I see people make assumptions about data and statistical analysis that just make me scratch my head. So, this is a rant of sorts. A short list of things that come up again and again when I post any type of graph on social media.

You better believe that I’m going to be dropping this link into my replies on dozens of occasions in the future.

1. The average rate of public opinion change is much slower than people anticipate.

I can’t tell you how many times I will post a graph that has data collected in 2024 and shows no major changes in public opinion and someone will reply, “If you had data from 2026, I’m sure it would show a lot of movement.” That kind of response is just so divorced from how public opinion actually works that I feel like I need to take some time to really lay out how wrongheaded it is.

For those aspiring data analysts out there who are just learning how to code and make graphs, let me give you a piece of advice — if you see a survey question that has been asked with regularity over a long time move more than 3-4 points in a short window of time (less than two years), then you need to check your code again. It is highly unusual for stuff to shift that much that quickly.

Take, for instance, the General Social Survey’s question about allowing a woman to have access to an abortion if she wants one for any reason.

If you track the individual points over time, there’s one really weird aberration between 1978 and 1980 when support for abortion jumped 8 points. In terms of actual percentages, it was 33% to 41%. But here’s the other part of the story — in 1982 it was already down to 37% and then it declined to 33% in 1983. So a complete reversal in a span of five years. You see this same thing happen between 1991 and 1998 — a temporary rise, then all those gains were given back.

You know what the average change is year to year in this graph? It’s 2.6 points. Which is inside the margin of error for most samples. But as you can tell in how I annotated the graph — something happened between 2016 and 2018. I think we all know what I’m talking about right now. In that two-year window, support for access rose from 44.8% to 49.9%. A full five percentage points. Now, that could have just been polling noise, right? We would be right to be cautious about writing a story with a headline about the huge surge in pro-choice views.

But here’s how we know that’s the real deal — the trend has continued. In 2021 it rose again to 53.1%, up nearly four more points from the prior survey. Then in the last two years it was 57% and 58% respectively. Which is more plausible: that every year of the GSS from 2018 through 2024 has just “missed” on this question in the same way? Or did an external event (or series of events) lead to a macro-level change in opinion?

Also, don’t let me forget to point out the fact that the last couple of surveys have “calmed down” on the movement after a whole bunch of rapid change. There was a total of ten points of change in four decades of polling. Then, we see a clear directional shift that can be easily traced to an external set of political events.

2. The average person thinks that THEIR world is THE world.

I get a version of this email about once a week. A reader, usually a thoughtful one, writes in with a question they’re sure is right in my wheelhouse. Here’s a recent example: “Have you worked on the question of whether the error of Christian Zionism and/or Dispensational theology is waxing or waning in American or global Christianity? Seems right up your alley.”

I understand why people think this is a survey question. It’s about religion. It’s about change over time. Those are the two things I write about every single week. The person that emailed is probably an evangelical pastor and went to seminary. They can read Hebrew and they listen to podcasts about minute details of Koine Greek. A lot of their friends are pastors and they stay up late at night debating pre-trib vs. post-trib.

In other words, they come from a very specific subset of the American population. Which makes a question like this basically unanswerable with survey data, and the reasons why tell you a lot about what surveys can and cannot do.

Think about what happens if I field a general population survey of 1,000 Americans (which would cost thousands of dollars) and try to say something about dispensationalism.

Start at the top: about 60% identify as Christian. We’re already down to 600.

But dispensationalism is a particularly evangelical phenomenon. Catholics don’t teach it. I don’t think that the local Episcopal Church is holding a mid-week Bible study on the rapture. The Orthodox have never heard of it. So we’re really talking about the quarter or so of the country that’s evangelical. Now we’re at roughly 230 (maybe even less).

But a belief system doesn’t transmit itself — somebody has to teach it to you, and that means being in the room. Less than half of self-identified evangelicals are in church on a given weekend, and regular exposure is what it takes to absorb an entire theological framework — not just watching Kirk Cameron in a low-budget movie about eschatology.

We’re now down to 110 people.

Now, of those 110, how many attend a church where the pastor actually teaches dispensational eschatology? Plenty of evangelical pulpits — and nearly all of the non-denominational megachurches absorbing everyone these days — are functionally agnostic about the end times. The pastor is vaguely premillennial, never preaches Revelation, and would rather talk about your marriage or how to raise decent kids. When’s the last time Joel Osteen preached about this topic? Without googling it, I’m going to say “never.” But I’m going to be generous and say half. We’re at 55.

And then the final cut: of the people sitting under that teaching, how many could articulate it well enough to answer survey questions about it? Knowing your church believes in the rapture is not the same as holding a position on the timing of the tribulation. Maybe half again, and I think that’s on the high side.

So my survey of 1,000 Americans contains perhaps 25 or 30 people capable of telling me anything meaningful about dispensationalism. The margin of error on a subsample of 30 is plus or minus 18 points. I could field that survey twice and “find” that dispensationalism doubled. Any trend I claimed to detect would be statistical noise wearing a headline.

This is the part of survey research nobody tells you about: the questions get harder to answer not because they’re more profound but because the denominators get smaller. I can tell you about Christians with great precision. Evangelicals, very good precision. Weekly-attending evangelicals, decent. Weekly-attending evangelicals in prophecy-teaching churches who know what pre-tribulationism is? At that point I’m not doing statistics anymore. I’m doing anthropology with extra steps.

3. The average American is really bad at guessing the size of religious groups.

If you’ve hung around here enough, you’ve heard me bellyache about this one on a regular basis. So much so that I wrote a full post entitled, “Why I Can’t Tell You How Quakers (or Unitarians) Voted in 2024.” I just can’t tell you the number of comments I get on a weekly basis that ask me to include groups like the Unitarians, Latter-day Saints, Anglican Church of North America, or Presbyterian Church in America. (And yes, I chose those specific groups on purpose, by the way, because those are the ones that appear in my replies with the highest frequency.) In contrast, it’s very rare to have someone ask me about Buddhists or Hindus. The social media crowd is not a complete representation of American religion.

Before I go any further, I want you to stop and in your mind just estimate the sizes of the four groups that I’ve mentioned in the previous paragraph. I know the answer to this because I get asked about it all the time:

UUs have 130,000 members. The Latter-day Saints are comparatively large at ~7 million. The ACNA is 125,000 and the PCA is 405,000. The only one that gets to 1% of the population is the LDS. The other three groups are rounding errors in the grand scheme of things. You want to guess how many Protestant denominations have at least one million members? It’s around 15 of them. How many are larger than 5 million members? It’s one — the Southern Baptist Convention.

A couple of years ago YouGov did this really fun polling experiment where they asked respondents to simply guess what share of the population met a whole bunch of criteria. For instance, what share of Americans are transgender? The YouGov sample said it was 21%. The actual percentage is certainly below 1%.

Well, there were a couple of religious groups that were asked as part of that battery.

You can see the only one that they got even close on was Christians. The median guess among the respondents was that the United States was 57% Christian. According to YouGov’s numbers back in 2022, it was 70% (which is not right, by the way). The Pew estimate in 2023-2024 was 62%. So, that’s not too bad at all. And even when it comes to Catholics, the public wasn’t off by that much. The median guess was that 39% of Americans are Catholic. The real number is closer to 20%.

But when it comes to really small religious groups, things just go sideways. The median guess as to the share of Americans who are atheist was 28%. It’s about 5% in actuality. Americans think that a quarter of our country is Jewish when it’s really about 1.5%. And then the median guess regarding Muslims was 20%, when it’s really closer to 1%.

This touches on something that I have been thinking a lot about lately — people LOVE the idea of outliers. Their minds fixate on the one person who bucks the trend or the one religious group that they don’t know much about but have seen a few TikToks that describe them. There are shows about the secret lives of Mormon wives. There aren’t shows about a bunch of Methodists living in Cedar Rapids, Iowa. The LDS seem exotic for some reason.

Whenever I post a graph that points to the fact that a significant majority of Americans will die with the same faith that they were born into (65-70% of us), you know what the first comment will be? “Well, I grew up a fundamentalist Pentecostal, then converted to Orthodox Judaism, and am now a pantheist who dabbles in Buddhism.” Americans hate being average. They want to be different and they are very proud of that fact.

My job, though, is to describe the median outcome, not the odd one. The work of quantitative social science is to generalize. That really rubs a lot of people the wrong way. Such is my lot in life.

After all that, I can only say one of two things: “You’re welcome” or “I’m sorry” for taking you through what the world looks like from my perspective on a day-to-day basis. It’s fun. And frustrating.

I’m probably going to be doing this for the rest of my life. Maybe one day I’ll get used to it.

Code for this post can be found here.

Ryan P. Burge is a professor of practice at the Danforth Center on Religion and Politics at Washington University.

Here are three different datasets, all collected in a pretty similar time frame, that just test that relationship.

You can see that in the Cooperative Election Study, the relationship is a positive one: an individual with a graduate degree is four percentage points more likely to attend a house of worship weekly compared to someone who stopped at high school. In the General Social Survey, there’s not a clear pattern at all. It’s 24% among high school grads compared to 25% among those with a graduate degree. That’s inside the margin of error.

In the 2023-2024 Pew Religious Landscape Study, the relationship is not clearly linear. Attendance is 25% among those with a high school diploma or less, then it dips slightly among those with some college credits, only to rise among those with a four-year degree and go even higher for folks with a master’s degree. Among the most highly educated in Pew’s sample, 28% go to a house of worship on a weekly basis.

Thus, it’s empirically false to say: people with less education go to church more. There’s zero evidence of that in the best data we have. It’s also accurate to say that the relationship is likely mildly positive — those with higher education are more likely to be frequent worship attenders. It’s not a huge difference, but it’s persistent.

Now let me simplify and expand this using Pew’s Religious Landscape Survey because they also included a question about frequency of attendance at online worship.

The bars on the left should come as little surprise based on the previous graph. Among folks who have at least a four-year college degree, 26.2% attend a house of worship weekly. For those who stopped at high school, it’s 24.8% — a difference of just 1.4 points. But look at the online attendance difference. For those whose education stopped at high school, 20.3% say that they watch a service online every week. The college-educated crowd is much less inclined to participate this way: 11.6%.

Now wherever I post either of the above graphs on social media, I get the same comment over and over again. People believe that the discrepancy can best be explained by the fact that those without a lot of education have to work on the weekends, while those with a college degree work a traditional Monday through Friday schedule.

I must admit that the data from the Bureau of Labor Statistics that addresses this is incredibly muddy. Which frustrates me to no end. And I don’t know of any survey that asks about this possibility directly. So that’s where the best part of my job kicks in — I have to get creative. One possibility was to just limit my sample to those who are retired. That would absolutely control for the “working on Sunday” confounding variable.

The problem with that is: people who are retired are old. When they were growing up, educational access was much different than it is today. Thus, people with a four-year college degree (let alone a graduate degree) were a very specific type of individual. It wasn’t just a general cross-section of the American public who went to college in the 1970s. I had to set that idea aside.

I wanted to attack the problem from a different angle. After doing data analysis for nearly twenty years, I’m pretty convinced that education really runs on the same causal tracks as being pro-social. Educated folks are just more likely to be engaged in all kinds of social activities. They serve on school boards. They volunteer at their local library. They’ve just built a thicker, more robust social network than those who stopped at high school. I’ve used this phrase a lot — “Dropping out begets dropping out” — and I really do believe it’s true.

But after poking around the data a whole bunch for this post, I’m convinced of something very important. If a person at the lower end of the educational spectrum can become a regular churchgoer, it opens a door for them to engage in much more pro-social activities. And it will actually make their lives better and close that “drop out” gap in some very meaningful ways.

Read more

The Catholic Church is one of the most doctrinally specific institutions in the world. The Magisterium — the Pope, Cardinals, and Bishops — has issued detailed guidance on everything from the mechanics of salvation to the appropriate use of birth control to how often a faithful Catholic should receive the sacrament of Reconciliation. These aren’t suggestions. They are, in the Church’s own framing, the authoritative teachings of an institution that believes itself to be guided by the Holy Spirit.

And yet, anyone who has spent time around American Catholics knows that the distance between official Church teaching and the lived experience of the average Catholic in the pew can be considerable. This isn’t a new observation — sociologists of religion have been documenting it for decades. But it raises a genuinely interesting empirical question: just how Catholic, in the doctrinal sense of the word, is the average American Catholic?

This is actually a pretty hard question to answer because most surveys of American religion are “general purpose.” They ask about things like religious attendance because that’s fairly universal across traditions, but they aren’t going to dig into the weeds of specific Catholic beliefs and practices. However, I managed to stumble upon a dataset called the American Catholic Laity Poll. It was conducted a while ago (2017), but I think it can still offer us insights into the specific practices and beliefs of Catholics.

For instance, this survey asked specifically about the act of Confession: “How often, if ever, do you participate in the Sacrament of Reconciliation (Confession)?”

The typical Catholic isn’t walking into the confessional booth with any regularity. In fact, just 4% of Catholics say that they go to confession once a month or more. For comparison, the percentage of Americans who attend religious services multiple times per week is in the 6-8% range. In actual numeric terms, there are about 2-3 million American Catholics who go to confession a dozen times per year.

In contrast, a clear majority (56%) describe their confession attendance as “seldom” or “never.” Another 16% indicate that they are going into the booth less than once a year. That means almost three-quarters of Catholics haven’t gone to confession for a very long time. It’s just not part of normal religious practice for most American Catholics.

Of course, I had to break that down by the Mass attendance rates of the Catholic laity. We could easily assume that more Mass attendance leads to more confession participation, but it’s nice to confirm a hunch with actual numbers. In this case, there is a positive relationship between both religious behaviors. Among weekly attendees, 36% say that they go to confession at least a few times per year, and a majority go at least once every twelve months.

But among those who go to Mass less often, confession is not a regular part of their religious routine. Among monthly Mass attendees, almost two-thirds indicate that confession happens seldom, never, or less than once a year. For those who don’t go to Mass much, confession is just not happening at all — 79% say that they never or seldom go. For those wondering, 42% of Catholics are in the never/seldom category for both Mass attendance and confession participation.

Another Catholic distinctive is the concept of the Magisterium. It’s the idea that the leadership of the Church is the authority on matters related to doctrine, practice, and views on controversial political issues. There are two intriguing questions on the survey worth exploring:

1. Catholic bishops often speak out about politics and elections, as well as about policy issues like healthcare, immigration, and foreign affairs.
2. Catholics can disagree with aspects of Church teaching and still remain loyal to the Church.

It’s pretty clear from the graph above that the average rank-and-file Catholic does not feel beholden to the teachings of the bishops on issues related to politics and policy. In the full sample, only about 1 in 10 Catholics said, “I try to follow the bishops’ guidance and instructions on political and public policy matters.” The modal position was clearly the middle path: “I consider what the bishops have to say about politics and public policy, but ultimately I make up my own mind.” However, note that among those who don’t go to Mass much, nearly half say the bishops’ views are completely irrelevant to them.

On the question about whether Catholics can disagree with the Church’s teachings and still be loyal, the differences between those who go to Mass every week and those who don’t go at all are pretty small. Even among the most religiously active Catholics, a strong majority (81%) agree that “Catholics can disagree with Church teaching and still remain loyal.” Among those who go less often, it’s north of 90%. So, it looks like the average Catholic extends a whole lot of grace and flexibility to their brethren.

Let’s take this a step further, though. I wanted to create a “conformity index” based on a whole bunch of questions from this survey of Catholic laity. I pulled together an additive index that touched on four broad areas of investigation: behavior, belief, moral authority, and institutional allegiance. I put them all in the table below so you can see the detailed question wording.

Some of these factors have a fairly high “conformity score,” like being a registered parish member (60%) or believing in the physical resurrection of Jesus Christ (64%). But there are other questions where only a very small percentage of Catholics conform. For instance, only 9% of Catholics say the Church has the final say on birth control issues, and only 13% say the same regarding abortion.

I indexed each respondent on a scale of zero (meaning conforming on no items) to 10 (conforming on all ten listed in the table above). I visualized the distribution in the graph below. The mean conformity score for the full sample was 2.9/10.

Nearly 30% of Catholics scored a zero or one on the Catholic conformity index. Half of them scored between zero and two. In contrast, only 6% scored at least a seven. As you can quickly tell from the bar graph, the modal score was 3/10, followed closely by 1 and 2. In other words, the average Catholic is not regularly engaging in practices or holding beliefs that are distinctly Catholic.

Of course, that varies noticeably when I break it down by the average level of Mass attendance. (I do understand that one of the items in the additive index is about Mass attendance, so this does confound things a bit — but I still think it’s a helpful way to think about committed Catholics vs. cultural ones.)

Among Catholics who don’t go to Mass much, the mean score on this scale is just 1.75 and the modal score is one. The average Catholic represented by the red bars at the top of the graph is really not that Catholic in any meaningful way. I would love to see how the typical Protestant scores on this scale and compare it to never-attending Catholics, but alas, I am data-limited here.

For Catholics who report monthly Mass attendance, the mean is a full point higher (2.9) and the modal score is 3 out of 10. You can still see that the bars are heavier on the left than the right, though. Over 40% of monthly attending Catholics score no higher than a two on this index.

What about the most active Catholics? Well you do clearly see a much higher mean (5.1). Which is nearly double the rate of the monthly attendees. This graph is fuller on the right side of the plot than the left side. Among weekly Mass attenders, 18% score at 7-10, three times the rate as the full sample of Catholics. There’s no doubt that this group has the highest level of conformity to Catholic doctrine and behavior.

I thought you all would appreciate some more granularity on this one, so I recreated the table from above but I did it for each level of Mass attendance.

In terms of behavior, you can clearly see that those most active in attendance are also the most likely to engage in regular confession and be registered as parish members. There’s also a clear gradient on beliefs. For instance, 74% of weekly attendees believe in transubstantiation, compared to 42% of those who attend Mass seldom or never. The gap on belief in the death and resurrection of Jesus was 25 percentage points as well.

But where Catholic conformity drops significantly is in the last two categories: moral authority and institutional fidelity. Even among weekly attendees, the scores are incredibly low. Only 26% of Catholics who attend Mass weekly say that Church leaders should have the final say on abortion, and it’s just 22% for homosexuality.

And because I am trying to be more innovative, I built a whole widget that allows you to “choose your own adventure” in determining what percentage of Catholics are conforming to the Church’s teachings.

That widget can be accessed by clicking on THIS LINK.

You can test combinations to your heart’s content. For instance, the widget indicates that the share of Catholics who attend Mass weekly and go to confession at least once a year is 17%.

How average is the typical Catholic? I think the answer depends, at least in part, on how often they go to Mass. If they rarely show up to Church on the weekend, they don’t hold to a lot of distinctively Catholic behaviors or beliefs.

If they attend church on an average Sunday, they are more inclined to go to confession and to believe in the resurrection of Jesus and transubstantiation. But even among these folks, individualism still dominates their decision-making. They don’t defer to the Church’s authority on matters of public policy, and they don’t look askance at other Catholics who question the Church, either.

What this data convinces me of is that in the phrase American Catholic, the most important word is probably the first one, not the second.

Code for this post can be found here.

Ryan P. Burge is a professor of practice at the Danforth Center on Religion and Politics at Washington University.

This post has been unlocked through a generous grant from the Lilly Endowment for the Association of Religion Data Archives (ARDA). The graphs you see here use data that is publicly available for download and analysis through link(s) provided in the text below.

There’s an old saying that bounces around from time to time: “Change is hard. Not changing is harder.” It’s often invoked by counselors helping clients alter their behavior, particularly regarding substance abuse. But outside that specific context, the opposite is often true. In fact, the strongest force in the world of social science is inertia. What’s the strongest predictor of where someone will live? It’s where they were born. The strongest predictor of one’s partisanship? The political affiliation of their parents.

And if one were to guess a random person’s current religion, the answer is almost always the same: whatever religion they were raised in. We stick with what we know.

The share of Americans who indicate that their current religious tradition is the one in which they were raised is 66%. Most people living in the United States will die with the same faith into which they were born, and that’s been true for decades. But how people switch and why they switch may be changing.

Let’s first tackle the first part of that question: what does religious switching look like in American religion? The alluvial diagram below maps the flow of Americans from their childhood faith to their current one, based on the 2023-24 Pew Religious Landscape Survey (hosted on the ARDA). The color of each band represents a person’s childhood tradition, and the thickness represents how many people travel a particular path.

The thickest bands tell the clearest story. The vast majority of folks who were raised Protestant stay in that same faith tradition throughout their lives, and the same is true for Catholics and smaller faith traditions. But notice the size of the blocks on the left compared to the right. More people were raised Protestant than are currently Protestant. Catholics show an even steeper drop.

The one group that clearly bucks this trend is the “nones.” Their bar on the right-hand side is over twice as large as the bar on the left. The non-religious have far more inflows than outflows, which is how they have grown to roughly a quarter to a third of the American population.

Switching happens, of course, but it’s certainly not the norm. In fact, about six in ten current nones came from Christian backgrounds—roughly a third from Protestant traditions and a quarter from Catholic ones.

Which groups are best at keeping their own, and which ones have a high rate of defection? It’s easier to get a handle on that by calculating retention rate, which is simply the share of each tradition whose childhood faith matches their current religious affiliation.

Jews lead the way here. About three-quarters of people who were raised in a Jewish household still claim Judaism as their present religion. The non-religious are statistically in the same spot: three-quarters of Americans who were raised with no faith tradition are still nones into adulthood. Couple that with the data from earlier indicating that 60% of current nones are Christian defectors, and you can quickly ascertain why the non-religious have exploded in size over the last three decades.

The Latter-day Saints had the lowest retention of any group—just over half—which should send alarm bells through church headquarters in Salt Lake City.

Let’s go one layer down now, by checking how retention rates have shifted based on the birth decade of the respondent.

For both Protestants and Catholics, retention has taken a nosedive among younger adults compared to their parents or grandparents. Among people born before 1980 and raised in a Protestant household, 75–80% still identify as Protestant. Catholic retention among that age group is noticeably lower (60–65%). However, defection among younger adults raised in those faiths has accelerated sharply. Now, half of young people raised Catholic are no longer Catholic; among Protestants, it’s closer to 40%.

The trend for the nones stands alone. Among the oldest Americans raised without religion, almost half found their way into a faith community as adults. Today, 80% of young adults raised as nones are still nones. For comparison, Amish retention has been estimated at 85%. The “none” identity is incredibly sticky now.

So we’ve spent some time poking around the “what” question. It’s time to pivot to the “why” question, and I want to zoom in on a single explanation for religious switching: politics. In 2002, Michael Hout and Claude Fischer published an article in the American Sociological Review titled Why More Americans Have No Religious Preference: Politics and Generations. For anyone who works in this field, this is considered canonical reading.

The abstract describes the conceit of the piece: “The increase in ‘no religion’ responses was confined to political moderates and liberals; the religious preferences of political conservatives did not change. This political part of the increase in the ‘nones’ can be viewed as a symbolic statement against the Religious Right.”

In other words: politics caused people to drop out of the church. Let me provide an updated version of the graph that appears on page 181 of that piece using data from the General Social Survey through 2024.

That result is about as clean as it gets in the world of social science. In the late 1990s, about 15% of liberals were nones compared to 5% of conservatives—a clear divide, but not an impassable chasm. That’s no longer true. Now, 43% of ideological liberals say they have no religious affiliation compared to just 12% of conservatives. A ten-point gap has tripled in size, and there’s no evidence these trend lines will converge anytime soon.

Does that thesis bear out with the switching data? I had a simple hypothesis to test: are people who grew up Protestant or Catholic but became non-religious more left-leaning than those who stayed in that same tradition?

Among Catholic “stayers,” 39% are Republicans and 36% are Democrats. For Protestants who stuck around, those numbers are 47% and 29%, respectively. But among those who switched from Protestant to Catholic or Catholic to Protestant, a clear pattern emerges: switchers inside Christianity are consistently the most Republican of any subgroup examined here.

But what about people who left Catholicism or Protestantism to become non-religious? Among the Catholic-to-none sample, 46% say they are Democrats and only 15% are Republicans. For Protestant-to-nones, it was basically the same: 44% and 16%, respectively. This is strong evidence that politics is wrapped up in why some people leave Christianity to become secular.

However, are people who defected from Christian groups more left-leaning than folks who have been nones their entire lives?

And here’s where I had a “holy cow” moment. The types of nones who are the least Democratic are those who have been non-religious their entire lives. Only 38% say they are Democrats, 42% identify as politically Independent, and 20% align with the GOP. In contrast, among ex-Christians who become nones, roughly 45% say they are Democrats—seven points higher than the “always nones.”

This is fascinating because it suggests a mismatch between what drives people to leave Protestantism or Catholicism and what it means to have never been religious at all. People who actively exit a faith tradition aren’t just drifting; they’re making a decision, and that decision appears correlated with a particular political worldview. Leaving a church isn’t religiously neutral; for many people, it seems to be a political act.

The “always nones” tell a different story. They weren’t pushed out of anything; they simply never had a religious identity to begin with. Without that experience of exit, there’s no corresponding political signal. They’re secular by default rather than conviction, and their politics reflect that—more Independent, less reliably Democratic, and harder to categorize.

Let me lay one more variable on top of this: age. I wanted to test whether politically motivated exit was widespread among older Americans while younger cohorts are switching for reasons that extend beyond the ballot box.

Among the youngest birth cohorts of “always nones” (those born 1990–2004), just a quarter identify as Democrats. For always nones from the Baby Boomer generation, 40% say they are Democrats. Protestant-to-none and Catholic-to-none defectors exhibit the exact same trend line. Among the youngest adults who left their faith behind, only about a third identify as Democrats. For Christian-to-none defectors in their retirement years, a bit less than half say they are Democrats.

This is where Hout and Fischer’s argument becomes both confirmed and complicated. Looking at the older cohorts, they were right. Among Baby Boomers who left Protestantism or Catholicism, nearly half identify as Democrats. These are people whose exits tracked closely with the culture war battles of the 1980s and 1990s. For them, leaving the church and voting Democratic were part of the same identity package.

But among those born in the 1990s and 2000s who left Christianity behind, only about a third identify as Democrats. The political premium that came with religious exit has nearly vanished. These younger “dechurchers” don’t look like political refugees; they look more like people who simply stopped going—gradually, quietly, without it meaning much ideologically one way or the other.

Hout and Fischer weren’t wrong; they were describing a real mechanism operating in a specific historical moment. What they couldn’t have anticipated is that mass secularization would eventually normalize religious exit to the point where it no longer carries a strong political signature. When leaving religion is something a third of your generation does, it stops being a statement and starts being a default.

The story of the nones, in other words, has changed. The first wave left with conviction. The second wave just drifted out the door.

Code for this post can be found here.

Ryan P. Burge is a professor of practice at the Danforth Center on Religion and Politics at Washington University.

But not this time. I noticed a single line in the article that caught my eye,

“As part of the process, the Detroit archdiocese released information online in March about parish finances and attendance data in workbooks that provide a detailed look at the conditions of churches, showing how some are doing well with revenue while others struggle.”

Jackpot. It’s time to get to work.

The Archdiocese of Detroit has pulled together a really impressive website about the process that the Church is undertaking over the next couple of years. They have a relatively new Archbishop - Edward Weisenburger. He was installed as the leader of the Detroit archdiocese in March of 2025. My read of the situation is that he was primarily chosen to secure the organizational and financial future of the Archdiocese. That’s the goal of this discernment process.

The real treasure trove for me is a series of workbooks for the 224 parishes that exist in Eastern Michigan. They are organized across 15 different planning areas and they are posted for anyone to download on this page. The PDF files are awesome because they provide unprecedented access into what is happening at the parish level across hundreds of churches in Michigan.

So here’s my plan. This post is going to just introduce you all to the 30,000 foot view of the Archdiocese of Detroit based on the metrics provided in these workbooks. If this article does well, I plan on writing a series of reports about Mass attendance, sacramental activities, and the finances of these 200+ parishes for my paid subscribers. I think there’s just so much to learn about what is happening in American Christianity through the granular study of this one Archdiocese.

First, I need to frame the rest of this discussion around a misconception that I had about what’s happening in the Detroit area. I, like many of you, have read dozens of stories about how the Detroit metro area was undergoing a population collapse. I mean, the city of Detroit went bankrupt. And that, statistically speaking, WAS the case for a long time. Between 1950 and 2010, the population of Detroit declined by 61%. But now there’s a new reality - metro Detroit has actually experienced some modest growth in the last fifteen years.

The six counties that make up the Archdiocese of Detroit grew by 2.2% between 2010 and 2024 — modest, but not certainly decline. Wayne County did lose residents, dropping about 49,000 people over fourteen years, but that’s a far cry from the hollowing-out narrative that defined Detroit a decade ago. The region has stabilized.

But there are counties in the suburbs of Detroit that have experienced solid growth recently. Oakland and Macomb counties are north of the city center and have seen increases in population of 7.9% and 5.4% respectively. Monroe county is just to the south of downtown, and it’s up 2.6% during this same time frame.

The crisis facing the Archdiocese of Detroit isn’t really about what’s happened in the last decade — it’s about an institution still reckoning with the massive population exodus of the mid-to-late 20th century. The infrastructure, the parish footprint, the number of churches — all of it was built for a city and a Catholic population that no longer exists at that scale. The most recent Mass attendance data makes clear that the reckoning isn’t over.

In 2011, the average weekend Mass attendance in the Archdiocese was about 231,000 people. That dropped below 200,000 by 2015, it declined to 163,000 by 2019. The most recent data that these workbooks provide is from 2024, when Mass attendance was just under 140,000. According to their own reports, the Catholic Church in Detroit is recording an attendance decline of 4% per year.

The PDFs contain this data nugget as well, “At the current rate of decline, one-third of today’s Mass attending Catholics will no longer be attending within the next decade.” That means that Mass attendance will drop below 100,000 by 2035.

If Mass attendance had simply kept pace with population growth since 2011, roughly 236,000 Catholics would show up to a Detroit-area parish this weekend. The actual number is 139,000 — about 40% lower. The archdiocese is currently filling about 29% of its available pew space on a given weekend. If current trends hold, that number could slip below 20% within a few years. That’s a church with an empty building problem.

That’s a common theme when looking through this data and reading the Archbishop’s letter: there’s too much capacity in the system. The average Catholic Church is not the right size to meet the needs of the congregations. There’s no simpler way to say it. They built these houses of worship to accommodate thousands of “butts in seats” every weekend. And now they have a huge problem with maintaining these buildings.

Their own data says that the average parish has under 600 regular Mass attendees. They have 62 parishes where weekend attendance is less than 300 and another 76 parishes where 300-600 folks show up on an average weekend. In contrast, 34 parishes have a Mass attendance that is north of 1,000 people.

For the Protestants that are reading this, these numbers seem huge, right? The median attendance in the Protestant world is about 70 folks on a Sunday. Catholics have always had fewer parishes with more people in the pews. That model requires huge sanctuaries. The median capacity of a parish in Detroit was about 650 people. I am going to guess that the average Protestant church could probably accommodate 300.

But here’s the number that should be keeping archdiocesan leadership up at night: the median parish fills about one-third of its seats at a typical Mass. Two out of every three chairs are empty. These buildings were designed for a Catholic population that no longer shows up — and no amount of consolidation changes that math if attendance keeps falling.

That causes two problems. One is logistical. Maintaining a huge building with a massive sanctuary is a money pit (ask me how I know). The heating bills in the cold Detroit winters for a room that has 75 foot ceilings has to be astronomical. And let’s not even discuss the maintenance and upkeep of buildings that are 80-100 years old. Boiler replacements can easily run into the six figures.

The other problem is less tangible - it feels really bad to be in a huge space that’s only 25-35% full. Again, ask me how I know. When my church got below 30 people we actually stopped worshipping in the sanctuary and moved to a larger classroom in the educational wing. You can get a sense of that in this NBC Documentary which contains some footage of our worship service back in 2019.

These aren’t easy problems to solve. These houses of worship are just being used for less…worship.

You can see that clearly by analyzing sacramental activities. The Catholic Church is very good at recording a bunch of metrics outside of Mass attendance. This report includes five key statistics: infant baptisms, first communions, confirmations, marriages, and funerals . The workbooks included data on these activities in both 2000 and 2024.

The percentage declines are noted above each bar. None of them are good. I write often about the basic math of religious retention — a church needs to be adding members faster than it’s losing them. Baptisms and confirmations have to outpace funerals. In Detroit, every one of those trend lines is moving in the wrong direction.

Think of Catholic membership like a pipeline. Infant baptism is the top of the funnel — if you don’t capture people there, you’re looking at a smaller committed Catholic population in 20 to 30 years. In Detroit, infant baptisms are down 67% since 2000. It follows logically that first communions are down 61% and confirmations down 54%. The pipeline isn’t just leaking, it’s running dry at the source.

What’s striking, though, is that those three numbers have converged. In 2000, infant baptisms, first communions, and confirmations formed a clear funnel — each step smaller than the last, with about 5,000 fewer confirmations than baptisms. In 2024, all three sit in a tight band between 5,000 and 6,000. The funnel hasn’t disappeared; it’s just gotten much smaller. Fewer people are entering at the top, but a higher share are making it through to confirmation. The pipeline is narrower but less leaky.

But notice which figure has declined the least? It’s funerals, by a long shot. Only down 38%. In 2024, there were 6,030 funerals conducted by priests in the Archdiocese. The number of first communions? It was only 5,983. It doesn’t take a fancy demographic model to convey this simple fact: that’s a very bad trajectory.

And what may be an even more pressing problem, as highlighted by the workbooks, is the age distribution of priests in the Archdiocese.

The Archdiocese of Detroit has 224 active priests serving its parishes. Three of them are under 30. Six are over 80. Just 14% are under 40, and nearly half have already passed their 60th birthday. The age cliff is coming fast — and it’s steeper than it looks, because the Catholic Church requires priests to step back from parish ministry at 70 and assume senior status at 80. The median Detroit priest is already on the back half of his active ministry. There is no cavalry coming.

The report included this projection model which estimated the number of priests in the next decade or so.

In 2024, the Archdiocese of Detroit had 224 parishes and 224 active priests — a ratio that sounds reasonable on paper. But the distribution tells a different story. Thirty-four parishes draw over 1,000 people to Mass on a typical weekend, and 71 see at least 600. A 1:1 priest-to-parish ratio assumes every parish is roughly the same size. They aren’t.

The trajectory makes it worse. The archdiocese’s own projections show the active priest count falling to 134 by 2034 — not in a generation, but in ten years. That’s a loss of nearly 40% of current clergy. With 224 parishes and 134 priests, roughly half of Detroit’s parishes will have no resident priest within a decade.

And I haven’t even touched on some of the other problems in the report. There’s one parish in the diocese where only 18% of those who attend Mass there live with the parish boundaries. In other words - the building doesn’t match the neighborhood anymore.

I applaud Archbishop Weisenburger’s courage and initiative to face these realities head on. I say this all the time: it’s pretty easy to lead an organization when it’s growing. Hiring new people, planning the construction of new buildings, and expanding programs is a lot of fun. You know what’s not so fun? Consolidating resources, closing parishes, and ending ministries.

But it must be done. The ledger doesn’t balance in the Archdiocese of Detroit (and in many other metro areas, as well).

And like I wrote earlier, if there’s some enthusiasm for this macro-level analysis, I plan on getting deep into the weeds of the problems facing many of these parishes. I have already scraped these reports for trends in baptisms, attendance and financial health.

It’s my best chance to trace religious decline in a meticulous way. I look forward to the challenge. And my prayers are with the over 900,000 Catholics who make up the Archdiocese of Detroit.

Code for this post can be found here.

Ryan P. Burge is a professor of practice at the Danforth Center on Religion and Politics at Washington University.

First, just a bit of background on the Seventh day Adventist movement. The SDA church traces its roots back to the mid-1800s, arising from the Millerite movement, named after William Miller, a prophet who believed that Jesus Christ would return to earth in short order. From the aftermath of that failed prophecy, a number of groups spun off, including what would become the Seventh-day Adventist Church. The SDA denomination has a few distinctives: worship on Saturday rather than Sunday, a strong emphasis on health and diet (a significant share of Adventists are vegetarian), and a robust global missionary infrastructure that has made them one of the fastest-growing denominations worldwide. They also operate one of the largest hospital systems in the United States - AdventHealth.

Let’s just start this whole thing by giving you a sense of the size of Seventh-day Adventists.

The data here is remarkably consistent - somewhere between .3% and .5% of all adults in the Cooperative Election Study identify as SDAs. I think there’s some evidence here that the trend line is moving in the downward direction, but I wouldn’t stake my career on that claim. When you are dealing with a group that is this small, it doesn’t take much to send a trend line downward. In fact, it probably only takes 20-30 fewer respondents identifying as Seventh-day Adventists (out of a sample of 60,000) to move the overall trends.

Of course, it’s always nice to have a type of “reference case” to compare the SDAs. They are about the same share of Americans as Hindus. They are likely half the size of Latter-day Saints or Muslims. And for every Seventh-day Adventist there are three Jews in the United States. So, pretty small in the grand sweep of American religion.

There’s another way to get at the size estimate, though: the Religion Census. It just goes to individual religious denominations and asks for a copy of their membership records. That provides another way at figuring out just how large they are compared to other well known religious groups.

Obviously the Catholic church dominates. I’ve written about this plenty. There are nearly 62 million Catholics in the United States. The non-denominationals are listed at 21.1M, but I’m almost certain that’s an undercount. About 14% of all Americans identify as non-denominational. That works out to 35-40M folks. Then we have the Southern Baptists and United Methodists below that.

But if you take your eye down to the bottom half of the graph you can see a whole bunch of denominations that are tightly clustered between 1M and 1.8M. You’ve got a lot of the most widely discussed groups here: Missouri Synod Lutherans, Episcopalians, PCUSA and American Baptists are all pretty similar in size. This is exactly where the SDAs land: 1.3M members back in 2020. They are essentially the same size as most mainline denominations. Again, not tiny, but also not that consequential in our larger understanding of Protestantism.

But where are Seventh-day Adventists in terms of geography? Let me show you one way to visualize this and then another way. I think both are helpful to get a more accurate picture of this religious group.

Read more

I feel like I got into a debate or a disagreement with someone about politics or religion on a near daily basis when I was between the ages of 13 and 25. I would fight with anyone about anything and I would often argue positions that I completely disagreed with because I just liked the intellectual gamesmanship of the whole exercise.

In short, I was a pretentious jerk. I’m very glad that I outgrew that phase of my life.

But if I look back on the last decade of my life, I think that the pendulum has clearly swung in the other direction and in a really severe way: I hardly ever get into a disagreement with anyone about anything. (Insert hacky joke with the exception being my wife and sons). I really don’t get into any type of verbal conflict anymore. And I can’t say that I am pining for my teenage years when I seemed to be geared up to be a disagreeable snot-nosed kid on an hourly basis.

There’s a practical reason for that, by the way. I generally don’t hang out with people that I strongly disagree with on important matters. And even if I do manage to find myself in that type of company, both parties are smart enough to know not to talk about topics that will raise our blood pressure. We are middle aged, after all.

I also teach on a college campus. One of the most prestigious in the United States, empirically speaking. And it’s got students from all over the country (and a bunch of international students, too). They come from about every background you can imagine and they have very different political and religious experiences.

But they are incredibly reluctant to actually disagree with me (or other students) in class. It’s super frustrating because my understanding of the classroom is that it’s exactly the space where free speech should flourish. It doesn’t. Everyone seems so afraid of saying something that will be recorded and reposted on social media. Thus, we sort of cosplay a rich debate. We don’t actually do it.

The Foundation for Individual Rights and Expression (FIRE) has been doing an annual survey of college students where they ask questions about which topics are particularly hard for them to discuss in various environments on campus.

They were given a list of twenty items and asked to indicate which ones were particularly difficult to talk about on their campus. Here’s the ordering of those twenty topics.

The “third rail” of discourse on college campuses is the conflict in the Middle East. Over half of all those who took part in the survey said it was hard to have an open and honest conversation about the war between Israel and Hamas. You have to remember that this survey was conducted about a year ago, when this was still a very ongoing war between those two entities. I don’t know if (and how much) that percentage would decline if FIRE includes this question again on this year’s questionnaire.

What else scores near the top? I think I would call them “the usual suspects”: abortion, the 2024 election, and transgender rights. They were all in the 41%-46% range. I don’t think that anyone would be particularly surprised to see those issues float to the top of the list. Adults don’t want to talk about them either.

I also need to mention the issues that don’t raise so much hesitation that appear at the bottom of the list. Students don’t seem to be wary of discussing: climate change, crime, the Supreme Court, or China. It’s interesting how there aren’t really any “culture war” issues at the bottom of this graph.

That got me thinking, though: I wonder if Christian college students have a different idea about what topics are unwise to go near compared to Muslims, or atheists. Below, I just visualized what items were in the top three for about a dozen religious groups.

This graph is really boring, honestly. I would have hoped that for one of these religious groups there would be a true “wild card” issue that they find pretty difficult to talk about. Instead, it pretty much follows what I would have expected. For all but two groups (generic Christians and LDS), the Middle East conflict scored at the very top. What’s also notable about those two aforementioned groups - the conflict doesn’t even appear in the top three for either.

But it’s basically the same issues showing up over and over again. Abortion is on every single graph. The 2024 election shows up in nine of the twelve. The transgender issue is there for five of these religious groups. But it doesn’t appear on the list for any of the non-religious college students in the sample. Again, I don’t know what to make of that one.

I really did want to dig into the most “unique” issue for each group. To do that, I just compared the overall scores for all twenty items to the scores for each religious group and then plucked out the one topic that deviated the most from the overall percentage. That’s what is in the table below.

For both Catholics and Protestants, abortion is the one issue that they are more reluctant to talk about than the average college student. They are the only two groups where abortion is the outlier. The other huge thing that jumps out to me is the Israeli/Palestinian conflict that appears a bunch of times in the middle of the table. For Jews, Muslims, atheists, and agnostics it’s the big outlier in their responses.

Notice the largest delta between the two figures, though? It’s Jewish students and the Middle East conflict and it’s in the expected direction. In the full sample, 53% of students found it difficult to discuss. For Jews, 74% found it uncomfortable. Agnostics were also in double digits on this one, too.

However, for a bunch of groups, the deviations from the norm are just non-existent. For instance, the biggest outlier for Latter-day Saints was climate change. But even then, they were less than three points away from the baseline. And for nothing in particulars and Hindus, they were about as average as you can get.

To this point, though, we haven’t discussed the fact that the political partisanship of respondents can have a big impact on whether they are willing to discuss an issue on campus or not. So that’s what I tried to visualize below: I rank ordered all 20 topics across the political spectrum. I tried to highlight the ones that were the most important or had the greatest amount of movement from left to right.

First note that abortion and transgender rights almost always rank at the top, no matter the personal politics of the respondent. But I do need to point out that the lines go up from left to right. That means that Republicans are more reluctant to talk about both compared to Democrats. In contrast, the Middle East conflict goes in the opposite direction: it’s the most difficult topic to discuss for those on the political left but it ranks sixth among strong Republicans.

I was also struck by the variation in a couple of other lines that I highlighted. For instance, look at “gay rights.” Among the strong Democrats in the sample, it ranked 11th out of 20 issues. For strong Republicans, it was 3rd. I think you can probably figure out why a strong Republican is less likely to bring it up. Then, sexual assault goes the other direction. It’s 7th among strong Democrats but 13th among strong Republicans.

You can probably assume the next logical step, right? Let’s try to parse out how political partisanship and religious affiliation impact reluctance to speak out on certain topics on college campuses. I wanted to begin by looking at the three largest Christian groups and how they felt about three topics that tend to be central for them: abortion, gay rights and transgender rights.

There’s a consistent finding across this analysis: Republican college students are more likely to hold their tongue compared to independents and Democrats. Among Protestant respondents, the gap was noticeably large. On abortion, 62% of Republican Protestants said it was difficult to talk about compared to just 49% of Democrat Protestants. The divide was 12 points on gay rights and 15 points on transgender rights. It doesn’t take a lot of wild guessing to figure out the reason: those Republican college students have an opinion that is outside of widely held views on college campuses.

There is a gap between left and right among Catholics, but it’s noticeably smaller. It’s five points on abortion, eight points on gay rights and seven points on transgender rights. I don’t know if I can easily explain why the divergence between Republican and Democratic Catholics is not so severe compared to Protestants, but it’s a very consistent finding in this data.

Let’s do the same kind of thing but for a topic that stood out on the left: the Middle East conflict.

Here the partisan gap is reversed from the previous graph. Instead of Republicans being more hesitant to speak out, it’s the Democratic students who are more prone to keeping their mouth shut. Think about this: 80% of Democratic Jews said that they found the conflict between Israel and Palestine to be hard to discuss on their college campus. That was 27 points higher than the full sample. I think I may have found the hottest of hot button issues for a subgroup of American college students.

But even among Muslims and atheists and agnostics, it’s pretty much the “third rail” of discourse at universities across the country. At least two-thirds of the Democrats say that they just don’t want to talk about it. I do find the partisan gap among atheists interesting, though: only 43% of atheist Republicans don’t want to talk about the Middle East Conflict.

Are colleges and universities bastions of free speech? In my experience, not really. They are certainly more open to the idea of honest discussion than the average workplace or sideline of a soccer field. But everyone is terrified of getting cancelled. Myself included.

Maybe there needs to be a concerted effort to introduce Chatham House rules in some form or fashion on college campuses. But I don’t think that’s even feasible.

Instead, the cosplaying of ideas will continue full speed at colleges across the United States for the foreseeable future.

Code for this post can be found here.

Ryan P. Burge is a professor of practice at the Danforth Center on Religion and Politics at Washington University.

This post has been unlocked through a generous grant from the Lilly Endowment for the Association of Religion Data Archives (ARDA). The graphs you see here use data that is publicly available for download and analysis through link(s) provided in the text below.

There’s this concept in social science that I’ve brought up a few times in this newsletter that is worth revisiting: boundary maintenance. It’s the often subtle, subconscious process where groups make a decision about who can be included in a group and who should be excluded. Put simply, it’s the ongoing movement of defining and redefining who is “us” and who is “them.” The concept can be traced back to Émile Durkheim, one of the founding figures in modern social science, when he described how societies reinforce norms by how they respond to deviance.

Christian Smith can be credited with adapting it to the religious context in what he termed subcultural identity theory. This is most evidenced in his book American Evangelicalism: Embattled and Thriving. He argues that religious groups are most successful when they actively engage with the larger culture while also maintaining clear boundaries about us and them. In evangelicalism, boundaries are typically not defined by a hierarchical leadership organization; they are often informal and seemingly haphazard. For an example of this, I would recommend reading about the “Farewell, Rob Bell” incident.

Catholicism is an interesting case study in subcultural identity theory because of how the Church is structured. Catholics believe that the Magisterium — the Pope, Cardinals, and Bishops — holds the God-given authority to interpret scripture. Beyond that, they also believe Church teaching can develop over time through Divine revelation, a process guided by that same hierarchical leadership. This would suggest that boundary maintenance in Catholicism should flow from the top down, defined and enforced by church leadership rather than the laity. But as I’ve widely discussed in this newsletter and in various outlets: the average American Catholic disagrees with the teachings of the Church on all kinds of issues like sex outside marriage, the use of birth control, and the legality of abortion access.

I’m always interested in exploring how the average Catholic thinks about concepts like authority and how open they believe Catholicism should be when it comes to welcoming people who don’t agree with the Church’s teachings on key issues. To accomplish that, the Association of Religion Data Archives has been hosting the Catholic Laity Poll, which is a semi-regular questionnaire conducted by Gallup of American Catholics. It was first collected in 1987, and the most recent survey was fielded in 2017. It contains a wealth of information.

What may be the most relevant piece of information to inform the rest of this article is that the average Catholic doesn’t go to Mass nearly as frequently as the average Catholic from thirty years ago. Not even close. The question asks, “Aside from weddings and funerals, about how often do you attend Mass?”

In the 1987 sample, the most popular response option was weekly attendance—44% chose this response on the survey. The rest of the sample was almost evenly divided between monthly Mass attendance (30%) and those who attended seldom or never (26%). By the time that survey was fielded again in 1993, attendance had already started to decline just a bit, and in 2005, the rare attendees had surpassed the weekly attendees for the first time.

From there, the never/seldom share of the sample continued to explode. It rose from 36% in 2005 to 47% in 2011 and landed at 53% of the sample from 2017. That’s a doubling of low attending Catholics in just thirty years. At the same time, the weekly attenders slowly eroded over time, dropping below 40% in 1999 and then dipping below 30% in 2017. There’s no doubt that Catholic attendance is down substantially in the last few decades.

But what’s interesting is that Mass attendance is just one way to gauge the attachment that the average Catholic feels toward their Church. There’s another question in the survey that offers additional insight: “On a scale from one to seven, with ‘1’ being ‘I would never leave the Catholic Church’ and ‘7’ being ‘Yes, I might leave the Catholic Church,’ where would you place yourself on this scale?” I simplified this into three categories: “I would never leave” (1), “middle” (2-4), and “I might leave” (5-7).

The most popular answer, both in 1987 and 2017, was that respondents had no intention of leaving the Catholic Church. Although this share has declined just a bit over time from 56% to 44%, it has stayed around that ~45% level since 2005. It’s still certainly the case that the average Catholic doesn’t have one foot out the door.

Instead, the percentage of Catholic laity who indicated that they might leave has never risen above 25% and has actually declined noticeably in the last few surveys. In 2017, just 18% of Catholic laity said that their response to this question was a 5, 6, or 7 on the scale. So, while Mass attendance has declined sharply, that doesn’t indicate that Catholics are going to leave the Church behind in terms of affiliation.

But how much do these rank-and-file Catholics feel like they need to listen to the moral teachings of their Church on contentious issues? The preamble to each of these questions is:

“Next, we are interested in your opinion on several issues that involve the moral authority in the Catholic Church. Is it Church leaders such as the pope and bishops, individuals taking Church teachings into account and deciding for themselves, or both individuals and leaders working together?”

They were given three response options: church leaders, individuals, or both. And they touched on topics like birth control, abortion, and sex outside of marriage. I plotted the share of Catholics who said that the individual, not the Church, has the final say on these issues.

One commonality across these findings is that the average Catholic has a more individualized view of morality now than a similar Catholic in the late 1980s. For instance, in 1987, just 32% of Catholics believed that a person’s individual morality should be the final arbiter on the issue of “A divorced Catholic re-marrying without getting an annulment.” In 2017, that had jumped up to 47% of Catholics. There was fifteen points of movement on the topic of homosexual behavior and 14 points on the topic of sex outside of marriage.

Thus, the evidence is strong here that the average lay Catholic is questioning the moral leadership of the Church on these issues at a much higher rate in 2017 compared to 1987. There is a confounding variable here that we just discussed: Mass attendance. If the average Catholic is going to church less, it makes logical sense that they would be less open to deferring to the Church’s teachings on these hot button issues. To check that I just broke the sample down into weekly Mass attenders and those who attended seldom or never.

It’s always been the case that weekly attenders were more open to the authority of the Magisterium than the laity. You can see a huge gap on every single issue in the survey, and those differences are numerically massive—often 25 percentage points between high and low attenders. On three of the five issues, there’s compelling evidence that individual morality has risen sharply among Catholics, regardless of how often they attend Mass. Those topics are: sex outside of marriage, homosexual activity, and the issue of divorce and remarriage in the Church.

Among weekly Mass-attending Catholics in 1987, 32% believed that individuals should have the final say on homosexuality. In 2017, that share was up to 44%. On sex outside of marriage, there was a ten-point movement in the same direction (33% to 43%). In fact, the movement for both attendance groups was essentially the same magnitude.

There are two outliers that are worth some reflection, however. On abortion, weekly Mass attenders didn’t budge over a thirty-year window. In 1987, 36% said it was an issue of personal morality; it was 35% in 2017. And on birth control, there was movement away from individual morality as well: 54% to 52%. It would appear that those issues are now “distinctives” among devout Catholics while issues like sex outside of marriage or remarriage are not seen the same way.

Speaking of Catholic distinctives, there was also an intriguing battery of questions that had this setup, “The following statements deal with what you think it takes to be a good Catholic. Please indicate if you think a person can be a good Catholic without performing these actions or affirming these beliefs.” And then there were options like: not going to Mass every Sunday or opposing the Church’s teachings on contraception.

The broad conclusion that I arrive at in looking at the answers to these questions is that the average Catholic has a lot of grace for their brethren. For instance, 80% of Catholics believe that someone could be a good Catholic without going to Sunday Mass. An even larger share (83%) says that good Catholics can ignore the Church’s teachings on birth control. There’s also not a lot of appetite for boundary maintenance around the issue of divorce and remarriage. At least 7 in 10 Catholics seem unbothered about fellow Catholics bucking the Magisterium here.

There are just two issues where Catholics draw the line, and they are both theological matters. For instance, just 45% of Catholics believe that you can be a good Catholic and reject the teachings of the Church regarding what happens when the priest blesses the elements. The Church believes in transubstantiation—the idea that the bread and wine literally turn into the body and blood of Jesus Christ. The other doctrine that is a bridge too far is a belief in the resurrection of Jesus. Two-thirds of Catholics say that one cannot be a good Catholic and reject the Easter story.

But what happens if we break it down into those two attendance categories, though? (Also keep in mind that not every one of the questions above was asked in 1987 and 2017, so some were excluded).

It’s certainly the case that weekly Mass attenders engage in stronger boundary maintenance than those who don’t attend so frequently. The gaps on many of these questions are 30-40 percentage points. When it comes to Mass attendance, half of the most devout Catholics believe that an individual can not go to church on Sunday and still be considered a good Catholic. That was true in 1987 and 2017. But on other issues, weekly Mass attenders seem to favor a larger Church, even if it’s less obedient to the Catholic hierarchy.

For instance, among weekly Mass goers in 1987, only 27% of them said that a good Catholic can reject the Church’s teaching on abortion. In 2017, that number had risen to 40%. There was a thirteen point jump in the share of who said that good Catholics can get remarried without seeking an annulment, too. There was a nine point increase around the topic of birth control, as well.

I just don’t get the impression that the most devout Catholics have gotten more “hardcore” about boundary maintenance in the last thirty years. In fact, the data points in the exact opposite direction. The boundaries have actually become more blurry and circumstantial. That finding stands in sharp contrast to the picture that emerges from the loudest voices in religious life today.

If one spends any time in religious corners of social media, the impression that emerges is that devoutly religious people have become strident and often belligerent over time. They often fight about what would appear to be obscure or esoteric parts of theology or worship practice. It’s easy to believe that these purists are indicative of the larger religious movement.

This data strongly rejects that notion. Among Catholics, even the ones who go to Mass every week, they seem more willing to reject the Church’s teachings on a number of controversial topics. And they are not willing to cut people off from the Catholic Church even if individuals hold views that fall outside of mainstream Catholic thought.

In other words - Twitter/Facebook/TikTok/Instagram are not real life.

Code for this post can be found here.

Ryan P. Burge is a professor of practice at the Danforth Center on Religion and Politics at Washington University.

Being around academics for my entire working career has taught me that most of them grew up in highly populated areas and want to live in close proximity to large population centers. I never really wanted to do any of that. I’m happy to live in a small town where housing is cheap, crime is low, and traffic is non-existent. But I also realize that my environment shapes my work. I understand American religion through the lens of a guy who has lived in the rural Midwest for four decades.

Many of the reporters I speak to have grown up in the Beltway bubble for their entire lives, so their understanding of religion is based on their experience living in a highly cosmopolitan, rapidly growing part of the country. Neither perspective is completely accurate, of course. That’s why data is great, though. It helps us move beyond our own bubbles and see the full sweep of American religion.

What I wanted to focus this time is that issue I just spoke of: geography. I wanted to explore whether religiosity has faded more quickly in densely populated areas of the United States or if rural America has secularized faster than suburban areas.

One methodological note before diving in. To measure geography, I’m using a variable in the GSS called srcbelt, which classifies respondents based on where they live relative to metropolitan areas. It has six categories that I’ve collapsed into four for this analysis:

Large metro — respondents living in or around the 12 largest metropolitan areas in the country, things like New York, Chicago, and Los Angeles.

Smaller metro — respondents in other metropolitan statistical areas that didn’t make the top 12. Think mid-sized cities and their suburbs.

Small town — respondents living in areas adjacent to a metro area but not actually part of one. Small cities and towns on the urban fringe.

Rural — respondents living in areas with no meaningful connection to a metropolitan area. The most genuinely rural category in the data.

One caveat worth considering: This is a geographic classification based on where someone lives, not where they grew up or how they identify. Plenty of people in “Large metro” consider themselves small-town at heart, and vice versa.

So, let’s track attendance over time based on these four geographies.

Those red lines really jump off the screen, don’t they? That’s the share of each sample that reports that they never attend religious services. Back in the 1970s, those kinds of folks were basically non-existent. I mean, look at the share of people living in smaller cities and rural areas that were never attenders: it was about 10%. It was about five points higher among folks living in big cities, though. This didn’t last that long. Around 1990, the lines for the three largest geographies began to rise in a noticeable way. It took rural areas another decade to see a discernible increase in never attenders. But from that point, it was just straight upwards for each line. Now, never attenders are 30-35% of each graph.

The other thing that I wanted to highlight was the trend lines for “monthly” and “weekly” attendance. In the top two graphs, those lines lay right on top of each other. That’s not the case for people living in small towns and rural America. In those two places, weekly attendance was always noticeably higher than monthly attendance. Also, I must point out that the only geography where weekly attenders outnumber never attenders are in rural America. Although that may come to an end in the next 5-10 years.

Of course, attendance is just one part of the equation – I was even more intrigued by the results about religious tradition.

While the “never attenders” line jumped off the screen in the previous graph - the non-religious also exhibit a very similar pattern in this visualization. They were pretty low in every geography through the early 1990s, but then they just took off. In every graph but the “rural” one, the nones are now the largest share. Secularization is absolutely happening in densely populated places, but it’s also common in smaller towns, too. Even in rural America, the nones make up a quarter of the population.

The other huge shift is the largest Christian group when comparing the two graphs on the top to the two on the bottom. In large and small metros, Catholics really dominate. That’s especially true in the “smaller metro” category where they went from nearly 40% of the sample in the 1970s to 28% today. Less populated places just don’t have that many Catholics, in comparison.

Instead, the small town and rural graphs show a similar pattern. The mainline was the largest group back in the early 1970s, but that didn’t last that long. Within the next 8-10 years, evangelicalism experienced clear gains and a lot of that came at the expense of mainline Protestants. By the mid-1980s, evangelicals outnumbered the mainline and that divide only began to widen. I mean, the mainline was nearly 40% of each geography back in 1972 and then it was only 15% in the 2022 data.

To make that easier to digest, I just made a bar graph that visualizes the distribution of religious traditions in 1972 and 2022 for each of the four geographies.

The trend lines help to give us a broad sense of how things are shifting but the bar graphs really sharpen our understanding of the size of these changes. For instance, you can see that evangelicals made modest gains in urban areas (up 4 points), but those were much larger in rural America - 9 points. It’s also readily apparent how much evangelicalism is a rural phenomenon - making up over 40% of the population. That’s seventeen points higher than any other geography.

The right side of the graph drives home the “rise of the nones” in pretty stark terms. They were in the single digits even in urban areas back in 1972. But now they are 28% of major cities. What’s even more notable is that cities aren’t as secular as those geographies in the middle of range. Among people living in smaller metros, almost a third are non-religious. This is suburban America, really.

One other thing that I have to point out is just how much the Black Church has collapsed in major cities. Back in 1972, Black Protestants were a larger share of urban areas than mainline and evangelical Protestants combined and they were basically the same size as Catholics. Now, just 5% of urban areas are Black Protestants, while those other Christian groups have held steady. There’s a doctoral dissertation to be written about this massive shift.

Another angle that is worth exploration is looking at how younger residents of these geographies have different religious compositions compared to Baby Boomers.

The big thing that I take away from this is that the nones are rising among younger generations in basically the same way whether we are talking about big cities or small towns in the United States. But, what really jumps out to me is just how non-religious young adults are in rural America. Among Millennials living in small towns, 40% are non-religious and 48% of Gen Z claim no affiliation. Those rates are just as high (if not higher) than younger adults living in urban areas. The rise of the young, rural none is not something that I’ve ever seen written about.

At the same time, you can clearly see how evangelicalism has lost ground with each successive generation living in rural areas. Among Boomers in these geographies, 30% are evangelical and 28% of Gen X. For Millennials it drops down to 22% and for Gen Z it’s even lower: 12%. Remember that evangelicals have dominated these kinds of places for decades now, but according to this data that may not be true going forward. If so, that could have a tremendous impact on the culture and politics of rural America.

Speaking of politics, I had to trace the partisanship of the three largest Christian groups over each of these geographies.

The top row of graphs are white evangelicals over the last fifty years. I think we can pretty easily see when the Republicans began to take over. The crossing of the two lines happened at nearly the same time for evangelicals living in big cities and small towns across the United States. However, the last ten years of data is radically different for evangelicals. Among those living in urban areas, the GOP has stopped making inroads and may have even lost some ground recently. In less densely populated spaces, evangelicals just continue to become more Republican.

For mainline Protestants, the differences between geographies is fascinating because the trend lines don’t really follow a discernible pattern. It’s generally the case that mainliners are more aligned with the Republicans compared to the Democrats. But for mainliners living in smaller metros, the lines have started to converge in recent years, when they have begun to move further apart in small towns.

What about white Catholics? Well, among those living in more densely populated areas, the Republicans haven’t quite overtaken the Democrats. But that is likely to happen soon. For those living in small towns, the GOP became more popular than the Democrats somewhere around 2016. In rural America, white Catholics began to favor the Republican as far back as 2002.

Here’s what I’ve learned from doing this fun little bit of analysis:

1. Secularism has risen basically everywhere in America. It’s taken hold in both major urban areas but also rural parts of the country, too.

2. In urban areas, the Black Church has gone from dominant to nearly non-existent since the 1970s. In other parts of the country, that same trend happened for mainline Protestant Christianity.

3. Can rural white evangelicals get any further away from the Democrats? In 1972, 45% of them said that they were Democrats. Today, it’s closer to 10%.

One thing to keep in mind in this analysis is that rural America is obviously not that many people. In fact, they made up just 10% of the full sample in 2022. So while I’ve talked about that geography a whole bunch in this piece, they only represent a very small sliver of the country. And there’s no reason to think that they will increase in importance in the years to come.

Code for this post can be found here.

Ryan P. Burge is a professor of practice at the Danforth Center on Religion and Politics at Washington University.

I’ve written a bit about mega-churches before based on a list provided from the Hartford Institute and there’s a lot to learn there. But, I have another data source now that gives me even more insight into the largest churches in the United States.

For the last decade, Outreach Magazine (in collaboration with Lifeway Research) has collected a list of the 100 largest churches in the United States. I am in no way responsible for the data collection for this project and I’m sure that the research team would tell you that this list is probably not 100% accurate. But guess what? It would be next to impossible to compile a list like this without leaving out a church or two. This is as good as we are going to get. If you are interested in how this list was pulled together, there’s a bit about their methodology to be found at this link.

This data has been collected with regularity all the way back to 2015. I wrote a scraping program to acquire the information from the Outreach website and now I’ve got an absolute treasure trove of information about these churches. In total there are 1,100 entries on my list (100 churches x 11 years of data), but here’s the first big revelation from this effort: there are only 209 unique churches to be found. Which means that a whole lot of these congregations are “frequent flyers” in that they appear every time. In fact, 20 churches show up every single year and another 18 show up in ten of the eleven years. Big churches stay big.

However, there is some real variation at the top of the list. I just visualized the largest 10 churches from the 2015 list and then the 2025 list.

One huge thing that jumps out to me is that the biggest churches in the United States have gotten a whole lot bigger in the last decade. In 2015, the ten largest churches combined for a total membership of just over 260,000. In 2025, that attendance number rose to over 427,000. The fifth largest church in the 2025 data (North Point), had 37,000 attenders. That number would have put it at the top of the list in 2015. Speaking of North Point - they are one of the churches that shows up in the top ten in both years. There are several on the list like this: Church of the Highlands (Birmingham), CCV (Peoria, AZ), Crossroads Church (Cincinnati), Christ Fellowship Church (Palm Beach Gardens), Southeast Christian (Louisville), and Saddleback (Lake Forest, CA).

There are some newcomers to the list in 2025, though. One is Life.Church. It’s based in Edmond, OK and was just added to the dataset in 2022. (Don’t ask me why it wasn’t there before that date, I don’t know). But you’ve also had the addition of Lakewood Church, which is run by Joel Osteen. Again, that one has been very large for a while but only recently was added to the Outreach list. The other newcomer was Eagle Brook Church in Centerville, MN.

But I know what you all really want to do - you want an interactive map where you can pan and zoom and look at where these churches are located. That’s exactly what I’ve done for you below. Each church is marked with a pin and you can click on that pin to give you the name, location, and the last reported attendance from each church. I’ve also colored in the circles with the last year they appeared on the list and the size of the circles denotes the last reported attendance.

Read more

And, for reasons that still escape me to this day, my third field became comparative politics. I have no earthly idea why I chose it because at that time the only other country I had visited in the world was Canada. But, I think I took about nine hours in comparative politics. The most memorable course was taught by the brilliant Fred Solt and was titled, “Comparative Political Behavior.” It was basically a class about how different societies engage with the government of their countries. And after poking around in my Dropbox, I found the syllabus. I still remember reading about elections in the Soviet Union that had a number of spoiled ballots, because every citizen had to vote and there was only one choice on the ballot. It was their form of protest.

Well, for the final project, I knew I wanted to write about religion and I knew I wanted to write an empirical paper. My only option was to use the World Values Survey. But here’s what I distinctly remember about the project – my little Dell laptop could not load the entire data file without crashing. There just wasn’t enough RAM to do the job. So I had to ask Dr. Solt to send me a curated dataset that only contained a handful of countries for analysis.

I don’t have that problem now, of course. I can load the full file. The most recent version of the survey is Wave 7, and it was fielded between 2017 and 2022. It contains 64 total countries, but some don’t have responses to religion questions, so we are left with 54 countries. Which is still a lot.

What follows is me just giving you all a whole bunch of “here’s some cool stuff I found in the World Values Survey when it comes to religion.”

One simple question in the WVS was: Do you believe in God? As you can clearly see above, there are a lot of countries in the dataset where almost everyone believes in God. In fact, there were a dozen countries where belief was at least 99%. There were 31 where belief was at least 90%, as well. The countries represented here are concentrated in several regions: Sub-Saharan Africa, Southeast Asia, and the Middle East/N. Africa shows up a whole lot on the top of the list.

What about the least “believing” countries? There are two that stand out on the bottom of the graph. Just 26% of people living in Czechia said that they believed in God and it was even lower in China, a paltry 17%. The regions that tend to appear down at the bottom are Europe and East Asia, with countries/territories like Macao, South Korea, Hong Kong, and Japan, all of which have pretty low belief in God.

There was another set of belief items on the survey that I always enjoy exploring that were focused on the existence of heaven and hell. They were asked independently, so we can actually parse out people who believe in both, neither or just one of the two. I made heatmaps of all 53 countries where respondents answered both queries.

In a lot of the countries that appeared at the top of the belief graph, almost everyone believes in both heaven and hell. It’s 100% of people living in Egypt, 99% of those living in Bangladesh, and 98% of Tunisians. But there are some interesting outliers when comparing belief in God to belief in the afterlife.

One that jumped out was Bolivia. In that country, 98% of people believe in God but only 60% of Bolivians believe in both heaven and hell. Nicaragua is even weirder: 92% of people believe in God but 64% of folks believe in neither heaven nor hell. I’m sure you guys can find other interesting examples like that, please point them out in the comments.

Of course, China scores the lowest on this graph – just 10% of Chinese people believe in both heaven and hell. Czechia is at 11%, but then Germany is only a few points higher than Czechia at 15% belief in both heaven and hell, even though over 60% of Germans believe in God.

You don’t see a whole lot of folks believing in heaven or hell but not both, but there are some countries where this figure is pretty high. For instance, 30% of respondents from the Philippines are in the top left square, and it’s the same share of Colombians. Puerto Rico is right behind at 25%.

There was also a question of how important God was to each respondent and they were given a scale that ran from 1 to 10. I futzed around with this question and it was frustrating because in a huge number of countries the vast majority of people who took the WVS chose 10/10. I was more interested in countries where there were large populations of people who chose 1 or 10. In other words, the countries below had the highest level of religious polarization.

The country with the highest ratio on this graph? It’s China. For every one Chinese respondent who said that their “God importance” score was 10, there were 15 who scored it 1 out of 10. The ratio in Czechia was about 5 to 1. So, there were five non-religious folks for every one deeply religious person. But in lots of other large countries the ratio was pretty modest: 3 to 1 in Japan and the United Kingdom, for instance.

Now there were places where the ratio went the other direction, almost all of them are at the top of the graph. For instance, in the country of Brazil for every person who scored themselves a 1, there were 34 who chose ten out of ten on the “God importance” question. It was 33 to 1 in Armenia and 31 to 1 in both Guatemala and Puerto Rico.

In other words, a deeply non-religious person is far more likely to be outnumbered in their country than a deeply religious one. That holds for almost every part of the world. Even in a largely secular country like Germany, 16% scored themselves as 10/10 and 26% were 1/10. So, it’s certainly possible to find a religiously like-minded individual in a place like that.

Moving on from belief, my curiosity was especially piqued by a statement included in the WVS: “The only acceptable religion is my religion.” I love this one because it’s basically asking people to make an exclusivist claim about their own faith group.

You can see that there are a lot of countries where pluralism is not super robust. In places like Bangladesh, Myanmar, Jordan and Pakistan, the vast majority (90%+) of respondents agreed that the only religion was their own religion. It should come as no surprise that the countries that appear at the top of this graph tend to be the most religiously homogeneous, by the way. It’s pretty easy to make the claim that your religion is the only acceptable one when it’s the only game in town.

What about the countries at the bottom? I think it’s fair to say that it’s a whole lot of first-world countries. Places like New Zealand, the United Kingdom, Germany, and Australia show up in this part of the graph. These are countries where religious pluralism is pervasive. Someone claiming that their particular faith is the only acceptable one is just incredibly rare, often being held by 10% of the population.

I am fascinated, though, by the countries where answers to this question are evenly divided. It’s an eclectic mix: Nicaragua, Lebanon, Ethiopia, Kazakhstan, Slovakia, etc. I certainly can’t make some kind of sweeping claim about divided countries being in one region of the world.

But that got me thinking about what factors led to more exclusivist thinking. To do that I made two simple calculations. The first was the mean score on the “God importance” question from earlier for each country. The second was the share of each country that agreed that their religion was the only acceptable one. I wondered if countries with higher levels of religiosity were also more likely to be exclusivist in their approach to other faith groups.

There is a strong positive relationship between national religiosity and religious exclusivism (r = 0.69), meaning that countries where God is seen as more central to daily life also tend to have higher shares of residents who view their own religion as the only acceptable one. However, the R2 value is just .48, which means that the God importance variable only explains about half the variance.

Notice which countries are below the line consistently – it’s places in Latin America. For a country like Brazil, the importance of God is incredibly high (over 9/10). However, at the same time, just 15% of Brazilians agreed that their religion was the only acceptable one. These are people who are deeply devout but have an incredible tolerance for other faith groups. Other countries in the region are in this quadrant of the graph, too: Peru, Colombia, and Guatemala are high on religiosity and high on pluralism.

But then look at the countries where the opposite is true: overall religiosity is low but exclusivism is relatively high. For instance, in Thailand, average score on God importance is 5/10 but almost half the country takes an exclusivist view. Vietnam is in a similar spot on the graph. Tunisia and Turkey also jump out to me. They are fairly religious but not at the top of the scale, but also very exclusivistic.

One final scatterplot before I close up today. This one is just a simple replication that is widespread in my corner of the world: specifically, the secularization hypothesis. This is the idea that as a country becomes educationally advanced and economically prosperous, it will be less religious. I created a composite index of religiosity based on a number of questions about religious belief, belonging, and behavior. Then I grabbed a measure of GDP per capita to assess economic prosperity.

As you can probably guess, the poorest countries tend to be the most religious and the most economically prosperous ones score the lowest on my composite index of religiosity. The correlation coefficient is r = -.70, which is strong but not completely determinative. The R2 is right around .50, which means that only half the variance can be explained by simple GDP.

The outliers are always worth flagging. The biggest one is, without a doubt, China. It scores the absolute lowest on the religiosity index of any country, yet its GDP is solidly in the middle of the pack. If China’s GDP were to actually reflect its religiosity score based on this simple model, it would need to be nearly $400,000. That’s over four times higher than any other country in the sample. The obvious explanation is not poverty but politics: decades of state-enforced atheism under the Communist Party have suppressed religious expression in ways that wealth and development simply cannot replicate.

What about outliers above the line? One big surprise is Puerto Rico. Its GDP is $40K but their religiosity index is .83. The combination of a strong Catholic Church and the rise of Pentecostalism probably explains a lot of that. Then you’ve got a bunch of Eastern Orthodox countries who are more religious than they should be based on their GDP: Cyprus, Greece, and Romania all fit this pattern.

Where does the United States fit? Well, it’s still an outlier. The religiosity index is .62, and GDP is nearly $70K. In order to fit the trend line in this scatterplot, GDP would need to drop to just $28,000 or religiosity would need to dip to .47.

When it comes to religion, no two countries are the same — not even two countries that share a border, or two regions within the same nation. This kind of work humbles me. I could spend my entire career studying the religiosity of a single country and still not fully understand it. The idea of doing that for 200 more countries is almost incomprehensible.

I will spend my entire life studying religion and still grasp only a fraction of it. That’s not a discouraging thought. It’s the reason I keep coming back.

Code for this post can be found here.

But, wait…there’s more. Someone asked me a while ago if I would occasionally drop in book recommendations. I just finished a book that I am still thinking about. It’s called Hell is a World Without You by Jason Kirk. It’s a novel about a kid who grew up deeply enmeshed in evangelicalism in the early 2000s.

The main character is basically a slightly fictionalized version of the author, you can pretty easily tell that after reading just a couple of dozen pages. And let me tell — Jason Kirk understands the world of Millennial evangelicals better than anyone I’ve ever read. He gets all the weird nuances, culture, and oddities that consumed my teenage years.

For instance:

Anyway, Wednesday church was usually fun. Obstacle courses! Eating live goldfish! A pretend funeral? This kid Zack read his friend’s suicide note, which listed times Zack should’ve shared the gospel. After Zack cried, “Why didn’t I tell him about Jesus?” the casket burst open. The dead kid screamed, “I’m burning in Hell, thanks to you,” and then we played red rover.

That last sentence made me laugh out loud because stuff like that happened ALL THE TIME when I was in youth group. You would have the most emotionally traumatizing moment of your life, then go to the fellowship hall and eat cold Domino’s pizza.

It gets a bit too cliche in a couple spots near the end, but I can easily overlook that in the name of drama and closure. If you grew up evangelical and are between the ages of 35 and 45, this book will hit like a sledge hammer of nostalgia.

Ryan P. Burge is a professor of practice at the Danforth Center on Religion and Politics at Washington University.

This post has been unlocked through a generous grant from the Lilly Endowment for the Association of Religion Data Archives (ARDA). The graphs you see here use data that is publicly available for download and analysis through link(s) provided in the text below.

Every once in a while, I will see a survey question that makes me say, “I wish I would have thought of that.” Not because the results are shocking, but because the question itself cuts right to the heart of something Americans have been arguing about for a very long time. Pew’s American Trends Panel recently asked a simple but theoretically loaded question: how important is it for the United States to have a culture based in Christian religious beliefs? Not whether you are Christian. Not whether you attend church. It’s: whether you think the country itself should be culturally Christian.

I like the way this is phrased, by the way. Because it even allows Jews, Muslims, and some non-religious folks to nod toward the importance of Christianity in the functioning of the United States. To me, this is a question akin to one about religious importance. It’s not asking how often the respondent personally goes to church, or what they believe about Jesus and the Bible. It’s tapping something more amorphous – do you think that having a Christian culture is a net positive for the country?

What’s nice about this data (hosted on the ARDA) is that it was collected not that long ago. It was fielded in August of 2024 and the total number of respondents was 9,201. So we have some pretty fresh data here and it’s a robust sample size. So, let’s just dig in now by looking at how the full sample responded to that question and then a breakdown among the largest faith traditions.

Here’s a big headline for me: the share of Americans who say that it’s “extremely” important for the United States to have a culture based in Christian religious beliefs is not that large at all: just 16%. It does seem that this contingent is really loud, but they aren’t that numerically large. In total, one-third of Americans say it’s very/extremely important. But nearly the same share (26%) say it’s not important at all.

Forty-two percent of Americans chose the not at all/not too important option, which was nine points higher than the very/extremely folks. And then, nearly a quarter of Americans chose the “somewhat” option. I just don’t get a really strong sense here that huge swaths of the country think that having a “Christian country” is that important.

Of course, that varies significantly by religious tradition. Evangelicals are, without question, the outlier on this one. Nearly 70% of them chose the very/extremely important options and less than 10% said it was not at all or not too important. However, among other Christian groups the numbers are much more in line with the national average: 35% of Catholics chose the very/extremely option and it was the exact same share of non-evangelical Protestants. So, the average non-evangelical Christian looks exactly like the average American on this one.

It should come as no surprise that the other faith groups and the non-religious lean much more toward the left on this question. Among atheists/agnostics, 90% indicate that it’s not at all/not too important for the U.S. to have a Christian culture, and it’s 68% of those who claim no religion in particular. I do need to point out here that the latter group is somewhat warmer to a general sense of Christianity than the former groups, though. As I’ve said many times, “The nones are not all the same.”

What I really want to lean into for the rest of this post is that huge canyon that exists between evangelical Protestants and the other two types of Christians: non-evangelical Protestants and Catholics. As I just mentioned, those evangelicals hold a view on this topic that is far out of step with their Christian cousins. I wondered if age was playing some kind of role in driving the difference.

The share of older evangelicals who say that it’s extremely important that the United States have a culture based in Christian religious beliefs was 45%. That was nearly thirty percentage points higher than the oldest non-evangelical Christians. The gap is also large when comparing those 50–64 years old (25–30 points). It does narrow when looking at younger respondents, and among the youngest adults, it’s in the 10–15 point range. But the statistical gap persists when comparing these groups inside each age category.

However, notice something really important here: younger evangelicals don’t feel so strongly about this issue compared to their parents or grandparents. I mean, a young adult evangelical is about 20 percentage points less likely to say that Christian values are “extremely” important to the United States compared to a retired evangelical. There’s definitely something going on with age here that I will revisit a bit later.

We also have to assume that church attendance is going to be driving up responses for these Christian folks, too, right?

I think that the answer to this question is, “yes, but not nearly as much as I would have guessed.” In essence, there’s a threshold effect happening in the data. When you compare Christians who never attend church to those who go about once a month, there’s no statistical difference in the share who chose the “extremely” important option. That’s true for all three types of Christians we are examining here. Across the board, evangelicals are much higher on this metric compared to non-evangelicals and Catholics. The gap is basically the same magnitude.

But then weekly attendance is something else entirely. It pushes up the “extremely” important response option for two groups: evangelicals and Catholics. However, it has zero impact on the non-evangelical Protestants in the sample. For evangelicals, the increase was 13 points and for Catholics it was basically the same amount: 14 points. That’s something worth flagging, by the way. The Catholic increase on this question from the bottom of the attendance spectrum to the top is a whopping 19 percentage points. That does not appear at all among non-evangelicals.

The proverbial “elephant” in the room, of course, is political ideology. Just a casual glance at the online discourse points to the fact that political conservatives are the ones that seem to be touting the importance of Christian values in the functioning of American democracy. Liberals don’t lean nearly as hard into that messaging, even if they are professed Christians.

This couldn’t be more apparent in the data — it’s politics that is doing a whole lot of the work here. An ideologically conservative evangelical Protestant is over twice as likely to say that it’s extremely important for the United States to have a Christian culture compared to an ideologically moderate evangelical. That huge gap is also there when comparing conservative and moderate non-evangelicals (29% vs. 12%) and Catholics (25% vs. 9%).

But I do need to highlight that there’s still a huge difference inside each ideological group, though. For instance, a moderate evangelical is still 14 percentage points more likely to have chosen the “extremely” important option compared to a moderate Catholic. It’s a twelve-point gap when comparing liberals on this measure. It’s both things working at once: religious affiliation and ideology.

Of course, the only way we have to untangle all of these pieces is to throw a bunch of them into a regression model. The thing that we are trying to predict is whether someone says that the United States having a Christian culture is “extremely” important to the country. I’ve got all the usual suspects here: age, church attendance, ideology, education, race, and gender. I ran two models: the red circles are the evangelicals in the sample and the blue circles are folks who identify as non-evangelical Protestants and Catholics.

The interpretation is much easier when it’s presented visually. If the circle or the error bars overlap with zero, that variable has no statistically significant impact on the question at hand. If the point estimate is positive and doesn’t overlap with zero, it drives up the probability of someone choosing the “extremely” important option. If it’s negative, it drives down that likelihood.

I think it’s important to note how many of these variables “work” the same for both subgroups I analyzed. For instance, being a weekly attender has statistically the same impact for evangelicals and other Christians. It drives up the log-odds by about 60% of choosing the “extremely” option. The impact of having a college degree is the same magnitude, but in a different direction. Holding all other factors constant, a Christian with a bachelor’s degree is about 50% less likely to choose the “extremely important” option.

I do want to highlight the ideology variable, though. The reference category is folks who were ideologically moderate. So an ideological liberal is less likely to say that having Christian values is extremely important for American culture, and ideologically conservatives are much more likely to choose this response. In fact, of all the variables in the model, ideology was the strongest predictor of all of them. This is another instance of politics swamping religion.

There are two instances when the variables work differently for evangelicals compared to the other Christians in the sample. One is age. For evangelicals, being over 50 years old drives up the belief that Christian beliefs are extremely important to American culture. In fact, the magnitude of the effect is the same as being a weekly church attender. But what’s interesting is that age has no impact for non-evangelical Protestants and Catholics.

The one other divergence between the models is gender, which really surprised me. For evangelicals, the gender of the respondent had zero impact on the likelihood of them saying that Christian values were important to American culture. For the other Christians in the sample, women were about 40% less likely to say it was “extremely” important compared to men in the same religious tradition. I don’t know what to make of all of that, really. But it’s worth thinking about some more.

It’s a constant refrain on this newsletter, but I feel like I’ve got to at least mention it here: the direction of the causal arrow is impossible to determine. Are conservative evangelicals more likely to want a Christian culture, or are people who want a Christian culture sorting into conservatism? This kind of data can’t really get us to an answer.

Code for this post can be found here.

Ryan P. Burge is a professor of practice at the Danforth Center on Religion and Politics at Washington University.

A quick plug: The Danforth Center on Religion and Politics at WashU has released a free public lecture course called 1776, Then and Now for the 250th anniversary of the founding. Big-name scholars (Danielle Allen, Eddie Glaude Jr., and more) on the Revolution and why it still matters — including a whole section on religion, secularism, and religious liberty in the founding era (lead by some of my esteemed colleagues at the Danforth Center). Worth your time.

“I don’t particularly like Donald Trump; I just voted for him because I really didn’t like the alternative.” That’s a comment that I feel like I hear on a nearly weekly basis in various environments. It’s usually during the Q&A portion of a talk about my newest book—The Vanishing Church. I totally understand what that person is trying to articulate, by the way: American electoral politics is a “this or that” phenomenon. If your disdain for Candidate A is an 8/10 and it’s a 7/10 for Candidate B, then the second person is probably going to begrudgingly get your vote when you walk into the voting booth. It’s just the nature of our “winner-take-all” system.

I also think that our binary choice drives this much bigger issue in American politics: affective polarization. This is the idea that a growing number of Americans feel negatively toward the other party not because of their policy positions on things like healthcare or immigration, but simply because they are the other party. It’s the “Cry more, libs” approach for Republicans. A Democrat will demonize or mischaracterize the other side of the argument because this helps them justify their own voting choices. It really goes both ways.

But now that we have data from the 2025 Cooperative Election Study, we can actually dig into how vote choices are relating to approval ratings of Donald Trump.

Let me start by just showing you the most basic analysis - here’s Trump’s approval rating in October of 2025 across a whole bunch of different religious groups.

I sorted these traditions in terms of the ones that were the most approving of Trump up top and those who didn’t like him as much near the bottom. You can pretty easily ascertain which groups are the most MAGA. Among white evangelicals, 71% said that they somewhat or strongly approved of his job performance. Next in line were Latter-day Saints at 57% job approval, followed closely by white Catholics at 55%. There was only one other group where Trump was above water: Orthodox Christians with a 54% approval rating. I mean, it doesn’t take a lot of deduction to figure out the common thread among the groups I just listed: white and Christian. That’s Trump’s base. It’s always been that way.

But he’s underwater with all kinds of other religious traditions. Among mainline Protestants, disapproval was 52% and approval was 46%. His standing with non-white evangelicals is twenty points underwater. It’s even worse with non-white Catholics: 65% disapproval against 31% approval. It’s pretty apparent from this that Trump is just not doing well with non-white voters. I mean, 79% of Black Protestants disapproved of his performance.

However, he is also doing incredibly poorly among non-religious voters, too. For atheists, strong disapproval is a whopping 84%. It’s just slightly better with agnostics: 75%. The “nothing in particular” crowd is 30% approval/62% disapproval. That’s pretty terrible, to be honest.

For a baseline, the full sample was 55% disapprove and 40% approve. Things are not going well for the second Trump administration.

I wanted to see what religious groups tended to have the most regrets about a Trump vote. To do that, I plotted two things: approval from 2025 and then vote choice from 2024 for all sixteen groups. The line below is helpful because any groups to the right of that line are folks whose approval of Trump is lower than his vote share.

The group that deviates the furthest from that line? Surprisingly, it’s white evangelicals. He earned about 84% of their votes in the 2024 election (voter validated), but then his approval rating in 2025 had slipped to just 71%. That’s a fourteen-point slide in just about a year in office. There are a couple of other groups with a double-digit difference here:

• Non-white evangelicals: 48% vote share vs. 38% approval

• White Catholic: 65% vote share vs. 55% approval

• Latter-day Saint: 69% vote share vs. 57% approval

It’s pretty clear to me that Trump did lose quite a bit of ground with some groups that made up the core of his coalition.

But for the rest of the groups, the differences between vote choice in 2024 and approval in 2025 are pretty small. It’s about five points for non-white Catholics. Just 3 points for agnostics, and for atheists, there’s no difference at all: 10% of them cast a ballot for Trump in 2024 and his approval rating in 2025 was 10%.

Notice just a handful of groups that are to the left of the line—these are folks who have a higher approval of Trump in 2025 than vote share in 2024. It’s three groups: Muslims, Buddhists, and Hindus. The biggest deviation is Hindus: he got 7.8% of their votes in 2024 and his approval rating was 19% with this group a year later. But we really need to mind the sample size issue: it’s just 57 respondents in 2025. It’s less than 100 Muslims, too. So we shouldn’t make too much hay about this.

I wanted to dig down one layer on some of these larger groups to try and sort out which factors are driving down Trump’s approval ratings. To start with, I just zeroed in on the age of the respondent. I wanted to know if it were younger white evangelicals who were souring on Trump faster than their older counterparts.

That’s exactly what the data indicates. The biggest divergence between vote choice and approval rating is among white evangelicals who are under the age of 40. In 2024, nearly 78% of this subgroup were Trump voters. But in 2025, Trump’s approval rating was 61% among this same group. That’s a seventeen-point slide. Among older white evangelicals, the difference is smaller but still significant: 86% down to 75%. That’s an eleven-point drop in just a year. That’s a really consistent finding here: white evangelicals were near-unanimous supporters of Trump; however, they aren’t quite as enamored with him now.

There are also a few other instances of a drop, most notably among white Catholics. Among the younger set, his approval rating stands at exactly 50% in 2025, while 60.5% supported him in the 2024 race against Kamala Harris. For older white Catholics, it’s a slide from 66.5% vote share to 57% approval. So, it doesn’t look like Trump is doing well among the two largest white Christian groups regardless of age.

There’s not much worth pointing out among the other subgroups. One interesting finding: older mainliners have slid 6 points away from Trump. We just don’t see anything notable among the “nones,” though.

What about if we break down the larger Christian groups by level of attendance?

Again, those white evangelicals have slid away from Trump in a noticeable way. That’s true for those who are going to church on a weekly basis as well as those who are seldom/never attenders. The magnitude of the move is nearly the same, as well: ~12 percentage points. What’s really interesting is how non-white evangelicals who go to church a lot have really turned against the president. He earned nearly 57% of their votes in 2024, and his approval rating was down to 44%. There’s been no real movement among non-white evangelicals who don’t go that much.

Two other results are worth flagging. Trump has lost a dozen points among Black Protestants who don’t go to church that much (but it’s not like they were big fans of his anyway). But the white Catholics who don’t go to Mass that much represent a much more consequential result. Among this group, a strong majority favored Trump in 2024: 58%. But his approval rating was down to just 48% in 2025. This is not a small group, either: 5% of all Americans are white Catholics who rarely go to Mass. That’s about the same size as atheists.

Now, here’s a fun question: is Trump doing that much worse nine months into his second term compared to the same period back in 2017? We can check that really easily using the Cooperative Election Study.

The answer is unequivocal: nope, not in any meaningful way. If I plot mean approval in 2017 and 2025 and add in those confidence intervals, what you see is a whole bunch of overlap between the two estimates. For many groups, the actual percentage differences are so small it’s hard to even see them on the graph. That’s the case for huge groups like white and non-white Catholics, “nothing in particulars,” atheists, and Black Protestants.

Now there are a handful of groups where the differences get outside the margin of error. For instance, Trump’s approval rating among white evangelicals was 70% in 2017. It was 73% in the 2025 data. That’s outside the margin of error, but is it substantively significant? I think we can debate that point. You can see that mainline Protestants are a bit less warm toward Trump in 2025 compared to 2017 (47% vs. 51%). There’s also been some movement in Trump’s direction among Jews, too: 27% vs. 35%. Agnostics have moved five points in Trump’s direction.

However, looking at all of this broadly I just don’t think that I can make a strong claim here that Trump is doing particularly worse in 2025 compared to 2017.

I thought I would end this by building a table that has a couple key metrics: Trump’s vote share in 2016 and 2024, along with his approval in 2017 and 2025.

I think the most helpful way to look over this data is comparing the gap in 2016/2017 vs. the gap in 2024/2025. For white evangelicals, there’s been more movement away from Trump in 2025 compared to 2017: 13 points vs. 8 points. The difference for white Catholics is even larger, though. In 2016/2017, the difference was nearly zero. In 2024/2025, it was 10.5 points. That’s entirely because of a shift in vote choice, not approval, among this group.

The movement among non-white Catholics is also worth pointing out. During Trump’s first term, they actually warmed up to him in his first nine months in office (+5). That’s not what happened during Trump 2.0: he went five points the other direction. And the reason is exactly the same as it was for white Catholics: a higher share voted for him in 2024 compared to 2016, but his approval rating stayed about the same.

After looking over all of this, the impression that I get is that Trump’s approval nine months into each term is basically the same in 2017 vs 2025. The big difference? He won a bigger share of a lot of groups, but then they quickly seemed to regret their decision. That’s what is leading to the big erosion in Trump’s support between November of 2024 and November of 2025.

Code for this post can be found here.

Ryan P. Burge is a professor of practice at the Danforth Center on Religion and Politics at Washington University.

This post has been unlocked through a generous grant from the Lilly Endowment for the Association of Religion Data Archives (ARDA). The graphs you see here use data that is publicly available for download and analysis through link(s) provided in the text below.

The amount of cynicism in the world is sometimes hard for my mind to grasp. I really struggle watching TV shows and movies that have a really bleak view of the world. I don’t want to watch The Golden Girls all the time (although it is a really great thing to put on the TV right before bed), but I just can’t spend a lot of my time meditating on the evilness of the human experience over and over again. I think it’s bad for my soul, to be honest.

The internet only makes it worse. During the college football playoffs, the Indiana Hoosiers went on an absolute heater. They were playing out of their minds. In the semi-final game against Oregon, they blew them out of the building, winning by 34 points. On the first play from scrimmage, an Indiana player (D’Angelo Ponds) intercepted a pass and ran it back for a touchdown. After the game had ended, an equipment manager from Oregon found Ponds on the sideline and presented him with the ball that he had intercepted to start the game. It was an absolute class act.

While most of the comments were positive, a few of them were so incredibly cynical.

“Staged. Only doing it because of the cameras and for clicks”
“He’s just doing that in the off chance that they get Ponds to enter the portal and transfer to Oregon next season”

It almost feels like there is a growing group of people who can see no goodness in the world. And that viewpoint is being celebrated and amplified on social media.

I get the foreboding sense that Americans just don’t trust each other anymore. They think that everyone is selfish, cruel, and without a moral compass. People don’t start a company to build a great product; they do it to make themselves a fortune. Politicians don’t run for office because they want to make life easier for their fellow citizens; they do it because they are power-hungry.

Well, the other day I was browsing the codebook for the 2024 General Social Survey on the Association of Religion Data Archives and was reminded of a really pertinent question to this discussion. It simply asks:

Generally speaking, would you say that most people can be trusted or that you can’t be too careful in dealing with people?

You know what, I am kind of surprised at how stable these lines are over time. It does look like there has been a decline in people saying, “others can be trusted,” but it’s a pretty subtle shift. For the Silent Generation, it declined ten percentage points across about five decades of data. For Boomers, it was rock steady from 1972 through 2010, then it dropped noticeably in the last couple of years.

But check out Generation X. If anything, their level of interpersonal trust has actually gone up over time, and now it’s at basically the same level as Boomers. Back in 2010, the gap between those two lines was ten percentage points. I need someone to explain that to me. And, while you are at it, figure out what’s going on with Millennials. I think trust is up, but it’s not a clear pattern.

Then, that Gen Z line is pretty easy to figure out: trust started out low and has only slipped from there. It was slightly above 20% in 2018. Now, it’s in the single digits. Oof. There’s no generation that is less trusting of other people than Gen Z.

I do feel like I need to zoom in on the last couple of years of data, though. Just to make this point plain.

I think this graph does a better job of showing the complete picture when it comes to interpersonal trust. Look at the middle three generations: Boomers, Gen X, and Millennials. The share who say that “you can’t be too careful in trusting other people” is really similar: 62%, 64%, and 67%. It’s not like these folks have a huge well of goodwill toward random strangers. In contrast, the share who say folks can generally be trusted is around 30% among Boomers and Gen X, but is clearly lower among Millennials at 24%.

Gen Z is just on a different planet, though. The share who can’t trust others at all is 74%; that’s seven points higher than Millennials. The share who say others can be trusted is a measly 13%. That’s 11 points lower than Millennials and 17 points lower than Gen X.

This is an empirical fact: unless they experience a significant increase in trust (which we haven’t seen in fifty years of polling data), Gen Z will be the least trusting generation in modern American history.

Okay, that’s firmly established, right? Now we get to move to the unpleasant task of trying to figure out what factors have an impact on interpersonal trust. One that absolutely looms large is education. I just divided the sample into those with a high school diploma or less and those who have earned a four year college degree.

These gaps are huge, folks! The relationship between education and trust is just impossible to deny. Among Boomers with a college degree, 41% trust other folks. It’s just 23% among those who stopped at high school. For Gen X, those figures are 38% and 30%, respectively. But then they balloon back up for Millennials. Among those who stopped at high school, just 15% say that others can generally be trusted. It’s 40% of Millennials who have a bachelor’s degree.

Now, what can we make of Gen Z? Well, the first thing to note is that the baseline of trust for them is just so much lower than other generations. But the education gap is still there. Among members of Gen Z who stopped their education at 12th grade, only 5% say that other people can be trusted. But even among those who have earned a four year college degree, it’s 10%. Part of the problem here is that Gen Z is still working their way through educational attainment. So it may take a while for these education gaps to fully solidify. But even among highly educated Zoomers, trust is in short supply.

To this point, I haven’t even mentioned religion, which is an absolute travesty. We need to rectify that immediately. I do think that there are ample reasons to believe that trust is related to religious attendance. I mean, you have to be a bit more trusting to show up at a house of worship with lots of other people voluntarily, right?

There are certainly some gaps in trust based on church attendance, but they really aren’t as large as I would have guessed. For instance, among Boomers who go to church every week, 29% say that other people can be trusted. It’s 26% of those who never go to church. The gap on the “can’t be too careful” response is a bit bigger, though: 9 percentage points.

But then if you compare Gen X and Millennials, I just have to scratch my head a bit. There’s a huge gap in trust among Gen X. Among regular church attenders, 49% say that others can be trusted. It’s only 30% of the never attenders. However, there is no gap at all among Millennials on this question. Which makes no sense to me. It’s the only generation in which religious attendance makes no difference on this question. In every other cohort, the gap is consistent and large.

I do need to highlight the huge disparity with Gen Z, though. This is the biggest gap of all. Among Gen Z folks who don’t go to church at all: 88% say that you can’t be too careful in dealing with people. It’s only 50% of Zoomers who attend church regularly. That’s a chasm. Non-religious Gen Z folks are incredibly wary of their fellow citizens.

There are a lot of moving parts here, though. Many factors can impact trust: education, income, ideology, church attendance, gender and race. I wanted to try and isolate them from each other by pulling together a regression. What I really wanted to concentrate on was comparing Gen Z to older generations to see if trust “works” differently for young adults. Man, I was blown away here.

There were two factors that I identified that worked the same for the whole sample: education and church attendance. In both cases, they led to higher levels of interpersonal trust. The effect for Gen Z was larger than for older generations; that may be due to a smaller sample size. I think we can pretty much parse why education and socialization drive up interpersonal trust. It takes trust to go away to college and it also takes trust to show up to a house of worship each weekend. Non-social, untrusting folks wouldn’t volunteer to do stuff like that.

But then, the other two factors just don’t work the way I would have expected. For instance, take income. Among older Americans, higher income individuals are more trusting. For Generation Z, it’s the exact opposite. As income rises, so does distrust. And remember, that’s controlling for education. I don’t know what to make of that at all.

The other one that I will think about for a very long time is political ideology. Among older Americans who identify as liberal, they are more likely to trust other people. Among Gen Z liberals, it’s exactly the opposite. In fact, I struggled to find anything that had a more deleterious impact on trust among Gen Z than identifying as a liberal.

Now, this one has an interesting thought experiment behind it. Older Americans who are liberal are more trusting. Which makes some kind of sense. Liberalism is based on collectivism. It embraces the idea that “together, we can achieve more.” Things like universal healthcare rely on a sense of trust among large groups of people. So, yeah, I can see why liberals are more trusting for Boomers, Gen X, and Millennials.

However, why does it go in the opposite direction with Gen Z? Well, maybe liberals feel very slighted by the fact that Donald Trump has been on the ballot for president in every election in which they’ve been eligible to vote, and he’s won twice. Or it could be that they are more impacted by the cynicism of the internet than conservatives? Or maybe it’s because they report higher rates of depression and anxiety than the rest of the ideological spectrum?

I am just spitballing here. But, in my mind, this may be the most important thing I’ve uncovered in my newsletter in the last few months. Trust is an absolutely essential component of a functioning democracy. Without it, we return to the Hobbesian view of the world where we can accomplish none of the higher things in life because we are constantly worried about being assaulted, kidnapped, robbed, or murdered.

If this is the worldview of Gen Z, then the future of America is in real peril.

Code for this post can be found here.

Ryan P. Burge is a professor of practice at the Danforth Center on Religion and Politics at Washington University.

Whenever I read a lot of chatter on the internet about people’s financial situations, the clear emotion that seems to come through is a creeping level of despair. People feel like they will never be able to get ahead. If their car breaks down, they won’t have the money to fix it. If their refrigerator stops working, they won’t have any way to replace it. They certainly can’t save for a down payment for a home nor can they really afford to have children. What a lot of them seem to believe is that their lives will be worse than their parents or grandparents.

I’ve always wanted to somehow measure despair, but I didn’t really know a good way to do that until Praveen Vissa, a bright student at John Burroughs high school in St. Louis reached out and wanted to propose that we work on a post together about a single question in the GSS aptly titled “goodlife.”

It simply states, “The way things are in America, people like me and my family have a good chance of improving our standard of living -- do you agree or disagree?” Before we dig into how religion may play a role in shaping economic optimism (or despair), let’s start by just tracking how answers to that question have varied in the last thirty years.

This is clearly a question where I had absolutely no idea how the answers were going to look over time. I tried to hypothesize some reasons why economic pessimism would be higher or lower, but I just couldn’t fully conceptualize how someone would have answered this question in 1994 (the first time it was asked) or in the 2024 General Social Survey. What I found was that economic optimism is still fairly popular but it has taken a noticeable dip in the last 15 years or so.

In the 1990s, about two-thirds of Americans agreed that they had a good chance of improving their standard of living. That actually rose in a significant way as we moved into the early 2000s, when about three quarters of Americans believed that their economic future was going to be better than their current situation. But that was the high water mark. From there, things began to slide. By 2010, the share who took a positive view was around 60% and then it just continued to drop. In the last three waves of GSS (2021-2024), the share who agreed that they had a good chance of improving their standard of living dropped below 50%.

I do need to point out, though, that the rise in the share who disagree with that statement hasn’t really skyrocketed. It went from 20% of the sample around the time Barack Obama won the White House to 30% in the 2024 GSS. So, the optimism still exceeds the pessimism by 15 percentage points even today.

You know what we have to do at this point, right? Poke around some demographic variables to see if there’s any evidence that things like age or education have an impact on how people view the possibility of moving up the economic ladder.

When I look at this question by generation of respondents, I don’t know if I’m blown away by the result. That’s especially the case when considering the confidence intervals. The only thing that really jumps off the screen to me is the clear difference between Baby Boomers and everyone else. Only 39% of Boomers are convinced that their financial situation can be improved. Among younger generations, it was about ten percentage points higher.

There are two things that are worth considering here. One is that there’s a good reason Baby Boomers don’t think their financial situation is going to get better: they have just retired or are going to retire in the near future. That’s not usually a period in life when money becomes more plentiful. I looked back at prior periods in the GSS and found this same pattern — the generation that was closest to retirement was less apt to agree with the idea that their standard of living was going to rise.

The other thing that jumps out is that Gen X still feels pretty good about their ability to improve their financial state — really the same as Gen Z. I don’t know why that is, really. Is it a function of Gen Z just not being that optimistic about things? Or is it that Gen X just feels much rosier about their salaries rising? I don’t have a good handle on that.

The other big factor that could impact this is, of course, education. I just restricted my sample to folks born in 1965 or later. This means that age really has no impact on these results based on the prior graph.

What really comes through for me is that there is a clear cut-point in this data between those with graduate degrees and everyone else. Whether one just stopped at a high school diploma or earned a bachelor’s degree, the share who thought that their economic situation would improve in the future was statistically the same: just about 45%. To me, that’s fascinating because the data is crystal clear on this point: folks with college degrees make hundreds of thousands of dollars more than those who stopped at high school.

Maybe it’s just lifestyle creep? Maybe having a white collar job means living in an area with a higher cost of living? Maybe it’s student loan debts that are dragging them down? It’s hard to know, honestly. But those with a graduate degree are incredibly optimistic about the future: almost two-thirds believe that they have a good chance of improving their standard of living.

Okay, we haven’t even touched on religion yet. But let me do that now. I’ve written before about how religious people tend to report higher levels of happiness compared to those who claim no religious affiliation, so maybe those people are also more optimistic about their ability to move up the economic ladder?

One possible reason lies in how religion can shape a sense of hope for the future. A study found that religiosity is linked to higher levels of optimism and expectations that life will improve, which may stem from how belief systems shape a person’s outlook on life. If that’s true, we should at least see some hint of it in how people answer this specific question about their economic potential.

But, I certainly don’t see a huge difference in responses here. Among people who claimed any religious affiliation, the share who agreed that their economic future could be better was just slightly above 50%. The lowest score was among Catholics at 49%. The highest percentage was among Black Protestants and those of other faith groups at 55%. However, the differences between any of these religious groups was not statistically significant. In other words, we can’t really say if evangelicals have a more positive outlook compared to Muslims. It’s all right about the same level.

Now, I do think that the data is suggestive of the idea that the non-religious are a bit less likely to see a better economic future. Only 46% of the nones agreed to the statement we’ve been focusing on in this post. That difference, though, is not statistically significant. Compared to any of the other individual groups, but when I just collapsed all Christians together that tightened up the confidence interval in a way that made the nones clearly lower than Christians.

But, again, I’m not really bowled over by this huge finding about how religious people have a more hopeful economic outlook compared to the non-religious. That led me to poke around a bunch of variables in a regression analysis. I controlled for things like education, age, political ideology, and religious affiliation.

You can quickly ascertain what I found when I plotted these coefficients and confidence intervals: not a whole lot of anything, really. I do find that having a graduate degree is predictive of having a more positive economic outlook. But that’s really all I could ascertain from this analysis. There’s really no difference between generations, the gap between political liberals and conservatives is non-existent. And, like I just discussed earlier, the question about religious affiliation doesn’t really do much, either.

In other words, a whole bunch of “null results.” But then I threw in a different religion variable into the regression model and I hit paydirt. Frequency of religious attendance does have a pretty clear impact on this question of one’s likelihood of raising their standard of living.

Again, I controlled for a whole bunch of factors that could matter here like age and education, but what I found was unmistakable — people who go to church more often are much more likely to believe that they will have a positive economic future. Among never attenders, just 44% agreed that they could raise their standard of living. Among those who attended church once a week, it was 58%. That difference is both statistically significant and substantively meaningful. A fourteen point increase on this question is the largest we’ve seen across any variable.

Why does attendance have such a clear impact? I don’t know if I can point to one single causal pathway. One thing that came to mind is that maybe people who are more social believe that they can always rely on their network of friends and associates to help them get a job or move up in their career? Maybe more socially active people are just more optimistic about everything else in life?

One possibility is that attendance captures what affiliation alone misses: being embedded in a community. Showing up week after week builds trust, connections, and support over time. One paper shows that frequent attenders report greater life satisfaction and purpose in life, and that attendance predicted these outcomes more strongly than other forms of community involvement, suggesting that benefits come from social integration.

If religious social networks function as a safety net when things go wrong, it makes sense that frequent churchgoers would feel more confident navigating setbacks and more optimistic about their future.

I don’t know for sure. But I really do think that this analysis is going to be the impetus for some future data exploration, as socialness and optimism seem to run on the same tracks.

Code for this post can be found here.

Ryan P. Burge is a professor of practice at the Danforth Center on Religion and Politics at Washington University.

I plan on the first mailbag post to run in late August, so get your questions in before August 1. If this goes well, it will become a recurring feature.

I’ve said offhand a few times that PR people are not my favorite folks on Earth and there’s a reason for that. Their job and my job are not the same job. Actually, many times they are diametrically opposed to each other. Folks who work in public relations want to cast their organization in the best light possible. I want to try and get to the unvarnished truth about what’s actually going on behind the headline numbers.

This is on full display when a university releases its enrollment numbers and they aren’t very good. The headline may be something like, “<UNIVERSITY> experiences a 7% increase in transfer students.” Then you actually read over the report and the first couple of paragraphs are about peripheral metrics that point in a positive direction. Buried near the end is the actual number that we all really care about: overall enrollment, which is often down by a significant percentage. The press release is not actually lying, it’s just burying the bad news under a mountain of words that could be seen as positive.

Denominational reporting tends to be less opaque, but it still tends toward emphasizing the positive trends. For instance when the Baptist Press released their annual article about updated Southern Baptist Convention numbers it had the headline, “Southern Baptists see attendance, baptism gains amid membership declines.” I can actually appreciate how this one was presented – the good news up front but also not totally minimizing the continued downward trend in SBC membership.

You can probably guess where I’m headed: I’m going to tell you the truth about what’s happening in the Southern Baptist Convention. I can summarize my views pretty succinctly – there’s nothing in the most recent numbers that changes my view of where the SBC is headed in the future.

Let’s start, as we always do, with the broadest look at SBC membership data. I started this chart right after the end of World War II.

I’ve said this on many occasions, but the rise of the SBC from 1945 through 1990 will never ever be replicated again in the history of American religion. The denomination recorded 5.8M members right as all those GIs were coming home from defeating the Axis powers. As the United States began the 1990s, the SBC had swelled to 15M members. In many years, they were adding a quarter million new folks per year. That’s five thousand people being added to the membership files every single week for a period of 45 years. Insane to consider that momentum.

But the good times didn’t last, of course. Religious historians will look back at 2006 as the absolute peak of denominationalism in the United States. The SBC enjoyed 16.2M members. Something that Convention, or no other denomination in the U.S., will ever see again.Then the cracks started to form. Slowly at first, and then the pinholes became the size of golf balls and the hull of the SBC began taking on water in an incredibly rapid fashion.

Read more

So, that’s a 4.4 percentage point shift between 2022 and 2025. The nones were back to their levels from 2018. That’s a big deal, clearly. In that previous post, I noted that the General Social Survey also recorded a drop in the nones: down about 3 points between 2021 and 2024. This isn’t just a single rogue survey, it’s now two different instruments (with different methodologies) arriving at a similar conclusion.

If the nones dropped by 3-4 percentage points over the last few years, how did that happen exactly? That’s the goal of this post: trying to zoom in on a specific trend in the data that needs some more careful exploration.

The graph below was included in the previous post—it’s the share of each generation that reported no religious affiliation in the Cooperative Election Study in the last few years.

The generation that really jumped out to me for this post was the Baby Boomers. I’ve written about this group several times in the last couple of months on this newsletter. They are the most important generation in American religion right now for a couple of reasons. One is that they are numerically large. The second is that they are religiously active. And the final reason is that they are rapidly heading toward their twilight years now (the youngest Boomers are in their mid-sixties).

Notice their trajectory in the CES—it looks like the first portion of a roller coaster. A slow ascent, a clear peak, and then a rapid decline. In 2016, 23% of Boomers were non-religious. That rose to 28% by 2023. But the last two surveys have this figure at 24% and 21% respectively. That’s a return to levels seen in 2010. What in the world is happening there?

If I break this down into the three different types of nones, we can see much more granularity in the data. The share of Boomers who were atheists has never really been that high. It was 3% of the sample for a stretch that ran from 2008 through 2015. It did slowly creep up in the last decade, though, and in some survey years it was 5%. Agnostics were on a similar trajectory. But that noticeably reversed in the last two years. Now, atheists and agnostics are 3% each, down from a total of 9% in 2023.

But the bigger drop in this graph is that third category of nones: the nothing in particular. Notice how they’ve grown from 15% of Boomers to 19% in the 2023 data. That’s reversed significantly. Now, just 14% of Boomers are nothing in particular. That’s a five point decline in just two years. That’s a level of movement that is uncommon in the world of religious demography.

Of course my wheels are spinning now, seeking out the reason why. One quick theory I had was that Republican Boomers are coming back to religion because they believe that to be a crucial part of being a member of the GOP. Republicans are the party of people of faith, so it makes sense to check that out.

This graph is not what I expected to see at all, honestly. Not even a little bit. In fact, it’s a graph that does not line up with my understanding of religion and politics in any meaningful way. The share of Republican Boomers who are non-religious is just 13%. That’s basically what it was four years ago. The baseline has been 12-13% for about a decade now. That’s not the story of how Boomers have shifted their religiosity.

Instead, the trend lines show a big shift in the lines for Democrats and the politically Independent Boomers. For Boomer Democrats, the rise in the nones is startling. It was just 21% back in 2008. By 2023, it had exploded to nearly 37%. That’s an increase of about a percentage point a year. But then it reversed itself and the share of Democrats who claim no religious affiliation dropped about seven percentage points in a span of two years. For Independents, it’s even larger: nearly eleven percentage points between 2023 and 2025.

If you even have a passing interest in demography you know that shifts like that should be met with skepticism. Change is best described as glacial. Two percentage points in a year is a lot of movement. Here we’ve got five points a year. It seems…off.

We have ways to validate the CES data, by the way. It’s pretty simple: just pull in another survey that asks similar questions and see if the trend lines there move in a similar pattern. That’s what I did below, just took the General Social Survey and calculated the share of Boomers who are non-religious over time. That survey was last fielded in 2024, by the way. So that’s as far as my sample can go.

In the GSS, the pattern is really stable over time. You can see that back in 2008, the GSS pegged Boomers at around 14% non-religious. That was about three points lower than the CES. The GSS trend line is exactly what we typically see doing this kind of work—a slow and steady change over a long time horizon. Even in 2024, the GSS estimate of Boomer nones is 16.5%. That’s not statistically distinct from the 2008 data.

The CES, in contrast, displays a consistent rise in the share of Boomers who claim no religious affiliation. The numbers do seem to see saw during a period that runs from 2012 through 2020, but the macro-level trend is an increase from 21% to 25%. Then a noticeable bump in 2023 and two years of decline. In other words, the lines don’t look that much like each other. CES shows a clear rise and fall. The GSS just shows a very gradual rise that’s not statistically significant.

Here’s the same analysis, but I broke it down by political partisanship.

Here, I just can’t see anything that resembles a consistent pattern across these two samples. I mean, the line for Democrats in the CES just has a completely different shape than that from the GSS. The GSS had a big spike in 2021, then a noticeable drop in the last two years. The CES had a spike in 2023 and then a drop in the two years after that. There’s just no consistency there. And you can see the same thing for Independents—the lines don’t make sense when compared to each other. The last few years of the CES is a massive decline. In the GSS, it’s a rise in the nones.

I think that the graph for Republicans is much less concerning. Both lines are definitely trending upward. They each show a little bump in the nones in close succession. But then both lines dropped back to baseline in the last few years.

Guys, this is why this job is hard! There’s not a consistent pattern in this data. Both surveys are not speaking with the same voice on a very basic question of religious demography. I can’t look at these last few graphs and make a clear pronouncement about the overall religiosity of Baby Boomers. I wish it was easy, but it’s just not.

There is one graph, though, that does seem to tamp down my anxiety a bit. It’s the share of Boomers who attend church nearly every week or more over the last fifteen years.

This graph looks much more plausible. Both samples have the weekly attendance rate around 30% over time. There is some separation between 2015 and 2020, but it’s not substantively large: maybe just three percentage points. Then the lines come back together for a couple of years before separating just a bit in 2024. But there’s a consistent voice that emerges from both surveys: the share of Boomers who report weekly attendance is likely around 30%. Is that changing over time? I don’t get a strong sense of direction from either instrument. In other words, attendance is pretty stable.

By the way, this last graph actually complicates the conclusions from the prior graphs even more. Remember, Boomers seemed to find faith again in large numbers in the last two years of the CES. If that were the case then why did Boomer attendance not change in any demonstrable way during that same period? We should theoretically see a jump in weekly attendance but it’s not there in the data.

That’s why this kind of work is maddening. There isn’t a clear narrative that emerges from this data. I’ve written recently about the two most prominent methods for survey collection recently: panels and probability-based surveys. The CES uses the former approach, while the GSS has employed the latter design for decades. That could be one explanation for the divergence. It’s just hard to know what’s the “right” answer here.

What both these instruments point to, though, is that the overall trajectory of American religion has changed over time. After decades of a consistent increase, that secularization era is over. The rise has turned into a plateau. Have the nones reversed? Yes, it seems like they have. But why has that happened? I just can’t put my finger on it.

And it’s driving me crazy.

Code for this post can be found here.

Ryan P. Burge is a professor of practice at the Danforth Center on Religion and Politics at Washington University.

Man, I have been asked that question at least a hundred times across various social media platforms over the last couple of years. It’s really funny being in my position, because I have this really weird vantage point on what groups people really want to see visualized in data sources. There are a couple that pop up on a regular basis: Latter-day Saints are very interested in seeing themselves in charts and graphs. Members of the Presbyterian Church of America are also very much “reply guys” (and they are almost always male). And then the aforementioned Unitarians are frequently invoked in a comment.

Just for some quick background on our UU friends: Unitarian Universalism is a tradition that asks almost nothing doctrinally of its members — atheists, Christians, Buddhists, and pagans share the same pews, united by shared values around human dignity and social justice rather than creed. The UUA, formed in 1961, has always been small but culturally influential, claiming six U.S. presidents and a leading role in progressive social movements from abolition to marriage equality.

I’ve actually spoken at two separate UU gatherings. One was only a few blocks from my house and the other was in Nashville, TN earlier in 2026. I thought that the Nashville community had a great acronym: FUUN (First Unitarian Universalist of Nashville). When I was with both communities, the impression that I got was that there’s a lot of demographic similarities between the UU and many mainline denominations. They are usually quite a bit older and don’t have a ton of racial diversity.

That’s also reinforced in the data. The denomination has some obstacles in the future: shrinking membership, fewer congregations, and an increasingly narrow geographic and demographic footprint. There are some bright spots, of course, but many of the key trends facing the UU look strikingly like what’s happening with the Episcopalians or the United Methodists.

The Unitarians are nice enough to publish a “Certification Report” on an annual basis and that’s where almost all the data below came from, although, as you will see - I’ve also included some data from the Census Bureau as well.

Let’s start with the most logical metric: membership. Unlike most denominations which only count formal members, the UUA tracks two distinct groups: dues-paying members who have officially joined a congregation, and “non-member friends” — people who regularly participate and financially support a congregation but haven’t taken the formal step of membership.

The trend for both groups is absolutely in the downward direction. In 2016, the UUs counted a total membership of just over 151,000 and then the “friends” added another 99K to that total. So, there were about a quarter of a million Unitarian Universalists in the United States a decade ago. That means that there is 1 UU for 1,340 Americans (give or take). It’s just not a numerically large group, a fact that I will return to in just a minute.

However, as you can clearly see in the previous graph: both types of membership metrics are in pretty steady decline for the UUs. Membership held relatively constant between 2016 and 2020 - losing just 7,000 folks. But the “friends” dropped by almost a third during that same time period (from 99K to just 64K). In the last few years, both metrics have continued to slide. In 2025, membership is down to 130,000 and non-member friends are at 54,000. In other words, there are 66K fewer Unitarians today than just a decade ago. That’s a drop of just over 26%.

But, can I just put the Unitarians in the larger context of American religion really quickly? This has become one of my hobby horses in the last couple of years: the average American just has absolutely no idea how large (or small) many religious groups really are in the grand sweep of the American religious landscape. Unitarians, numerically speaking, are a rounding error in the bigger picture.

The Religion Census pegged them at just over 201,000 members when they did their data collection a few years ago. For context: for every UU in the United States there are sixty Southern Baptists, there are thirty Jews, and there are three hundred Catholics. The only survey that I’ve ever seen that really tries to assess the size of UUs using population based sampling is from PRRI, and they place them at .5% of the population.

In numeric terms, that’s 1.5-2M folks. Why the discrepancy? Well, it’s the classic case of members vs people who say that they are affiliated with a group. I have a sneaking suspicion that a whole bunch of Americans like the idea of Unitarianism but just can’t or don’t want to join a congregation. But according to their current membership data: UUs are less than half the size of the aforementioned PCA and are marginally larger than the Anglican Church of North America (who are in the 125K member range).

Read more

This makes a lot of anonymous people on Twitter very angry, of course. I’m guessing that many of them are Catholics who believe that Catholicism, correctly measured, will exhibit results similar to their evangelical cousins. The replies are always something like, “No, show us what REAL Catholics believe on this issue.” Which I think is shorthand for: I want you to only compare weekly attending Catholics to weekly attending evangelicals. Their assumption is that if I do that, the statistical differences will disappear.

Well, let’s just put that to the test today, shall we?

The first thing we need to establish is that the reply guys are kind of right in one way: the average white Catholic doesn’t go to church with nearly the same regularity as the average white evangelical.

Back in the halcyon days of the early 1970s, Catholics were more devout than evangelicals — and by a fair amount. About 60% of white Catholics were going to Mass every week in 1972 compared to only 45% of white evangelicals. That’s not a small difference. But it didn’t last long. By 1980, the two lines had clearly crossed around 45%. But then the Catholic line continued to move downward while evangelical attendance rates held steady. By the late 1990s, evangelicals were more likely to be weekly attenders than their Catholic counterparts.

From there, the gap only widened. It was about ten percentage points by 2000 (45% vs. 35%), and the chasm has only increased in the last twenty years. By the 2010s, over half of white evangelicals went to church nearly every week compared to less than 30% of Catholics. In the data from 2024, 57% of white evangelicals were weekly attenders compared to 25% of white Catholics. So not controlling for attendance gives us a much different sample when analyzing evangelicals and Catholics. Keep that in your back pocket for now.

Now, there’s another variable at play here that needs to be considered: political conservatism. Are white Catholics more politically moderate than white evangelicals? The answer is: absolutely.

The graph on the left does not contain an attendance filter, so you’ve got an evangelical sample that is much more religiously active than the Catholic subgroup. There are clear differences in the political ideology of these two groups. I think it’s fair to say that evangelicals have always been more conservative than Catholics, although those gaps just weren’t that big back in the 1970s (less than ten percentage points).

That gap has widened over time in a very noticeable way. You can clearly pick out an inflection point, too: somewhere in the early 2000s the evangelical line began to shoot up while it rose much more slowly for Catholics. Today there’s just a huge ideological difference between the two. Almost two-thirds of white evangelicals identify as politically conservative vs. just 38% of white Catholics.

But remember how evangelicals are much more likely to go to church than Catholics? I controlled for that in the right panel by only including people who indicated that they attended church nearly every week or more. I don’t think this alters the general trend I just described in the “all attendance levels” graph. White evangelicals have always been more ideologically conservative than white Catholics. The gaps between the two lines on the left and the two lines on the right are almost exactly the same magnitude as well.

So let’s test out this idea that Catholics are just as socially conservative as evangelicals by pulling in both things that we’ve learned. I am going to show you three graphs: the entire sample of evangelicals and Catholics, only weekly attenders of those two groups, and finally only weekly attenders who also identify as politically conservative. That way we can control for those differences as much as possible.

Let’s start with a question about abortion. This one asks if respondents think it should be possible for a pregnant woman to obtain an abortion if she wants it for any reason.

In the full sample, a pretty big gap emerges between evangelicals and Catholics on this issue in just the last ten years. Prior to that, evangelicals were about ten points more pro-life than Catholics. After 2012 or so, the Catholic line began to rise. Now, it’s very fair to say that the average white Catholic is twice as likely to support abortion access as the average white evangelical. But remember: we need to control for the fact that Catholics aren’t as likely to go to church and aren’t as conservative. That’s what I did in the next two panels.

If we restrict the sample to weekly attenders, the gap narrows. In fact, until just recently I don’t think we can make a strong claim about which group was more pro-life. But in the last couple of years, weekly attending white Catholics have moved in the direction of permissiveness on abortion while evangelicals haven’t budged. When we control for both ideology and attendance, the lines are basically right on top of each other. The sample sizes aren’t huge here (especially for Catholics), but when we compare just weekly attending conservatives, it’s fair to say that Catholics are neck and neck with evangelicals.

Okay, so let’s switch to a different dependent variable: same-sex marriage. The question is simply: Do you agree or disagree with the following statement? Homosexual couples should have the right to marry one another. This is the share who strongly agree or agree.

The reason the lines look funny on the left-hand side is that the question was asked for the first time in 1988 and then not again until 2004, so I interpolated the trend line to fill the gap. Again, the top graph exhibits a strong divergence between white Catholics and white evangelicals, with Catholics being a whole lot more likely to support same-sex marriage. In fact, the average Catholic today is twice as supportive as the average evangelical (69% vs. 34%).

When we just look at weekly attenders, the overall level of agreement is lower but the gap between the two groups persists. For weekly attending Catholics, agreement reached 50% around 2016 or so and has stayed there consistently in the last decade. For evangelical weekly attenders, it’s never risen above 30% and shows signs of retreat in the last couple of years — down to 18% in 2024.

What happens if we just look at weekly attending conservatives? Well, for evangelicals, support is almost non-existent: just 9% in 2024. For Catholics in the same category, it’s much higher: 40%. Although the sample that fits this criteria is very small, so there’s a lot of uncertainty in the estimate.

One more question I wanted to highlight: If a man and a woman have sexual relations before marriage, do you think it is always wrong, almost always wrong, wrong only sometimes, or not wrong at all? This is the share who indicated that premarital sex is always wrong.

This is where a huge difference emerges. In 1972, 40% of Catholics said that premarital sex was always wrong. Today, that is down to just 7%. For evangelicals, there’s been a dip too, though it’s much smaller (from 58% in 1972 to 45% in 2024). Of course, that includes a whole lot of folks who say that they are Baptist or Catholic but don’t go to church that much.

The second graph is only weekly attenders, and the same pattern repeats. The drop for evangelicals is almost exactly the same: twelve percentage points between 1972 and 2024. For weekly Mass-attending white Catholics, a majority said that premarital sex was always wrong back in 1972. Today, that’s down to just 21%.

Even in the most restrictive sample of weekly attending political conservatives, there’s a chasm between white evangelicals and white Catholics. But what’s different here is that the trajectory of both lines is basically flat. About 70% of evangelicals thought premarital sex was always wrong. It was around 45% for Catholics.

Of course, there’s one more way to test this: a regression. I just threw in the most basic controls — gender, age, attendance, and ideology. The key thing to know here is the reference category: anyone who is white but does not identify as evangelical or Catholic. So the coefficients below are compared to that benchmark.

The further left the point estimates, the more conservative the view on these three social issues. On abortion access, an evangelical is about 10 percentage points more likely to oppose a woman’s right to an abortion than a Catholic. When it comes to same-sex marriage, an evangelical is also much more likely to disagree with the practice compared to a similar Catholic. And remember — we are controlling for the things that should matter most here, particularly their level of church attendance and their political ideology. So this is apples to apples, as best as we can manage.

The one that really blows me away is the question about premarital sex. Compared to white folks who aren’t evangelical and aren’t Catholic, evangelical respondents have a much more conservative sexual ethic. But look at the Catholic coefficient — it’s actually positive. That means the average Catholic is actually more permissive of premarital sex than the average white person who is neither evangelical nor Catholic. That’s staggering.

I think this is the best test I can devise to really compare devout, conservative Catholics to evangelicals in the same segment of the population. I just can’t look at these results and say that “real” Catholics are just as socially conservative as “real” evangelicals. They aren’t — empirically speaking — as conservative on these three core issues that I think do a good job of tapping morality questions.

The reply guys on Twitter aren’t entirely wrong to push back. Comparing all evangelicals to all Catholics really is an apples-to-oranges exercise given the massive differences in attendance rates and political ideology between the two groups. That’s a fair methodological critique. But when you actually do the work of controlling for those things — and I think this regression does that about as well as observational data allows — the differences don’t disappear. They persist, and on some issues they’re quite large.

What I take away from all of this is that evangelical identity carries something that can’t be fully explained by how often you show up or how conservative you vote. There’s a theological and cultural foundation to evangelicalism that shapes how adherents think about the body, sexuality, and the family in ways that Catholic identity simply doesn’t replicate — even among the most devout and politically conservative Catholics. The Church may teach the same things on paper, but the people in the pews aren’t internalizing them the same way. And that gap between official teaching and lived belief is, frankly, one of the most interesting stories in American religion right now.

Code for this post can be found here.

Ryan P. Burge is a professor of practice at the Danforth Center on Religion and Politics at Washington University.

This post has been unlocked through a generous grant from the Lilly Endowment for the Association of Religion Data Archives (ARDA). The graphs you see here use data that is publicly available for download and analysis through link(s) provided in the text below.

Measuring happiness is a real problem for social scientists. It’s an area of intense interest because I think all of us would like to be just a little bit happier. But it all seems pretty subjective and transient at some level, right? I mean, if you took part in a social survey that asked you: “Taken all together, how would you say things are these days — would you say that you are very happy, pretty happy, or not too happy?” How would you answer that question right now, and if asked again in four hours, would you give the same answer?

Take the often hyped-up “World Happiness Report,” published annually by Oxford’s Wellbeing Research Centre in partnership with Gallup. The report asks respondents to imagine a ladder with the best possible life at the top and the worst at the bottom, then place themselves on it. It’s elegant, but notice what it actually measures: not happiness in any rich sense, but a cognitive life appraisal. And each respondent is comparing their life to a “best possible life” that could be completely different from another respondent. When a Dane rates herself a 7.5, she’s not reporting a feeling so much as rendering a verdict. It’s a valid measure, certainly. But is it measuring the same thing as happiness? Are satisfied people always happy?

The report also tries to explain its rankings using six factors: GDP per capita, social support, life expectancy, freedom, generosity, and corruption. Notice what’s missing? Things like religiosity, family structure, community belonging, a sense of transcendent meaning. Build the index around those variables instead, and the map of global happiness looks very different. For what it’s worth, the Human Flourishing research is doing a better job of leaning into a wider variety of variables. I’m excited for that data to become publicly available.

With that preamble, I want to focus on probably the best long-term measure we have of self-described happiness: the data from the GSS. The Association of Religion Data Archives has been hosting the General Social Survey files for a long time now. Lucky for us, the happiness question has been asked in every single wave of the GSS, all the way back to 1972.

We should start by just looking at how responses to that question have changed over the full time series.

There was a whole lot of long-term stability in these measures. It’s startling to see how flat these lines are across several decades. The share of Americans who said they were “very happy” was basically the same from 1972 all the way through 2000 — about one-third of the population. The “not too happy” line held in the 10-15% range, and the “pretty happy” line edged up just a bit, rising from the low fifties to the upper fifties between 1972 and 2000. I don’t think we can make strong claims about a big shift in happiness during this period.

But that’s certainly not the case in the last fifteen years. Happiness has clearly declined recently. The share of people who said they were very happy was 32% in 2004. It was 22% in 2024. The share who said they were not too happy rose about ten percentage points over the same period. There used to be a 20-point gap between those two lines. Now the gap is statistically zero.

There is a methodology shift worth mentioning here -- the GSS had to pivot to online administration because of COVID, and that subtle change can have big ripple effects on a question like happiness. But as Sam Peltzman argues in this paper, that shift can only explain part of the “happiness crash.” There really has been a happiness dip in the United States.

Let’s spend a bit of time trying to parse through the factors that would contribute to this pretty significant bump in the share of “not too happy” people. Obviously, we should start with religion. All I did was divide the sample into religious people — meaning they claimed any affiliation — and the nones.

The first thing I want to point out is how these lines tend to move in near concert with each other. When one starts trending downward, so does the other. That tells us that macro-level conditions in the United States tend to shift happiness for everyone. There’s really no evidence in the data backing the claim that non-religious people grew much less happy while people of faith remained more upbeat.

But the other conclusion is hard to miss: non-religious people are more likely to say they are “not too happy” compared to religious people. That gap is statistically and substantively significant, and it’s persistent — averaging about five percentage points across the full time series. In 1972, 18% of nones were not too happy compared to 14% of the religious sample. In 2024, the numbers were 25% and 19%, respectively.

So, a pretty plain result: religious people tend to express higher levels of happiness than non-religious folks. But we all know there are confounding factors that can complicate that. Non-religious folks tend to be more politically liberal, and a persistent finding in the data is that liberals are just not as happy -- though some recent work has complicated that narrative.

Let’s check with our GSS data.

And this is the absolute definition of a null result. The gaps between liberals, moderates, and conservatives are basically non-existent across the full 52-year sequence. In 1972, 13% of conservatives said they were not too happy — the exact same share as liberals. Moderates were just slightly lower at 11%. I do find it interesting that the conservative line dropped noticeably over the next few decades, with the “not too happy” share falling into the single digits from about 1980 through 2005. That wasn’t the case for moderates and liberals, who stayed in double digits throughout.

The last fifteen years are really about convergence. By the late 2010s, there was no substantive difference in the share of each group saying they were “not too happy.” The trend line is unmistakably upward. Between 2010 and 2024, the share of liberals who were not too happy basically doubled — but the same thing happened for conservatives and moderates. Everyone just got grumpier in the last decade, regardless of politics.

So we have a mixed bag. Religious people are noticeably happier than non-religious people. But we can’t chalk that up to politics. Let’s keep digging by tracing how each generation answered the happiness question as they aged through the life course.

Again, stability is the big takeaway. When the Greatest Generation aged through the back half of their lives, the percentage who said they were “not too happy” was remarkably consistent: 12-14%. For Silents, the share actually dipped a few points as they aged, then jumped noticeably in the last life stage — which makes sense. Many were probably dealing with health concerns and the loss of a spouse, both of which tend to alter happiness.

Boomers are a fascinating pattern. They really have gotten less happy as they’ve aged. When they were between 30 and 44, only 10% were not too happy. By the time they reached their sixties and seventies, that rose to 17%. The back half of life has not been kind to Boomers. But Gen X seems to be speed-running this. At 45-59, they are already less happy than Boomers were at that same age.

For Millennials and Gen Z, we obviously don’t have as many data points since they haven’t reached as many life stages yet. But it’s unmistakably clear that Millennials are less happy than their parents or grandparents were at similar life stages — though we haven’t seen a big sudden bump. They just started at a higher baseline. And speaking of baselines, look at those Gen Z numbers. Oof. Among the youngest adults, 29% say they are currently not too happy. That’s nearly double the rate of Millennials at the same age. Gen Z is starting out adult life as the unhappiest generation in American history.

There are a lot of moving parts here, though — age, religion, political ideology. I need to hold some of that constant while testing the impact of something like church attendance or education on happiness levels. That’s exactly what I’ve done below. I built a model with controls for gender, partisanship, education, religious affiliation, religious attendance, and income — all factors that could have a big impact on happiness. I’m showing just the two ends of the spectrum: the predicted probability of saying you are “not too happy.”

The variables at the top don’t pop. There’s really no difference in happiness levels between men and women — just one percentage point. The same is basically true for Democrats vs. Republicans. A Democrat is about two percentage points more likely to be “not too happy” compared to a Republican, holding everything else constant. So we can’t chalk it up to gender or religious differences.

As we move down the list, though, the gaps widen. A person with a graduate degree is about five percentage points less likely to be “not too happy” compared to someone who didn’t finish high school. The biggest gap is on income. Holding everything else constant, about 31% of people in the bottom third of incomes said they were “not too happy,” compared to just 13% of people in the top third. Does money buy happiness? It does seem to help quite a lot.

What about religion? A non-religious person is about five points more likely to be unhappy compared to an evangelical — and remember, we’ve controlled for ideology, so we can’t attribute that to political differences. But an even bigger factor is attendance. Among those who attend services at least weekly, only 23% report being “not too happy.” For those who never or seldom attend, it’s 31%. The data paints a pretty clear picture: religious folks — however you measure the concept — are less likely to report being “not too happy.”

Of course, measuring happiness is difficult, as I noted at the top. Maybe religious people are just conditioned to put on a happy face in ways that non-religious people aren’t. If your tradition teaches you that suffering has meaning, that gratitude is a virtue, and that contentment is next to godliness, you may simply be primed to report greater happiness, regardless of how your life actually feels from the inside.

But here’s an interesting wrinkle: the World Happiness Report consistently ranks Finland — one of the least religious countries on Earth — as one of the happiest. Yet religious Americans report greater happiness than non-religious ones.

Social science is hard. More job security for me, I guess. Which makes me…happy?

Code for this post can be found here.

Ryan P. Burge is a professor of practice at the Danforth Center on Religion and Politics at Washington University.