Smart people are more varied than other people

I showed in an earlier post that, according to the General Social Survey (GSS), the smartest category of people are also the most diverse in terms of income.  Some smart folks are very rich, some are very poor.  They are all over the map.  Nicholas Taleb takes advantage of the wide income dispersion among the intelligent in order to make IQ look like it cannot predict income among smart people and thus is a worthless concept.

If he were biased against the concept of social class like he is against IQ, he could use the same tactic: According to GSS data, the correlation between father's education and offspring income is a paltry .13, and the standard deviation (sd) for people with the most educated dads is $42,000, while it is only $33,000 for those with dads who only finished high school.  For non-nerds who don't know stats, sd is a measure of how spread out the scores are for a sample of people, in this case, individual incomes. 

Turns out, people in the smartest category are diverse on many variables, not just income.  I'll give you the sd for the middle and highest IQ categories to show what I mean.


Standard deviation for average intelligence people and those in the smartest category

Education
Middle IQ  2.4
Top IQ  2.5

Job Prestige
Middle IQ  12.6
Top IQ  14.4

Political Orientation
Middle IQ  1.3
Top IQ  1.6

Church Attendance
Middle IQ  2.7
Top IQ  2.8

Confidence in the Existence of God
Middle IQ  1.3
Top IQ  1.7

In every case, intelligent folks are more diverse than ordinary people.  Compared to the middle, smart people are simply all over the place in terms of education, job prestige, politics (higher scores mean more conservative), church attendance, and belief in God.  Humans beings are so complicated, it's hard to predict which box a person will fall into, and it's even harder when the person is very smart.

This reminds me of the argument that Michael Levin made in Why Race Matters that, because whites are more intelligent, on average, than blacks, they have greater free will, and thus perhaps greater responsibility for their actions.

(You might argue that the sd is higher for smart people simply because their means are higher, thus one can expect the sd's to also be larger. Actually, the means for political orientation, church attendance, and confidence in God are lower for the smart group.)

UPDATE:  I am also reminded of what is said about Jews, a famously smart group:  two Jews, three opinions.

Church attendance and fertility cross-nationally

I was able to get total fertility rate (TFR) estimates from the CIA World Factbook. I subtracted 2.1 from each score--the rate necessary to keep a country's population stable. Positive scores reflect the extent to which a society is growing. Negative scores show the degree to which a society is dying.

I correlated this measure with the percent of the population that attends church weekly (World Values Survey). For 91 countries, it is .68--a very strong relationship. Societies with low numbers of people going to church are very likely to be dying.

By the way, I noticed that few countries are really close to 2.1. Societies tend to be either growing or dying.

Church attendance, alcohol, divorce, and suicide

Here are a number of additional cross-national correlations that are supportive of a socially conservative orientation (the church data is from here and the divorce data from here):


Pearson Correlations

Alcohol consumption--crude divorce rate .61 (strong relationship)
Crude divorce rate--suicide rate .49 (medium)
Weekly church attendance--alcohol consumption -.40 (medium)
Weely church attendance--suicide -.42 (medium)
Weekly church attendance--crude divorce rate -.34 (medium)

Measures of religiosity

I often use attendance at religious services as a measure of religious commitment (because it is usually the only question available) but there could be many reasons to go church: to please your parents, because you like some girl there, to find business opportunities, etc. How well does the measure actually tap sincere commitment?

Teenagers--individuals who are often forced to go to church--were asked in the Add Health Study: 1) frequency of church attendance, 2) frequency of prayer, and 3) importance of religion to self. I correlated the measures for 6,504 individuals:

Pearson Correlations

Church attendance-Importance .69
Church attendance-Prayer .66
Importance-Prayer .76 

Getting correlations this high from survey data is rare (frankly, it feels good when it happens). Cronbach's alpha--a measure of how well the measures hang together--is .88 which indicates that the measures are tapping an underlying concept--religiosity.

The numbers are this high even though some of the teenage religious involvement is coerced. I conclude that any one of these questions is an adequate measure of religiosity.  

Church attendance and infidelity

Whites N = 13,431

Blacks, N = 1,627

GSS respondents who have ever been married were asked if they ever cheated on their spouse. The graphs shows that both blacks and whites who often attend religious services are less likely to be unfaithful. The rate of infidelity for never-goers is about double that of those who go all the time among whites and about 1.5 times higher among blacks. This, in spite the fact that one frequently runs into hot chickies at church.

Then there's the moral recession

The General Social Survey has put up its 2008 data. When it provides new stuff, I like to look at the latest trends among young people in order to give us a glimpse of the future. Looking at the graphs above for 18-29 year olds, you can see that everything is moving in the wrong direction (at least if you think like I do). Sample sizes are between 6,300 and 11,300.

The first half of the decade made me think things were getting better (you can see that in the graphs) but trends reversed, and 2008 data are supporting the view that more young Americans approve of homosexual sex, abortion-on-demand, and never go to church. Throwing in Congress and Obama, the Left is kicking our asses at the moment.

Trends in church attendance

This chart shows the 1972-2006 trend in church attendance (GSS, N = 50,496). 2006 saw the highest share of people in 35 years who never attend--22.7%. You can see that the decline is in the flesh-colored area (?, I'm a man--I don't know colors, plus "flesh-colored" is racist) which is the percent who attend weekly. It has fallen from 28.8% in 1972 to 18.7% in 2006.

If the same trend continues, in another 35 years over 30% of Americans will be unchurched.

Rap and going to church: Readers have suggested that the decline of black singing and the rise of rap is due to declining church attendance. Church is a great place, especially for poor people, to learn how to sing and perform.

One of my favorite things about mass is the chance to sing and to listen to the organist play classical music. If you're a singer like I am, you know how satisfying it is to exercise those pipes once in a while, but if you're not a professional and think that the karaoke circuit is a little pathetic, where do you ever get a chance? In the car? You look like a crazy person. Even in church, a guy who is in the choir or who sings in the pews feels a little funny, but at least it's legitimate.

My friend got me started in high school. He tried to persuade me to join the school choir--I told him it was gay. He twisted my arm and reminded me that Bono is a singer. I agreed to visit one time. We walked into a large room, and there were about 40 girls and 5 guys. I signed up immediately and was awash in females for the next two years.

But I digress. Back to the question. I looked at General Social Survey to see if the share of blacks ages 18 to 25 who go to church at least almost every week--you're going to have to go at least that often to be in the choir--dropped in the late 70s or early 80s:


Percent of young blacks attending church nearly every week or more

1972-1976 22.5
1977-1981 23.3
1982-1986 28.5
1987-1991 24.3
1992-1996 20.3
1997-2001 29.7
2002-2006 29.9

No evidence here. The numbers bounce around a little because sample sizes are not large, but the story here is one of stability over the past four decades. But since rap was created in the ghetto, let's try to focus on urban blacks (those living in the largest 100 SMSAs):


Percent of young urban blacks attending church nearly every week or more

1970s 23.7
1980s 28.1
1990s 21.2
2000s 30.0

Same basic story. Along with rap, the crack-related crime wave of the 80s was not preceded by a drop in religiosity.

Let's look at America's scIQ: In 2006 the General Social Survey asked 437 respondents eleven basic science questions. The first one, for example, was whether the earth's center is hot. I gave each person one point for answering a question correctly, and then summed the scores. My next step was to convert these totals so they resemble IQ scores. I set the white mean at 100, and the standard deviation at 15. Here are some averages:


ScIQ mean scores

Whites 100.0
Blacks 80.3
Hispanics 90.0

Males 99.2
Females 94.9

Less than high school 82.4
High school 94.2
Junior College 96.8
Bachelor 101.2
Graduate 103.1

Protestant 94.1
Catholic 98.7
Jewish 109.3
None 102.8

Never attends church 101.1
Attends more than once a week 87.4

New England 101.4
Middle Atlantic 98.0
East North Central 96.8
West North Central 103.6
South Atlantic 92.8
East South Central 95.7
West South Central 94.3
Mountain 101.8
Pacific 100.9


Interesting stuff. Blacks are well over one standard deviation below whites. This doesn't quite square with all the black doctors and scientists we see on TV. Women are about one-third of a SD below men--score one for Larry Summers.

Educational degree is no surprise; neither is the high Jewish mean. Anyone who completes high school should have been exposed to the information needed to answer the questions correctly. Many of those who finished high school and even attended college missed some of the questions, indicating that people are not understanding or retaining the information.

ScIQ differentiates churchgoers from non-attenders better than IQ does. I calculated Pearson correlations between scIQ and church attendance, and then the latter with IQ: the estimates are -.26 and -.08, respectively. Scientifically-minded people are less likely to go to church than people who are smart in a more general way, which supports the notion that there is intellectual friction between science and religion. By the way, scIQ correlates with IQ at .46.

It is also interesting that the West North Central region (Iowa, Kansas, Minnesota, Missouri, Nebraska, North Dakota and South Dakota) is at the top of the scIQ list, considering that New England tops the IQ list based on GSS data. Is it all the nerdy white folks who live in those parts? The uncool Mountain States are not far behind.

Gay men on their knees: In a previous post on violent victimization, I displayed data showing that gay men are like women and lesbians are like men. Let's look at another characteristic--religiosity. Going to church is more of a woman's thing (unless you're a leader). According to the General Social Survey, the following percent of people attend religious services at least once a month:


Percent attending church at least once per month

Straight women 51.4
Straight men 41.7
Gay men 38.5
Lesbians 32.4

Lesbians end up being the most "masculine" of the bunch. You might respond that lesbians don't attend church because most religions condemn homosexuality, and there is certainly some truth to that, but then why do so many gay guys go? They go almost as much as straight men.

I won't list percentages of church membership, since the only significant differences are that more gay men and lesbians belong to no church, and compared to straight women, fewer lesbians are Catholic. Maybe lesbians drop out because they feel the Catholic Church is more patriarchal. Or maybe more Catholic women refuse to identify themselves as lesbian.

I'm surprised by the high numbers for gay men. They attend church a lot, although I'm sure some people make them feel uncomfortable there. Evidently, they're a pretty motivated, religious group. That's news to me (the stereotype is that a church is the last place you'd find a gay guy on his knees) but I shouldn't be surprised when I think of all those gay Catholic priests.

Global church attendance trends: In the last post, I looked at trends in global church attendance over two time periods--1981 and 2001. The problem with the analysis is that two samples included different countries, and it might be that the 2001 sample included more of the developing world which tends to be more religious. I'm very busy right now, but I will update this post with additional countries, as I find the time.

My first choice of countries is China since it is the largest. It was surveyed in 1990 and 2001. The percent never attending religious services declined from 94.1% to 89.6%. The largest growth was in the more frequent attendance categories. Of course the numbers are still low, but hey, this is China.

India: Attendance in India is down bigtime over the same period. The biggest drop was among those going more than once a week from 29.4% to 15.3%. Almost everyone goes, but not so often.

USA: America remain unchanged between 1982 and 1999.

Nigeria: Between 1990 and 2000, the biggest change is more people going more than once a week: from 55.5% to 63.7%. Now that's religious.

Russia: During the 90s, the trend was for those who go never or rarely to begin to go for Christmas and Easter and other holy days: this category rose from 7.9% to 22.6%.

Mexico: From 1990 to 2000, the percent attending weekly went from 31.6% to 42.5%.

The Drunken Deist: General Social Survey respondents were asked if they sometimes drink too much. I looked to see if this varies with belief in God and church attendance:

Percent that drink too much

1. I don't believe in God. (34.0)
2. I don't know whether there is a God and I don't believe there is any way to find out. (40.9)
3. I don't believe in a personal God, but I do believe in a Higher Power of some kind. (46.8)
4. I find myself believing in God some of the time, but not at others. (43.6)
5. While I have doubts, I feel that I do believe in God. (40.9)
6. I know God really exists and I have no doubts about it. (27.3)

1. Never attends religious services. (47.3)
2. Less than once a year. (44.1)
3. Once a year. (46.4)
4. Several times a year (38.1)
5. Once a month (39.0)
6. 2-3 times a month (32.8)
7. Nearly every week (28.1)
8. Every week (24.4)
8. More than once a week (12.8)

In terms of beliefs, the confident folks seem to have the least problem with drinking. The atheists and especially the non-doubters have the lowest numbers. The people who are not sure, as well as the group who don't believe in a personal God, are most likely to be drunks. This suggests that having conflicted or vague beliefs, or perhaps a wavering personality, may not be healthy.

The pattern is even clearer with church attendance. The more often you go, the less likely to drink too much.

There is a "God Gap"...for whites and Mex-Ams, anyway: Razib links to a survey which reports that the "God Gap" between Democrats and Republicans is not as large as is conventionally thought. The survey fails to discuss differences by race, so let's help them out. I'll cut to the chase and only look at how often people go to church, and how they vote. If they attend religious services at least once a month, I categorized them as religious. The General Social Survey does not have voting behavior more recent than the 2000 presidential race.


Percent religious

Whites
Gore 40.5
Bush 58.8
Nader 21.8

Blacks
Gore 74.0
Bush 75.0

Mexicans
Gore 49.0
Bush 73.8

Church attendence differentiates white voters (especially Nader fans from the rest) and it differentiates Mex-Ams even more. On the other hand, African American Gore voters are just as religious and three blacks who voted for Bush. When I addressed previously the question about what does differentiate black GOP from Democrat voters, the only thing I found was that men are more often Republicans. (I don't present Nader numbers for the two minority groups since sample sizes are too small).

While there are differences here, people like Ann Coulter who say that Democrats are godless exaggerate wildly, even if speaking of whites: two-fifths of whites, one-half of Mexicans, and three-quarters of blacks who voted in 2000 for Gore go to church at least once a month.

The secular creep: Using the General Social Survey, I averaged the percent of Americans who never attend church:


Percent never going to church

Whites
70s: 14.0
80s: 14.8
90s: 17.3
00s: 20.4

Blacks
70s: 6.8
80s: 8.3
90s: 8.9
00s: 10.7

So the increase in church avoidance is up 46% for whites and 57% for blacks over (roughly) 3 decades. One in five whites never attends religious services. America is still one of the most religious countries in the world (with indicators similar to Ireland, Poland, and India) but the drift is in a European direction. I suspect that the trend has less to do with growing disbelief than with diminishing appeal and sense of obligation, and later I'll see what the data say.