June 22, 2007

Birth order is back

Lots of people are excited by the Norwegian study showing 2 or 3 point higher IQs for first-born than latter-born sons in the Norwegian army conscription test.

According to the paper "Explaining the Relation Between Birth Order and Intelligence" by Petter Kristensen and Tor Bjerkedal, the first scientist to notice the higher-level of accomplishment among first-born sons was ... you guessed it, Sir Francis Galton.

The Norwegians have a huge sample size to work with, so their result sounds pretty reliable (although the effect isn't that big, so it might not be generalizable to other countries). They appear to be comparing brothers to brothers within families, so that eliminates hereditary average genetic differences (assuming they are only looking at full, not half, brothers).

They also try to answer the Why question: is it because of social effects (e.g., first-borns get more alone time with Mom and Dad) or because of those obscure gestational effects such as the development of male hormone antibodies. So, they look at conscripts whose older brothers died as infants and find their IQs are almost as high as first-borns, arguing against the gestational wear and tear on mom argument.

I suspect there is a subtle problem with this that reduces the confidence level, besides the much smaller sample size. For the primary How Much question, they should be able to compare living brothers' IQ scores directly to each other. For the secondary Why question, however, they can't compare a younger brother's IQ scores to those of his older brother who died in infancy. They have to estimate what the dead brother's IQ would be based on various demographic factors. And that adds another level of uncertainty to their secondary finding that social factors are more important than biological ones.

Anyway, birth order is an interesting topic. It's kind of odd how it has been out of fashion to talk about it for some years, even though it doesn't seem to be all that politically incorrect. Here is my 1996 review for National Review of the last major book on the subject, Frank Sulloway's Born to Rebel.


My published articles are archived at iSteve.com -- Steve Sailer

16 comments:

Anonymous said...

goof for Norwagians.I wish i was one of them high breeds.lol

Anonymous said...

Purely anecdotal, but I have a twin brother who is older than me by about 3 or 4 minutes. His IQ is certainly higher than mine (by how much I can't say). To sum it up, he went to Dartmouth and I didn't.

Anonymous said...

...showing 2 or 3 point higher IQs for first-born than latter-born sons...

For a 'failed idea', IQ testing seems to attract a lot of interest.

agnostic said...

A 3-point difference implies that first-borns will be twice as likely than later-borns to reach 145 IQ, if the pattern is real (p = 0.0026 vs. 0.0013).

I think people are anxious about doing birth order stuff since Frank Sulloway's idea was totally eviscerated once scholars started to really dig deep into it. See:

This editorial

The link to the PDF is at the bottom ("PERG.Johnson.pdf"). In it, a journal editor explains at length what's wrong w/ Sulloway's research and why a roundtable discussion of the issue was set back 4 years due to legal threats from Sulloway. And that's just an overview editorial: there are substantial articles in the journal issue.

Anonymous said...

Wow. This article makes me feel guilty as a dad. My 2 year-old son got a lot more mom and dad time when he was a baby, including reading, teaching him new words, playing with blocks, etc., than my baby girl does now (I am actually trying my best to change this).

In fact, just before this article came out I was thinking about this issue.

The thing is, having two children still in diapers is more than twice as hard as having one. Doesn't sound logical, but that sure is how it feels. So the second child, if close in age to the first, is going to get a whole lot less attention. That is why I'm not surprised to see that as the age between children increased, the effect diminished.

However, I am surprised that the difference is so small. Given the huge advantage firstborns have in terms of personal attention during the most crucial years for brain development, a couple of IQ points isn't that much of a difference.

What this article says to me is that environmental conditions do make a difference, but not as much as you might expect.

Stopped Clock said...

Not politically incorrect? Try telling this to a sixth-born child. Which brings to mind another question: is it just boys that have this effect? I suppose no one knows.

Anthony said...

THe obvious way to test biological versus social effects would be to find children who were firstborn and given up for adoption, and compare them to the next sibling in cases where the second was not given up for adoption. But that sample size might be too small in Norway. (There would also be confounding effects if the firstborn were adopted by a family which already had children, was two-parent vs single-parent, or had a significantly different IQ level than the biofamily, etc., etc.)

Anonymous said...

Did they test if they were brothers?Or did they just take their word. Chances of a child not being related to their dad increase with the number of children a woman has had.

Anonymous said...

Perhaps birth order explains even more. I've only known a few homosexuals including my brother. All were last born. Has anything been done on this?

Steve Sailer said...

Yes, men with older brothers (but not older sisters) are a little more likely to be homosexual. It's not a big difference, but it's the best established influence on male homosexuality yet found, which otherwise seems pretty random.

Ray Blanchard theorizes that it's caused by the build-up of mothers' medical problems caused by the reaction of male fetuses' male hormones.

Anonymous said...

I wonder to what degree this effect holds with families having various numbers of children. Do we see the same decline in IQ among first-, second-, and third-born siblings in families with five or six children that we see in families with only two or three children, for example? Perhaps, if congenital effects are to blame, families capable of having more children don't show the same decline.

It would also be nice to know if spacing between births has any effect.

Anonymous said...

The "nurture" people are excited because they think this might give them a leg up over the "nature" people, but look at the downside:

All of the industriousness, drive, ambition, and fire-in-the-belly, which presumably comes from being the first-born male, only increases your intellectual output by a point or two.

Kinduva downer, really, if you think about it.

In fact, all of this IQ stuff can be just one great big downer if you dwell on it too much.

Anonymous said...

Just from observation, I've seen mixed results with families having boys and girls. Two in particular interest me. One, birth order, b, g, g, b (b boy, g girl), all four were valedictorian in their respective high school graduations. Also, the mother had miscarriages though I don't know when they happened. The other, birth order, b, b, g, the first born and last born were valedictorians, the middle child didn't score well at all on his SAT but was a good student. The other two must've scored well because they both went to Rice.

If there is some sort of rhesus factor affecting intelligence and/hormones, maybe girls in between boys help restore balance?

Anonymous said...

Of course, personal experience is only anecdotal, but if you have a big family, it may count for something. In my experience, the difference in evironment between first borns and the rest, is gigantic. Have you ever looked at the baby-book sequence in your wife's or your mother's archives. My God, it's a wonder the nth baby makes it at all. ALso, isn't the standard error in statistics always greater than one or two points out of a hundred? AND IQs measure something, but it isn't intelligence. Lots of room for an infinite number of studies here.

Anonymous said...

Now I wonder if Charles Murray will go back and re-analyze the SAT-V
test score decline off its high in the early 1960's, relative to the high concentration of firstborns at that highwater point.

Anonymous said...

Chances of a child not being related to their dad increase with the number of children a woman has had.
If true, that would tend to raise the IQs of later children since, IIRC, women tend to have affairs with someone of higher SES than their putative mate.