I think that last bit of analysis was extra telling. I think it dovetails Hank Green's razor: "Anything that could possibly be influenced by socio-economic status, probably is" with your previously stated theory that going to church/being religious is something for already successful people in the US.
I've been thinking about writing this comment for a while. I have been seeing your posts for some time showing the number of nones who have high-school education or less. I don't have any statistics or surveys to cite; what I am about to say is based on my observations over 75 years of life (and for most of that time, I was a weekly attender). And I have lived in rural areas, city neighborhoods, and suburbs. I strongly suspect that the departure of the less-educated from the churches is the fault of the clergy of my own generation and later.
Before WWII, college was not a common thing for most people. That began to change with the passage of the GI Bill, allowing veterans to go to college with the government paying their tuition. (My own father didn't qualify for that--when called up for the draft after Pearl Harbor, he flunked the physical; he spent the war working in a factory making radios for the military.) Before the GI Bill, in most places, the only ones most people knew who had gone to college were teachers, lawyers, doctors, and ministers. It might have been different in Manhattan and wealthy neighborhoods, but that was how it was in most of the country. So local pastors, by necessity, had to get along with less-educated people. Older ministers I knew, men who were in their 60s when I was in my 30s, mostly did quite well at it.
Even after WWII, the separation was not as great. The GI Bill graduates and the blue-collar people lived in the same neighborhoods, sent their kids to the same schools, shopped at the same stores, and often went to the same churches. But as housing costs began to rise in the '70s and '80s, separation rose with it. I saw ministers of my own generation who tended to associate with the educated people in their congregations, and had less time for the blue-collar ones. Having more college-educated people in their churches made that possible. I suspect that may not have been possible in the church you served for years; but in the city and suburban churches I saw, it would have been easy for them. And in plain English: if you give people the impression you don't care about them, they may just drift off and not come back. Snobbery is a very human trait throughout history, but it will be damaging for a church. And after all these years since WWII, the majority of working people in the US do not have college degrees.
As I said, I don't have any surveys or statistics to back up my opinion, and I have no idea where you could find any. My background has influenced my thinking on this: I do have a college degree, but I am of Appalachian stock on both sides of my family. Both of my parents grew up poor during the Depression. My father was the first in his family to finish high school; I was the first to finish college.
Commenting on a Pew survey on levels of education and religious belief and practice, Emma Green, “Why Educated Christians are Sticking with Church,” Atlantic Monthly, 4/26/2017
“Among Christians, the pattern of educated people being more involved in their religious communities makes sense. As I’ve written before [https://www.theatlantic.com/politics/archive/2017/03/religiously-unaffiliated-white-americans/518340/], communal involvement of all kinds is increasingly becoming a luxury good of sorts, with higher levels of income and education making people more likely to participate in activities like church, book club, parent-teacher association, and more. It could be that high-school-educated Christians feel less able to find and connect with a religious community in a broader context of financial strain, family stress, and geographic isolation. Or it could be that college-educated Christians put more of a premium on connecting with their brothers and sisters in the church.”
Your analysis of income levels suggests that the ‘luxury goods of sorts’ hypothesis is on the right track at least for US whites and African Americans.
Does degree of access to online social media also track with income levels? Is online contact regarded as a poor substitute for the contact afforded by attendance?
A significant percentage of Asians belong to very different religions from other groups and they don't function in the same way. You can be a Hindu or Buddhist and never attend temple at all. You can have a worship room for puja in your home.
But a Hindu or Buddhist wouldn't identify as "Atheist/Agnostic/Nothing in Particular", would they? I'd expect them to identify as that (or "Something Else"). I guess maybe there's some that would opt for Nothing in particular.
My low confidence guess WRT Asians is the sample is being thrown off by Chinese, both FOB and ABC, who, having come from a Communist country, really don't have a religion, and tend to be a pretty high performing group in the education attainment level.
I'd be curious if this pattern holds for a long time and if you see it holding the 3rd gen, and how it would compare with pre-1949 Chinese immigration
Also this entire question is a reminder of what a cursed census category "Asian" really is; are we talking Chinese? Indonesians? Indians? Uzbeks? All of these are on some level "Asian"!
A Chinese, Japanese, Korean or Vietnamese Folk Religion person absolutely does and it's normative for them to identify as a None even if they're a full on polytheist just because it isn't an "official category" like being a Buddhist.
I know this because I was raised in this tradition. I always write in Chinese folk religion just so that I can be counted on a survey. Most Chinese people in the US came here after 2005, FWIW so the "3rd gen" people barely exist.
And as for attendance, you don't have to attend temple ever to be a good Hindu.
I'm surprised you didn't separate atheist/agnostic from "nothing in particular". I'd guess that the second group fully accounts for the trend in belief, and that the underlying factor is social disengagement, also reflected in church attendance. That is, at the top end of the education spectrum I'd expect lots of atheists/agnostics, lots of church-attending believers, and fewer non-attending (stated) believers and nothing in particulars.
Interesting point! I'd be curious to see this as well. I've always heard that atheism/agnostic increases with attendance, and was surprised this survey seemed to be show complete opposite.
It makes intuitive sense that higher education leads to less spiritual indecision/apathy. Burge's previous surveys have shown that the NINOs are the biggest bloc of "nones" right now, so it also makes sense that these people becoming more active believers would skew the survey results.
Edit: Another recent post (https://www.graphsaboutreligion.com/p/the-nones-project-ninos) seems to track with this. NINOs are the biggest sub-group of nones with HS or less, by a wide margin, and are then the smallest group in both 4yr degrees or grad school.
Proverbs 22:6 says, "Train up a child in the way he should go; even when he is old he will not depart from it."
I think parents are the single biggest factor into not only if their child becomes a Christian, but remains an active Christian later in life. Also, (and this is obviously just a generalization) educated parents want educated children. Educated parents are also more able to help their children financially through college. So, (again, in general) parents who care both about their child's spiritual and earthly education will make sure they get it.
On the other hand, high school dropouts usually have parents who did not place a high value on education. Most likely, they didn't take their child to church, either.
Of course, there are always exceptions in both cases, but this is the first thing that popped into my head while reading this article.
You think correctly - Barna and others have repeatedly found that, more than leadership, modern vs. older forms of worship, more than church size, the parents faith is the single biggest predictor of children's faith.
(I'll come back to edit this comment when I can find the source.)
The part below 'Onderwijsniveau' (= Education level) is what is important. Google translate:
(begin quote)
"Religious affiliation decreases with educational level. Of those who have completed only primary education, 61 percent consider themselves a member of a religious denomination or philosophy of life, compared to 34 percent of university graduates. Here too, there are differences by religion. Among Catholics, three distinct groups can be distinguished. Of those who have completed primary or pre-vocational secondary education (VMBO), over a quarter are Catholic; of those with secondary or higher professional education, this is approximately 18 percent, and of university graduates, 13 percent. Among Protestants, on the other hand, there is hardly any variation: at almost all levels of education, approximately 16 percent are Protestant. University graduates are an exception, with 11 percent.
Of the group with only primary education, 12 percent are Muslim. This share is significantly higher than at other levels of education, where it is 3 to 5 percent. Within the "other denomination" group, there is little difference by educational level.
When it comes to regular attendance at religious services, there's a three-way split. Of the two groups with the lowest educational levels, almost one in five attends a prayer service at least once a month; among those with secondary and higher professional education, this figure is one in eight, and among university graduates, one in ten"
Hi, Trev! Ryan does not do much analysis of religion outside of the USA, but he did use European Social Survey (ESS) data to query some of the same questions for Europe (including the Netherlands) as he does here. Here's that post:
Inspired by Ryan's research, I used Statistics Netherlands (CBS) data and ESS data to look at what relationship exists between religiosity and educational level both in the European Netherlands and in the Caribbean Netherlands, which is where I live. Short summary: There is no relationship between church attendance and educational level in the European Netherlands (where, uh, basically no one goes to church, of any educational level), but the Caribbean Netherlands has an American pattern of a positive relationship between educational attainment and religiosity. Also way more people are religious in the Caribbean Netherlands compared to the European Netherlands. Then I talk about some historic and demographic features of the Caribbean Netherlands that might explain those differences. Here's a link to this essay of mine:
You measure religiosity by how many times people are actually going to church and yes, basically here in the European Netherlands virtually no one goes.
For catholics that's certainly true, here are numbers from Kaski, an institute that's part of the Dutch Radboud University. Dutch link
Not sure but I guess Statistics Netherlands doesn't use that but instead uses some other question to measure religiosity?
If you haven't already found it, Nature Communications a few weeks ago had an article about religiosity all over the world measured by participation, identification and belonging:
I guess church visit is a question about participation and Statistics Netherlands uses a question to measure if people identify as a Christian or belong to christianity.
Statistics Netherlands measures religiosity two ways: by self-identified religion and by religious attendance. I wanted to look at religious attendance rather than just self-identified religion, so I could compare to Ryan's work. However, CBS queries "what's your religion" much more frequently than "how often do you attend religious services." That's why I had to go to ESS data to query religiosity per religious attendance; CBS hadn't asked that question in the European Netherlands since 2009. ESS doesn't ask anything about the Caribbean Netherlands (presumably because of the "Caribbean" part), so there I used CBS data. CBS only asks fine-grained behavioral questions like that of residents of the Caribbean part of the Kingdom in the Omnibus survey that's done about every 7 - 8 years--so the data I have for church attendance in the Caribbean Netherlands is from 2018. (The next Omnibus survey is planned for this year.)
Thank you very much for the links. I hadn't seen either of these, and they're both quite interesting. One gripe I have about the Pew/Nature writeup is that it assumes that there's a "secular progression" that inevitably will happen to all countries as they develop more--that everywhere follows the ideologic trajectory of the West, and it's just lack of education/wealth that explains why cultural beliefs that secular Westerners don't like persist in other places. I do not think the data supports this assumption. If you're not already familiar, the sociologist Alice Evans at The Great Gender Divergence has interrogated that idea (through the lens of gender equality, not religion) at her Substack, here: https://www.ggd.world/
Thank you for re-posting this link. I had not scrolled over to column 3.1, so I hadn't seen this data when you'd posted it before. Now I did, and took this opportunity to read through a bunch of the rest of this report, too.
It looks like CBS is defining "regular church attendance" as once per month. ESS has a more fine-grained analysis for their 2023 Dutch data, similar to the US data that Ryan analyzes, and similar to CBS Omnibus survey from 2018, allowing us to look at church attendance as frequently as >=1x per week. I wonder if the 2023 CBS data doesn't look at church attendance >=1x per week because that would be so few people, so 1x per month for "regular attendance" is about the best you can do.
It's also interesting to me how CBS has changed their questions over time to capture more religious diversity--and that their changed questioning approach led to more people being identified as religious compared to their earlier approach. I'm fascinated that even in the hyper-secular Netherlands more people have some belief in God/a higher power than those who don't. I'm also fascinated that based on these definitions, there can be conclusions like "10% of Dutch Catholics do not believe that God exists," which is, um, not my understanding of Catholic doctrine.
I, too, am learning new things today. So thanks again!
I've been following this subject for some time now.
I'm not a sociologist or more in general particularly good at statistics, but I think CBS / Statistics Netherlands has a methodogical problems they need to deal with.
For quite some time CBS and SCP (Sociaal Cultureel Planbureau) used a DIFFERENT question to measure religiosity. One of the two (not sure which) asked the question if people were religious, and then asked which religion they were in. The other however just asked one question mentioning different belief systems where one of the possible answers was 'no religion' or something like that. Here's a link on how CBS tried to solve the issue and the results they got.
An easier to solve issue was the following. In 2004 three Dutch churches merged into what is now the Protestantse Kerk Nederland (PKN). They were the Hervormde kerk, the Gereformeerde Kerken in Nederland and the Evangelisch-Lutherse Kerk. The problem is a lot of members of those churches still identify as eg member of the Hervormde Kerk instead of PKN. CBS decided to keep the 'outdated' church names in the questionnaire and just add the numbers to PKN.
They explain this in the link, below the table that's presented.
It is painfully obvious that what is being captured in these data isn't that people with more education are more religious. What is being captured is that the more religious a person's family is, the more stable it is, the more likely it is that people from that environment will be able to succeed. This isn't a case that the the most educated are religious. It is likely that, because of religion's benefits to families and their finances, the most religious are the ones most likely to value education, and the most able to afford it.
Very interesting analysis, but I would like to see the statistics behind the findings — sample size and significance, control of other variables (e.g., age), … etc.
I think that last bit of analysis was extra telling. I think it dovetails Hank Green's razor: "Anything that could possibly be influenced by socio-economic status, probably is" with your previously stated theory that going to church/being religious is something for already successful people in the US.
I like how you take us on a narrative trip through the possibilities in your pieces like this. Love the twists and turns.
I've been thinking about writing this comment for a while. I have been seeing your posts for some time showing the number of nones who have high-school education or less. I don't have any statistics or surveys to cite; what I am about to say is based on my observations over 75 years of life (and for most of that time, I was a weekly attender). And I have lived in rural areas, city neighborhoods, and suburbs. I strongly suspect that the departure of the less-educated from the churches is the fault of the clergy of my own generation and later.
Before WWII, college was not a common thing for most people. That began to change with the passage of the GI Bill, allowing veterans to go to college with the government paying their tuition. (My own father didn't qualify for that--when called up for the draft after Pearl Harbor, he flunked the physical; he spent the war working in a factory making radios for the military.) Before the GI Bill, in most places, the only ones most people knew who had gone to college were teachers, lawyers, doctors, and ministers. It might have been different in Manhattan and wealthy neighborhoods, but that was how it was in most of the country. So local pastors, by necessity, had to get along with less-educated people. Older ministers I knew, men who were in their 60s when I was in my 30s, mostly did quite well at it.
Even after WWII, the separation was not as great. The GI Bill graduates and the blue-collar people lived in the same neighborhoods, sent their kids to the same schools, shopped at the same stores, and often went to the same churches. But as housing costs began to rise in the '70s and '80s, separation rose with it. I saw ministers of my own generation who tended to associate with the educated people in their congregations, and had less time for the blue-collar ones. Having more college-educated people in their churches made that possible. I suspect that may not have been possible in the church you served for years; but in the city and suburban churches I saw, it would have been easy for them. And in plain English: if you give people the impression you don't care about them, they may just drift off and not come back. Snobbery is a very human trait throughout history, but it will be damaging for a church. And after all these years since WWII, the majority of working people in the US do not have college degrees.
As I said, I don't have any surveys or statistics to back up my opinion, and I have no idea where you could find any. My background has influenced my thinking on this: I do have a college degree, but I am of Appalachian stock on both sides of my family. Both of my parents grew up poor during the Depression. My father was the first in his family to finish high school; I was the first to finish college.
Commenting on a Pew survey on levels of education and religious belief and practice, Emma Green, “Why Educated Christians are Sticking with Church,” Atlantic Monthly, 4/26/2017
[https://www.theatlantic.com/politics/archive/2017/04/education-church-attendance/524346/] concludes:
“Among Christians, the pattern of educated people being more involved in their religious communities makes sense. As I’ve written before [https://www.theatlantic.com/politics/archive/2017/03/religiously-unaffiliated-white-americans/518340/], communal involvement of all kinds is increasingly becoming a luxury good of sorts, with higher levels of income and education making people more likely to participate in activities like church, book club, parent-teacher association, and more. It could be that high-school-educated Christians feel less able to find and connect with a religious community in a broader context of financial strain, family stress, and geographic isolation. Or it could be that college-educated Christians put more of a premium on connecting with their brothers and sisters in the church.”
Your analysis of income levels suggests that the ‘luxury goods of sorts’ hypothesis is on the right track at least for US whites and African Americans.
Does degree of access to online social media also track with income levels? Is online contact regarded as a poor substitute for the contact afforded by attendance?
I think you have if backwards. It is because of religion, and it's effect on the stability of families, that people are able to succeed.
I'll see if I can find relevant research on the direction of causation and welcome pertinent citations.
A significant percentage of Asians belong to very different religions from other groups and they don't function in the same way. You can be a Hindu or Buddhist and never attend temple at all. You can have a worship room for puja in your home.
But a Hindu or Buddhist wouldn't identify as "Atheist/Agnostic/Nothing in Particular", would they? I'd expect them to identify as that (or "Something Else"). I guess maybe there's some that would opt for Nothing in particular.
My low confidence guess WRT Asians is the sample is being thrown off by Chinese, both FOB and ABC, who, having come from a Communist country, really don't have a religion, and tend to be a pretty high performing group in the education attainment level.
I'd be curious if this pattern holds for a long time and if you see it holding the 3rd gen, and how it would compare with pre-1949 Chinese immigration
Also this entire question is a reminder of what a cursed census category "Asian" really is; are we talking Chinese? Indonesians? Indians? Uzbeks? All of these are on some level "Asian"!
A Chinese, Japanese, Korean or Vietnamese Folk Religion person absolutely does and it's normative for them to identify as a None even if they're a full on polytheist just because it isn't an "official category" like being a Buddhist.
I know this because I was raised in this tradition. I always write in Chinese folk religion just so that I can be counted on a survey. Most Chinese people in the US came here after 2005, FWIW so the "3rd gen" people barely exist.
And as for attendance, you don't have to attend temple ever to be a good Hindu.
I'm surprised you didn't separate atheist/agnostic from "nothing in particular". I'd guess that the second group fully accounts for the trend in belief, and that the underlying factor is social disengagement, also reflected in church attendance. That is, at the top end of the education spectrum I'd expect lots of atheists/agnostics, lots of church-attending believers, and fewer non-attending (stated) believers and nothing in particulars.
Interesting point! I'd be curious to see this as well. I've always heard that atheism/agnostic increases with attendance, and was surprised this survey seemed to be show complete opposite.
It makes intuitive sense that higher education leads to less spiritual indecision/apathy. Burge's previous surveys have shown that the NINOs are the biggest bloc of "nones" right now, so it also makes sense that these people becoming more active believers would skew the survey results.
Edit: Another recent post (https://www.graphsaboutreligion.com/p/the-nones-project-ninos) seems to track with this. NINOs are the biggest sub-group of nones with HS or less, by a wide margin, and are then the smallest group in both 4yr degrees or grad school.
Proverbs 22:6 says, "Train up a child in the way he should go; even when he is old he will not depart from it."
I think parents are the single biggest factor into not only if their child becomes a Christian, but remains an active Christian later in life. Also, (and this is obviously just a generalization) educated parents want educated children. Educated parents are also more able to help their children financially through college. So, (again, in general) parents who care both about their child's spiritual and earthly education will make sure they get it.
On the other hand, high school dropouts usually have parents who did not place a high value on education. Most likely, they didn't take their child to church, either.
Of course, there are always exceptions in both cases, but this is the first thing that popped into my head while reading this article.
You think correctly - Barna and others have repeatedly found that, more than leadership, modern vs. older forms of worship, more than church size, the parents faith is the single biggest predictor of children's faith.
(I'll come back to edit this comment when I can find the source.)
"Those with the most education are the least likely to identify as atheist, agnostic, or nothing in particular."
This came as a huge surprise...
Pretty sure you're capturing conscientiousness personality trait and the correlation is spurious.
Here in the Netherlands religiosity is low. But it's fascinating to see how higher degrees here mean less (not more) religious involvement.
The below link is from Statistics Netherlands, a government agency that keeps score of a LOT of things. Link is in Dutch, results are from 2020.
https://www.cbs.nl/nl-nl/longread/statistische-trends/2020/religie-in-nederland/3-religieuze-betrokkenheid-naar-achtergrondkenmerken
The part below 'Onderwijsniveau' (= Education level) is what is important. Google translate:
(begin quote)
"Religious affiliation decreases with educational level. Of those who have completed only primary education, 61 percent consider themselves a member of a religious denomination or philosophy of life, compared to 34 percent of university graduates. Here too, there are differences by religion. Among Catholics, three distinct groups can be distinguished. Of those who have completed primary or pre-vocational secondary education (VMBO), over a quarter are Catholic; of those with secondary or higher professional education, this is approximately 18 percent, and of university graduates, 13 percent. Among Protestants, on the other hand, there is hardly any variation: at almost all levels of education, approximately 16 percent are Protestant. University graduates are an exception, with 11 percent.
Of the group with only primary education, 12 percent are Muslim. This share is significantly higher than at other levels of education, where it is 3 to 5 percent. Within the "other denomination" group, there is little difference by educational level.
When it comes to regular attendance at religious services, there's a three-way split. Of the two groups with the lowest educational levels, almost one in five attends a prayer service at least once a month; among those with secondary and higher professional education, this figure is one in eight, and among university graduates, one in ten"
(end quote)
Hi, Trev! Ryan does not do much analysis of religion outside of the USA, but he did use European Social Survey (ESS) data to query some of the same questions for Europe (including the Netherlands) as he does here. Here's that post:
https://www.graphsaboutreligion.com/p/does-education-have-the-same-impact
Inspired by Ryan's research, I used Statistics Netherlands (CBS) data and ESS data to look at what relationship exists between religiosity and educational level both in the European Netherlands and in the Caribbean Netherlands, which is where I live. Short summary: There is no relationship between church attendance and educational level in the European Netherlands (where, uh, basically no one goes to church, of any educational level), but the Caribbean Netherlands has an American pattern of a positive relationship between educational attainment and religiosity. Also way more people are religious in the Caribbean Netherlands compared to the European Netherlands. Then I talk about some historic and demographic features of the Caribbean Netherlands that might explain those differences. Here's a link to this essay of mine:
https://doctrixperiwinkle.substack.com/p/the-smart-set
You measure religiosity by how many times people are actually going to church and yes, basically here in the European Netherlands virtually no one goes.
For catholics that's certainly true, here are numbers from Kaski, an institute that's part of the Dutch Radboud University. Dutch link
https://www.ru.nl/over-ons/nieuws/kerncijfers-rooms-katholieke-kerk-2024
Not sure but I guess Statistics Netherlands doesn't use that but instead uses some other question to measure religiosity?
If you haven't already found it, Nature Communications a few weeks ago had an article about religiosity all over the world measured by participation, identification and belonging:
https://www.pewresearch.org/short-reads/2025/09/02/how-religion-declines-around-the-world/
I guess church visit is a question about participation and Statistics Netherlands uses a question to measure if people identify as a Christian or belong to christianity.
Hi again, Trev, and thanks for reading!
Statistics Netherlands measures religiosity two ways: by self-identified religion and by religious attendance. I wanted to look at religious attendance rather than just self-identified religion, so I could compare to Ryan's work. However, CBS queries "what's your religion" much more frequently than "how often do you attend religious services." That's why I had to go to ESS data to query religiosity per religious attendance; CBS hadn't asked that question in the European Netherlands since 2009. ESS doesn't ask anything about the Caribbean Netherlands (presumably because of the "Caribbean" part), so there I used CBS data. CBS only asks fine-grained behavioral questions like that of residents of the Caribbean part of the Kingdom in the Omnibus survey that's done about every 7 - 8 years--so the data I have for church attendance in the Caribbean Netherlands is from 2018. (The next Omnibus survey is planned for this year.)
Thank you very much for the links. I hadn't seen either of these, and they're both quite interesting. One gripe I have about the Pew/Nature writeup is that it assumes that there's a "secular progression" that inevitably will happen to all countries as they develop more--that everywhere follows the ideologic trajectory of the West, and it's just lack of education/wealth that explains why cultural beliefs that secular Westerners don't like persist in other places. I do not think the data supports this assumption. If you're not already familiar, the sociologist Alice Evans at The Great Gender Divergence has interrogated that idea (through the lens of gender equality, not religion) at her Substack, here: https://www.ggd.world/
I learn something new every day. Thanks for the link. :)
For the record: at the link I gave above
https://www.cbs.nl/nl-nl/longread/statistische-trends/2020/religie-in-nederland/3-religieuze-betrokkenheid-naar-achtergrondkenmerken
Statistics Netherlands also gives numbers for people visiting church by education level.
It's the last column in the table "3.1 Religieuze betrokkenheid naar achtergrondkenmerken, personen van 15 jaar of ouder, 2019 (%)"
Of those with only primary education 19.5% visits church, of those with a masters of doctoral 9.8%. These are 2019 numbers.
Thank you for re-posting this link. I had not scrolled over to column 3.1, so I hadn't seen this data when you'd posted it before. Now I did, and took this opportunity to read through a bunch of the rest of this report, too.
It looks like CBS is defining "regular church attendance" as once per month. ESS has a more fine-grained analysis for their 2023 Dutch data, similar to the US data that Ryan analyzes, and similar to CBS Omnibus survey from 2018, allowing us to look at church attendance as frequently as >=1x per week. I wonder if the 2023 CBS data doesn't look at church attendance >=1x per week because that would be so few people, so 1x per month for "regular attendance" is about the best you can do.
It's also interesting to me how CBS has changed their questions over time to capture more religious diversity--and that their changed questioning approach led to more people being identified as religious compared to their earlier approach. I'm fascinated that even in the hyper-secular Netherlands more people have some belief in God/a higher power than those who don't. I'm also fascinated that based on these definitions, there can be conclusions like "10% of Dutch Catholics do not believe that God exists," which is, um, not my understanding of Catholic doctrine.
I, too, am learning new things today. So thanks again!
I've been following this subject for some time now.
I'm not a sociologist or more in general particularly good at statistics, but I think CBS / Statistics Netherlands has a methodogical problems they need to deal with.
For quite some time CBS and SCP (Sociaal Cultureel Planbureau) used a DIFFERENT question to measure religiosity. One of the two (not sure which) asked the question if people were religious, and then asked which religion they were in. The other however just asked one question mentioning different belief systems where one of the possible answers was 'no religion' or something like that. Here's a link on how CBS tried to solve the issue and the results they got.
https://www.cbs.nl/nl-nl/longread/statistische-trends/2019/religie-en-sociale-cohesie/2-vraagstellingen-religie
An easier to solve issue was the following. In 2004 three Dutch churches merged into what is now the Protestantse Kerk Nederland (PKN). They were the Hervormde kerk, the Gereformeerde Kerken in Nederland and the Evangelisch-Lutherse Kerk. The problem is a lot of members of those churches still identify as eg member of the Hervormde Kerk instead of PKN. CBS decided to keep the 'outdated' church names in the questionnaire and just add the numbers to PKN.
They explain this in the link, below the table that's presented.
https://www.cbs.nl/nl-nl/cijfers/detail/83288NED/
Both examples show the pitfalls we might (and sometimes do) fall into. :)
It is painfully obvious that what is being captured in these data isn't that people with more education are more religious. What is being captured is that the more religious a person's family is, the more stable it is, the more likely it is that people from that environment will be able to succeed. This isn't a case that the the most educated are religious. It is likely that, because of religion's benefits to families and their finances, the most religious are the ones most likely to value education, and the most able to afford it.
Very interesting analysis, but I would like to see the statistics behind the findings — sample size and significance, control of other variables (e.g., age), … etc.