Bryan Caplan  

Where Dysgenics Goes Wrong: Comparative Advantage Strikes Again

Confirmation Bias confirmed... Mozart: An Ultimate Resource...

A number of smart people, Charles Murray included, are worried about "dysgenic pressure." The story, in brief, is that:

1. Intelligence is highly heritable.

2. The more intelligent have fewer kids than less intelligent.

3. Our average IQ is declining, or (here's the Flynn effect caveat) at least rising at a slower rate than it otherwise would.

4. We've got to get low IQ people to have fewer kids.

The response is predictable: People who find (4) offensive put their heads in the sand about (1), (2), and (3), and people who like (4) insist that it follows straight from the facts.

Once again, both sides are wrong. Yes, there is ample evidence that (1) and (2) are true. I've checked (2) using the NLSY myself, and found that the smartest people really do average almost one child less than people at the other end of the scale. And while all the data says that IQ is rising, it's hard to deny that IQ would have risen more if there were no relationship - or a positive relationship - between IQ and fertility.

But, by an argument parallel to my critique of eugenics using the Law of Comparative Advantage, (4) simply doesn't follow. What happens when low IQ people have more kids? It encourages greater specialization and trade. High-IQ people have a stronger incentive to focus on brainy work, because there are more low-IQ people to handle the non-brainy work.

Of course, another route to the same result would be for high-IQ people to have more kids. And it's plausible that if we could have one more person, it would be better for the world if he had a high IQ. But that's a trade-off we virtually never face in the modern world! There's plenty of food, and if low-IQ families have fewer kids, high-IQ families are not going to "take up the slack."

If you are really worried about dysgenic pressure retarding the advance of civilization, there are two sensible solutions. The first is to encourage high-IQ people to have more kids to increase the supply of brains; the second is to encourage low-IQ people to have more kids to increase the demand for brains. Urging either group to have fewer kids "for the good of society" is not smart.

P.S. If someone has zero or negative productivity in everything, the consolation of the Law of Comparative Advantage is admittedly hollow. But people this incompetent or malevolent are so rare that this counter-example is scarcely worth mentioning.

Comments and Sharing

COMMENTS (14 to date)
dearieme writes:

"people this incompetent or malevolent are so rare..": there is the fear that an ill-designed welfare state could manufacture them in numbers.

Patrick writes:

This is all well and good in a truly libertarian economy, but we don't have that. We have a welfare state which chews up about 30% of GDP (50% in Western Europe).

Anyone with positive net productivity is a benefit to the economy, but a large welfare state makes positive net productivity harder to achieve.

Suppose we decide as a society that it's unconscionable for anyone to live on less than $8,000 a year and we decide to give them cash or in-kind benefits to make up the difference if they can't earn that much. Let's also assume nobody that poor pays taxes. If so, anyone who can't add $8,000 or more to GDP each year is a net drain on the economy.

The more generous the benefits, the larger the fraction of the population has negative net productivity. In a country like Sweden, I'm sure all of the bottom decile falls into this category.

Dog of Justice writes:

The simplest reason to not worry about dysgenics is the imminent arrival of at least limited human germline engineering. It takes multiple centuries for dysgenics to inflict a significant toll (or, on the flip side, traditional eugenics to achieve anything); history is now moving too quickly for that sort of thing to be relevant.

Stefano writes:

One of the reasons people have children is as insurance for the future, to have somebody who will help in their old age.

Rationally, low-IQ (and thus poor) parents expect their children to have a low probability of success in life. Therefore, they compensate increasing the number of children, hoping that at least some of them will be successful enough to support them in old age.

High-IQ, rich parents know their children will be successful: they have their first-rate genes, can afford the best schools, etc. Every one of them will be able to support the parents in old age. No need to have many of them.

So, to encourage high-IQ people to have more children, and low-IQ people to have less, we have to make high-IQ people less confident about the future, and low-IQ people more confident.

Looks like a case for progressive taxation to me. ;-)

olivier writes:

IQ is not 'highly' heritable. Smart parents ALSO have smarter kids because they have more financial, intellectual and social capital resources to distribute to their children.

A large number of the study of twins reared apart was undertaken by Thomas Bouchard of the University of Minnesota starting in 1979. He “collected” pairs of separated twins from all over the world and reunited them while testing their personalities and IQs. Other studies at the same time concentrated on comparing the IQs of adopted people with those of their adopted parents and their biological parents or their siblings. Put all these studies together, which include the IQ tests of tens of thousands of individuals, and the table looks like this:

Concordance rates of IQ scores: [The number is a percentage correlation.]

* Same person tested twice 87%
* Identical twins reared together 86%
* Identical twins reared apart 76%
* Fraternal twins reared together 55%
* Biological siblings reared together 47% (studies show that reared apart about 24%)
* Parents and children living together 40%
* Parents and children living apart 31%
* Adopted children living together 0%
* Unrelated people living apart 0%

Ridley, 1999, p.83 from Genome: The autobiography of a species in 23 chapters. London: Fourth Estate Ltd.

Conchis writes:

I agree that (1), (2) and (3) do not necessarily imply (4). But nor does comparative advantage necessarily imply ~(4).

If we confine our analysis to the effects on the current population, population growth is essentially analogous to immigration. Is low-skilled immigration good? There's a vast literature devoted to this, but the essential questions are these:

(1) Is having more low-skilled people better than having no more people at all? This depends on:

(a) whether newcomers improve the productivity of the existing population (comparative advantage says it does); but also on

(b) whether newcomers receive transfers from the existing population that outweighs these productivity benefits, either through: (i) having a negative fiscal impact; or (ii)crime etc. (a la Steve Levitt)?

(2) Even if having more low-skilled people is better than having no more people, does it crowd-out increases in the high-skilled population? Although other factors are probably more important, those mentioned in (b), which effectively decrease the return to being high-skilled, probably do have this effect at the margin.

Tom writes:

To take an example from your earlier post, suppose Brains can make 5 Computer Programs or 10 Bushels of Wheat per day, and Brawns can make .1 Computer Programs or 5 Bushels of Wheat per day. As you point out,

Brains and Brawns can still trade to mutual benefit: Just have one Brain switch from farming to programming (+5 Programs, -10 Bushels of Wheat), and three Brawns switch from programming to farming (-.3 Programs, +15 Bushels of Wheat), and total production rises by 4.7 Programs and 5 Bushels of Wheat.
But that's true only for a given population of Brains and Brawns. On the other hand the addition of a Brain increases total output by 5 programs or 10 bushels of wheat, whereas the addition of a Brawn increases total output by only .1 program or 5 bushels of wheat. It seems to me that I would always prefer the addition of a Brain to the addition of a Brawn. I'm not making an argument for eugenics -- just an observation about the mathematics of the problem.

Eli writes:

Bryan, I fail to understand why a rational agent should care about the "advance of civilization" after he is long gone. (Eu|Dys)genics takes centuries to work. I am young, but I will almost surely not be around in, say, 2080. My children probably don't have much beyond 2110. Beyond that, should I care? If so, why?

Max writes:

I think you are dead right, Sir! But there is one problem in all this. What if brainy work goes up and non-brainy work goes down, because of automatisation?

In this case, a society is always better off with more kids of High IQ than low IQ and it has a problem when it has too few High IQs.

This assumption also brings a problem to the Comparative Advantage, because it distorts the available work and the division between the brainies and the brawns.

Daveg writes:

To test your theory you should go live in a country with a low average IQ and see what you quality of life is like.

You can pretty much pick any african nation south of the Sahara for this test, except for SA, although they might join that club soon enough.

Feel free to include me in your will.

quadrupole writes:

Competetive advantage *does* handle even situations in which people are a net drain on society. Imagine that Person A can produce 5 programs or 10 bushels of wheat per day, and Person B can produce 0.1 programs or 5 bushels of wheat per day. Imagine further that a program is valued at $100 and a bushel of wheat at $4. So given a 250 work year, Person A could make $125,000 a year writing programs, or $10,000 a year growing wheat. Person B could make $2500 a year writing programs, or $5000 a year growing wheat.

Say we as a society decide that no one should have to live on less than $10,000 a year. Person A could do that at either job. Person B can do that at neither. Imagine that two new instances of Person B pop up freeing up a Person A to write programs. Person A experiences a net gain of $115k in income. If we capture $10k of this gain to top off the two person B's income, person B is still $95k better off.

All of this of course ignores capital effects (ie, maybe for the cost of thousands of person B's at a negative $5k per year contribution, we'd be better off buying a LOT of capital. If we could invest in capital to allow person A to produce 250 bushels a day, then we'd be better off doing that than bleeding into a large number of person Bs).

Cold Water writes:

(a) I would prefer to change step (4) to: encourage smart people to have more children. One easy way to do this: even out taxes and subsidies. Since we literally pay feckless people to produce children (EITC and TANF), why do we levy extra taxes (e.g., child-care-deduction limits & phaseouts) on productive people with kids?

(b) Very many less-intelligent people in our society do exhibit negative productivity in everything! We keep a lot of those people in prisons, but those who are still on the loose cost the US on the order of $600 billion in 1993, according to the best estimates (reviewed here:
). The costs of crime are orders of magnitude larger than the productive value of criminals' occasional, desultory labor. I won't even mention welfare costs.

Patrick writes:

quadrupole, you're correct in saying that specialization would reduce the burden on the welfare state, but that wasn't the issue at hand. Caplan was saying that the nation is wealthier with more people, even if their IQs are low. He would be right in the special case of a libertarian economy. In a welfare state, though, the nation becomes poorer whenever a person shows up with productivity below a certain threshold determined by the generosity of welfare benefits.

In a welfare state, the case for proposition 4 ("We've got to get low IQ people to have fewer kids") is strong. The more generous the state, the stronger the case.

John S Bolton writes:

If comparative advantage can be applied to classes within a country, does it follow that a gain of that kind will not be a net gain, if a loss of comparative advantage occurs from immigration causing us to be that much less specialized?

Comments for this entry have been closed
Return to top