Bryan Caplan  

Goaded

PRINT
The Fiscal Singularity is Near... Monterey Tea Party...
I'm touched to see Tyler publicly defending me and my clone, and think I ought to respond to his only reservation:
If I have any criticism of Bryan, it's that he's pro-natalist (fine in my book) but I've never heard him promote the idea of adopting a child or defend the idea of raising a biological child who is, for whatever reason, very different from his or her parents.  (Don't overreact here and interpret his silence in a negative way, I'm simply goading him to take up these issues, which I think will force him to revise his thought.) 
I probably haven't addressed these issues because my views are conventional.  On adoption: I think that adoption is a noble, generous act, and admire those who do it.  But I personally don't want to adopt.  On raising a biological child very different from myself: Of course I'd still love and raise him/her.  My post on "parenthood as the trump of all past regret" is predicated on this endowment effect.  Still, I'm honest enough to admit that I'd be happier if my child and I had a lot in common.

Tyler goes on:
Furthermore I think his intuitions about similarity, and child-rearing, will change once (some of) his kids start rebelling against him.
I'm puzzled.  Since my current view is that nurture has little long-run effect on kids, why would their rebellion make me rethink my position?


Comments and Sharing





COMMENTS (18 to date)
Taimyoboi writes:

Postulate: The only reason to be pro-natalist is to create genetic diversity.

Blackadder writes:

Since my current view is that nurture has little long-run effect on kids, why would their rebellion make me rethink my position?

Because rebellious children are no fun regardless of whether their rebellion has any long term impact on them (I think the implicit assumption here is that a clone of yourself would be more rebellious than a child who shared only half of your genes).

kebko writes:

You think adoption is noble & generous, and since you hold the conceit that genetics has something to do with your familial connections, you are psychologically pre-ordaining yourself to this view.

Most adoptive parents will tell you that they are not noble or generous, but in fact are just as selfish as you are. Confirmation bias & all the other biases that bond us in tribal affiliations are mostly what is a work - with them & with you. You have just never experienced the process in a way that forced you to question your conceit about the genetic connection.

Just as the ickiness factor is blocking many of the commenters from seeing the possibilities of your cloning plans, your conceit about genetics is blocking you from seeing the non-genetic basis of familial bonding.

I would hope that if you have a clone, those commenters would overcome their biases to treat your child respectfully, and I would hope that if you are ever in a parental role with no genetic connection, that you would overcome your biases in order to be a full parent to your child.

Badger writes:

Blackadder makes a good point: one of the reasons why I'd hate to have myself as a father is because I'd probably feel like I'd have to rebel against what he represents much more than with a typical, "not like me" kind of father.

To be more precise, I think Caplan's mistake is to underestimate the value some people put on their individuality (pretty valuable to me).

Dan Weber writes:

The year is 2150, and Bryan's clone has just written the best selling book "Why to be Anti-Natalist." Discuss.

Philo writes:

"Since my current view is that nurture has little long-run effect on kids, why would their rebellion make me rethink my position?" Tyler must have been thinking that it is the inherent *difference* between parents and children that causes the child to rebel; if your clone rebelled against you, it would show that there was substantial difference between you, in spite of genetic identity. But you probably reject this account of child rebellion. Your point is that *you* are an inherently rebellious person--you would rebel against your parents, no matter what they were like--so you would expect your clone to be and do likewise.

Loof writes:

Perhaps Bryan was somewhat spoiled (like most American children, I believe) and, so, self-centered; though not a rebellious spoiled brat. And, if he weren’t rebellious; he implies, with selfsame nature, neither would his clone. Problem: even if nurture is only a tiny-weeny bit influential, I see a self-centered father and a selfsame son as a potent recipe for culturing a rebellious brat.

jack writes:

Bryan's belief that nurture has little long-run effect probably stems from having little knowledge of psychology aside from a few twin studies(or sociology or biology for that matter).

His probably limited imagination cannot grasp how different he would be had he had different parents, different experiences, molested by his priest, etc.

Tracy W writes:

Jack, may I recommend Judth Harris's The Nurture Assumption? She makes a rather good case that normal parenting behaviour, within a very wide range (so excluding those parents who starve their kids, or lock them in the basement for 10 years, or beat them about the head) does not have much of a long-run influence on their parents, once you control for genetic relationships and child-to-parent effects.

And I don't know how you think being molested by a priest says anything about the effect of parents on their kids.

Loof writes:

Harris has developed an innovative case for the influence of peer groups on character development. Her critique of studies on parental influence and birth order is relevant. However, the degree of influence with each type (parenting, birth order, early childhood experience, peer group, fashions and fads) is unresolved. Also in dispute are the influences in kind: nurture vs. nature as well as nurture-and-nature. We’ve a lot to learn still.

tom writes:

Bryan, I asked this question in comments at Overcoming Bias: what if your wife said no, she'd like to have the two of you raise her clone instead?

zeljka buturovic writes:

"Harris has developed an innovative case for the influence of peer groups on character development."

harris' theory in "the nurture assumption is innovative but flawed. much of the compelling evidence for the importance of peer groups explains what makes children, including identical twins, alike, rather than different. she herself admitted as much in the "no two alike" which is why she there proposed a different theory.

"Her critique of studies on parental influence and birth order is relevant."

first of all, this is not her critique. those data on heritability have been around forever. the interesting thing about harris is the peer group stuff - not the trait heritability stuff.

in addition, the heritability stuff is not without problems. those data are based on a certain conception of personality that many people find lacking. in fact, in "the nurture assumption" harris herself implicitly criticizes it when she talks about children behaving differently with their peers and their parents and studies showing that personality tests administered in different environments produce different results. so, she senses that trait-based personality is problematic. but she needs data based on just such understanding to discredit importance of parental influence. so, she can't get rid of it.

Tracy W writes:

zeljka buturovic - well there are multiple things going on in terms of how people behave, aren't there? For a start, there is socialisation going on, eg most people acquire the accents of their childhood peer group and then as adults keep those accents in all situations, we know that's not genetic, we know that's not parents. (There are some people who still pick up accents quickly as adults, and apparently some autistic kids keep their parents' accents even if their peers have different ones, and there are probably a few other ones). And there are some other ways in which people within a culture tend to be more alike than otherwise.

Judth Harris's peer group hypothesis in The Nurture Assumption seeks to explain the socialisation part of people, the way in which different people within a culture are similar.

In No Two Alike, she proposes a set of hypotheses to explain how people within a culture are different to each other.

As for the genetics, I am not a psychologist but my understanding was that there do tend to be correlations between how people act in different settings, those correlations are however quite different from 1.

zeljka buturovic writes:

"Judth Harris's peer group hypothesis in The Nurture Assumption seeks to explain the socialisation part of people, the way in which different people within a culture are similar."

no, the book doesn't seek to explain socialization. the point of the book is to explain individual differences (such as those among identical twins) that are not due to genetics. this is why harris talks so much about studies in behavioral genetics - otherwise those studies would have been irrelevant.

however, harris fails in that goal, while nevertheless providing many important and original insights along the way. she mentions that there exist different roles within peer groups that children choose and gradually perfect, but, she doesn't develop this enough and, in any case, the role-based idea of personality is incompatible with the trait-based model that she otherwise must use to argue against parental influence.

"As for the genetics, I am not a psychologist but my understanding was that there do tend to be correlations between how people act in different settings, those correlations are however quite different from 1."

i am not a specialist in the area, but from what i have read, those are rarely above 0.3. few people contest that trait scores are poor predictors of behavior in any particular situation.

Loof writes:

Thank you, zeljka buturovic, for qualifying Harris’ hypothesis and critiques.

An anecdote on birth order effect regarding male sexual orientation: https://www.msu.edu/~breedsm/pdf/BogaertCommentary2006.pdf

Wonder what it'd take for Bryan to change his current view that “nurture has little long-run effect on kids”? Perhaps never, if absolutely preoccupied on nature having almost total long-run effect.

Tracy W writes:

zeljka buturovic - if the Nurture Assumption does not seek to explain socialisation, why does Judith Harris spend so much time talking about things like accents, or how when children change their membership of a peer group at school, their attitudes towards homework change? Or how, based on a Danish study, whether adopted children turn out to be criminals is correlated more with their genetics and the crime rate in the area they were brought up in, not with the criminal nature or otherwise of their adoptive parents?

I do not have my copy of the book to hand at the moment, but I find this claim of yours remarkably at odds with my memory of the book.

You also assert that the idea of role-based personality is incompatible with the trait-based model. But you also agree that there is some correlation across settings. Doesn't this contradict? You dismiss this by saying that's a poor predictor.
There are several things going on. Let's take alcoholism, which I understand has been established to have a genetic link, and a social link. (Eg if someone with the genes for alcoholism lives somewhere where there is very little alcohol around socially, eg, Utah, they are much less likely to become an alcoholic than someone with the genes in a culture where there's lots of alcohol drunken socially).
Now an alcoholic can change how they behave considerably in different situations, eg a early-stage alcoholic might keep away from drinking at work, keep it light and social at the bar, and then soak up the booze at home. So the correlation between the behaviours in different places should be weak. But the genetic link should predict more drinking overall during the course of a year than someone without the genes, for someone in the right culture.

So, say, if you were to predict drinking and could only pick one factor, social situation (bar in New York versus courtroom in New York versus Mormon church gathering in Utah) would be a better predictor than genetics. But if you held the social setting constant, and then added in the genetic influence, the accuracy of your prediction should go up. At that New York bar, the alcoholic presumably is statistically less likely to be sipping a mocktail because they've got an important court case tomorrow.

Am I missing something here?

zeljka buturovic writes:

"You also assert that the idea of role-based personality is incompatible with the trait-based model. But you also agree that there is some correlation across settings. Doesn't this contradict?"

when i mentioned compatibility, i meant theoretical compatibility. it is not easy to reconcile context-dependent conception of personality with the trait-based one.

some people (e.g. michel et al.) tried to develop a context-dependent model of personality by aggregating behavior across types of situations ('behavioral signatures'). in one well-known study, they did get more consistency within situational types. however, the problem they ran into when trying to generalize this approach was that it is very hard to figure out what types of situations are there. this problem stems from the fact that there is no obvious lexical equivalent of situations that could be used to boil them down to types the way adjectives were boiled down to traits. in other words, lexical hypothesis might (still) apply, but it is much harder to execute.

on the other hand, empirically speaking - to take the boring middle road - there is some truth to both. traits do capture something important and they are not entirely useless predictors of behavior, especially across situations. but they are not very good predictors either and fail to capture many aspects of individuality that people care about.

Tracy W writes:

, i meant theoretical compatibility. it is not easy to reconcile context-dependent conception of personality with the trait-based one.

I think we must differ on that - I personally find it rather easy to reconcile the two ocnceptions. When I spelt out that scenario with the alcoholic, I spent far more brain power thinking about the wording than about the basic theoretical setup.

Of course all this may simply mean that I'm being stupid in missing something obvious that complicates the reconcilliation.

Your explanation of why it is hard, that you give here, is a problem that springs from definitions - how do you know which situations should be treated similarly and which treated differently. I agree that working out which situations are similar is a serious problem from a philosophical viewpoint, and this has implications for testing theories, especially for personality traits that, unlike alcoholism, don't have a relatively simple way of aggrating the effect (ie amount of alcohol drunk per year, or number of days a year drunk to amnesia, or etc). I don't however see how this implies that trait-based theories of personality are theoretically in conflict with situation-based theories of personality.

I agree with you that traits are not very good predictors.

Comments for this entry have been closed
Return to top