September 23, 2020

Michael Bell: Looking at the Pew Forum’s “Changes in Religious Affiliation” Data

mikeprofileInternet Monk First Officer Michael Bell returns with a look at some of the recent Pew Forum data on changes in American religious affiliation.

The line it is drawn
The curse it is cast
The slow one now
Will later be fast
As the present now
Will later be past
The order is
Rapidly fadin’.
And the first one now
Will later be last
For the times they are a-changin’.
– Bob Dylan 1963

For years I have heard about the many changes that have been taking place within the Christian World. Churches and denominations growing, churches and denominations shrinking. We have had a pretty good idea of who has been growing, and who has been shrinking, but with birth rates, death rates and other factors, it has been pretty hard to pin down the source of the growth and decline. Have Evangelicals been growing? If so, has the growth come from the non religious, Catholics, Mainline Protestants, or other religions? What sort of outflows have they experienced that have offset the inputs? Is the back door larger or smaller than the front door? How are the Catholics, the Mainline Protestants, the non religious and others doing?

Well now we know.

A few days ago, The Pew Forum on Religion & Public Life released the results of a survey entitled Faith in Flux: Changes in Religious Affiliation in the U.S. This was followup to their U.S. Religious Landscape Survey that they released last year.

I have been busy in the last week doing a bit of reverse engineering on the numbers to represent the shifts in belief as best and as clearly as I can. Here is the resulting chart. You can click on it to see the full sized version.


What you are looking at is changes in American adults, from their childhoods to present day. As such it eliminates such factors as birthrate and death rate, and strictly looks at who is changing to what. We should note that immigration is a factor in this chart as present day Americans may have been born elsewhere, and so their childhood would have been in a different country. More on this later in the post.

The NONE group

The red color, or None, stands for those with no particular faith. The temptation is to think of this entire group as atheistic, but that is not the case. It is currently comprised of atheists (1.6% of total adult population up from 0.5%), agnostics (2.4% up from 0.2%), and those of no particular faith (12.1% up from 6.6%). I like to think of them as the “no God, don’t know, or don’t care group”. Of those who have no particular faith, roughly half of them (6.3%) would classify themselves as secular, and half (5.8%) would call themselves religious.

The “None” group, now makes up a total of 16.1% of American adults today, a huge increase from the 7.4% who were in this group in their childhood. The interesting paradox is that of those who were raised in this group half now have a religious affiliation, with 1.6% of American adults moving from None to Evangelical (yellow) and 1.0% moving to Mainline Protestant (orange) religious beliefs. Yet, while the None group has had significant outflows they have had much more significant inflows. 4.4% of American adults have switched into this group from Catholic (green). This is 11 times greater than the move from None to Catholic. 3.5% have moved to None from the Evangelicals, more than double the outflow, and 2.7% from Mainline Protestants, almost triple the outflow. There has also been an inflow of .8% of American adults to None from people who classified themselves simply as Protestant, without being willing or able to be more specific. This is represented by a blank space on the chart as we don’t know specifically from where these moves came.

One of my co-workers, who is an atheist, had this interesting question: “If 8.7% of Americans have switched from some type of religious belief to None is one generation, how many generations will it take for religion to be extinguished in America?” This question can’t really be answered, it assumes too much, that rates of change will remain constant, that all religious groups will have the same rate of loss, that help won’t come from other countries, and that God won’t intervene. But the fact that this sort of question has some sort of validity must be of concern to Christians everywhere.


Catholics (green) have the most to be concerned about. They are experiencing huge outflows of 4.4% to the None group, 2.8% to the Evangelicals, and a smaller amount of 1.6% to the Mainline Protestants. This is without any real significant moves to Catholic from any other religious group. In total, 10.0% of adult Americans have moved from Catholic to one of the other groups, and only 2.5% have moved to Catholic from these groups. From other surveys we know that most of this loss has occurred in the North East. While other surveys, like A.R.I.S have shown that Catholic numbers have not fallen significantly overall, this is because the one thing that has kept Catholic numbers from falling significantly overall has been Hispanic immigration in the American Southwest. If you were to visualize this on our chart, you would need to imagine that part of the green column that extends from the top to the bottom of the chart is in fact an input to American Catholicism from Catholicism in other countries.


Evangelicals are at best currently treading water. Their inflows have been matched by outflows, albeit coming from different sources. Much to my surprise, Evangelicals and Mainline Protestants have in fact been swapping members, I had expected much more of a move from Mainline Protestants to Evangelicals. This has not been the case as 2.6% of Americans have moved from Evangelical to Mainline and 2.5% have moved back the other way. More on this later in the post.

Evangelicals have gained 2.8% in moves from the Catholics, and 1.6% in moves from the None group, but have also had 3.5% of their group move to the None group and 0.8% move to Catholics. These moves have largely offset each other as well.

A total of 1.1% has moved to the Historical Black Protestant group (blue), which I will discuss in greater detail when discussing this group. 1.1% have also moved to the “Other Religions” (purple) group which has only been replaced by 0.5% coming back the other way.

As an Evangelical, I find the moves to and from the None group and the Other group quite disconcerting. One of the attributes of being Evangelical is being willing to share the good news of Jesus Christ to the “lost”. With apologies to the Eastern Orthodox (who I grouped with others solely for numerical reasons), the target for Evangelicals is primarily the None and Other groups. Yet, twice as many Evangelicals are moving to these two other groups than are moving from these other groups into Evangelicalism. Clearly Evangelicals have failed mightily in their call to be Evangelical.

Treading water is a dangerous place for Evangelicals to be, because studies like A.R.I.S. show that they are an aging group facing a serious generational horizon, similar to that which has already been experienced by the Mainline Protestants. With a slow leaking away into the None and Other groups and a soon to be experienced generation decline, the Evangelical group looks to be quite different and smaller a generation from now. This is not necessarily a bad thing, but it will require the next generation of evangelicals to have a focus and a vision that the current generation does not seem to possess.


This is the first time that I had worked with numbers that has a separate classification for Historical Black Protestant. This group is made up of a Black Churches from a number of different backgrounds, the largest of which currently are Baptist at 4.4%, Pentecostal at 0.9% and Methodist at 0.6%. A significant majority of the Black Protestants could also quite easily be also classified as Evangelical, so it is not surprising that their largest influx, 1.1%, comes from the Evangelical group. All other inflows were 0.3% or less. From the data that I was able to obtain I was not able to determine the outflows from this group, but I believe them to be less than the inflows. At 6.9% in total they are a significant part of the American landscape, and it will be interesting to see what role they play in future American society.


As mentioned above the Mainline Protestant group has been holding up much better than I expected in terms of its inflows and outflows. Their inflows and outflows with Evangelicals have been virtually identical. They have gained twice as many from the Catholics (1.6%) as they have lost (0.8%). On the other hand they have had 2.7% move to the None group, and only gained 1.0% from that same group. They have also lost 1.1% to the Other group, but only had 0.4% coming back the other way.

It is clear then that the declines that the Mainline Protestants have experienced over the last 40 years have not largely been because of people switching to other groups. Instead it is because they have gone through a very large generational horizon. I recently read an older study that argued quite convincingly that most of the decline in the Mainline Protestant group could be attributed to the birthrate within that group. What we have seen here in this chart would tend to bear that out. While the worst of their declines may be behind them, like the Evangelicals they have significant work to do.


For those who are fans of the TV show “Lost”, this group does not refer to those who battled the Dharma initiative, although they too were called “The Others”. Instead this group is composed of those who did not fall within any of the other classification, and whose numbers were small enough that individually it did not make much sense to show the moves in and out, even if that were possible. When taken as a whole however we are able to look at the moves in and out. Both inflows and outflows occurred proportionally from across the religious spectrum. With the exception of the Black Protestants, inflows were in the range on 0.7% to 1.1% from the Non religious, Catholic and Protestant groups, and outflows were roughly about half that number.

This group is composed of the following:

Eastern Orthodox, currently at 0.6%, and unchanged from the childhood numbers. My apologies to Internet Monk contributor, Father Ernesto, for grouping his church family in this group.

Other Christian, not contained in any other classification, 0.3%, unchanged.

Jehovah’s Witnesses, currently at 0.7% and up 0.1% from the childhood numbers. This group had relatively large inflows and outflows, but ended up with very little change.

Mormons, 1.7%, down 0.1%.

Other religions, 4.7%, up 1.2%. It is this Other religion sub-classification that is of most interest here because it is the only one that has had significant change. When we look at the details within this sub-classification we see that of the other major world religions, only Buddhism had an increase of 0.3% of Americans The other increases (0.9%) came in other religions which were not specified, but were not among the world’s major religions. All we know is that half of this change came from Protestants. The Protestant sub-grouping was also not specified. This is an area in which I would like to know a little more about.


On the expanded chart, you will see a little white column that represents the 0.7% (childhood) to 0.8% (current) who could not or would not complete the survey.


There is a lot to digest here, and I am afraid that the tendency will be to jump to conclusions. Clearly some serious introspection needs to occur within the Christian community. Part of what will aid that introspection will be my next post where I will continue my analysis. The “Flux” survey includes what must be one of the largest series of exit interviews ever done, and I think this information is vital to where we need to be focusing some of our attention. Therefore, in my next post I will summarizing the reasons why all these changes have occurred.

As usual your thoughts and comments are welcome.


  1. MB –

    Who were being considered as Protestants, which differed from Evangelicals?

    And, by the looks of your EC blog and this post, it seems that you are back in the blogosphere world? Is that correct?

  2. Hi Scott,

    I am still taking a break from my regular blog at I will be posting here about once a month as iMonk permits. I will go back to the EC blog once time permits.

    As to the makeup of Evangelical, Mainline Protestants and Historical Black Protestant here are the major subdivisions.

  3. What a great chart. I stared at it for ten minutes. Thanks!

  4. Memphis Aggie says

    In polls many folks who self identify as Catholics are “cultural Catholics” (to borrow’s term) who don’t attend weekly mass. This is especially true in the Northeast. If you consider that large sub-population then you could argue that Catholic numbers are inflated. I used to be “culturally” Jewish, meaning I identified myself as a Jew although I never actually practiced the religion seriously. You could reasonably argue that I would’ve been more properly categorized as “none”, which I suspect is underestimated.

    Knowing the degree of classification error in each category would be very informative. In studies on Catholics opinion and voting and other such things strongly correlate with frequency of attendance at mass. The actual number of practicing Catholics may be as little as 1/3 of the self identifying number. Loosing people who don’t show up is quite different from loosing people on questions of faith or dogma.

  5. Memphis Aggie,

    Absolutely! Self identification always gives the highest numbers, typically double that of regular attendance numbers. I think that what we see here raises some important questions about discipleship.

    “Loosing people who don’t show up is quite different from loosing people on questions of faith or dogma.”

    I will address this directly in my next post as I look at the reasons for the switch.

    Is one more important than the other? I would rather have someone disagree with me and leave, having thought about what he believes and acted upon it, rather than slip away because I haven’t taken the time or energy to build the relationship that would get him to church and learning about his childhood faith.

  6. Bob Brague says

    I find your written narrative very interesting and informative, but I can make neither hide nor hair of the colorful chart. The best thing you could do with it, I think, is send it to the Metropolitan Museum of Art in New York City and suggest they replace some old, worn-out Jackson Pollack painting with these new, vibrant colors.

    If you got all your written narrative from that chart, then I think you have been in the chart-interpreting business a little too long.

  7. From Landscape Survey: How often do you receive a definite answer to a specific prayer request — would you say at least once a week, once or twice a month, several times a year, seldom, or never?

    Jehovah’s Witnesses stated they had a definite answer most often of all groups surveyed!

    A lot of information – but what exactly does it all mean? Do you know of anyone who is trying to “digest” all of this information in specifics? Or is that possible? Thanks for giving this a go in some manner. Look forward to the next post.

  8. Keith B says

    I love the chart!!

    I haven’t kept current with software, what did you use to make it, or is it hand done?


  9. Hi Bob,

    It was actually the other way around. I created the charts from the numbers. Once you realize how the chart works, it is really quite simple.

    Currently there are 228,182,000 adults in the U.S.A. The graph is 1000 pixels wide, so each pixel width represents 228,182 people or one tenth of one percent of the entire adult population.

    Let us focus in on just one move. Towards the left side of the chart is a green stripe that moves from top to bottom and is 44 pixels wide. You will notice at the top of the chart it is below the green bar that runs along the top, and at the bottom of the chart it is above the red bar.

    44 pixels represents 4.4% of American adults, or just over 10 million people. This means that 10 million Americans who currently have no religious affiliation (the red bar at the bottom) were raised Catholic (the green bar at the top).

    So quite simply the wider the stripe, the bigger the move from childhood faith at the top, to current faith at the bottom.

    Hope this helps.

  10. Keith B.

    I used Microsoft Excel to calculate the numbers, and Microsoft Paint to create the chart! I suppose if I had access and experience with Illustrator or Photoshop, then creating the chart would have been a lot easier. As it was, any time I made a mistake, (and I made a few) it would take me quite a bit of time to erase and redraw the line.

  11. FollowerOfHim says

    Steve Jobs visited the Xerox campus in the early 80s, and afterwards said that once he saw their prototype GUI, that was all he “saw” the rest of the visit. The Mac was born.

    As a statistical analyst, I had the same reaction to your chart. I can think of lots of applications where such a chart would be useful. I hope I can sleep tonight, thinking about how to automate something like that… and I think Jackson Pollack would have like it, but would have thought it a little too simple.

    The Historical Black Protestant category brings up a subject I’ve often wondered about. Most postings on these blogs appear to draw a delightfully diverse crowd — Eastern Orthodox Cubans-Americans, Irish Catholics, and Memphis Aggies (I realized recently it wasn’t a name), etc., but very seldom any self-identified African-American voices. If you have any further analysis of this segment, I’d enjoy reading about it.

    Expanding a bit on this topic — and this of couse would best be pursued in a SEPARATE THREAD someday when the subject arises organically — white evangelicals appear to react differently when a black person says, e.g., “Thank you Jesus!” than if another white person says exactly the same thing. I suspect it has something to do with the sense that such expostulations are viewed as simply standard expressions within African-American culture, while they serve to identify one as a probable evanglical if, like me, one is a bit melanin-challenged. It’s the sort of division that Churchill spoke of regarding his own country and America: “Two peoples separated by a common language.”

    In any case, thank you for the analysis; you are quite skilled at getting your points across without snowing everyone with too many numbers.

  12. All by hand!


    The chart reminds me of how Charles Joseph Minard portrays the losses suffered by Napoleon’s army in the Russian campaign of 1812.

    I’ve always wanted to take Edward Tufte’s class on presenting data.

    I think you have done an excellent job of visually telling the story of the numbers.

    On the other hand, being a bit picky, I see a couple blanks that I would wonder about:

    1) A small grey area on the right end of the bottom None bar.

    2) A wider grey area on the left side of the top Black Protestant bar.

    I’m guessing you just got tired of fixing it in a paint program 🙂 (I guess you set the brush width to the number of pixels each stroke should be, then started drawing)

    I could see doing it with lines in Illustrator, Freehand (if it still exists) or even Corel. Start with a line with a midpoint, then duplicate it across as many times as you need, (1000 if you’re using your numbers) then color the correct lines, probably set up some layers for ease of use, then drag the bottom of each line or number of lines to where you want it to go, then move the midpoints to get a nice smooth curve.

    A lot of work!

    But the brilliant part is the idea.

    Great job!

    And thanks iMonk for this post, it’s beautiful.

  13. Mike Bell, great job! Your stats seem to reflect what I am seeing in the North East. It is no doubt a blessing that we can’t see how many of those churches are spiritually dead.

  14. Hey Michael,

    You provide a good, thorough breakdown of the new Pew survey results, and I am interested to hear the analysis of these data from an informed Christian perspective. Your next post will have a lot of fun hypotheses, I’m sure: the full report of the survey shows how deep the Pew Forum folks went in trying to determine exactly why a person would “switch faiths,” which, when analyzed critically, should provide a better picture on the relationship between one’s faith and all the other things that are going on in one’s life, which I think is what you may be trying to get at in this project of poring over the survey results.

    I thought your co-worker’s question was interesting, about whether based on this data can we ask when religion will be extinguished in America (extinguished is kind of a harsh word, but I guess it fits the perspective). I agree with your take, that this question makes many assumptions: of course rates of inflow and outflow will not remain constant, and I have to think that there are widespread cultural signifiers that act on these flow rates in ways that might not be clear, even given this survey. It would be interesting to see what these rates were like 25, 50, 100 years ago; surely then somewhat significant predictions could be made as to what the religious landscape in America will look like in the future.

    As someone who was raised Presbyterian, fell out of it in my teens, and now am unaffiliated (under which category would “Jedi” fall?), seeing data informed by paths similar to mine not only piques my interest, but makes me enthusiastic about gaining insight towards why we believe the things we do, and how much control we do or do not have over our beliefs. All that fun stuff. Keep up the good writing and analysis.

  15. I quite agree with Maggie about the Catholic numbers. Any kind of research that purports to measure the number of Catholics based on self-report is going to be hopelessly inflated. Catholicism has become a pseudo-ethnicity for some people, and they will identify as Catholics despite the fact that they do not attend weekly mass and disbelieve in many of the Church’s core teachings.

    The institutional church cannot be relied upon to correct these numbers, as there seems to be a pastoral policy to keep these folks registered in their parishes in order to facilitate easier return to the fold.

    All considered, it seems likely that things are even worse for the Catholic Church in American than the Pew research would suggest. As a Catholic layman, I can understand the decision to “leave the light on” for the fallen away who wish to return. The reality, however, is that dioceses that are returning to a more orthodox stance and emphasizing Catholic identity are growing in both laity and clergy, while the rest of us are watching our people die off or drift away.

    Still, I have hope that I’ll see these numbers turn around in my lifetime. If the number of priests falls off to critical levels, we might just be mission ground for African and Latin American priests. In my experience, priests from these areas combine orthodoxy with evangelism and energy in a way that might just turn things around.

    Thanks for the analysis. Keep up the good work.

  16. Keith B.

    Two gold stars for being observant. You wrote:

    On the other hand, being a bit picky, I see a couple blanks that I would wonder about:

    1) A small grey area on the right end of the bottom None bar.

    2) A wider grey area on the left side of the top Black Protestant bar.

    I did address both of those in my post. In regards to number one I wrote:

    There has also been an inflow of .8% of American adults to None from people who classified themselves simply as Protestant, without being willing or able to be more specific. This is represented by a blank space on the chart as we don’t know specifically from where these moves came.

    Regarding the Historical Black Protestant I wrote:

    From the data that I was able to obtain I was not able to determine the outflows from this group, but I believe them to be less than the inflows.

    If either of this information was known it would have a small impact on the rest of the chart, but not of a very significant nature.

  17. Keith B.

    Paint allowed me to draw lines and S curves up to 5 pixels wide. For S curves that were longer than that I drew a 1 pixel wide S curve to represent the left side of the line and a 1 pixel S curve to represent the right side of the line, and then filled inbetween.

  18. Teenage Mutant Ninja Tertullian says

    The answer to your atheist friend’s question–how soon before Christianity will be extinguished in America?–is never. Just not on. Many, perhaps most, of the unaffiliated actually do have a religious orientation, if not belief set or affiliation. If not them, then their children. Every human population groups that we know anything about seem to have something like religion.

    Memphis Aggie: “Knowing the degree of classification error in each category would be very informative.”

    This is impossible in principle to decide. Can people be “wrong” about what religion they are? Is that even possible in principle? If so, then what must one do or believe to be Catholic, or Jewish? A good case can be made for saying that the Jew who goes to the synagogue once a year despite not believing, is MORE strongly tied to the religion than the Baptist who goes to church twice a week because he believes! (The Baptist might change his mind or slack off, but the Jew would have to be actively converted away.) Many religious groups have their own criteria, which cannot be meaningfully compared.

  19. sue kephart says

    Regarding self-reporting Roman Catholics: As a Protestant I did not realize the RC Church has a different way of reporting members than most Protestant denominations.

    If you were baptized RC you are RC. If you join another Church you would be excommunicated from the RC Church.

    Most Protestant churches purge their membership rolls at some interval in some fashion.Thus someone not showing up after years and years would not longer be a member. Not so RCs.

    We had an exchange student from Italy. She ‘self reported’ she was RC. This was because her father told her she was as she had been baptized as an infant in the RC Church. Other than that she had never been to a worship service of any kind and know no one who had.

    She attended worship services with our family and Sunday school with our daughter. The teenage kids were amazed that this girl knew absolutely nothing. Even kids they knew who never went to church knew something about God and religion. But not our exchange student.

    She is counted as a RC by the Roman Church and considers her self as one. She said she wished sometime she could go to church again but didn’t know anyone at home who went. So sad.

  20. Teenage Mutant…

    “Every human population groups that we know anything about seem to have something like religion.”

    That of course would be the biggest reason. Note that half of those raised with no particular faith end up in some type of faith group. There is always going to be that sort of ebb and flow. Also as a group gets down to its committed core, the outflows from that group will diminish both in terms of numbers and rate.

  21. Just for Quix says

    @ Mike Bell

    Compliments for hand creating the chart. Some of my thoughts:

    Many people like to deal in percentages when discussing growth or decline. One useful caution is that the percentages shown in “childhood faith” and “adult/current faith” categories are represented as extrapolated percentages of _the American population_. Therefore, in your summation, when discussing growth/decline percentages within a subgroup, the study numbers must be re-computed to be a bit more illuminating in print. (The graph certainly helps show this point visually.)

    For example, it is not quite as eye-opening to say Catholics have lost 4.4% to the None category; it would be better said as 4.4% of Americans who were childhood Catholics now identify as Not Affiliated. This represents a 14% proportion of childhood Catholic subgroup volume is now non-affiliated. That represents 40% of Catholic losses over this timeframe is toward non-affiliation. When you look at it that way, it really puts a point on the arrow.

    As a former Mormon I was pleased to see that Mormons have not declined a tenth of a percent, except if we’re only talking market share. Their decline has been about 5.5%. While not as big a subgroup share decline as Catholics (-24%) considering the high LDS proportionate proselytization emphasis, and that Mormons have 77% of their membership that were born Mormon, and that they have the second highest rate in religiously homogenous marriages, these are not encouraging data at all for that group.

    But that’s nothing like the unaffiliated None subgroup that has had nearly a 50% negative retention rate. Granted, inflow has been significant in percentages — and a cause for concern in Christianity — these are hardly the data to encourage firm atheists who would like to think there is reason to think religion will be extinguished in America in a few generations.

    Contrary to the secular evolutionist “vision of the future” it is much more probable to suspect that the pendulum will swing back, especially as a younger Christian generation finds new ways to reach back to the denominationally disenfranchised. Change is certainly afoot, and not all encouraging change. But if anything these data and worldwide historical data don’t prove is that modernizing, democratic, technologically advancing societies ‘progress’ inevitably toward non-religiousness. Tim Keller addresses this point very cogently in his “A Reason for God”.

  22. Just for Quix writes:

    For example, it is not quite as eye-opening to say Catholics have lost 4.4% to the None category; it would be better said as 4.4% of Americans who were childhood Catholics now identify as Not Affiliated. This represents a 14% proportion of childhood Catholic subgroup volume is now non-affiliated. That represents 40% of Catholic losses over this timeframe is toward non-affiliation. When you look at it that way, it really puts a point on the arrow.

    You certainly stated it well in the first part of your response. One reason why I did not get into percentages of groups or percentages of changes was to cut down on the amount of the numbers in the post so as to not overwhelm people. Your points however are very valid.

  23. sue kephart,

    Self identity is always much higher than actual attendance. This is true for Evangelical, Mainline, and RC churches. Southern Baptist attendance in the U.S. is roughly 30% of its membership. In Canada where I live, I know of a number of people who would identify as Anglican, or United Church of Canada, or Presbyterian, who don’t attend and probably haven’t even made it on to the membership rolls. The younger generation with the same lack of attendance patterns is much more likely to claim no affiliation.

  24. My first thought when looking at the chart was “is this some kind of a joke?”

    My second on looking at it more closely was “thanks!”

  25. Hey Michael, I don’t know if you’re answering direct questions, but I’m curious about something. I didn’t see anything in your data about noninstitution simple and home church types. Were we included in the “other” category? I know we’re probably hard to chart since we tend to fly under the radar, but I know in my own community we’ve gone from nonexistent (absolute zero, as far as I know) to several dozen in about five years. Are we an unusual case or is this kind of growth happening all over. I know George Barna gave some stats about organic, simple, home, and other emerging noninstitutional forms of church, but his numbers seemed a little high to me.
    Just curious.

  26. Ron,

    As this was a self-identification survey it would depend on how the person responded. I know that I was a member of a house church for many years, but the house church was associated with 1. Evangelicals and 2. Plymouth/Christian/Open Brethren, so I would have been slotted in the Evangelical side of things.

    There was a significant number of people in the survey who identified themselves as Evangelical non-denominational or non-specific. People in a house church that is not affiliated with any denomination and one that is evangelical would probably identify itself with one of those two categories. Mainline and Historical Black groups also had those two categories.

    As as side note I should mention that for years the two denominations with which I have spent the most time, the Brethren and Christian and Missionary Alliance, insisted that they were not denominations, but just groups of like minded Christians meeting together. Whether the home church movement goes this way remains to be seen. Already we see examples of house church ABC being a member of the XYZ house church network.

  27. The question your atheist friend poses is sort of absurd. 8.7% of Americans aren’t going to leave religion every generation, especially when religion drops to below 8.7% of the population.

    At current rates, 50% of non-religious people become religious every generation, and 9.4% of religious people quit every generation. Equilibrium is reached when the 9.4% who quits religion is equal to the 50% who become religious.

    At current rates, the next generation would actually be ever so slightly more religious:

    50% of 16.1% is 8.05% of Americans.
    9.4% of 83.9% is 7.89% of Americans.

    So both groups would be treading water. The rate of defection from religion would need to accelerate again next generation for non-religiosity to make a gain.

  28. Excellent point Fearsome Comrade.

    I do feel that what you have presented here is a best case scenario, whereas my friend is hoping for a worst case scenario. My fear is that as more people become non-religious it becomes more of an attractive option.

    Time will tell.

  29. Dear Michael,

    Thank you so much for this thoughtful first take on the Pew Forum data. I am really looking forward to your further analysis. That graph is spectacular.


  30. Why is the third-largest category still called “mainline”?

    [this is not actually re:BHT]