The Decline of Religion May Be Slowing

By Paul A. Djupe and Ryan P. Burge

In a companion piece published today on Religion in Public, Melissa Deckman of Washington College finds that the probability of being a religious none in Gen Z (born after 1995) is the same as for Millenials (born between 1981-1994). This bombshell finding sent us running for other datasets. Like all good scientists, we trust, but verify. In this post, we run through evidence from the General Social Survey, 2018 Cooperative Congressional Election Study (a RIP favorite), and the recent release of the Voter Study Group panel. The takeaway is that the finding is validated – the rate driving up the religious nones has appeared to be slowing to a crawl. We then discuss some reasons why the rate might be slowing.

It is conventional wisdom at this point that the incidence of religious nones is on a steady rise after 1994. Driven by a mix of politics, scandal, and weak parental religious socialization, non-affiliates have risen from about 5 percent to 30 percent. That trend appears to be accelerating by generation, so the rate of being a religious none is much greater among Millennials than it is among Greatest, Silent, and Baby Boomer generations as the figure below shows using the General Social Survey time series. Those older generations are still experiencing some secularization (the rates are rising across time), but not nearly as rapidly as the young. From this evidence, we expected that the rate of being a none among Gen Z might be even higher, leading to a bump above Millennials. The initial, small sample estimate from the General Social Survey, however, suggests that Gen Z is not outpacing Millenials and may have even fallen behind.

Even though it is highly reliable, the GSS is just one dataset and needs to be confirmed, especially with data sources with a larger number of cases. Therefore, we turned to the 2018 CCES, which has 60,000 cases and 5,000 Gen Zers – plenty with which to generate reliable estimates. The figure below shows the probability of being religiously unaffiliated for each generation in the data (we combined the few remaining Greatest with the Silent generation). The lesson is clear – the rate has drastically increased with each generation through to Millennials and has since slowed so that Gen Z is so far no more unaffiliated than Millennials. In 2018, 42.8% of Millennials were nones (combining atheists, agnostics, and those ‘nothing in particular’), while 42.9% of Gen Zers were nones.

Ideally we should be able to confirm these findings with other data. The Democracy Fund’s Voter Study Group panel was released recently, including survey data from 2019. The VSG panel is smaller than the CCES, but still enormous compared to most surveys with almost 6,800 respondents in 2019. We followed the same procedures in coding, made easy because the VSG adopted the same question wording as the CCES (which comes from Pew). The results below look slightly different than the CCES since here the Gen X none rate is also indistinguishable from the two younger generations. But this analysis generates the same lesson – the rate at which Americans are becoming nones has slowed so that Gen Z is not distinguishable from Millennials.

These results are only suggestive. Gen Z is still very young – only a few have graduated from college at this point and may be gaining distance from their parents and religious institutions. However, the college years are when we tend to find religious socialization gaps (though see Burge’s post about the life-cycle effect), so maybe these early estimates are sending a true signal. Only time will tell, though the general population rate of the religiously unaffiliated will continue to climb to meet the younger generations.

The upshot is that the population rate of the religious unaffiliated is at least leveling off and may be contracting. That is the remarkable conclusion reached with data from 4 years of PRRI’s American Values Atlas based on hundreds of thousands of interviews. This is shown in the figure below, which zooms in to the peak of the nonreligious in the United States seen over the last few years. The drop from 2018 to 2019 is statistically significant, though huge datasets allow for even very small differences to be statistically distinguishable. The data themselves are of the highest quality – RDD on both landlines and cell phones with interviews conducted in English and Spanish. To give context, the size of the religious nones has grown steadily from 1995 without much of a break until 2019. However, without more data, we cannot be sure whether 2019 is a blip or part of a new trend.


Why would the rate be slowing? In one view, the population is diversifying fast, which is of course why white Christians are adopting a politics of fear and threat at the moment. Non-whites in older generations show lower rates of disaffiliation as whites, though they have caught up in Gen Z as the final figure shows. This suggests that diversification is not holding back the rate of nones. On the other hand, the threat of diversification may be driving whites to find or stay in churches. White evangelicals do exhibit much more racial resentment and a forthcoming book by Robert P. Jones of PRRI details how deeply these attitudes are built into white Christianity’s foundations.

Another theory about the nones is that organized religion has been shedding marginal identifiers for several decades. As the social desirability of being religion has eroded, people have stopped identifying just to seem respectable. However, in some religious groups, like Catholics and mainline Protestants, the average level of attendance has continued to atrophy. So has their share of the population, though. It is possible that there are many fewer marginal identifiers left to shed.

People are sorting their religion on the basis of their politics. Political disagreement certainly causes people to leave particular houses of worship, and there are a number of pieces showing how the controversial politics of Trump and the Christian Right has backfired on religion in general as well as on evangelicalism in particular. It is remotely possible that this sorting process has been effective and is nearing completion, though people still report political diversity in the pews.

We also shouldn’t make the mistake of thinking that all nones are the same. As we’ve shown repeatedly (and recently reinforced by a piece in JSSR by Phil Schwadel), those who call themselves ‘nothing in particular’ are not the same politically or religiously as atheists and agnostics. NIPs fall in and out of religion but are mostly just disengaged. One possibility is that the great disengagement is slowing as well.

Exactly why the rate appears to be leveling off remains for future work to investigate. It will be a fruitful line of work.

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 (posts). 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.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s