The American Religious Landscape is Volatile

by Ryan P. Burge, Eastern Illinois University

I was browsing a political science message board a few years ago and an anonymous poster commented that he/she wondered why those who study religion and politics think it’s somehow special compared to other group based identities. I am always trying to be aware of blind spots in my own thinking, so that argument has stuck with me. I think I have constructed a rebuttal. Put simply, religion is not an identity in the way that race/ethnicity is an identity because religion is a choice. It’s not immutable or unchangeable. While someone may grow up evangelical, they can choose to leave that identity at some point in their life (and many do). Others are born into a faith and never leave. That’s something that doesn’t really exist in other contexts.

Ever since I read “Switching Faith” by Darren Sherkat (who was on my dissertation committee) I’ve always wanted to explore religious switching in more depth, but never really had the data or the inclination to wade into the problem. However, I became aware that the Cooperative Congressional Election Study not only has a longitudinal survey that they conduct every two years, but they also conducted a panel survey in 2010, 2012, and 2014. For those who are not privy to survey methodology, a panel survey is when the same people are asked questions over an extended period of time. They are invaluable to track actual change in opinion or behavior over time. They are also perfect to track religious migration in the United States. The CCES Panel is especially valuable because it started with 9,500 respondents. That allows for a lot of subgroup comparisons. Let’s start as broadly as possible and look at the entire dataset and how much change occurred between 2010, 2012, and 2014.

Each of the three vertical bars represented the distribution of religious identities in each year. This visualization also contains an alluvial diagram, which are the bands that represent the flow of individuals from one tradition to another across the three panel waves. The thicker the band, the greater number of individuals migrated from one religious tradition to another. There is obviously a lot going on in this graph, but two things become apparent. One is that there is a tremendous amount of religious stability amongst some religious traditions, but there is also a great deal of movement from one tradition to another across these two year time periods.


Let’s clean up some of the noise and zero in one specific dimension: religious defection. I am defining the defection rate as the percentage of individuals who reported a religious tradition in the 2010 wave, but did not claim the same religious affiliation in the 2014 survey. The overall defection rate for the entire sample was 18.9%. This defection rate aggregated by religious tradition is displayed in the graph below. It’s apparent that there are tremendous differences in defection across American religious groups. For instance, approximately 10% of Protestants, Catholics, and Jews defected over the four year period. Note, however that this defection rate only includes Protestants who left the Protestant tradition. That means that individuals who switched from a Southern Baptist affiliation to a United Methodist membership do not count as a defection in this analysis. For atheists, the defection rate is about 18%, but for those who responded “Nothing in Particular” and agnostics, this rate is much higher at 41.6% and 48.4% respectively. That’s a staggering amount of flux in those two groups. To state it simply: about half of the agnostics in 2010 were not agnostics in 2014. That needs further exploration.

Let’s begin by looking more closely at those who stated that they were “nothing in particular” (NIP) in 2010. Their religious change is tracked in the graph below. About 6 in 10 NIPs stay that way in 2014. However where the nothings migrate to is noteworthy. About 20% move to other “none” traditions like atheism and agnosticism. But nearly the same percentage move toward Christian religions with 13.3% becoming Protestant and 4% claiming Catholicism in 2014. How does this compare to agnostics?


About 55% of agnostics did not defect, compared to 62% of NIPs. However, defectingagnostics were much more likely to migrate to the “none” religious groups. 22.5% become atheists in 2014; while another 18.9% identified as “nothing in particular” in 2014. On the other hand, the movement of agnostics toward Christianity was basically non-existent. Just 4.5% of agnostics in 2010 claimed that they were either Catholic or Protestant in 2014. This migration pattern is in stark contrast to the “nothing in particular” group. Recall that about 20% of NIPs became other nones, and 20% moved to a Christian tradition. This ratio for agnostics was drastically different: for every agnostic who became Christian, ten agnostics moved to another “none” tradition. It’s clear from these results: agnostics are much more likely to move away from organized religion, while NIPs stand as a kind of bridge between some faith and no faith.


Honestly, I’ve not been so interested in a dataset in a long time. There are lots of angles to explore using panel data and I will hopefully put together some additional posts that look at defection rates in more detail. However, there’s a lot to consider from just this small piece of analysis. It seems that the line of demarcation between having a religion or not is the “nothing in particular” category. Those individuals are just as likely to tilt toward returning to a Christian tradition as they are to become even more entrenched in having no religion. On the other hand, once an individual affirmatively denies a belief in a specific deity (i.e. describing themselves as agnostic or atheist), it’s very hard for them to go back. This is not insignificant, in the grand scheme of things.

In fact, religious nones are the second largest religious tradition in the United States. According to the CCES 18.6% of respondents are “nothing in particular,” which translates to 55 million American adults. To extrapolate from this data: about 12 million of NIPs will drift to atheism/agnostics, while another 12 million will change their affiliation to Catholic or Protestant. That’s a tremendous amount of volatility. What drives these switching decisions? It could be demographic factors like age or it could be driven by political concerns. That’s a great topic for another post.

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

Full coding syntax for this analysis is available on my Github.


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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s