Tuesday, December 30, 2008

Religious denominations ranked by making most of what is given

A paper by Helmuth Nyborg tracking white adolescents by religious denominational membership, IQ, and income by way of the NLSY, reveals that people who belong to more dogmatic religious traditions tend to be less intelligent than those who belong to more liberal traditions are, with atheists and agnostics falling on the higher-IQ end of the liberal spectrum.

That isn't surprising. Adherence to fantastic, empirically unverifiable beliefs does not denote intelligence. The more literal and less metaphorical those beliefs become, the less they are going to appeal to intelligent people.

That religiosity and intelligence are inversely correlated obfuscates the potential value of religion by suggesting that religious belief causes a drop in IQ. Half Sigma suggested as much a couple of years ago:

Doesn’t anyone think that it’s at least as important to tell us that religion also brings stupidity? I guess not.
IQ stabilizes around age seven. I am unaware of data suggesting that an increase in religiosity corresponds with measured IQ over the course of an individual's life. So I am comfortable in presuming that to the extent the two are related, the causation arrow points from intelligence to religiosity.

From this perspective, I find religion potentially beneficial, for those of modest intelligence anyway. If some belief system is going to be uncritically followed, better for it to come from Billy Graham than from Us Weekly.

Nyborg's paper allows for an evaluation, by way of household income, of how members of various denominations fare relative to their average IQ levels. Unfortunately, he pulls the income data from the households of the teenagers, so the comparisons made below are based on the income of adults and the IQ and beliefs of their children (which are presumably mostly shared with that of the parents). Some of these children will have switched to different denominations upon becoming adults, just as some of their parents have brought the teenagers in the study up in a different denomination than they were raised in themselves. A study out of the University of Chicago finds that around two-thirds of Protestants, more than four in five Catholics, and nearly nine in ten Jews retain the same religious tradition they were raised in. Consequently, the table should only be seen as suggestive. If Nyborg had tracked adults in both cases, it would be superior.

Anyhow, following is an index created by subtracting a denomination's average household income (adjusted for sample size) in comparison with all groups, by standard deviation, subtracting it from the denomination's average IQ in the same way, and multiplying by 100 for ease of viewing. So a group that is .5 SDs above the mean in income and .3 SDs below the mean in IQ scores an 80 ((.5- (-.3))*100)--given their modest intelligence, they are pretty economically successful. Also included for reference are marriage-plus-cohabitation rates as reported by Pew's US Religious Landscape survey:

1. Jewish69.663%
2. Mormon31.274%
3. Presbyterian16.765%
4. Disciple of Christ12.658%
5. Baptist8.365%
6. Roman Catholic5.865%
7. Personal philosophy3.458%
8. Unified Ch. of Christ0.758%
9. Other religion(0.3)60%
10. Methodist(0.4)66%
11. Lutheran(6.7)66%
12. Episcopalian/Anglican(9.7)60%
13. Holiness(12.0)64%
14. Pentecostal(13.3)64%
15. Bible Church(13.5)67%
16. Agnostic(16.3)51%
17. Muslim(17.2)60%
18. Other Protestant(30.0)58%
19. Atheist(36.9)50%

Even with the second highest IQ scores of the 19 groups measured, Jewish incomes far surpass what intelligence alone predicts. This suggests the story behind Jewish success is more than the history of Ashkenazi intelligence alone. The stock explanation that Jews put a great deal of emphasis on learning and focus on scholarship comes to mind. So do Bernie Madoff's investors. It would be interesting to see a full study with adjustments for IQ comparing the life outcome of Jews with goyim. It should also be noted that Nyborg made no cost of living adjustment for incomes, so Jewish population concentration in the relatively expensive Northeast causes some score inflation.

Mormons shine by this measure as they do by many others. A remark by a medical student and Steve Sailer reader seems right on the money:

I don't think Mormons in general have higher IQ's, but I think that since they are not allowed to drink, smoke, party, gamble, or any other fun stuff, they are all ultra productive, and even the mediocre ones are able to channel their hard work into success.
Mormons are the least likely of the 19 denominations to live alone, but I suspect among the married, they are among the most likely to have a single breadwinner household.

Atheists and agnostics, by contrast, come in at the bottom. The low rates of multiple person households is part of the explanation, but the high number of lone wolves among their ranks illustrates their social marginality in another way relative to the cognitive endowments they enjoy. This does little to dispel stereotype I hold of atheists as cynical, single white guys who live in apartments downtown, work at used record stores, love George Carlin, and watch Adult Swim.

Nyborg's paper might give smug atheists justification for superciliousness, but unless the argument is that atheism increases intelligence, there isn't much to brag about, save being more worthy in the eyes of Socrates. They don't achieve as much as believers do given the hand they're dealt.

Speculatively, among Christian groups, those devoting relatively less attention to the Gospels and more to the entire Bible (Presbyterians and Catholics) and that also put greater relative emphasis on what is done by the individual in this world do a little better than others like Lutherans and Methodists.

Parenthetically, Razib uses the GSS to extensively compare Episcopalians and Jews, groups consistently found to perform similarly on IQ tests, averaging around two-thirds of one standard deviation above the white mean. Does Jewish liberalism relate to the group's material advantages over Episcopalians in the secular world?

The data, via Swivel.

* For lack of better alternative, marriage plus cohabitation for "Other religion" comes from Pew's "Unaffiliated, Religious" category. Pew's findings for the Presbyterian Church of the USA are used with Presbyterian in the table, since Nyborg classifies Presbyterianism as liberal. Pew's findings for ELCA members are used with Lutheran above for the same reason. The Bible Church marriage-plus-cohabitation percentage comes from Pew's findings on historically evangelical independent Baptist churches. Nyborg's Other Protestant rates come from Pew's findings on "Unitarians and other liberal faiths". The rates for both Pentecostal and Holiness are derived by averaging Pew's findings on Assemblies of God and Church of Christ, as it is unclear what parameters Nyborg uses to classify Pentecostals, and Pew does not report data specifically on the International Pentecostal Holiness Church.


Anonymous said...

"This does little to dispel stereotype I hold of atheists as cynical, single white guys who live in apartments downtown, work at used record stores, love George Carlin, and watch Adult Swim."

Best thing I've read this month. BTW, did they change the scheduled time for Space Ghost?

Peter said...

If this analysis is limited to whites, the figures for Muslims make little sense because there are very few white Muslims in America (or, for that matter, the world). Just Albanians and Bosnians, who for the most part aren't very religious, and a handful of converts.

Anthony said...

The Muslim numbers are probably depressed by the recency of much muslim immigration, and the relatively poor knowledge of English or Spanish of much of the muslim population.

I wonder how much higher the Catholic index would be if poor English skills weren't strongly concentrated in Catholic immigrants.

BGC said...

My guess is that society secularized because traditional religions were hostile to modernization (especially science and economic growth). Most still are.

This meant that more-secular individuals and societies both tended to become more powerful and prosperous.

However, the more secular individuals and societies are less fertile (etc).

So in terms of world views there is mostly a choice between either fertility or prosperity.

The Mormons seem to be unique in being as fertile as traditional religions but also pro-modernization - so they are both fertile and prosperous.

Stopped Clock said...

If white Muslims have low IQs and low incomes even relative to those low IQs they must be pretty poor. But if this white Muslim category excludes Arabs, I would have to agree that they're probably basically Turks and southeastern European immigrants of whom many are ESL's.

And if Jews have high IQs and high incomes even relative to those high IQs, there are a few factors that could be responsible:

1) The IQ figure given for Jews could be too low. Since there has been quite a lot of debate among IQ researchers as to what the real average figure is, I think the margin of error on any all-inclusive Jewish IQ estimate should be rather large. Also, I can't read the paper because I don't have access, but if the IQ scores given in the paper are the average of the visuospatial and verbal IQ's, then it doesn't account for the Jewish verbal IQ anomaly. I agree with La Griffe's hypothesis that verbal IQ matters more than visuospatial in terms of getting someone into a high-paying job, which would mean that the effective Jewish IQ would be higher than what is listed in the paper.

2) American Jewish culture discourages Jews from pursuing low-income careers, so even Jews of moderate intelligence struggle hard and often find themselves in relatrively high paying jobs. I wouldn't be surprised if Jewish students study harder and get better grades than other whites even of the same IQ levels.

I'm disappointed that his paper didn't have Orthodox on its list. Greeks make up less than 0.5% of America's population, and Armenians are less than 0.2%, but both groups seem to be over-represented among the famous and wealthy. You could add the Lebanese to this list as well; when you consider that they make up only 0.2% of the population as well, they might even outrank Jews. The Lebanese and to a lesser extent the Armenian success stories could be explained as a result of selective migration, but most Greeks and many Armenians come from families who have been in America for more than 100 years and started out working low-wage jobs in the cities just like all the other immigrants. By contrast, non-Brazilian Portuguese are about as numerous as Greeks, and according to Lynn have higher IQs, but are nearly invisible in the world of American politics and finance. Italians probably come in about average; the overachieving Mediterraneans seem to be confined to the Middle East. Whether this is due to a some common characteristic of Eastern Mediterranean cultures or a wider Jewish-like intelligence profile distributed around this area of the world is beyond my ability to hypothesize.

The Undiscovered Jew said...

"Whether this is due to a some common characteristic of Eastern Mediterranean cultures or a wider Jewish-like intelligence profile distributed around this area of the world is beyond my ability to hypothesize."

According to Bauchet, Jews genetically cluster closely with Greeks, Italians and even more closely with Armenians rather than Southern Levantines:


Also, anthropologists in the late 19th Century claimed that their measurements of ancient Jewish skulls were much more similar to the measurements of skulls from Armenians and the Hittites of Southeastern Anatolia than to Arabs.

To me Jews look more like an ethnic group from the Caucusus mountains than Arabs.

Peter said...

By contrast, non-Brazilian Portuguese are about as numerous as Greeks, and according to Lynn have higher IQs, but are nearly invisible in the world of American politics and finance.

I grew up in a part of Connecticut that has a substantial Portuguese population, and there are many in the area of New York where I now live, and it's been my impression that they're heavily represented in the skilled trades. Many Portuguese-origin contractors, for example.

Audacious Epigone said...


I wouldn't know, as I'm not an atheist ;) Adult Swim did introduce me to Family Guy five years ago. That's been culturally beneficial.


Nyborg broke denominations up into three categories: White, black, and Hispanic. Presumably Middle Easterners are considered white (they generally are for Census purposes--virtually all of the "other" race category is comprised of Hispanics) here as well, although he doesn't specifically address it. The denominational category is still the small though--smaller than the atheist category.


I'm not sure I follow. Nyborg only included non-Hispanic whites in his analyses. The US receives more immigrants from Canada, the UK, and Germany than it does from the largest sending white Catholic country, Poland, does. And the total representation of foreign-born white immigrants is tiny--less than 2% of the US' white population.


Which is why I'm an admirer of Mormons. They present a model for people of European descent to follow if we want to derail the track we're currently on--25% of the world's population in 1960, 18% in 2000, and 10% by 2050.

Re: fecundity versus modernization, Steve Sailer's VDare article The Return of Patriarchy? is a must-read.


It would be helpful to have the ASVAB sub-test scores broken out individually, or amalgamated into verbal (word knowledge and paragraph comprehension) and non-verbal (the rest) categories. I'm not competent enough yet with the NLSY online database to check on my own, if the data are even available in that level of detail. Anyway, I sent you Nyborg's pdf if you're interested.

Stopped Clock said...

It's strange that that PDF contains a small but obvious error and doesn't seem to have been noticed by anybody: on page 88, Table 9, it seems to claim that all 5722 atheist 13-year-olds in his study are boys, and none of them are girls. Yet there is no explanation of why this is so; whether it is due to a contaminated sample or an error in the table itself. It doesn't really affect his theory, but why didn't the author, editor, peer reviewers, proofreaders, or the publishers notice the discrepancy?

I have to hope it's a transcription error that didn't appear in the published version but somehow cropped up when this PDF was being made.

Anyway, the only other thing I have to add is that it appears the Muslim IQs are slightly bimodal, unlike all the other samples which have a single peak. This could be due to the difference between two kinds of immigrants: children of Muslims who came to America to make money in business, and children of Muslims who came to America to escape wars in their home countries, and are just glad to be here and not really worried about education.

Thanks again for sharing this information; I appreciate the knowledge.

Audacious Epigone said...


It looks like 12 and 13 yo females were grouped under the 12 column. For age 12, the table shows 3x as many girls as boys, which, if correct, would be anomalous, as most atheists are male (and at every other age category for atheists in the table, there are more boys than girls).

Melinda Barton said...

It's highly unlikely that there is any real causative relationship between IQ and religion. IQ varies little with time, except due to testing peculiarities and environmental effects. (The same person can have two different IQ's on two different dates due to stress, distraction, nervousness, etc.) Religion varies greatly over a single lifetime, usually based on psychosocial factors.

As for atheism and the single lifestyle, demographic profiles show that people tend to become more religious once they've reproduced for a variety of reasons. Atheist loners in their 20's may be married religious people in their 30's. This doesn't make them "losers," just people in a certain stage of life.

Melinda Barton said...

Oops. I did not mean to imply that those who remain atheists are losers either.

Neuroskeptic said...

"White" is a broad category. If there are IQ differences between "white" ethnic groups and if there is a substantial genetic contribution to this, then that could explain some of the differences because this data must be confounded by ancestry. (At a guess: Mormons & Episcopalian are largely of British ancestry, Orthodox are Greek or Slavic, Catholics mostly Southern or central European and Irish).

Audacious Epigone said...


Denominational membership is more malleable over one's lifetime than IQ is (not sure about piety, though--it'd make for a good GSS query), but it doesn't vary that greatly over one's lifetime. As the U of Chicago study reports, around two-thirds of Protestants, more than four in five Catholics, and nearly nine in ten Jews retain into adulthood the same religious tradition they were raised in.


Right. You've probably seen these maps at GNXP that offer visual insight re: Mormons. To add to your list, Lutherans are mostly German and Nordic in ancestry.

Peter said...

As for atheism and the single lifestyle, demographic profiles show that people tend to become more religious once they've reproduced for a variety of reasons. Atheist loners in their 20's may be married religious people in their 30's. This doesn't make them "losers," just people in a certain stage of life.

It would mostly be women who undergo that sort of change. Being a loner or otherwise antisocial is a far bigger drawback for men than for women in the dating and relationships market (just as being overweight or otherwise unattractive is worse for women). A man who is an atheistic single loner is his 20's is probably going to be a single loner in his 30's and beyond, whether he wants to or not. Now, it might be that his atheism will moderate with age, even as he remains single.

Mark Plus said...

I wish you theists would get your stereotypes about atheists straight. What happened to the claim make by christians for centuries that men become atheists because it supposedly makes it easier to engage in sexual promiscuity? (This accusation suggests that christians must feel sexual envy towards their fantasy construct of "atheists.")

I also find it ironic that the same christians who tend to preach abstinence before marriage have now started to ridicule single male atheists as losers for their alleged disability in getting laid.

Melinda Barton said...

Audacious Epigone,

When I said "varied greatly," I meant the change can be drastic. From fundamentalist Christian to atheist, from Orthodox Jew to Bhuddhist, etc. This is a much greater shift in worldview than is possible for IQ, which changes drastically only due to neurological injury or disease.


While men in general are more likely to be atheists, all of the scientific data I've seen doesn't draw any great distinction between attrition rates for male atheists or female atheists after childbearing. Although, this difference would be plausible simply because of the woman's role in childrearing in most circumstances.

John said...

The paradox is that most "Jews" are in fact atheists.

Audacious Epigone said...


Not according to Pew's US Religious Landscape survey--it finds 10% to be atheistic and another 11% to be agnostic (with 7% not answering). Anyway, Nyborg separated self-described atheists from other religious and spiritual denominations, so the Jews in question should at least be theistic.

Anonymous said...