What can parents influence?

I had been meaning to pen a longer response to Bryan Caplan (he is the one with "a theory of everything" in this area, not I, his theory just happens to have few variables), but I'll focus on two of his claims, as they are indicative of the larger disagreement:

Parents strongly affect what you say your religion is, but have little long-run effect on your intrinsic religiosity or observance.  I don't discuss language, but it's pretty clear how a twin or adoption study would play out: You can make your kid semi-fluent in another language with a lot of effort.

Both claims are false, at least at many commonly available margins.

Take Jews.  If a group of children are born to Orthodox, Conservative, Reform, or liberal parents, their later religious observance will be predicted by both peers and parental upbringing.  (Perhaps genes are a factor too.)  An Orthodox Jewish boy, with Orthodox parents, growing up in an otherwise non-Orthodox Jewish community of peers, is more likely to stay Orthodox than a Reform Jew from the same community is likely to become Orthodox.  And that of course correlates with levels of observance.  I have no formal study to cite, but can I just stamp my feet and scream this is true?  Because it is.  You can imagine numerous variants on this tale, even if it isn't true for all religious denominations.

Bryan has a tendency to concede environmental factors by noting something like "Of course parents can lock a kid in the closet and affect him that way."  He is less likely to admit that a lot of less extreme influences can matter too and that those influences are missed by twin adoption studies, for whatever reason.

Or take language.  Yana speaks Russian.  She learned Russian from Natasha (her mother, and it wasn't hard for her to speak Russian at home), and note that Yana left Moscow before she was two years old.  This is again a common pattern.  The parents matter, even though in most American families you won't see enough cross-sectional variation (most people speak English at home) to always pick this up.  Travel around India for more examples of this phenomenon.

Presumably the twin studies have in their data sets Jews and possibly some Russian immigrants as well.  And yet the twin studies, with their ultimately macro orientation, miss micro mechanisms such as these.  Parents can matter more than the studies suggest.

By treating those studies as an epistemic trump card, Bryan is led to make claims which are indefensible on the face of it.  I stick by my earlier points.  The evidence Bryan is citing for twin adoption studies is simply…the studies themselves.  Where is consilience when you need it?  


We show that genetics makes people more different than parenting. But the purpose of parenting is to make people similar.

For twin studies and some of their defects, you might want to check out Richard E Nisbett's "Intelligence and How to Get It: Why Schools and Cultures Count" (Ten bucks on Kindle)

He argues rather convincingly that the evidence of twin studies has been exaggerated, mainly by unaccounted for correlations.

Also, I don't recall seeing mention of David Cesarini's recent work casting doubt on (some aspects of?) twin studies as a solid approach to identification of treatment effects.
See: http://ideas.repec.org/e/pce28.html

In other words, the twin studies have a very low signal to noise ratio. Bryan is assuming absence of evidence in existing studies means absence of effect. In contrast you're arguing that the studies show large, crude effects but are poorly designed to pick out important but weaker effects for smaller subgroups with more complicated interactions. This is a common problem in double blind studies of audio and wine and leads to the same kind of shouting matches between true believers and skeptics.

One methodological issue that isn't mentioned much in all the talk about twin and adoption studies is the impact of sibling rivalry. For example, knowledge that the other sib is slightly better at something often drives a kid in the opposite direction. Thus, strange as it may seem, in some cases siblings might wind up more similar if they weren't raised in the same home.

There is a little nuance lost in Cowen's comments about religous communities. What Judith Rich argues (and Caplan apparently rehashes) is that the impact parents have is primarily in choosing the peer environment of their children. Thus, there will always be a regression to the mean of the children's peers away from the cultural idiosyncracies of the parents' religion, language, and culture. So, yes, of course a child with orthodox parents will be more like his peers than his peers will his orthodox parents. This is pretty well accepted in practice, that's why catholic parents and jewish parents spend so much on religous schools so that their kids grow up in an environment of nothing but like minded kids. If parents really had significant influence they would be able to raise orthodox kids in the middle of WASP suburbs. They don't because they can't.

"He is less likely to admit that a lot of less extreme influences can matter too and that those influences are missed by twin adoption studies, for whatever reason."

I think a good, convincing hypothesis for how the twins studies fail is needed. It's not generally good form to reject something because you don't like the conclusions but can find no flaw in the methodology.

> He found out the "Harvard number" is currently $5 million, which he found
> profoundly discouraging.

I suspect he was misled. For HBS (admittedly, not undergrad), the number is much, much lower. Orders of magnitude lower.

Tyler, you and Bryan are using different definitions of "religion". You're focused on the "publicly stated" affiliation, and he's focused on what people actually believe.

You do know that there are many Jewish atheists/agnostics, right?

I am comforted by the evidence that genetics could lead any potential progeny of mine to atheist/agnosticism, regardless of my parenting skills/effort.

The idea of any child of mine being a "true believer" of anything is terrifying, and I will sleep a little better tonight.

Where is consilience when you need it?

That families of immigrants lose their old language, except when there was a lot of emmigration. That the deaf children who are the offspring of hearing parents participate in Deaf culture (not just in terms of language, but in other cultural matters like attitudes towards timeliness). Studies showing that babies who are insecurely attached to their mothers don't show any repeatable difference in their attachment to other people than babies who are securely attached to their mothers. That adopted children are not more like their biological siblings than other kids that grew up in the same neighbourhood but different houses.

@Cournot, there are a number of problems with your analogy.
Firstly, the correct analogy would be that, in numerous studies, the drug showed no significant effect on fever. While there was a number of anecdotal claims that it did have an effect, it's not visible there in the clinical data once an effort has been made to control for confounding factors.

Secondly, for the analogy, there are significant side effects from taking the drug - for example, parents have been held legally responsible for what their teenage children do, on the assumption that parents can control their teenage kids. Another example would be the blaming of autism on refrigerator mums, which apart from the suffering of the parent, wasted a lot of research effort.

Thirdly, there are also theoretical reasons to believe that the drug won't work, in the case of parenting, given that humans evolved in social situations, why should we be evolved to learn from our parents, rather than from a broader group, including the people whose help we will need to survive and raise kids of our own if our parents die before us? (And given that everyone alive today is exclusively descended from those who lived long enough to reach sexual maturity, that was a very common situation for our genetic ancestors).

So, by analogy, believing in parental influence as a default is like taking a drug:
- Even though clincial studies show it's ineffective
- Even though there have been some bad side effects from the drug
- Even though there are theoretical reasons why the drug shouldn't work.

Would you really take a drug under these circumstances?


I reject your restatement. Genetic influence is one strong effect. Parenting involves a multitude of treatments with a lot of nonlinearities. To say that parenting doesn't have large visible effects on the average doesn't mean it doesn't matter that some sorts of parenting matter under specific and idiosyncratic conditions. Parenting is like talking about many other drugs/treatments while genetics is one big class of drugs.

The fact that it can't be picked up clearly in the studies is in my view exactly like the side effects we tend to observe.

This is especially true given that peer groups matter and that parents work hard to select the peer groups of their kids. To test otherwise, I'd like to see a study where one group of parents with biologically similar kids went out of their way to select peer groups designed to produce the opposite outcomes on kids.

And as for religion, Caplan treats identification as irrelevant. But if you were orthodox I doubt you would think it irrelevant if your child were kidnapped and raised as a fundamentalist evangelical. To Caplan, these might both involve equivalent religious temperament. To most parents they would be vastly different.

Finally, to push the analogy further, if there were a drug with statistically insignificant results (say at the .85 not .95 level) but with few side effects and large if uncertain observed benefits, I would say YES, I would want to take that drug. Indeed, most people should say the same.

"Yeah and look at how great those societies living by Sailer-style ethics are doing."

And so why does Tyler promote importing millions of illegal immigrants from cultures with what you call "Sailer-style ethics"? And why does Bryan promote importing millions of illegal immigrants from populations with many generations of poor achievement both in their own countries and in America?

Mr. Caplan errs by only evaluating "how kids turn out" by the metrics he cares about, which are primarily economic or status-driven. I work with people mostly my same age. We are all at roughly the same level of intelligence, with the same job and the same income. But I was raised Mormon and they were not. Ergo, none of them are married and will not be for many more years. They go to bars and sleep around. When they do marry, they might have two kids at most. Whereas I am in my 20s and working on kid #2 and plan to have many more. We read the scriptures and don't drink.

Most normal people would consider factors like 1) how many kids you have 2) how many sexual partners you have and 3) what you eat and drink, to be pretty big determinants of "what kind of person you are" for good or ill. Thus, by raising my kids as solid Mormons, I may not raise their income or intelligence, but I sure can affect "how they turn out" in the ways I care about most.

And if we want to worry more about the genetic profile of our population, let's get rid of anyone of German descent first. The Germanic mindset has created our bureaucratic society. Ben Franklin didn't want the Germans here and I don't want them either. Perhaps libertarianism would have a chance if we got rid of those Teutonic shitbags.

MC - were you raised by your biological parents, or by non-related adoptive parents? Did you attend church groups with other Mormon children (sorry I don't know if Mormons have Sunday School or the equivalent). We might be seeing peer effects or genetic effects, eg say that you and your ancestors have a brain that tends to value the longer-term benefits of sobriety and faithfulness over the short-term benefits of drinking and casual sex.

t that happens to everything, doesn't it?

Comments for this post are closed