Asymmetrical Information

Could you please link to the Caplan piece?

Found it.

It's up at EconLog

And the point that Caplan was actually trying to make is not that IQ matters--it's that the question of whether IQ matters is policy-relevant.

One ex-neurosurgical researcher I know thinks intelligence is almost entirely environmental; I disagree, but I do think one common thread that should be looked at closely is how children are handled at home between birth and Kindergarten.

During that period, I had the privilege of a stay-at-home parent (mother), no electronic entertainment in the household beyond radio and pre-recorded audio, and plenty of books and toys on which to train and exercise imagination. What would my intelligence be today if I had been dumped in front of a television instead?

I agree with the above.
My kids did not have TV and neither went to High School.
One has a Ph.D. in Physics and the other has a degree in Music theory and composition.

Both are involved with bleeding edge technology on a worldwide basis.

IQ and its excersise is what determines the outcome for children.
The greater the excercise of the inherent IQ the more use the posseser will get out of it. Just like muscles the more you use them the more you can use them.

All the money and programs and interventions will not overcome a defective, dysfunctional upbringing.

So the liberals can rest easy--they have nothing to prove, because none of anything they advocate works, they are better off making beleive.

thedaddy

I agree with the above.
My kids did not have TV and neither went to High School.
One has a Ph.D. in Physics and the other has a degree in Music theory and composition.

Both are involved with bleeding edge technology on a worldwide basis.

IQ and its excersise is what determines the outcome for children.
The greater the excercise of the inherent IQ the more use the posseser will get out of it. Just like muscles the more you use them the more you can use them.

All the money and programs and interventions will not overcome a defective, dysfunctional upbringing.

So the liberals can rest easy--they have nothing to prove, because none of anything they advocate works, they are better off making beleive.

thedaddy

According to some theories, omega-3 fatty acids and the amino acid tyrosine can improve mental functioning. So if we can get everybody to eat lox with cream cheese...

$23,000 per year seems good value, seeing as though incarceration likely costs much more.

"So the liberals can rest easy--they have nothing to prove, because none of anything they advocate works, they are better off making beleive."

Speaking of 'make-believe', how's Iraq coming along?

Speaking of 'make-believe', how's Iraq coming along?

Well enough, thank you. You have been following Chenkoff's Good News From Iraq, haven't you?

Steven Pinker in The Blank Slate has an excellent non-technical discussion of the relative effects of environment (understood as upbringing) and heredity. In short, upbringing has an effect on academic performance, but essentially none on IQ by adulthood.

The Perry Preschool Project was very small (around 60 children if I recall), extremely intensive (no guarantee it could be scaled up), and improved results for only a fraction of the children, so the cost per child favorably influenced has to bear the costs of the majority of children who did no better. And those are 1960s' dollars, too.

Even adoption -- surely the most intensive intervention in a child's life we can readily imagine -- has essentially no effect on adult IQ.

Even if you were convinced that Perry Preschool is cost effective as against no preschool, you don't know if it's cost effective against non-Head Start-oriented ideas. For example, I would venture to guess that it's radically cost ineffective against some incredibly successfuly middle school programs, such as KIPP (Knowledge is Power Program) and Amistad Academy which are operating in various urban school systems and funded as charter schools. Amistad and KIPP schools (just two examples of many successful programs cropping up as a result of charter funding) are (usually) no more expensive than regular public schools and are consistently transformational for three quarters or more of their students, ie, a greater percentage of kids benefit and more profoundly than at Perry Preschool. Another question is whether results get even better by combining Head Start with the KIPP/Amistad model. I haven't a clue, but it underscores the folly of the treasure hunt mentality that leads to obsessions with the likes of Head Start at the cost of seeking out the true -- and, it turns out, more humane -- least cost alternative.

If they can create a KIPP school system from pre-school to high school, combined with ensuring the kids eat lots of omega-3s, play games like chess or Go (and other games that work the brain) almost every day, you'd probably have a great deal of success. Imagine if all schools were built in such a way. How much better would education be for all students?

Jane,

While Caplan's comments do seem to be essentially correct, it strikes me that there might be a methodological issue with the conclusions. Specifically, given that IQ itself may be a function of some of the other variables he uses (education, parental income, etc.), I think trying to look at this through straight multivariate regression might be wrong econometrically. I think simultaneous specification of multiple equations might be a little more appropriate.

Yeah. What he said.

I'm curious, how do you measure cognitive capability? How do you know that the items you include on such a test actually measure that? How do you know that the items are inclusive, and that you haven't left out measurement of even more critical areas of intelligence?

IQ only matters if it measures something. What if it measures how rich your parents are?

Interventions to "correct" IQ may be nearly useless, but there are some early-life influences which can depress IQ; interventions to reduce or eliminate those factors may be significantly beneficial.

Murray's "The Bell Curve" addressed all the above issues ten years ago. The suggestion that maybe "IQ measures how rich your parents are" reflects profound ignorance of a hundred years of scholarship and practical results in every segment of dozens of populations.

I grew up in front of a TV, had a single-parent mother, was a latch-key child, lived off of food stamps and government subsidized housing. Yet apparently I'm still smarter than thedaddy who thinks anecdotes prove anything.

the project did not turn poor kids into middle-class adults; it turned poor kids into more functional low-income adults.

And what of the next generation down? It seems likely that there might be an escalating effect with each successive generation - at a lower marginal costs (as each successive generation improves its lot, less intervention is required). The question is, do you get a savings over multiple generations? My purely speculative, wild-assed guess is that you would - IFF you can maintain progress over that time period. That's admittedly a pretty big hurdle, but it's still worth considering.

The fact that you are smart and you didn't watch TV when you are growing up doesn't really mean anything. Maybe if you had watched TV you would have been smarter.

If watching TV is a serious concern, it isn't too difficult to design a study where you take a set of parents, interview them every year about how much TV their kids watch, then when they are 12 (or 18 or whatever) you see how smart they are. If no study like that is available, then the best you can say is, "we really don't know".

Dunno much about what's written above. From practical experience and my own observations of the world (anecdotes, eh?), it seems that, genetically, we are predisposed to certain conditions. I can't see any reason why intelligence wouldn't be one of those conditions. And it's the environment we grew up in that either triggers a given condition or depresses it. Money helps trigger intelligence when it buys the right education AND you have a predisposition for intelligence. But I've seen a lot of dumb people pass law school with the help of their parents' money. Alternatively, I've seen a lot of smart people never wake up to their potential because mom was too busy working two jobs to offer anything that would trigger her child's intelligence, had he been born with any. If I am right about intelligence being a predisposition that may or may not be triggered by the right environment, then don't we have a duty - just as human beings, not neocons or liberals or any of those other false, manmade labels - to implement programs that give everyone the opportunity to wake up to his/her intelligence potential regardless of his/her social/economic background? I'm not saying we don't have a right to the spoils of our own hard work, but tax dollars spent on education benefit all of us, not because we become safer when poor children grow up to be middle class adults, but because the more intelligent people we have out there: 1) the richer the marketplace of ideas, 2) the safer we are because more people are making more intelligent choices (ie, robbery is a bad choice to poverty, finding a way to cut expenses in addition to seeking financial help is a good choice to poverty), and 3) the more people who are exercising their intelligence, the more people are passing on the ability to exercise good choices to their children.

Nobody has questioned the nexus between intelligence and earning capacity - the data above seems to ignore any other measure of intelligence. Perhaps an intelligent person doesn't want to be the next Bill Gates? Perhaps these criticized educational programs are working better than we give them credit for, but we're not asking the right questions of them?

If I am right, then isn't how well adjusted poor children become as adults just as real a measure of the success of these programs as whether or not poor children are growing up to command six-figure salaries? I'm not certain that this debate can be neatly quantified into a blackline rule. I think the real mistake we're making is not in spending money for educational programs for the poor, but in attempting to equate money with intelligence and then using that to quantify the success of any of these programs. What results is the quantifying of human characteristics when we are all so individual and unique. I'm poor and I'm intelligent. Does that mean that the education I've received has been a failure or a waste of money?

While you're at it, ask yourself if an intelligent person would trust any kneejerk reaction he might have to what I suggest?

Belle -- I realize this is going off on a tangent, but I'm curious about your statement "I've seen a lot of dumb people pass law school with the help of their parents' money."

My question is, How does parents' money help a person pass law school? The only direct form of help I can think of is paying for tutors, and I'm not sure if that could really remedy dumbness (as opposed to laziness).

Belle,

Re; "...then don't we have a duty... to implement programs that give everyone the opportunity to wake up to his/her intelligence potential regardless of his/her social/economic background?"

I just saw the movie Serenity. I was impressed and delighted to see a movie with such a strong libertarian theme. To paraphrase my favorite scene...

Question; "Why would they [referring to the rebels in the outlying worlds] resist us when all we wanted to do was to make them happy?"

River; "Because we meddled. People don't like it when you meddle."

So to answer your question - no, we have no such duty - nor the right to meddle.

I wonder about the evolutionary implications of this.

If IQ is hereditary and the distribution in the general population is stable, then IQ does not convey any evolutionary advantage, yes?

Alternately, consider this; some statistics indicate that IQ scores (non-renormallized) have been steadily rising by (IIRC) ~1 point/decade since we started measuring in the late 1800s...

The fact that you are smart and you didn't watch TV when you are growing up doesn't really mean anything. Maybe if you had watched TV you would have been smarter.

Strictly speaking that's a logically correct statement, but since (a) I pointed to the range between birth and Kindergarten and (b) a majority of television programming is merely entertaining pap, a bit of common sense suggests that no, that's not likely -- unless your proclivity was toward Discovery and the History Channel (but even then it is questionable how much useful information a child that young would absorb).

Although I was generally favorable towards its thesis, the Bell Curve was based on dodgy correlations between IQ and SAT scores. Until we conduct an experiment (that is completely unethical) where we separate randomly selected children from their parents at birth and raise them in controlled environments, our conclusions will remain weak and controversial.

Mr. Speirs,

No disrespect, but my issue wasn't raised regarding the role of parental income. It was raised to the issue of model formulation. Its been a few years, but if memory serves me correctly, a model that has multiple variables in which one variable itself plays a role in specifying other variables, doesn't lend itself to a straight multivaraite approach. You have to use a model of simultaneous equations. Now if Caplan is assuming that IQ isn't influenced by things like education or parental income, no problem; but, he should say as much. Otherwise, its not clear that there's not a source of bias here, especially when you start doing significance testing.

There's smart, then there's s m a r t.

The problem, as Bill says, is when you've got variables in your model that have high correlation with each other. For example, IQ scores may correlate highly with parental income (it could be because IQ is hereditary, and higher IQ may cause higher income... or it could be because richer parents tend to be more focused on children's education... or, or, or...) If you try to do multilinear regression with both IQ and parental income, without realizing the mathematical effect high correlation between variables has on their coefficients, you will invariably get just numbers for your results. You may see a huge coefficient for income, and small for IQ... just due to the way the regression was calculated.

In any case, I don't see why IQ's hereditability and/or improveability should have anything to do with social policy. If the point to is make sure people are productive, then measure the outcome of what they do with their lives after whatever training/education they received. I see nothing wrong with aiming for a result where one can support one's self as an adult, maybe as a nursing assistant or a janitor, if not a brain surgeon. It's much better than being homeless or in prison.

People act like if the kids can't get into Harvard after whatever intensive efforts were put in, then it's a waste.

If IQ matters, then analyses that ignore IQ will typically overstate the marginal effect of other variables.

The above is Caplan's key point. His point is reinforced by the Jones & Schneider: 1330 IQ regressions paper:

we show that in growth regressions that include only robust control variables, IQ is statistically significant in 99.7% of these 1330 regressions. A 1 point increase in a nation’s average IQ is associated with a persistent 0.16% annual increase in GDP per capita...

We also evaluate the explanatory power of IQ in growth regressions that include Sala-i-Martin's education measures. Among these 56 education-related regressions, IQ was statistically significant in every one, thus passing not only Sala-i-Martin's robustness test, but also Leamer’s [1983, 1985] extreme bounds test. While one might expect that at least some linear combination of primary, secondary, and higher education measures could eliminate the statistical significance of IQ, we did not find this to be the case...

Finally, for an overall assessment of how IQ compares to other common growth variables, consider Sala-i-Martin's original results, which used combinations of 62 growth variables in over two million regressions. Among his top 21 regressors--the ones which he considered robust--the median regressor was statistically significant in 76.4% of cases, with a range from 100% (for fraction Confucian) to 2.81% (for revolutions and coups). Fraction Confucian was the only regressor that passed an extreme bounds test. Only eight of his top 21 had coefficients over three standard errors from zero, while in our full-sample results using his top 21 growth variables, IQ’s coefficient is over five standard errors away from zero. For his overall best performing variable, equipment investment, the coefficient estimate was 5.32 standard errors away from zero. IQ would thus appear to fit comfortably in the top half of Sala-i-Martin's top 21 growth variables.

Until we conduct an experiment (that is completely unethical) where we separate randomly selected children from their parents at birth and raise them in controlled environments, our conclusions will remain weak and controversial.

This "completely unethical" experiment has been conducted -- it's called adoption. The environments are about as well-controlled as you can ever expect in social science, since adoptive parents are carefully screened.

That said, even with such experimental data, any conclusions about the relation between genetics and IQ will remain controversial until people start using genetic engineering services to actively enhance IQ.

IQ only matters if it measures something. What if it measures how rich your parents are?

If it measured how rich your parents are, identical twins raised separately in families with different incomes would have different IQs. In fact their IQs tend to be highly similar. Also, if IQ measured wealth, decrease in wealth would cause a decrease in IQ. This, also, does not happen.

It does appear that there is an IQ/wealth correlation, but the best evidence we have is that this is primarily because smarter people tend to make more money, and dumber people less.

Until we conduct an experiment (that is completely unethical) where we separate randomly selected children from their parents at birth and raise them in controlled environments, our conclusions will remain weak and controversial

That's not necessarily true at all. We could very well find the genetic origins of intelligence simply by looking at correlations of genes to measurable intelligence test results. That could produce testable, reproducable results without violating ethics rules.

His policy implications simply do not follow.
Let's take the example of an orphan with low enough IQ that that person is totally incapable of taking care of themselves -- what are the policy implications?
There are policy implications, and it may have to do with marginal value, but first you have to define what the goals are.

A key variable is not IQ but literacy, and there are programs that can teach that reliably, such as Phono-Graphix, that are widely ignored.

If the argument is that because of IQ you should not support public education wouldn't the argument apply equally to private education.

Basically, in trying to argue against government he is making the argument that all education does not matter.

If the argument is that because of IQ you should not support public education

The argument is that you shouldn't expect education to increase IQ. We should be satisfied that education cures ignorance, and not expect it to cure stupidity too.

Jane, I think you miss Bryan's point a bit. You say:

As far as I can tell, interventions to correct IQ are nearly prohibitively costly.

True, because as Bryan writes in a follow-up, IQ is nearly perfectly inelastic. That is, we haven't found anything that can affect it in a positive way.

His point was a different one. If you don't control for IQ, its effects are distributed amongst the other variables you're studying. That can make some policy interventions look cost-effective when in fact they are not.

Jim

The long-term increase in IQ over the past century is plausibly attributed to better nutrition and childhood health care...I don't think most of us realize just how bad the diet and health of poor and semi-poor Americans was a hundred years ago. But you can see it in their faces in old photographs if you look for it.

In my opinion this points to: since the characteristic factor of the brain that underlies IQ, known as "g", must represent a physical characteristic of the brain, we should find a way to improve that characteristic through medical intervention. I am sure this will come during this century.

By "medical intervention" I mean drugs, genetic methods, nanorobots...whatever it takes.

"What would my intelligence be today if I had been dumped in front of a television instead?"

From all available evidence, it'd be about the same.

"But I've seen a lot of dumb people pass law school with the help of their parents' money. Alternatively, I've seen a lot of smart people never wake up to their potential because mom was too busy working two jobs to offer anything that would trigger her child's intelligence, had he been born with any."

See, this is the thing: No, you haven't. You haven't seen "dumb" people pass law school. You've seen people that you thought were dumb. Odds are they were, like you, somewhere above the 100 mark.

If you ever work with a group of people with IQs well below 100, you'll know how idiotic it is to say that most of them could make it through law school without the entire law school determined to lie on their behalf--and sadly, the schools do just that.

In the Thernstroms' book, No Excuses, the formerly liberal couple insist that the low-performing kids just don't have the "right culture". They accuse suburban kids of having the "wrong culture", too--in fact, everyone except the Asians have the "wrong culture".

But what works with these groups that have the "wrong culture"? Well, up through 8th grade, at least, what works is 12 hours of education a day. Constant discipline. No extracurricular activities. Continual reminders of expected behavior, signs, slogans, immediate crackdowns on any slips. While their test scores do increase considerably, no mention is made of how they do in high school. I suspect that the added abstractions of advanced ed caused some problems.

Suburban kids, with their lousy culture, seem to do just fine on their test scores with just the normal 8 hours, extracurricular activities, and slothful ways.

But no, the Thernstroms insist. It must be culture.

Fine, whatever. In any case, the answer seems to be that for lower intelligence kids, the best way to improve their educational results is to treat school like a boot camp, don't give them extracurriculars, forcefeed them education, and probably come up with some form of voc-ed in high school rather than pretend that everyone is fit for college.

That shouldn't cost too much money, and it would be money well-spent. But the students would be disproportionately poor and minority and it's never going to happen. Instead, we'll lie about their potential, pretend they're fit for college, and give them a dreadful education that the institution can use to pat itself on the back and the kid can use as toilet paper.

"In any case, the answer seems to be that for lower intelligence kids, the best way to improve their educational results is to treat school like a boot camp, don't give them extracurriculars, forcefeed them education, and probably come up with some form of voc-ed in high school rather than pretend that everyone is fit for college."

Wasn't treating education like boot camp the standard? The conventional wisdom. The way it was always done before we decided that, because we could blow up the whole world and us in it, we knew better than all our elders and could reinvent said world.

If IQ is indeed an accurate predictor of performance, then the best thing we could do for the poorest kids would be to administer IQ tests and give the smartest all the support in the world. Fixing education for the rest of them would undoubtedly save tens of thousands of lives from prison and be the difference between a live of deprivation and a life of self-respect and hard work.

Here is a link to a debate on gender differences in IQ (whether it is caused by biology), although that is not the exact topic at hand, the types of evidence presented to bear on the more general subject of how much of IQ is inherited: http://edge.org/3rd_culture/debate05/debate05_index.html

The observed phenomenon that IQ is rising is called the Flynn effect. It has been happening in the U.S. as well as countries around the world. I don't think most people think that nutrition is enough to account for it. Here are some links for that:
http://www.indiana.edu/~intell/flynneffect.shtml
http://pespmc1.vub.ac.be/FLYNNEFF.html

Also there is a Scientific American article on the Flynn effect, I believe from 1999.

In terms of how much IQ matters, it is one factor of success. Other factors, such as emotional well-being and work ethic. Maybe some of those other factors would be more cost effective to improve. These other factors, by the way, do not seem to be improving on their own over time :)

Wasn't treating education like boot camp the standard? The conventional wisdom.

Not even remotely. The average number of hours children spend being educated has increased since the "good old days". Extracurricular activities aren't anything new either.

The big change between then and now is that these days vocational education is discouraged; people are put on the college track whether they're bright enough for it or not.

"Not even remotely. The average number of hours children spend being educated has increased since the "good old days". Extracurricular activities aren't anything new either."

No, I'm pretty sure it was. If you think back to the way schools used to be, they were extremely strict, the teacher could hit the kid, it was very structured. Hours don't have much to do with it so much as what happens in those hours. I'm reminded of a City Journal article about how the poor in the UK during the Industrial Revolution, even the coal miners, etc, were highly educated, more educated in some ways than your average Ivy League student. The article discusses what school was like then, and today your average person would probably deride it as fascist but it worked. It's almost like the students of that education grew up and rebelled against it so we got the crap that passes for education today.

No, I'm pretty sure it was. If you think back to the way schools used to be, they were extremely strict, the teacher could hit the kid, it was very structured.

I was told by a couple UK residents that in Irish public schools, there is still a lot of traditional religious influence and it is implemented with a degree of strictness. In those same schools, the students keep quiet during lectures and regularly use genteel titles when addressing adults. In Britain proper, by contrast, the education has been secularized as heavily as anywhere else, with the predictable results.

No, I'm pretty sure it was. If you think back to the way schools used to be, they were extremely strict, the teacher could hit the kid, it was very structured

Aside from the fact that teachers were allowed to hit children, in what way are you saying that it was more structured?

even the coal miners, etc, were highly educated, more educated in some ways than your average Ivy League student

Well, that's certainly an interesting assertion. I don't believe it for a minute, but it is certainly interesting. :)

Here's the article in question. "The Classics in the Slums" Read it and weep.

Ok, I read it. It didn't answer any of my questions, I'm afraid, nor did it support your claim that 19th century coal miners were more educated, in any way, than Ivy League students. A handful of anecdotes is no substitute for actual hard data.

It isn't a handful of anecdotes. That was the education system back then.

I've read the article; as mariana points out, it's not merely "a handful of anecdotes", though anecdotes about particular working-class people of wide reading in the classics do make up a fair proportion of the article. It seems to present a fair bit of evidence that these people were typical rather than exceptional, or at least a large minority of working-class people in late 19th & early 20th century Britain (there's less evidence presented about other countries at the same period).

I would hesitate to say on the basis of this article and other sources I've read that "19th century coal miners were more educated, in any way, than Ivy League students". But I would venture to say that in literature, at least, they were probably better educated than the typical working-class student of the present day who doesn't go beyond high school, and arguably better educated than a fairly large subset of those who do go on to college.

"A 1 point increase in a nation’s average IQ is associated with a persistent 0.16% annual increase in GDP per capita..."; and presumably vice versa, which could explain part of the drift upwards in IQ?

Comments are Closed.