By Paul A. Djupe and Ryan P. Burge

The most momentous change in American religion over the last 25 years has been the growth of the religious nones from 5 percent in 1994 to 34 percent in 2019. Perhaps the dominant explanation for that rise has been the contentious and extreme politics of the Religious Right, which gained prominence in 1994 through the Republican takeover of Congress for the first time in 40 years. The Republican Party has grown more extreme from that point, fueling the negative image of religion for many marginal affiliates.

While the growth of the nones has been steady across this time period, recently we observed what seemed like a plateau in 2018 and suggested, tentatively, that the decline of American religion may be slowing. In particular, the youngest adult generation (“Z”) was not irreligious at a higher rate than Millennials. We noted this using the Cooperative Congressional Election Study (CCES) data and then confirmed it with PRRI’s American Values Atlas and GSS data, so it was not a one off. Now that the 2019 CCES study has been released, we can update the story and reassess the growth rate of the nones.

Let’s get right to it. The figure below shows the percent of each generation in 2018 and 2019 who identify as a religious none, which is a composite measure of those picking atheist, agnostic, or ‘nothing in particular.’ It is clear to see that the plateau in 2018 was temporary and that the growth of the nones has resumed. In just the past year, each generation from the Boomers on down has gained 2-5 percent nones, with the largest gain among Gen Z – they top out now at 46.7 percent. This shift mirrors the gains across this year as the nones went from 31.3 to 34.3 percent. (These analyses use the supplied weight in these datasets called “commonweight”).

However, this is not the finding from the most recent data from PRRI’s American Values Atlas based on hundreds of thousands of interviews. According to these data, the number of nones dropped for the first time in 2019 – over one point from 2018. The data themselves are of the highest quality – RDD on both landlines and cell phones with interviews conducted in English and Spanish. The PRRI measurement of the nones is the same as the CCES (atheists, agnostics, and nothing in particular), but the notable difference is the survey mode – PRRI uses a telephone-based sample with live interviewers. And speaking to live interviewers can be transformative. As PRRI researchers found in a research paper, “social desirability bias” affects the extent people report a religious affiliation and religiosity. Both are more common with live interviews. In particular, they found 5 percent fewer nothing in particulars and almost 20 percent fewer never attenders in live interview samples. That probably accounts for the difference between PRRI’s the CCES numbers. Of course, it doesn’t resolve which is the truth on the ground.

Given the origin story of the nones, we suspect that the partisan tilt should be leftward toward the Democrats. The figure below shows evidence to support that expectation – more strong Democrats were nones in 2019 (3% more), and 7 percent more who lean Democratic were nones by 2019. But it is notable that Republicans did not decline in the rate of nones among their ranks. This is not a simple sorting story. Instead the decline of religion is broad based with many driving factors. One of them is simply that irreligious parents have irreligious children, so that new voters are entering the electorate without ever being religious. Moreover, leaving congregations because of political difference occurs across the partisan spectrum.

Note the 5 percent uptick in independent nones – nones are also less likely to be affiliated with any civic institution, which has little to do with politics. In our post from last year, we acknowledged that many nones are simply disengaged and wondered if the “Great Disengagement” had likewise slowed. It’s too early to make that call, but 2019 at least shows a resumption in the disengagement with partisanship.

As American religion continues to fade, there are bound to be times when apostasy is accelerated and times when it slows. Some of this could be sampling error, of course, but there may be patterns in the data that suggest the role of elections or the economy in driving the connection with social institutions like congregations. Now, of course, on top of those forces, we look forward to assessing how the coronavirus pandemic has affected religion and organizational involvement.

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

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