The Pietist Schoolman

I was out of town over the weekend at one of my daughter’s softball tournaments, so I didn’t get a chance to curate my usual set of interesting links. If I had, I’m sure several of them would have collected around a surprising finding from the Public Religion Research Institute. Its 2020 Census of American Religion reported that white evangelicals now account for just 14.5% of the population (down from the 2006 peak of 23%), while white mainline Protestants rose three points since 2016, to 16.4%.

I’m sure such growth was news to just about every mainline Protestant leader, since all of the “Seven Sisters” have reported significant declines in membership and attendance in recent years, extending trends that often go back decades. Just last year I wrote a post about the “Dying of the Mainline.” Not because I’m an evangelical Protestant celebrating the death of a theological rival (the post was a lament entitled “What’s Lost with the Dying of the Mainline”), but because I now attend an ELCA church and had heard two of our pastors mention a study predicting the virtual disappearance of the denomination by the year 2050… a study conducted by the ELCA itself.

So what do we make of all this?

Critics were quick to point out that PRRI wasn’t actually asking if Americans belonged to any of the traditional “mainline” denominations, such as the United Methodist Church, the Presbyterian Church USA, the Episcopal Church, the United Church of Christ, and the ELCA. Instead, they asked any white respondents who identified as Protestant if they thought of themselves as “evangelical” or “born-again.” If the answer was no, those Christians were classified as “white mainline Protestant.”

By contrast, Ryan Burge pointed to two long-running surveys, often used by scholars of religion, that ask questions allowing respondents to be categorized according to denomination: the General Social Survey (GSS, conducted since 1972) and the Cooperative Election Study (CES, since 2006). By that technique, the GSS shows mainline Protestantism declining from its mid-1970s numerical peak of nearly a third of all Americans to just over 10% in 2018. In the newer CES, the mainline is currently around 13% and has never exceeded 16%.

Now, I take religion reporter Jack Jenkins’ point that “mainline Protestantism” is a complicated phenomenon, not simply a set of denominational membership reports. And it may be that PRRI head Robert Jones is right that mainline churches are drawing some disaffected evangelicals.

But even if you use the “non-evangelical Protestant” definition that PRRI equates with “mainline Protestant,” Burge found that the 2020 CES mainline number was 12%, down seven points from where it was ten years before.

So what do we make of all this?

Whether you’re a mainliner looking for good news or an evangelical inclined to doubt it, I’d heed some words of advice from Burge, who (as a political scientist) hesitates to make too much of the PRRI finding but (as an American Baptist pastor) is hardly predisposed to look for bad news about mainline Protestantism.

First, he emphasized (in an article for evangelical magazine Christianity Today) that the bigger story is not about Protestantism — evangelical or not — adding adherents. Quite the contrary. Here’s how he starts his article for CT:

If there’s one overarching conclusion that comes from studying survey data of American religion over the last several decades, it is that fewer people identify with an established religious tradition every year. The ranks of religiously unaffiliated, also called the nones, have grown from just about 5 percent in the early 1970s to at least 30 percent in 2020.

Religious demography is a zero-sum game. If one group grows larger that means that other groups must be shrinking in size. So that rise in the nones is bad news for churches, pretty much across traditions. 

(Which is why you should pick up Burge’s short book, The Nones — put out by Fortress, the ELCA publisher.)

But second, this kind of analysis is difficult. Here’s another recent Burge lede, in Religion Unplugged:

It’s almost become a cliche at this point, but it bears repeating again – measuring religion is hard. No matter if you’re a social scientist who uses qualitative methods like interviews and focus groups or have a quantitative approach that relies on survey data. Religious classification is nearly impossible. Creating categories that are meaningful enough that they make sense to the average person, but not too many categories that people get easily confused is a herculean task.

Consider another of Burge’s endless discoveries in the CES data: the share of respondents in each denominational category that also identify as “evangelical.”

• Among mainline groups, that ranges from just 13% of Episcopalians to over 60% of Burge’s American Baptists. (In fact, my Anxious Bench colleague Dan Williams recently suggested that those interested in evangelicalism pay more attention to the ABCUSA.) Meanwhile, I’m not all that unusual in being an evangelical within the ELCA: about a third of those Lutherans so identify. And as I’ve written before, I’m struck by the extent to which evangelical culture can be found within our ostensibly mainline church.

• But about 20% of Southern Baptists, 50% of PCA Presbyterians, and over 60% of Missouri Synod Lutherans don’t describe themselves as evangelical in the CES study, even though their denominations have participated to varying extents in evangelical networks and would never be classified as “mainline,” let alone “liberal,” Protestant.

• And since 2008, that label has plummeted among respondents who are placed with “interdenominational churches” (down 20 points). While a similar drop could be found among nondenominational churches that aren’t themselves classified as evangelical, nearly half those respondents do call themselves evangelical.

Of course, part of the problem here is that if scholars, pundits, pastors, and journalists endlessly debate what “evangelical” means, then “non-evangelical” must be as hard to figure out.

I often wonder, for example, what to make of the fact that few of my students ever identify themselves as evangelical. Are they “mainline (non-evangelical) Protestants,” despite choosing to attend a university that any observer of religious higher education would describe as evangelical? Are they actually theological liberals or postliberals… who overwhelmingly attend churches that emphasize something like David Bebbington’s evangelical quadrilateral? Or are they avoiding a term that they see as having more political than religious meaning?

Given that many of them don’t realize that they grew up attending Baptist churches that have dropped that term from their name, maybe my students simply don’t know the word evangelical to use it. (Hard as it may be to believe, it’s possible they don’t pay close attention to the lecture I occasionally give — in a team-taught gen ed class taken by 70% of our students — on early evangelicalism.) Or they just want to call themselves “Christ-followers” and see further classification as irrelevant or divisive.

So again I’ll ask: what do we make of this?

As difficult as religious measuring always is, that work would be particularly hard — and prone to contradictory results — if I’m right that what’s actually happening is a realignment of American Protestantism that is rendering these words less and less meaningful as differentiators for religious belief, practice, and identity.

“20th century religious categories are increasingly inadequate to describe 21st century religious realities,” I wrote for The Anxious Bench in March. The mainline/evangelical echo of the modernist/fundamentalist divide — plus older splits that go back to the Reformation and its aftermath — “haven’t gone away, but I’m not sure they’re as important as other fissures in shaping how Protestants identify and organize themselves. Isn’t it possible that we’re seeing realignment around other fault lines?”

If so, I think we should accustom ourselves to seeing even more confusion about religious statistics in the years to come.