiliation questions that place them in evangelical denominations. This is not just from one survey in one month of one year, but from a decade of surveying captured at various points of time in the year (the GSS is done in the spring, the CCES in the fall).

2. “One study of the 2016 GOP primaries showed that these nonchurchgoing evangelicals were more likely to support Donald Trump — around 53 percent of Trump supporting evangelicals marked that they seldom/never went to church. That percentage dropped to around 36 percent for Trump supporting evangelicals who went to church weekly. Of course, a strong majority of self-identified evangelicals went on to support Trump in the general election.”

Yes, there was a gap very early in the primary season, though a plurality of frequently attending Republican evangelicals still favored Trump, a fact often overlooked from that article. As Republican support coalesced around Trump, so did evangelical support, including committed evangelicals. As Andrew Lewis finds in a chapter in the forthcoming volume The Evangelical Crackup?, in late Spring 2016 evangelicals were less likely than other Republicans to consider defecting from the Republican nominee should Trump be the choice, either to vote for a Democrat or some third option. This was borne out as evangelicals came together in the fall and supported Trump very bigly.

Was there a pattern of Trump support based on church attendance? Was Trump support higher among the nominal evangelicals that Kidd reports? The figure below shows that support for Trump in November was much higher among frequent attenders compared to non-attenders – an almost 20 percent spread. It was the exact opposite of what Kidd insinuates. The most committed white evangelicals are the most Republican.

3. “Other than political allegiances, we don’t know much about what the term evangelical means anymore… [D]oes our country largely misunderstand what it means to be evangelical in America today?”

Evangelicalism is a big tent in the United States, with many denominations and non-denominations, many languages, and many theological variations. And Kidd is right that when we talk about evangelicals, perhaps the clearest signal is their partisan allegiance, though this association is no pollster-journalist conspiracy. Evangelicals simply have become well known for their politics, and that is much of their own doing.

One of the things we know is that “evangelical” is synonymous with “Republican” so that saying “evangelical Republican” is effectively redundant according to Strato Patrikios’ paper in Politics & Religion. We see repeated evidence of this through exit polling where 81% of white evangelicals supported Trump and 80% of white evangelicals supported Roy Moore in Alabama in 2017. Yes, there is some variation depending on how evangelicalism is captured. But whatever else a white evangelical is, the American public knows by their actions (e.g., 2016 presidential and 2017 senate vote) that they are Republicans and position themselves in the party system accordingly.

Of course, that’s not uniformly so, as the following figure shows. This lopsided pattern is the result of long run change among evangelicals that is primarily rooted in regional residence – the South became steadily more Republican, while Northern evangelicals moved in reverse (as Ryan Claassen shows) since the Civil Rights era. If their voting patterns were to change, there would surely be a period of time during which popular, sticky stereotypes would lag. With white evangelicals doubling-down on the politics of cultural conflict, backing even deeply troubled Republican candidates, there is no indication we are in such a transition period in regard to electoral politics, though other elements of “culture war” politics may be changing.

We suspect that this is the rub: It is clear the religious arguments of “Never Trump” and “Never Moore” evangelicals’ did little to alter the partisan attachments that would lead rank-and-file evangelicals to a cast a vote for someone other than the Republican standard bearer. We know this in great detail regarding Trump in 2016, and the pattern appeared to repeat itself in the morally-challenged candidacy of Moore in 2017 (80% of white evangelical voters supported Moore, though turnout may have been slightly depressed).

This pattern is troubling to many elite evangelical observers and has fueled attempts, like Kidd’s (and not only Kidd’s), to reform and even redefine what an evangelical is. Whether support for Trump and Moore should occasion self-reflection is none of our business. But polling is not the problem here.

Paul A. Djupe, Denison University Political Science, is an affiliated scholar with PRRI, the series editor of Religious Engagement in Democratic Politics (Temple), and co-creator of religioninpublic.blog (see his list of posts here). Further information about his work can be found at his website and on Twitter.

Ryan P. Burge teaches at Eastern Illinois University in Charleston, Illinois. He can be contacted via Twitter or his personal website.

Andrew R. Lewis is an Assistant Professor at the University of Cincinnati. His research covers the intersection of religion, law, and politics in the U.S. He is the author of The Rights Turn in Conservative Christian Politics: How Abortion Transformed the Culture Wars (Cambridge, 2017). You can follow him on Twitter and see more of his research on his personal website.