Does education increase IQ? The General Social Survey says no

The question was raised on Twitter the other day about whether more education causes increases in IQ. The General Social Survey can help a little with this. Below we have a chart of the mean years of schooling completed since 1977. The mean starts out at 11.71 and ends in 2018 at 13.89 years of education. Over the period, that is an increase of 19%. 

By contrast, here is the trend in IQ. In 1978, it was 97.45, and in 2018 it was 97.80--an increase of 0.3%. Basically, no change. US trends are consistent with the view that keeping people in school longer does not make them smarter. 

Data: The big move among American Jews is not to Orthodoxy, it's to no affiliation

Since Orthodox Jews have more kids than other types of Jews, some assume that they are quickly becoming the modal category.  Not so fast. This table show trends since the 1970s (GSS data, sample size = 737):

 

While the share of Jews who are Orthodox has increased over the past four decades (1.9% to 7.8%), the real movement has been toward being a Jew person with no affiliation. Over this period, unaffiliated Jews rose from 15.1% to 27.5% of the total.  Also--look at the huge drop in Conservative Jews: 43.4% down to only 17.6%.

This seems to be part of the larger social trend of some Americans moving away from religious affiliation of any kind. 

In a previous analysis, I found that less religious Jews are less ethnocentric. 

Data: The most athletic women have the fewest kids

Now that I've got a little more confidence that self-rated athletic ability has some validity to it (I also observed that it drops markedly with age), I looked to see if athletic people have fewer children. Here's the graph for women ages 40-55 (General Social Survey, N = 401):

American women who say that "athletic" describes them very well average only 1.58 children, while all the less athletic groups have roughly 2.1 kids.

What about men?  Here's a graph for them (N = 317):

With men, we see a U-shaped relationship: the very athletic group has a mean of 2.39 offspring, and the least athletic men average 2.31 kids.  The average athletic group has the smallest families: a mean of only 1.66 kids.

We could interpret this pattern to mean that highly athletic men are more attractive to women and consequently have more mating opportunities, while the least athletic tend to be low testosterone men who are highly committed to family, which is an alternative path to a large family.

The overall results suggest a mixed trend.  What sticks out to me is that athleticism among women shows that same dysgenic trend with see with traits like IQ, education, and health: the top scoring females consistently have the fewest kids.

And let's not forget the study I linked to that found that athletic performance is highly influenced by genes (heritability = .66).

Data: Survival of the fattest?

We saw in a recent post that the women with poor health have the most kids.  Is this true for heavy women, too?  Yes, it is: look at this graph (General Social Survey, self-reported weight, N = 200):

Women with one child are the lightest group with a mean of 165.6 pounds.  Compare this to the heaviest group: Women with seven children weigh an average of 208.5 pounds. The difference is roughly one standard deviation -- huge, no pun intended.

Of course, it's hard for a woman to lose the weight she gains from being pregnant, but it is also possible that genes underlie both weight and fertility.

For example, perhaps women who score low on conscientiousness (self-discipline, long-term planning) lack control over both eating and fertility.

Americans are likely to get fatter and fatter as the generations go by.  If civilization eventually collapses under all the dysgenic trends I have been documenting lately, these fat Americans might finally lose weight due to food shortages.

UPDATE: Readers might suspect that a respondent's weight is due to her height, and perhaps tall women are having more babies.  I checked: fertility does not vary by height. 

No surprise: Americans in the best position to have a large family are least likely to do so

A good income is associated with positive characteristics like industriousness and intelligence, traits that are strongly influenced by genes.  High-income adults are obviously in the best position to have large families.  Do they?  Look at the graph (General Social Survey, 2010-2016, women ages 40-55, household income in 1986 dollars, N = 1,325):

While women with two children have higher incomes than those with no children, income tends to fall as family size increases beyond two kids.  Compared to families with eight or more kids, two-child families earn more than double the household income.

Here's the graph for men ages 45-60 (N = 1,170):

We see a similar pattern for men, although the income drop beyond two kids is perhaps not as steep as for women. (Don't make much of the high bar for men with seven kids: it's based on only four cases.)

We're seeing the same kind of pattern again and again: Americans who are in the best position to have a big family are least likely to do so.

With these trends, the long-term future will go to the people on the bottom of American society -- the people who have the least genetic potential.

Data: Mentally ill women have the most kids

We saw in the last post that women with poor overall health have more kids than healthy women.  This is a bad sign for the long-term health of America.  But what about mental health?

The General Social Survey asked respondents, "Now thinking about your mental health, which includes stress, depression, and problems with emotions, for how many days during the past 30 days was your mental health not good?"

Here is the mean number of bad mental health days listed by number of offspring for women ages 40-55:

Childless women average 3.5 bad days, while women with eight or more kids have a mean of 6.0 bad days. 

Now it's possible that having more children to raise might make one crazy -- I have six kids myself and can testify that it ain't easy -- but it might be the case that people with mental health issues might have less control over their fertility.  The large family size means the next generation has a greater number of genes linked to mental illness, and over many generations, you get a pretty crazy population.  

In pre-modern times, mental illness would work against survival and mating success, but under modern conditions, we see the mentally healthy making every excuse against having kids, while people who simply don't have it together are reproducing like bunnies. 

A perfect example of the current mess is a colleague of mine.  She is tall, beautiful, athletic, healthy as a horse, and very smart (as least as smart as a progressive can be).  She is also nurturing and is exactly the kind of person who should be having kids.  But since the planet is going to melt at any moment, she doesn't want children.  So instead of having a family, she and her husband help the environment by jetting to exotic places every other weekend.  Not good.   

Data: The healthiest women have the fewest kids

Research shows that more than one-third of the variation in self-reported health is due to genetic differences.  Do healthy Americans have most of the kids, so future generations will trend toward good health?  The answer is, no.  Look that this graph that shows the current number of offspring for women ages 40-55 (General Social Survey, N = 965):

Women with excellent health average 2.08 children.  Those with poor health have a mean of 2.32 kids.  Now the gap is only small (one-sixth of a standard deviation), but it's enough of a difference to have an impact over the long-term.

So now we've documented negative fertility trends for IQ, education, and self-reported health. 

Liberals obsess over about how people on the bottom of society have got it so bad (they exclude poor, straight white men, of course), but in evolutionary terms, the bottom is made up of winners.

Data: Educated immigrants ain't impressive

In a recent tweet, Ann Coulter claimed that H-1B visa holders are not highly skilled like President Trump says but are simply slave labor.

While I can't use GSS data to identify this group of workers, let's look at the vocabulary test scores of immigrants with at least 16 years of total education. I won't call it IQ scores since English is a second language for most immigrants, but I will argue that an immigrant's English vocabulary says something meaningful about him.

Here are the mean vocabulary scores by decade (N = 725):

Mean vocabulary (using an IQ scale)

1970s  109.5
1980s    98.7
1990s  100.9
2000s    97.6
2010s  100.7

We can see that while educated immigrants in the 1970s had good vocabularies, they dropped to the average white score by the 80s, and it has remained in that area since.

This pattern is similar to what we have seen with immigrants in general, and Americans admitted to college or graduate school: mean scores were quite good in the 1970s, but the US then got more egalitarian and lowered standards. Now we take anyone into the country or into the university.  This shows that our elites no longer put a premium on quality, but really care about giving goodies (American citizenship, access to universities) to any schmuck who shows up. 

UPDATE: If we divide the educated immigrants from this past decade by race, the mean scores are: white 104.5, black 99.5, other 96.6.

Data: Nonwhite IQ is falling

If you believe that people basically can't be changed and children are like their parents, you should be concerned with the kinds of people who are coming into the US with plans to stay here for good.

The picture does not look good when we focus on IQ.  I typically use a vocabulary test given to participants of the bi-annual General Social Survey.  One problem with this measure is that it is biased against people who weren't raised speaking English.  For this blog post, let's focus on non-whites who also not black as a rough way to capture people born in this country whose family came to the U.S. in recent decades. It's crude (for example, American Indians get included), but we do what we can. Here is the mean IQ for this group listed by decade:

Mean IQ for nonwhites (blacks excluded, N = 764)

1980s  88.4
1990s  93.0
2000s  94.0
2010s  92.4

(If you're familiar with the GSS, you know surveys were conducted in the 70s, but not enough native-born non-whites were surveyed to be included.)  Mean IQ improved for this group for 30 years (80s, 90s, 00s) but that has reversed in this decade.

This is consistent with most of the IQ trends I've looked at recently. This past decade has seen a downturn for every group I've looked at (all whites, English/Welsh, German, Italian, Mexican, East Asian) except for blacks and the Irish.

The cause of the fall in IQ is, I imagine, different for different groups, and I assume that the drop for non-whites (excluding blacks) is due to changes in the mix of immigrants.  Whatever the reasons, most trends do not look good for America. 

Despite what Nassim Taleb thinks, average IQ predicts quite well how a country performs, and the US seems headed for mediocrity.  This, of course, could be reversed if we were picky about who gets to come to America, and if intelligent citizens had more babies, while dull ones had fewer.

Data: Has racism declined over the past few decades?

Despite decades of militant effort to root out racism in America, blacks continue to be concentrated at the bottom rungs of society.  Now we all walk on egg shells when it comes to race, so where is the supposed hatred that is keeping blacks down?  To save the argument, liberals have shifted and now claim that racism isn't cross-burnings and lynchings; it's sneaky.  It's so sneaky, you can't even find it. It's like the Devil: invisible but powerful.  It's either hidden in institutional arrangements, or whites are simply faking their goodwill. 

There is clear General Social Survey (GSS) evidence that contradicts the view that whites don't talk or act so racist, but the same hatred still lurks in their hearts.  I've long argued that one of the few survey measures of racism that has some validity asks participants how warmly they feel toward blacks.  At least it's better than BS measures like whether you support racial preferences or not.

GSS participants were asked the warmth question in 2002.  Look how mean "coolness" increases with age for southerners (blue) and northerners (red):

   









People in their 20s would have been born in the 1970s, those in their 30s in the 1960s, etc.  People aged 70 and above are noticeably colder than those in their 20s.  It's about one-half of a standard deviation difference. And even the elderly group averages a "warm" response.  A five would be neutral, but a four is positive.  Even whites born in the South in the 1920s like blacks.  The US is a racist country?  BS.  Just the opposite.

Notice how young southerners are a little colder than northerners, but the two regions converge and even flip among the oldest group. In other words, elderly northerners like blacks less than elderly southerners do.

Anyway, I think even most liberals would admit that whites act better now than they did in the 1950s, which is all anyone should worry about anyway.  Who cares what's in someone's heart as long as they treat you well?  But if we are worried about private sentiments, there is a significant difference between younger and older Americans.  (Of course, I'm assuming people develop their feelings when young and then tend to hold on to them throughout adulthood. It is possible that people are warmer when young and get colder as they age.)

By the way, I'm trying to figure out why my cohort, those born in the 60s, appear to be colder than those who are both younger and older.  I was in college during the Rodney King riots in 1992, and it did make an impression on me.  But I was a liberal, so I found myself making excuses for all the destructive behavior.  Perhaps all the violent crime of the late 80s and early 90s left its mark. 

Trendy trends in identifying as a sexual minority

This Gallup poll is from a survey of over 300k Americans in 2017. You can see a dramatic uptick from 5.8% to 8.2% among Millennials identifying as LGBT over the past 5 years. Kinda dumb to lump the sexes: for all generations, 3.9% of men and 5.1% of women now identify as a sexual minority (table not shown). This is the first time I've seen women with higher numbers than men. Trendy trends on display.

Increases in suicide by sex

These graphs show 1999-2009 suicide trends for U.S. men (top) and women (bottom) separately (CDC) :

Both sexes ages 45-64 show large increases over the past decade.

More suicide among people in 40s and 50s

The graph below shows 1999-2009 suicide rates by age group (CDC data):

Rates (not shown) are stable for children, adolescents, and younger adults, but suicides have been increasing for middle-aged people. By contrast, they have been dropping among those 75 and over.

How do we explain these trends?  Have advancements in medical care improved the quality of life for the elderly? Are people in their 40s and 50s increasingly feeling like failures at work?  More relationship, marriage and family disintegration?

Trends in Catholic religiosity

In his new book Suicide of a Superpower, Pat Buchanan claims that since Vatican II, American Catholics have gotten less religious and more culturally liberal. The graph (GSS data, sample size = 13,432) shows that he's right. Attending more than weekly (gray) or weekly (pink) is down. The following are up: 2-3 times a month (magenta), once a month (gold), once a year (green), and never (red).

This graph shows that Catholics are slightly more approving of abortion for any reason than they were in the 1970s.

Trends for Mexican Americans (data not shown) are more conservative. Attendance has basically held steady over the past four decades, and support for abortion on demand has actually dropped from 38 to 23 percent. My guess is that the constant infusion of new immigrants from Mexico keeps Catholic Mex-Ams religious and comparatively conservative on abortion.   

Decline in confidence that God exists

Once every two years the Inductivist gets giddy with excitement when the new GSS comes out. As a first plunge into the new data, let's look at trends in confidence in the existence of God.

The shrinking purple section (if I were a girl, I'd know the precise name for that color) indicates that "knowing that God exists" has been on the decrease over the past decade. On the other hand, I wouldn't make any long-term predictions. Most of the experts sixty years ago predicted the imminent death of religion.

Alphas and Betas

Alphas, N =360



Betas, N = 384

These graphs show the trends in number of offspring over the past three decades for white men ages 30-39 having a lifetime total of 20-700 sexual partners (Alphas) and those with only one or two (Betas). There are only about 30 cases for each group in the 1980s, so we should pay more attention to the last two decades. First, it's clear that Alphas were more likely to be childless prior to this decade, but the gap has almost disappeared. Now, Alphas are more likely than Betas to have one kid but less likely to have two. Twenty-percent of both groups have three or more.

If we look at means instead, we get:


Mean number of children

Alphas
1980s 1.11
1990s 1.03
This decade 1.23

Betas
1980s 1.64
1990s 1.66
This decade 1.44


The story is similar here. Betas average more kids, but the gap is narrowing. Standard deviations (not shown) are increasing for both groups too, indicating greater variation in this decade.

So the two groups are converging: more Betas are now childless, while Alphas are reproducing at greater numbers. It looks like fewer Betas are able to find a partner now, and perhaps women are becoming more willing to have an Alpha's baby. Single motherhood is becoming more common and acceptable for whites, so women might be more willing to risk having a baby with an unreliable man.

Predictions are a very dicey business, and Alphas are still less prolific, but if current trends continue, Alphas could surpass Betas in the baby race.

Trends in attitudes toward premarital sex

All ages

Ages 18-29

Ages 30-44

Ages 45-59

Ages 60 plus

From studies I've read, the consensus that premarital sex is wrong had collapsed by 1970 in America, but I wonder how attitudes have trended since then among different age groups. The top graph includes all ages, and it shows that the percent saying it's wrong was flat and roughly about 50% through the 80s and 90s, but the attitude grew quite a bit in the 70s and then again in this decade.

Among young adults (second graph) Americans thinking it's not wrong have been more constant, but those thinking it is sometimes wrong has shrunk, while those saying it's always or almost always wrong has actually grown. This shows that young people are more likely to have extreme views one way or the other.

The group ages 30-44 moved in a more accepting direction in the 70s, while the next age group (45-59) gradually became more accepting over the four decades. The oldest group has changed the most since 2000.

The percent of 60 and older who think premarital sex is always fine had been a small share of the total for 30 years. I assume it has been changing in the past few years as Boomers are entering the age range. It looks like this attitude has gone from the minority to majority position because of changes in the 70s, and more recently the dwindling of the greatest generation. This leads to the prediction that acceptance will continue to grow.

It looks like it's been old and religious people versus everyone else. You can't bring back the dead, so I guess folks like me need to work for greater religiosity. Tall order.

Race and IQ trends

Whites, N =19,726

Blacks, N = 3,441

As a follow-up to the last post, let's look at IQ trends by race since the GSS vocabulary test was first given in 1974 (WORDSUM). The test is 10 questions. The overall mean for whites is 6.3. Over the 35-year span, the percentage of whites with below-average scores has dropped about 5 points, and the top two categories have shrunk about the same. The average-to-above-average has picked up the slack.

For blacks (bottom graph), the overall mean is 4.9. The share scoring 4 or fewer has dropped over the period from around 45% to 30%. The improvement began in the late 80s. The percent scoring 5 or 6 grew from 38% to 58%. That's the good news. The bad news is that those scoring 7 or higher appears to have shrunk. (It's not easy to tell since few blacks score so high, so there is a lot of year-to-year fluctuation).

So, the main story seems to be that both blacks and whites have expanded their middles at the expense of their extremes. Next, let's look at mean changes by year:

Mean IQ (standard deviation in parentheses)

Whites
1970s 6.2 (2.2)
1980s 6.1 (2.2)
1990s 6.3 (2.0)
2000s 6.4 (2.0)

Blacks
1970s 4.6 (1.9)
1980s 4.7 (2.1)
1990s 5.0 (1.9)
2000s 5.2 (1.9)

The past three or four decades have been good for blacks. Their mean has increased more than whites, and so the IQ gap has narrowed from 1.6 in the 70s to 1.2 in this decade. Another way to look at it is the gap in standard deviation units (SD). The gap fell from .72 SD in the 70s to .60 SD now. The narrowing is significant, but the gap is still large.

It's also important that, while the white mean has gone up a bit since the 70s, the standard deviation is now smaller. This is due to fewer whites having low scores, but it is also due to a smaller share of smart people.

It looks like America's efforts may have raised the scores of the bottom half, while ignoring the high end. As La Griffe has shown, a nation's wealth depends on the fraction of the population that is intelligent. So it makes little sense to obsess about the bottom, but that seems to be what we're doing. (I'm not sure what has brought up low scores. Getting rid of really bad schools still around in the 60s? Nutrition? All that TV watching improving vocabulary?).

People do not age out of their secularism: Recently, Razib and I both used GSS data to show what other surveys have been showing: compared to older Americans, the young are less religious. But it is possible that this is due to an aging effect. One might tend to grow more religious as one gets older and finds a secular life unrewarding. I have seen research before (don't recall where) that found support for this.

I did not. Taking 18-22 year olds in 1990-1994, I followed them up until 2008 when they would have been in their mid-30s:


Percent who never go to church

1990-94 14.7
1995-99 21.1
2000-04 17.4
2005-08 21.0

Any changes just look like noise to me. It's possible that religiosity might develop in one's forties or older, but I doubt it. If the current cohort of twenty-somethings stay consistent like this older group has, their share staying completely away from the pews will be almost 30%.

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.