Overall, this is really interesting, great work. In particular, it really highlights that India is a world unto itself.
But another one that stands out strongly to me is Turkey. In fact, it stands out so strongly that it makes me question the data there. From everything we know elsewhere, Turkey should really stand out as being a more religious country than Western Europe. But here it appears to be less religious than Belgium. Something isn't adding up.
For example, Turkey is run by an Islamist party that rose to power democratically, overcoming various secular elements of its constitution to do so. There are lots of other anecdotes to suggest that, while Turkey isn't as religious as the Arab world, it's still a fairly pious Muslim country. But we also know the country has a sharp divide, that the Turkish Republic has its roots in the hard secularism of Kemal Ataturk, and that Kemalist ideas still hold sway over a large swathe of the population, especially in the urban cores of Istanbul and other parts of Western Turkey. So what it feels like is that, with that sharp divide, the more pious part of Turkey's population was under-sampled here.
For the record, here's the description of the sampling:
"Participants were recruited from university student samples, personal networks and representative samples accessed by panel agencies and online platforms (MTurk, Kieskompas, Sojump, TurkPrime, Lancers, Qualtrics panels, Crowd-panel and Prolific)."
On the point about weird sample percentages, the data are not at all representative; rather, they are derived from convenience samples: https://www.thearda.com/data-archive?fid=MARP
So, this statement is a bit odd and not that helpful,
"Participants were recruited from university student samples, personal networks and representative samples accessed by panel agencies and online platforms (MTurk, Kieskompas, Sojump, TurkPrime, Lancers, Qualtrics panels, Crowd-panel and Prolific)."
Because Qualtrics, MTurk and Prolific are all survey providers that use panels. So it's not exactly a convenience sample. But did they recruit folks to take surveys on those platforms with a convenience approach?
The university student thing does make a lot of sense when it comes to this data, though.
"The countries were convenience sampled (i.e., through personal networks)."
Having participated on some of these platforms, I have been asked to forward the surveys I took to others like me. I can imagine other instances where the authors encouraged such sharing to other students in their courses and other similar courses. But yes, this does make a lot of sense when seeing some of the head-scratching percentages.
Thanks for bringing attention to another interesting data source.
Surprising to me how irreligious Israel is, despite the rapid growth of the Orthodox to the point that they now have to serve in the military.
The 1% well-being effect for 14% increase in religiosity might seem small with 100% well-being as the denominator. But the Center for Positive Psychology surveys suggest that only 25-30% of well being is circumstantial at all. With that denominator we'd be looking at a jump which is more in keeping with other happiness studies (where marriage and religiosity are consistently the most significant circumstantial factors).
That's a good bit of context - but I am continually frustrating with anything related to psychological self-assessment. Feels like you would get a different answer from folks if you asked them the same questions once a week for a year.
Yeah there's pitfalls in every direction. My favorite is how "the pursuit of happiness" reliably leads to misery.
What you're suggesting has been done. It's called the "ecological momentary assessment," and it's run more than weekly but often multiple times per day. It also has limitations, such as the need to look at your phone. This is one place where the findings that Americans are happier working than at home, that most U.S. parents hate child care, and that most happiness is non-circumstantial, came from.
The "Harvard Study of Adult Development" and Gallup's long-running annual satisfaction assessment both point in the same direction though. Personal connection with others (debatably including God) matter more than anything, and they matter a lot.
The one I find most interesting is the book "The Geography of Bliss" which concludes very little (cross-national self-assessments are a messy business) but does conclude that "knowing what's expected of you" in society does really matter.
It' striking to me how much noise there is in the subjective well being data. Those data points are all over the place leading to correlations so weak that I don't know that I'd even want to call them correlations for places like the USA, where the effect size (such as it is) is the strongest.
Ryan, given that the data for the “Many-Analysts Religion Project – Data on Religiosity and Well-being” report was collected in 2019, do you think the percentage of U.S. adults who identify as ‘Atheist’ or ‘Not Religious’ would be higher or lower in 2025 compared to 2019? Also, since the 56% cited in the report is significantly higher than the 28% typically reported for the religiously unaffiliated in the U.S., do you think this discrepancy is due to differences in methodology—such as self-reporting versus actual religious service attendance—or do you believe other factors are at play?
Not the main goal, but this provides me with lots of fodder to call out the Canadian tendency to assume that what takes place among American Christians is relevant here!
Interesting, Ryan. Question: How does the categories of "non-religious" and "religious" capture the ever-growing number of people who (at least in the West) define themselves as "spiritual but not religious" (SBNR)? Not sure it's possible to know, but so many people have moved themselves out of the "religious" category ... that said, I'm not sure they'd go so far as to categorize themselves as "non-religious." What's your take?
I'm looking forward to digging into this further (thanks for sharing all the links, by the way!).
Very intriguing analysis. A few obvious questions. One would be India, where two of their three large religions are not monotheistic. Not sure how polytheistic Hindus would believe in God or how Buddhists with no formal deity would create one. though each group has people who live by the rituals. The Islamsits there would be straighforward on this. Unfortunately the survey does not subdivide by religion. Parallel comments on China. The Communists discouraged formal religion, so it is expected that atheism would dominate. Again, there are pockets where Buddhism sort of thrives for individuals though under siege as public policy. Would be interesting to know if Buddhists have different well being than the residual Maoists or those younger Chinese now fully exposed to global pluralism, though still probably without a personal religion.
I see America and to a lesser extent Europe a little differently. In both places, affiliation with churches is on the decline. The survey, for better or worse, takes a snapshot, where are you religiously and with well-being right now, showing affiliation with religion in America has a minor correlation with well-being. I might politely challenge that snapshot format, at least in America where religious affiliation and practice is more fluid than other places. A 2x2 grid format might be more accurate. People who were always religious, never religious, defectors from their religion such as the Dechurched, and seekers of religion that they previously did not have. I would suspect that those four quadrants would have very different correlations of religion and personal inner peace.
1. It's a legitimate question about how Chinese Buddhists interpret the question, but Buddhism is tiny in India (although Hinduism was heavily influenced by it). There are more Christians than Buddhists in India.
2. Not an expert on Hinduism, but I know that it's characterized by a dizzying array of religious philosophies. As I understand it, plenty of Hindus are monotheists, or monists, or I suppose what we might call monolatrists or henotheists. I'm not entirely sure how a strict polytheist would interpret the question. But I recall from my college Greek translations that even the classical-era Greeks would refer to "God" (theos) all the time: (https://oxfordre.com/classics/display/10.1093/acrefore/9780199381135.001.0001/acrefore-9780199381135-e-6379).
Overall, this is really interesting, great work. In particular, it really highlights that India is a world unto itself.
But another one that stands out strongly to me is Turkey. In fact, it stands out so strongly that it makes me question the data there. From everything we know elsewhere, Turkey should really stand out as being a more religious country than Western Europe. But here it appears to be less religious than Belgium. Something isn't adding up.
For example, Turkey is run by an Islamist party that rose to power democratically, overcoming various secular elements of its constitution to do so. There are lots of other anecdotes to suggest that, while Turkey isn't as religious as the Arab world, it's still a fairly pious Muslim country. But we also know the country has a sharp divide, that the Turkish Republic has its roots in the hard secularism of Kemal Ataturk, and that Kemalist ideas still hold sway over a large swathe of the population, especially in the urban cores of Istanbul and other parts of Western Turkey. So what it feels like is that, with that sharp divide, the more pious part of Turkey's population was under-sampled here.
For the record, here's the description of the sampling:
"Participants were recruited from university student samples, personal networks and representative samples accessed by panel agencies and online platforms (MTurk, Kieskompas, Sojump, TurkPrime, Lancers, Qualtrics panels, Crowd-panel and Prolific)."
I passed that result by an expert in Turkish politics and religion and she was pretty dumbfounded by it, too.
On the point about weird sample percentages, the data are not at all representative; rather, they are derived from convenience samples: https://www.thearda.com/data-archive?fid=MARP
So, this statement is a bit odd and not that helpful,
"Participants were recruited from university student samples, personal networks and representative samples accessed by panel agencies and online platforms (MTurk, Kieskompas, Sojump, TurkPrime, Lancers, Qualtrics panels, Crowd-panel and Prolific)."
Because Qualtrics, MTurk and Prolific are all survey providers that use panels. So it's not exactly a convenience sample. But did they recruit folks to take surveys on those platforms with a convenience approach?
The university student thing does make a lot of sense when it comes to this data, though.
Adding to the confusion is this sentence:
"The countries were convenience sampled (i.e., through personal networks)."
Having participated on some of these platforms, I have been asked to forward the surveys I took to others like me. I can imagine other instances where the authors encouraged such sharing to other students in their courses and other similar courses. But yes, this does make a lot of sense when seeing some of the head-scratching percentages.
Thanks for bringing attention to another interesting data source.
Do you ever think that the evangelical emphasis on Christianity as “not a religion, but a relationship” could impact these surveys in any way?
Yeah, but probably nowhere outside the United States. Americans just have a different conception of religion than anywhere else on Earth.
How much does this affect the “religious” rate in the US do you think?
Surprising to me how irreligious Israel is, despite the rapid growth of the Orthodox to the point that they now have to serve in the military.
The 1% well-being effect for 14% increase in religiosity might seem small with 100% well-being as the denominator. But the Center for Positive Psychology surveys suggest that only 25-30% of well being is circumstantial at all. With that denominator we'd be looking at a jump which is more in keeping with other happiness studies (where marriage and religiosity are consistently the most significant circumstantial factors).
That's a good bit of context - but I am continually frustrating with anything related to psychological self-assessment. Feels like you would get a different answer from folks if you asked them the same questions once a week for a year.
Yeah there's pitfalls in every direction. My favorite is how "the pursuit of happiness" reliably leads to misery.
What you're suggesting has been done. It's called the "ecological momentary assessment," and it's run more than weekly but often multiple times per day. It also has limitations, such as the need to look at your phone. This is one place where the findings that Americans are happier working than at home, that most U.S. parents hate child care, and that most happiness is non-circumstantial, came from.
The "Harvard Study of Adult Development" and Gallup's long-running annual satisfaction assessment both point in the same direction though. Personal connection with others (debatably including God) matter more than anything, and they matter a lot.
The one I find most interesting is the book "The Geography of Bliss" which concludes very little (cross-national self-assessments are a messy business) but does conclude that "knowing what's expected of you" in society does really matter.
Fantastic analysis, Ryan.
It' striking to me how much noise there is in the subjective well being data. Those data points are all over the place leading to correlations so weak that I don't know that I'd even want to call them correlations for places like the USA, where the effect size (such as it is) is the strongest.
Ryan, given that the data for the “Many-Analysts Religion Project – Data on Religiosity and Well-being” report was collected in 2019, do you think the percentage of U.S. adults who identify as ‘Atheist’ or ‘Not Religious’ would be higher or lower in 2025 compared to 2019? Also, since the 56% cited in the report is significantly higher than the 28% typically reported for the religiously unaffiliated in the U.S., do you think this discrepancy is due to differences in methodology—such as self-reporting versus actual religious service attendance—or do you believe other factors are at play?
Not the main goal, but this provides me with lots of fodder to call out the Canadian tendency to assume that what takes place among American Christians is relevant here!
Interesting, Ryan. Question: How does the categories of "non-religious" and "religious" capture the ever-growing number of people who (at least in the West) define themselves as "spiritual but not religious" (SBNR)? Not sure it's possible to know, but so many people have moved themselves out of the "religious" category ... that said, I'm not sure they'd go so far as to categorize themselves as "non-religious." What's your take?
I'm looking forward to digging into this further (thanks for sharing all the links, by the way!).
Working on a big project about spirituality vs religion right now. Huge Templeton grant - we surveyed over 15,000 folks.
Book writing will start in the new year.
Very intriguing analysis. A few obvious questions. One would be India, where two of their three large religions are not monotheistic. Not sure how polytheistic Hindus would believe in God or how Buddhists with no formal deity would create one. though each group has people who live by the rituals. The Islamsits there would be straighforward on this. Unfortunately the survey does not subdivide by religion. Parallel comments on China. The Communists discouraged formal religion, so it is expected that atheism would dominate. Again, there are pockets where Buddhism sort of thrives for individuals though under siege as public policy. Would be interesting to know if Buddhists have different well being than the residual Maoists or those younger Chinese now fully exposed to global pluralism, though still probably without a personal religion.
I see America and to a lesser extent Europe a little differently. In both places, affiliation with churches is on the decline. The survey, for better or worse, takes a snapshot, where are you religiously and with well-being right now, showing affiliation with religion in America has a minor correlation with well-being. I might politely challenge that snapshot format, at least in America where religious affiliation and practice is more fluid than other places. A 2x2 grid format might be more accurate. People who were always religious, never religious, defectors from their religion such as the Dechurched, and seekers of religion that they previously did not have. I would suspect that those four quadrants would have very different correlations of religion and personal inner peace.
1. It's a legitimate question about how Chinese Buddhists interpret the question, but Buddhism is tiny in India (although Hinduism was heavily influenced by it). There are more Christians than Buddhists in India.
2. Not an expert on Hinduism, but I know that it's characterized by a dizzying array of religious philosophies. As I understand it, plenty of Hindus are monotheists, or monists, or I suppose what we might call monolatrists or henotheists. I'm not entirely sure how a strict polytheist would interpret the question. But I recall from my college Greek translations that even the classical-era Greeks would refer to "God" (theos) all the time: (https://oxfordre.com/classics/display/10.1093/acrefore/9780199381135.001.0001/acrefore-9780199381135-e-6379).
How much do you enjoy your life?
To what extent do you feel your life to be meaningful?
How well are you able to concentrate?
Are you able to accept your bodily appearance?
How satisfied are you with yourself?
How often do you have negative feelings such as blue mood, despair, anxiety, depression?