Arnold Kling  

The Topic du jour

PRINT
Video Victory for Vouchers... The Engineering Mindset...

Lots of stuff showing up on IQ and genetics these days. The New York Times has a story.


“Let’s say [hypothetically] the genetic data says we’ll have to spend two times as much for every black child to close the achievement gap,” said Jason Malloy, 28, an artist in Madison, Wis., who wrote a defense of Dr. Watson for the widely read science blog Gene Expression. Society, he said, would need to consider how individuals “can be given educational and occupational opportunities that work best for their unique talents and limitations.”

I've referred to Malloy's long post previously--I recommend reading it if you have not already.

The Cato Unbound discussion continues, with an essay by Linda S. Gottfredson. She writes,


people often confuse the yardstick (IQ scores) with the construct (g) actually measured. The manifest content of ability tests items provides no guide to the ability constructs they actually succeed in measuring. The active ingredient in tests of intelligence is the complexity of their items, and it is also the ingredient—“processing complexity”—in functional literacy tasks that makes some more difficult than others (more abstract, more distracting information, require inferences, etc.). To oversimplify only a bit, as long as two tests have similar g loadings, both will predict the same achievement equally well (or poorly), no matter how different their content might seem (Gottfredson, 2002).

You probably need to read her whole essay to grasp the point. Flynn theorizes that environmental influences have increased some test results more than others. It sounds like he is saying that there are different components of intelligence, and some of those components have been affected by environmental influences. She seems to be saying that the individual tests are all indicators of the same thing, not components of anything.

Economists used to construct--and I believe the NBER still maintains--indexes of leading, coincident, and lagging indicators. Things like stock prices, pig iron shipments, and wholesale prices supposedly behave differently depending on whether you are heading into, undergoing, or heading out of a recession. If it turns out that the components of, say, the index of leading indicators have separate trends, that does not tell you that something is changing about the frequency and duration of recessions. A recession is still a recession, no matter what set of indicators you use to predict it or how those indicators behave in terms of secular trends.

Finally, the American Enterprise Institute held a forum on a new book by Jon Entine on Jewish IQ. I did not attend, but I listened to the audio of the event which is available at the link. I got the sense that the evidence for a genetic basis of high average Jewish IQ is speculative and that various theories are inconsistent with one another. For example, Entine's story has only Jews of Eastern European origin having the higher IQ, while Charles Murray wants to include Jews from Spain and Northern Africa.

Is it possible to talk about IQ and lose the we of group identity? There are all sorts of indicators of intelligence that have a high signal-to-noise ratio at an individual level. Ethnicity has a pretty high noise-to-signal ratio at the individual level. I do not see any justification for discrimination based on ethnic identity, given how easy it is to obtain other indicators.

Tim Harford's next book has an interesting chapter on the dangers of using racial indicators. He cites work by Roland Fryer, Jacob Goeree, and Charles Holt. If I know that employers use color as a signal, and my color is a negative signal, then it may not pay for me to invest in education. Of course, if everyone of my color looks at it the same way, then none of us will invest in education, and now more employers will find color to be a useful signal. The authors created an experiment that shows how this sort of self-reinforcing discrimination can result in a world where the colors are purple and green, assigned to individuals at random.

So ethnic identity is a dangerous thing. But we seem to be really attached to it, and we seem to be really attached to using it as a signal.

It's striking that popular discussions of IQ seem to quickly become discussions of race. Holding one's individual intelligence constant, is there something happy about being a member of a high-IQ ethnic group and something sad about being a member of a low-IQ ethnic group? That's the way a lot of people seem to react. I'm not sure why.


Comments and Sharing


CATEGORIES: IQ in Economics



COMMENTS (12 to date)
Brandon Berg writes:

Is it possible to talk about IQ and lose the we of group identity?

Sure, it's possible, and it's probably the best thing to do at the individual level. But it's still important to acknowledge the black-white IQ gap, because it's the only way to refute the left-wing assertion that the black-white achievement gap is due to the pervasive effects of racism and discrimination.

There are all sorts of indicators of intelligence that have a high signal-to-noise ratio at an individual level.

Which ones? Employers are barred from using IQ tests in hiring decisions (although sometimes there are ways around this: in the software industry we interview using g-loaded programming questions). Affirmative action limits the usefulness as a proxy for intelligence when comparing people of different races (i.e., having graduated from an elite university doesn't say the same thing about a black person that it does about a white person).

Ironically, these measures, which were designed to combat racism, actually create incentives to discriminate on the basis of race. Without the ability to measure a person's IQ directly, it can sometimes make sense to discriminate on the basis of crude indicators like race.

William Newman writes:

I see the principle of the Fryer work, but I don't buy that it applies very strongly in the current economy. It seems to me that all it takes to make it break down conspicuously is a few niches where people see talent systematically being undervalued and take advantage of it. And there are a *lot* of niches in our economy. E.g., a few weeks ago I skimmed _Poker Face of Wall Street_, which (among many other things) tells the story of how poker players and other gamers ended up being recruited heavily enough that they were seriously overrepresented on CBOE for a while. (That book is my only source for the story, but I assume if it were just a pack of lies, someone on the web would've noticed, and instead the book seemed to be favorably recommended.) (And regarding your "invest in education," note that "ability to play X is the sign of a wasted youth" seemed to apply at least somewhat to the population described. More grist for the mill of to what extent today's education is really a signalling mechanism...)

I'm really unconvinced that we are seeing a genetic pattern in IQ performance. Yes, intelligence is heritable. But if effects as large as the Flynn effect can result from just the cultural and/or environmental differences between 1950s USA and today, differences which we don't know how to control for, then it's nearly impossible for differences in IQ scores between (e.g., in Watson controversy) today's Subsaharan Africa population and populations elsewhere to convince me that there's a underlying genetic difference. Learn to control for stuff better, then we'll talk. (I expect we'll be talking within about a decade, as data comes in faster and faster about the correlation between individual genes and mental performance.)

I do think that IQ seems to be testing something real and important, and I'm unimpressed by the arguments that it's just some culturally biased shallow property. Sports, and mind games like Chess and Go, are culturally biased too, and it's reasonably common for people from alien cultures to beat people from the culture which developed a game. And "IQ is just measuring something which depends on the culture" should come with a good followup response to the predictable "and it seems that it is present in those cultures which succeed in using modern technology to feed themselves efficiently." So, for example, I'd expect that despite genetic similarity there's likely a significant difference in IQ between North and South Korean populations today, and whether it is dependent on the difference between Communist and consumerist Korean culture is less important than whether it is something real which'd continue to affect economic performance of the two populations for decades even if the political repression in the north were suddenly removed.

Also, while I'm extremely skeptical about people claiming to see through all those confounding factors to see racial patterns in differences IQ, they're at least less unbelievable than people who try to argue that it's technically impossible that such racial mental differences could exist. We have no shortage of examples of genes having nothing to do with physical appearance which are significantly correlated with race, and we don't understand the links between individual genetic variation and mental performance in very useful detail yet, but we have abundant evidence that the links are strong. It may yet turn out that there are no significant inborn mental differences between the races, or it could easily turn out that real differences aren't particularly correlated with current stereotypes, but claiming that we know there are no differences between the races seems outright dishonest.

Steve Sailer writes:

Actually, the opposite is more true: the anti-discrimination jihad has led to things like the Supreme Court's Duke Power 1972 case making it hard for employers to use IQ-like tests, so we have fewer tools available for allowing high IQ individuals from lower average IQ groups to distinguish themselves.

The SAT, which is more or less an IQ test, is still legal, so hiring decisions have more and more shifted toward which college you got into based on the average SAT scores of freshmen there. This means that people who didn't have their act together from age 14-18 to get into an exclusive college (which includes a lot of black and Hispanic males) have a harder time showing they really are smart later in life.

Steve Sailer writes:

Think about politically correct censorship from the perspective of the economics of information. Why should Less Information be better for humanity than More Information? Obviously, the burden of proof should be on the shoulders of the censors, not of those who want to increase knowledge.

bernard sapak writes:

thang u

[Awaiting email address validation--Econlib Ed.]

Buzzcut writes:

Just wondering, maybe Sailer can answer this question:

Regarding the Flynn Effect, why are "experts" so quick to discount the "height effect" with regards to rising IQ?

Height is genetically determined. But the genetic dependence of height wasn't clear until the larger population obtained proper nutrition.

Something similar may be happening with obesity. Obesity may be genetically determined, and our current obesity epidemic may just reflect that we, as a society, have hit the point nutritionally that allows that obesity gene to be fully expressed.

Now, why couldn't we say the same thing about the Flynn Effect? Nutrition improves to some point, and IQ's take off, along with height and obesity.

General Specific writes:

What you guys plan on doing with this IQ data?

Medic writes:

The difference in IQ is probably real and is quite explainable in terms of disease burden. Malaria is the most common cause of brain damage in children in the world. Virtually all sub Saharan children have had malaria. Its effect on intelligence have been known since at least 1900.

TGGP writes:

What you guys plan on doing with this IQ data?
Whatever we want. Knowledge is good.

General Specific writes:

TGGP: That's not an answer. Every pursuit has a cost. What's the benefit? I'd really like to hear what Kling and Caplan plan to achieve with their IQ interests. Everyone talks about it, but nobody talks about what they want to do with it.

Someone must have an answer.

9160 writes:

IQ is not as important as some people make it out to seem. I do not believe that it has anything to do with your ethnic group or your genes. Yes, some people who have a high IQ also have children with high IQ’s, but I don’t think that it has anything to do with the genes in the family. I think that it all just depends on what/how you are taught to think while growing up. Along with determining what/how you think about certain things, I believe that your environment has a strong influence on what the results of the tests would be. Every person grows up in a different environment, which will alter the way that they interpret things and with the influence of some other things, this environment can either increase or decrease their IQ scores.

Troy Camplin writes:

1/3 of the 30,000 genes we have are expressed exclusively in the brain -- and you don't think genetics has anyhting to do with intelligence?

Comments for this entry have been closed
Return to top