Bryan Caplan  

How Twin Research Changed My Life

PRINT
School Nurses and Econ 101... Posner on Means-Testing...
My autobiographical essay, "How Twin Research Changed My Life," is now up at the Wall St. Journal's Idea Market.  They changed my title to "Twin Lessons: Have More Kids.  Pay Less Attention to Them," which isn't exactly my style or advice, but close enough.  Opening grafs:

Nine years ago my wife had her first sonogram.  The technician seemed to be asking routine questions: "How long have you been pregnant?"  "Twelve weeks." "Any family history of genetic diseases?"  "No."  "Any family history of twins?" "No." Then she showed us the screen.  "Well, you're having twins." My wife and I were scared.  We were first-time parents. How were we supposed to raise two babies at the same time?

Strangely enough, I already knew a lot about twins. I'd been an avid consumer of twin research for years. Identical twins (like ours turned out to be) share all their genes; fraternal twins share only half. Researchers in medicine, psychology, economics, and sociology have spent decades comparing these two types of twins to disentangle the effects of nature and nurture. But as our due date approached, none of my book learning seemed remotely helpful.

Only after our twins were born did I gradually realize how much I was missing. Twin researchers rarely offer parenting advice. But much practical guidance is implicit in the science.

The WSJ will feature critical responses and online debate later this week.  Stay tuned.


Comments and Sharing





COMMENTS (22 to date)
Mariana writes:

I have recently discovered that I'm pregnant and so this post has especially touched me. I haven't had my first sonogram yet but I do hope I'm not expecting twins... it has to be really overwhelming!

Ignacio Concha writes:

Great column. I will show it to my wife so she understands the point I have trying to make after reading your posts. I live in Chile, but next time I am in the USA, I will buy your book to give it to her to reinforce this point. She needs to relax with our lovely girls.

English Professor writes:

OK, Bryan, you gotta help me here. I've been reading your posts like this for a long time now, and every time I read one, I am convinced you are right. Then I look around the condo building where I live, and experience seems to disagree with you.

First, culture APPEARS to play a much larger part than you suggest. Several families from India live in the building, and most of the parents are professionals of one sort or another. (The parents were born in India and generally came here for graduate or professional school.) The children have been raised with a lot of discipline, and they appear to respond without rebellion. All the children are well behaved, and all have done well in school. Most want to become doctors. Several of the children are not as bright as their siblings, but they seem to work just as hard, even though they will probably never make it to medical school. There is nothing like this sort of uniformity among the children of parents of European descent born in the USA. You keep asserting that studies say that nurture means almost nothing in the long run, but at least in this condo complex, the disciplined up-bringing of the Indian children appears to have them headed to careers as professionals. There is no similar likelihood for many of the children of American parents who have similar jobs and similar IQs, but who don't apply the same discipline.

Second, if nurture is unimportant, why does divorce appear to have such devastating effects on the lives of children?

I acknowledge that you have "evidence" for every claim that you make, but why should my personal experience (and that of almost EVERYONE ELSE) be so at odds with the these claims?

John writes:

I read your article. It's odd that no one seems to take the "oh, that's an interesting take" stance to your claims. It's almost like it's a political issue rather than a scientific one. I guess only passionate disagreement causes one to comment. But Bryan, if you're right, doesn't it make you wonder why people so systematically overestimate the benefit of parenting when they will have to pay for it?

fructose writes:

John: I thought of the same thing. I assume it is because the real reason people overparent their kids is because it is the stylish, socially applauded thing to do, amongst yuppies and the upper class.

This goes a long way toward explaining why upper middle class and upper class people have fewer kids, even though they can afford more. Their friends and neighbors will look down on them if little Madison or Cody isn't taking violin classes, learning French and Mandarin, and writing narrative essays by the time he or she is 9.

GT writes:

Brian,

It appears you are putting FAR too much weight on the ability of twin studies. Even most authors themselves will tell that it is difficult to distinguish which is the result of genetics. And identical twins even whenever they are raised completely separate because they are still likely to face similar environments even if they are both adopted. Adoptive couples will probably provide a similar environment(as opposed to what they may get in a foreign country) they live in the same country and often are faced with the same cultural pressures. They will have identical appearance and there is long list of studies that show the role appearance plays in the way people are treated.

Identical twins even when separated face similar environmental circumstances. I say this as a repentant twin study advocate. There is quite a bit of material published on the issues with twin studies. One example is below.


Sidney J. Segalowitz (1999). Why Twin Studies Really Don't Tell Us Much About Human Heritability. Behavioral and Brain Sciences 22 (5):904-905.

57 writes:

I'm not sure the editors at WSJ understood your article: the caption for the photo accompanying the article reads, "Parents of twins should do less parenting."

Simon K writes:

English Professor - The research says there's not much impact of parenting on *adult* outcomes. Obviously parenting has an impact while the kids are still living with the parents!

lemmy caution writes:

English Professor's anecdote would be more convincing if the Indian patents had well behaved European kids and the European parents had ill behaved Indian kids. Being well behaved and hard working is the big 5 physiological trait "conscientiousness"


Conscientiousness has a moderate heritability (i.e., the proportion of variance due to genes; h2). Loehlin (1992) analysed biological and adoptive parent–child and sibling/twin correlations re-ported from previous studies and showed that variation in conscientiousness could be explained by additive genes (0.22), gene by gene interactions (0.16), shared environment (0.07), and unique environment (0.55).

The parents provide the shared environment which doesn't contribute to a lot of the variance in conscientiousness.

Bo writes:

Bryan, I would love to buy your book for my Barnes & Noble Nook, but the B&N website says that the "NOOKbook (eBook) - not available."

Are there plans to release the book in this format?

Tracy W writes:

GT: Your criticisms of twin studies don't really affect Bryan's suggestions. Firstly, yes, twin studies generally can't distinguish between direct genetic effects and indirect genetic effects (eg, a pair of separated twins have the genes to be tall, so they do a bit better at basketball in informal games, so they practice more, and they persuade their adoptive parents to buy them hoops, and eventually as adults they wind up both being quite good at basketball, regardless of their adoptive parents' interests.) But those indirect genetic effects don't mean that high-effort parenting has an effect on children's adult outcomes.

Secondly yes, most twins who are separated at birth are raised by adoptive parents who are similar in terms of culture. That controls for a lot of variation. That makes the twin studies *more* useful for working out what parents should do, than if there was a lot of variation in culture.
Most parents are well aware that they should refrain from hitting their children about the head with frypans, or locking them in the cellar for years (sorry to bring up such unpleasant topics, but from experience if I don't, people will point out that the extremes of bad parenting do have effects), and that if they want their children to grow up speaking and being French, sending them to a French school is the way to do it. What Bryan is advising about is more minor things, like shuttling kids to activities that even the kids don't enjoy.

eccdogg writes:

English Professor,

It is interesting that you bring up those questions because just last night I finished reading the section of Bryan's book where he addresses them.

In that section Bryan points out that parents absolutely CAN have short run affects on their children while they can control them, but that that these changes are not long lasting once the kids are on their own. So your anecdotal evidence does not contradict what Brian claims.

I would also question whether the Indian families are truly similar to the European ones in genetics. I have found that because of language or cultural barriers an immigrant doing the same job as a native will generally be smarter and more hard working (but not necessarily more productive due to culture/language differences).

English Professor writes:

Thanks to all for the comments, but why has no one commented on the issue of divorce? Everything I've ever read about divorce suggests that it is bad for the children. Why should this aspect of environment be important if others are not? From what I've read, the children of divorced parents do worse in school (below the level of expected performance based on IQ) and are more likely to have behavioral problems and get involved with drugs, etc. Or is the key once again "long run" effects? Do these effects disappear over time?

Finch writes:

It's probably a combination of "it disappears over time" and "people who's parent's get divorced have genes that incline them to bad outcomes". So the effect is smaller than you see at a young age, and partially explained by genetics. But I've not seen a study on this precise issue.

Even parental death has only a small effect on lifetime earnings:

http://www.iza.org/index_html?lang=en&mainframe=http%3A//www.iza.org/en/webcontent/publications/papers/viewAbstract%3Fdp_id%3D5425&topSelect=publications&subSelect=papers

lemmy caution writes:

Divorce may be more devastating than parental death. The effects of parental divorce on children's lifespans may be significant:

The early death of a parent had no measurable effect on children's life spans or mortality risk, but the long-term health effects of broken families were often devastating. Parental divorce during childhood emerged as the single strongest predictor of early death in adulthood. The grown children of divorced parents died almost five years earlier, on average, than children from intact families. The causes of death ranged from accidents and violence to cancer, heart attack and stroke. Parental break-ups remain, the authors say, among the most traumatic and harmful events for children.


Some of this may be due to genetics, but this children of twins study suggests that there may be some effect of the divorce itself:

The association between parental divorce and offspring substance use problems remained robust when controlling for genetic and environmental risk from the twin parent associated with parental divorce, and measured characteristics of both parents. The results do not prove, but are consistent with, a causal connection. In contrast, the analyses suggest that shared genetic liability in parents and their offspring accounts for the increased risk of internalizing problems in adult offspring from divorced families

My best guess is that parental divorce today is less devastating than parental divorce back when the longevity study parents divorced (and less indicative of genetic issues) as a result of divorce's increasing popularity. But, yeah. Don't get a divorce if you can avoid it.

tracy w writes:

Divorced children also often move homes as a result of the divorce, breaking peer links. And, as maintaining two households is more expensive than maintaining one, the new home/s are often in a lower socio-economic area, affecting available peer groups.

Ted Craig writes:

"Identical twins (like ours turned out to be) share all their genes"

No, they don't.

Scott F writes:

Sorry guys but, libertarian fantasies notwithstanding, there is only half a truth here. Once you focus on twins raised in poverty, as Eric Turkheimer and several French studies have, you find that the heritability of IQ plummets. Perhaps there is an "enough" level where intellect is free to grow and thrive but it is clear that this level is not universally achieved.

Finch writes:

Thank you for the links lc.

> Don't get a divorce if you can avoid it.

Yeah, while it might be hard, this is good advice for a lot of reasons.

Tracy W writes:

Scott F - where people live is affected by how much they earn, people tend to cluster by income to some extent. So twins raised in poverty have different peer groups to twins raised by middle-class parents. We know that peer groups influence adult outcomes, eg your accent as an adult is likely to be that of your peers, not your parents, if the two are different.
So the French study, which went out of its way to find adopted twins into low-socio-economic groups, might be picking up on something the adoptive parents did in the home, or it might be picking up on the location of the home.

Of course bad enough environmental inputs wreck any result, as do bad enough genetic inputs (no matter how many resources you lavish on a rock, it's never going to grow into a human adult). The discussion here is about variations in ordinary parenting.

Kevin writes:

Very interesting, but how do you reconcile these twin studies with research on early childhood education? It seems that better nurturing early on makes a huge difference in outcomes, at least for preschoolers-kindergardeners.

Maybe parents should lighten up as kids get older, but still invest heavily when their kids are very young? I'm not sure how the twin data would resolve this.

Tracy W writes:

Kevin - early childhood education typically involves taking a bunch of kids from several different families and doing stuff with them. As normal kids obviously learn some things from their peers (eg language, accents), then the effects seen from early childhood are entirely possibly down to the group effect on the peers, rather than what the parents are doing.

Comments for this entry have been closed
Return to top