January 19, 2021

Religious Switching 2.0 – Analysis


ReligiousSwitching2014-legendAs promised, my commentary on the Pew Forum’s Religious Landscape Study. Prepare the skewers!


If you run a business, you are concerned with cash flow. If the money coming in to your business is less than the money going out of your business for an extended period of time you are going to be in trouble. Conversely, if the money coming in to your business in greater than the money going out, that will provide some growth opportunities. The graph represents people flows, both in and out of religious traditions. It shows a generational change and answers the questions of: “What faith group did you identify with as a child, and faith group do you identify with today?”

I believe that most of the changes can be explained by two things:

1. The faith groups ability to keep its own people.
2. The likelihood that people in other faith groups will choose a particular faith group over other options.

I think it is important to note that of those who leave a Christian tradition, typically three to four percent will chose a non-christian religion, and the rest will be evenly split, with half becoming Unaffiliated and half choosing another Christian tradition.

Here are some observations from the data:

The Evangelicals

The Evangelicals are the only stable group of the “Big Three” religious traditions. They do relatively well at keeping their own people (65%). While they have two significant outflows (to Unaffiliated and Mainline churches), they have three significant inflows (from Mainline, Catholic, and to a lesser extent, the Unaffilated. Lets look at the outflows first.

Evangelical –> Unaffilated and Mainline Protestant.

People leaving Evangelical churches go primarily to two destinations: The Unaffiliated or the Mainline Protestant religious tradition. These moves are roughly the same size. If an Evangelical is turned off church entirely, they move to the Unaffiliated. If they are interested in a different form of worship or Christian expression, they move to the Mainline Protestant faith group. There is little incentive for an Evangelical to move to the Catholic church as much of what the Catholic church has to offer can also be found in Mainline Protestant churches. (Our own Jeff Dunn and Denise Spencer being the exceptions to the rule.) There is little incentive to move to the Historical Black Protestant churches as both their key similarities (primarily evangelical) and differences (race based) would make moves less likely.

The Mainline Protestants

The Mainline Protestants were the worst of the big three at hanging on to their own members with only 45% of childhood adherents continuing on with the church into adulthood.

Mainline –> Unaffiliated and Evangelical

If you grow up in a Mainline church and do not find liturgical worship appealing, you really only have two options, the Unaffiliated and the Evangelicals. Just slightly less than half of those who leave mainline churches end up in the Unaffiliated group. Of the remainder who leave, roughly two thirds end up at Evangelical churches with the rest scattered across other traditions. So, those going to the Unaffiliated are slightly larger than those going to the Evangelicals. The Evangelicals biggest gain comes from the Mainlines.


The Catholic tradition falls somewhere between the Evangelical and mainline tradition in that they hang on to 59% of their childhood adherents. This is probably a good time to mention that the graph does not accurately represent what is happening through immigration. Some Catholic churches are booming in some parts of the country as a result of immigration. This however has been more than offset by declines in the other parts of the country.

Catholic –> Unaffilated and Evangelical

Of those who leave Catholic churches, roughly half end up Unaffiliated. The other half is evenly split between Evangelicals and everyone else. While they are the largest group entering the Unaffiliated in terms of absolute numbers, of the big three, the Mainlines send more as a percentage of their base.

The Unaffiliated

There have been several questions about the makeup of the unaffiliated. In short they are those who claim no religious affiliation. They currently make up 22.8 percent of the population and can be broken up into the following sub classifications:

Atheists 3.1 %;
Agnostics 4%;
Nothing in particular and religion not import 8.8%
Nothing in particular and religion important 6.9%

As mentioned previously, nearly half of those who leave another tradition end up in the Unaffiliated. At the same time nearly half of those who were raised unaffiliated end up in some other tradition. Because the other traditions started with significantly larger numbers, the ranks of the Unaffiliated have grown rapidly.

Some other observations:

Of the smaller groups, Hindus, Muslims, and Jews are most likely to hold on to their adherents.
Buddhists and Jehovah’s Witness are least likely. The rise of Communism in China and neighboring countries probably had a significant role in Buddhist loss of faith from childhood until now.

The Future?

So what does the future potentially look like? Leah Libresco had this interesting statistical analysis:

Pew’s survey data makes it possible to calculate where all 12 of the religions for which Pew has data — including some non-Christian religions — settle into a steady demographic distribution. This equilibrium is the logical extension of present trends, fast-forwarding until a new, persistent normal emerges. This kind of analysis gives us a chance to zero in on one factor that drives religious change — recruitment of new members and retention of old ones

She did her calculations both with and without fertility rates and came up with the following data.

Buddhist	       1%	             1%	                   1%
Catholic	      21	             8	                   6
Evangelical Prot.     26	            32	                  25
Hindu	               1	            <1	                  <1
Black Protestant       7	             8	                   8
Jehovah’s Witness      1	             1	                   1
Jewish	               2	             2	                   1
Mainline Protestant   15	            15	                   9
Mormon                 2	             2	                  22
Muslim	               1	             1	                   8
Orthodox Christian     1	            <1	                  <1
Unaffiliated	      23	            29	                  19

I must admit to being quite surprised by the results. Though she doesn’t provide a timeline, this is the direction we are eventually headed.

As always your thoughts and comments are welcome.


  1. First off, I think your analysis of present religious switching trends is spot on, at least from my own experiences.

    Now, about the future projections… I am something of an amateur futurist (and I must admit, somewhat of a pessimistic one overall) and look at futures reports a lot. Based on that experience, when I see someone trot out straight-line projections of current trends (no matter what is being discussed), I get twitchy. Straight-line projections are fine – if one assumes all other factors affecting the outcome are stable. In this case, the factors I can see affecting the outcome of religious demographics in this country are internal theology and culture, the milieu of the wider culture, and the economic and social environment. From where I sit, none of those three factors in America are anything close to resembling stable. If we believed the internal theology and culture of American evangelicalism (and many other Christian traditions here) was stable, well, we wouldn’t be regular customers here. The secular/progressive milieu of the culture seems pretty set in its ways. And the economic and social environment for most Americans is creaky at best, and I fear will get much worse in the next decade (if not sooner). Any projection of future religious demographic trends must take those factors into account.

    Now of course, taking those factors into account means that you can’t make a spot-on projection of what is going to end up happening, because those three factors are liable to sudden shifts, or “tipping points” in the current jargon. The trends that lead to tipping points can lie quietly for years, then suddenly explode. And the resulting shifts can be quite dramatic (at least from a demographic/generational timeframe). Look at Europe as an example. The churches there had all the advantages that American evangelical/conservative Christians covet for themselves – state sponsorship, cultural hegemony (varying from country to country), cultural veto on mores. But within two generations of the end of World War II, that hegemony has evaporated, and the churches of Europe struggle to achieve single-digit attendance demographics overall. The cultural factors, internal and external, forced a change. The same thing (or a somewhat similar thing, we are not an exact copy of Europe, of course) could happen here – WILL happen here, if my instincts (and IMonk’s thesis of “evangelical collapse”) are correct.

    Anyways, here endeth the rant. Time to scrounge up some morning caffeine.

    • Robert F says

      I tend to think it was the close relationship between state, nation and church, the state church and cultural hegemony that you mention, that was a primary catalyst for the quick demise of European Christian institutions. Disbelief was already widespread in England by the beginning of WWII, especially among the working classes, who gained no cultural advantage by continuing to be active in the Church when they no longer had faith, unlike their social superiors who continued to frequent religious institutions due to cultural expectations. Even in the 19th century, English Christian institutions had lost the working class.

    • Robert F says

      But if it does come, which it may, it will not just be an evangelical collapse, because it will include the American Roman Catholic Church; it will be a Christian collapse (the mainlines are already there).

      • The Finn says

        I don’t see the support for the “collapse” narrative. It looks like erosion to me. And what that looks like depends on how you interpret the numbers. All this data is based on self-reporting. Self-reporting of an identity is a *LAGGING* indicator – I believe we are much further down this curve than the data indicates [as the data is a look at a snapshot of history, possibly up to half a generation old].

        When does someone become disaffiliated? When the Boomer, or even older Gen-X, moved into their subdivision and their church attendance declined… (and actual attendance has been falling for a long time) but by self-identification they did not disaffiliate [I know so many of these people – they still ardently self-identify as what they *WERE*, their children almost to the one do not]. The children of that mobile population by default have that *historic* self-Identity, but then drop it as they reach adult hood, and possibly adopt another. Who in this story [told about 80 million times] *functionally* disaffiliated or took up a new affiliation.

        • Robert F says

          It’s pretty clear that America is going down the secularization road, not by the same route as Europe did, to be sure, but headed to the same destination.

          Otoh, as Peter Berger has noted, desecularization is happening in the east. Russia and China, each in its own way, are rapidly becoming more religious, despite the decades long totalitarian effort to eradicate religion that occurred in each nation in the 20th century, and is still going on to a certain extent in China.

          Ironies abound.

  2. Thanks for including the reference to Leah Libresco’s analysis. It is the most sophisticated analysis I’d seen. (And even it leaves out the effect of immigration.) I strongly suspect the equilibrium points are generations downstream, especially with fertility considered.

    Looking at that analysis, the story of the “rise of the nones” is significantly reworked . Yes, the nones have risen sharply – but they are near or past their peak from current trends. Without fertility they’ll grow another quarter of their current numbers to 29%, considering their low fertility they would shrink about a sixth of their current numbers to 19%.

    Another story that hasn’t been told is the low attraction rates of Christianity. Evangelical, and to a lesser extent other Christian, traditions put a lot of emphasis on reaching out and attracting converts. There seems to be a low rate of attraction. The only curve going from outside Christianity with significant thickness is the “none to evangelical”. And it is smaller than “catholic to evangelical” and “mainline to evangelical”, and not a whole lot larger than “black protestant to evangelical”. Is evangelical outreach inside America primarily a matter of sheep stealing?

  3. “Is evangelical outreach inside America primarily a matter of sheep stealing?” Yep, pretty much. My church has a strong evangelistic outreach particularly to the disaffected. We often see two types of converts. The first is someone with a cultural memory of church or a church background as a child. Mom and Grandma keep that memory alive. Often they were turned off by legalism. They drift away until they come to a life crisis and then return. The second are people who may or may not have a cultural memory; but they descend into the drug/alcohol culture, shipwreck their lives, often end up in jail (we have a very active and extensive jail and prison ministry) and like the Prodigal Son come to the end of themselves and want a way out.

    • flatrocker says

      So are you actually “stealing sheep” or are you simply seeking the lost sheep and bringing them home?

      I think the more interesting question becomes what happens next? If the lost sheep are brought back to the fold but subsequently at some point feel a new calling to move to another flock, will you joyfully let them go?

      • Yes, we send them off with well wishes. We realize that oftern our church is a way station. The “Yep, pretty much” refers more to what I see around me than what our church does. We don’t concentrate on the already attending somewhere folks with trendy programs, light shows, and fog machines. Also the church was founded by an ex 1% biker and is mostly blue collar workers, self-employed construction workers, factory workers. We don’t have much of an outreach to the upper middle class and they are not much interested in us. Doesn’t pay as well, of course.

    • Atheists are 0.07% of prison population:


      (Yeah, I know: Alternet. Probably evokes in some people the reaction I would admit feeling if they posted a link to WorldNetDaily. Still.)

      Oh and yes, I’m aware that if we’re talking about who’s in prison in America, we MUST needs talk about race. Still.

      We are, however, 93% of the membership of the National Academy of Sciences.

      • Rick Ro. says

        That could be attributed to the same concept behind “there are no atheists in foxholes.” Maybe when folks get to prison they realize they need God.

        • You do realize that there’s an organization specifically called Atheists in Foxholes, right?

          • The Finn says

            Yep, foxholes don’t cure atheists, they make\ them. I’ve been told that, and listened to enough people’s stories supports this – someone in a fox hole situation… and God didn’t show up to help, he couldn’t even spare the time of a minor hall-sweeper angel to deliver a memo of support. And worse, almost universally, his church didn’t show up either, if they did it was likely with a “tough love” message about how you were so stupid to end up in a fox hole, that doesn’t happen to better people.

            And I would be hesitant to use prison populations to prove any points; that is en environment rife with coercive pressures.

  4. Lotsa LDS.

  5. OldProphet says

    Sorry Brethren For the last 40 years surveys continue to show Evangelical churches either growing or holding steady and mainline churches diminishing. The Coming Evangelical Collapse is no more true than Hal Lindseys’ The Late Great Planet Earth.Maybe its time to come in from the “wilderness” and have a cup of ecumenical coffee. With creamer.

  6. The Finn says

    I am surprised nobody has keyed in on this text in the analysis:

    “As a rising cohort of highly unaffiliated Millennials reaches adulthood, the median age of unaffiliated adults has dropped to 36, down from 38 in 2007 and far lower than the general (adult) population’s median age of 46.4 By contrast, the median age of mainline Protestant adults in the new survey is 52 (up from 50 in 2007), and the median age of Catholic adults is 49 (up from 45 seven years earlier)”

    36 vs 52

    It would be interesting to see the proportions of the population if calculated as (count * (lifespan – averageAge)) to account for years-of-influence of an individual.

    “The percentage of college graduates who identify with Christianity has declined by nine percentage points since 2007 (from 73% to 64%). ”

    Those with the highest earning potential and thus, likely, the highest societal impact have significantly declining Christian affiliation.

    related: “Religious “nones” now constitute 24% of all college graduates (up from 17%) and 22% of those with less than a college degree (up from 16%).”

  7. Mike, I was wondering if those who participate in home or simple church fellowships factored in anywhere in the stats. This group may be too small to show up on the radar at all, but looking around at the make-up of my own community, I would guess that they represent two or three percent and growing. But many home and simple churchers may be self-identifying as evangelical or as nones. I’m just curious if there’s any data on this group.

  8. Earning a very good living from DoE and trend modeling, I have to say that I don’t expect anyone with a background in these sciences to take this seriously. In fact, trying to model any trend based on two points is frankly absurd. To give you an idea of how this data has no ability to describe a narrative, let me give you an example. What if in between the beginning and end point, people were switching? What if the average length of stay in any stream was 6 months?What if the rate of switching was nearly identical between streams? Then any shift in current affiliation would only be the result of compound variance. All that to say that I have no idea what the Pew poll was trying to do, but I can say with certainty what it is incapable of doing.

    • The Finn says

      Agree, it is much more useful as a snapshot than for creating projections.

      However, some of the trend lines sit on very similar data I see from other sources. Particularly a decline in affiliation of younger people, across the board. It is interesting to me how much this “religious” data lines up with other data – indicating these trends CANNOT BE EXPLAINED, at least not completely, from a religious perspective [not that the church obsessed care about any of those other trends].

  9. Robert F says

    Let’s settle this once and for all.

    * * * * *

    Oh, Magic 8 Ball! Is there a coming evangelical collapse?


    * * * * *

    Even the Magic 8 Ball is uncertain of the answer to this question! Best we leave it alone for now.

    • Rick Ro. says

      Oh, silly Robert. Everyone knows Magic 8 Balls don’t work. Instead, let’s do this…

      * * * * *

      Behold, I am laying a fleece of wool on the threshing floor. If there is dew on the fleece alone, and it is dry on all the ground, then I shall know that you shall save all the evangelical churches and there will not be a collapse.

      * * * * *

  10. OldProphet says

    My Ouija board told me this is so

  11. Robert F says

    “The rise of Communism in China and neighboring countries probably had a significant role in Buddhist loss of faith from childhood until now.”

    I don’t understand the connection here. Chinese Communism has adopted a far less harsh policy toward traditional Chinese religions in the last few decades, so I don’t think this observation is correct.

  12. Very interesting results there from a former evangelical to Anglo-Catholic who is now on the way towards becoming Catholic (via the Personal Ordinariate). I wonder if these results also translate to here in Australia (where I live).

Speak Your Mind