Bryan Caplan  

Does Asian Parenting Cause Asian Success?

McArdle Waves the White Flag... Temptations to Cruelty...
Scholars familiar with twin and adoption research will be sorely tempted to summarily dismiss Yale Law professor Amy Chua's recent defense of Chinese parenting.  It's hard to find a stauncher defender of what Judith Harris called "the nurture assumption."  Chua:
A lot of people wonder how Chinese parents raise such stereotypically successful kids. They wonder what these parents do to produce so many math whizzes and music prodigies, what it's like inside the family, and whether they could do it too. Well, I can tell them, because I've done it.
Chua then lists all the fun things she denied her kids, the thousands of hours of academic and musical drill, and her generous helpings of shame. 

My initial reaction is exasperation.  Yet another essay on parenting that doesn't even contain the words "genes" or "heredity"?  A vast literature finds that heredity is not merely part of the reason for family resemblance, but virtually the whole story.  How can a professor at Yale act as if this consensus doesn't even exist?  Nevertheless, there are two big reasons why Chua's piece deserves a closer look.

First, Asian parenting techniques seem so extreme, and Asian success seems so pronounced, that most people find it counterintuitive to deny causation.

Second, and more importantly, twin and adoption researchers have largely neglected Asian populations.  The vast majority of twin and adoption studies focus on largely white samples in largely white countries.  Bruce Sacerdote famously studied the effect of (largely white) American parenting on Korean adoptees, but to the best of my knowledge this social experiment has never been reversed.

When you put these two points together, the defender of the efficacy of Chinese parenting could easily say, "Aha!  So you can't disprove our intuition that the extremely strict discipline typical of Asian parents causes Asians' adult success."  And taken literally, this defender of Chinese parenting is right.  Existing twin and adoption evidence can't "disprove" their claims.  But the same holds for all empirical research.  Even in a double-blind experiment, the nay-sayer can still stonewall, "Your results work for your sample.  But my sample is slightly different, so who's to say?"  The reasonable approach isn't to demand decisive disprove of your initial position, but to calmly weigh the available evidence.  By this standard, Chua's claims about Asian parenting fare poorly.

1. There are plenty of strict non-Asian parents.  Chua warns us that, "[W]hen Western parents think they're being strict, they usually don't come close to being Chinese mothers," but she also admits that:
I'm using the term "Chinese mother" loosely. I know some Korean, Indian, Jamaican, Irish and Ghanaian parents who qualify too. Conversely, I know some mothers of Chinese heritage, almost always born in the West, who are not Chinese mothers, by choice or otherwise.
So Chinese-like parenting styles are already in the data after all.  If strict parenting worked the wonders Chua claims, existing twin and adoption research should detect big effects.  They don't.  Educational and financial success does run in families, but the reason is almost entirely heredity.

2. Why does the power of Asian parenting seem so intuitive to Asians and non-Asians alike?  The reason, most probably, is that people make a big distinction between intelligence, where they admit that heredity plays a major role, and character, which they imagine is entirely environmental. 

They're very wrong to make this distinction.  Not only do genes have a strong effect on character, but upbringing does not.  By the time they grow up, adoptees' work ethic and discipline moderately resemble their biological parents' - and barely resemble their adoptes parents' at all.  See Loehlin's chapter in Unequal Success for the best single summary of the evidence.

3. Even more importantly, twin and adoption research shows that heredity has a stronger overall effect on educational and financial success than existing measures of intelligence, character, and everything else predict.  Identical twins have much more similar incomes than fraternal twins - far more than their extra IQ and personality similarity can explain.  The lesson: Genes demonstrably affect success in more ways that we currently understand.  It's cheating to give parenting the residual.

4. Before we marvel at Asians' success, it's worth getting a handle on how successful they really are.  They definitely earn more than whites, but only about 15% more.  Yes, that pools all Asians together, including recent immigrants.  But even if we double this figure to 30%, it's a modest difference that genetically-influenced differences in IQ, personality, and the like can easily explain.

5. Chua doesn't mention a minority that has been far more successful than Asians in general and the Chinese in particular: Jews.  How would she explain Jews' vast educational and financial success?  Yes, Jewish parents have been known to stress education and nag their kids to become doctors and lawyers.  But very few are strict enough to meet Chua's standards.  How do they pull it off?  If you'll buy a genetic explanation for the Jews, why not for the Chinese?  And if Jewish parents were far stricter, wouldn't we be quick to falsely attribute Jewish success to Jewish parenting?

The upshot is that the tough love that Chua heralds is not just pointless, but cruel.  The defender of Chinese parenting might retort, "Well, at least it does no lasting damage."  But only massive future benefits could conceivably justify the truly sadistic things that Chua proudly admits she did for her children's alleged benefit.  Here's how she once taught her daughter piano:
I used every weapon and tactic I could think of. We worked right through dinner into the night, and I wouldn't let Lulu get up, not for water, not even to go to the bathroom. The house became a war zone, and I lost my voice yelling...
To my mind, the mere memory of this experience is lasting damage of a heinous kind.

Comments and Sharing

COMMENTS (39 to date)
Hyena writes:

I doubt the research because it's highly gratifying to the researchers--all of whom are highly-educated, successful and carrying high IQ proxies.

There's just so much room for cognitive bias.

Milton Recht writes:

Ms. Chua's reporting of her parenting techniques is self-reported. Self-reporting is a notoriously biased description of actual events. She is remembering her behavior and projecting her parenting to the world as America's image of a stereotypical Asian parent. A video of actual events between herself and her daughter in her household may find that her parenting behavior is less severe and more typical of other American parents.

The acceptance of Asians into American society is recent. One only has to think back to the WWII internment camps of 70 years ago. Americans' image of Asian parents may be a protective signal and camouflage created by American Asians that does not reflect actual private Asian household behavior.

Assuming Asians do act as Ms. Chua describes, does it make evolutionary or economic sense? Ms. Chua is investing enormous amounts of time and energy into the skill and intellectual development of her daughter. If genetics determine her daughter's outcome, then Ms. Chua's efforts are wasted and her opportunity costs are very high. She could be using the time to write a legal textbook, do legal consulting, have a private legal practice or for her own leisure. All Asian (and also non-Asian) parents, if they invest unnecessary time in their children's intellectual and skill development, could use that foregone effort to improve the household income and well-being of their families. Evolution and economics would favor efficient families that do not waste effort over inefficient households. The continued existence of parental time investment into childhood development creates a necessity to justify its continuation. Is there research that explains the continued existence of this supposedly unnecessary parental behavior?

I would treat Ms. Chua's book and excerpt as her undocumented and unverified belief of her parenting techniques, and not as an indication of actual Asian parental behavior. If her self-reported behavior is indeed verifiable behavior then its justification, maybe more along the lines of a necessary self-defense camouflage that Asians have adopted to protect themselves in our extremely xenophobic US society. An Asian parent is insuring dissemination of the Asian parent image. Her daughter's reports to her friends of her mother's extreme techniques will find a way to her daughter's schoolmates' parents. It will establish and reinforce American parents' stereotypical image of an Asian parent-child relationship.

Mik writes:

I am amazed that someone, even more a Yale professor, admits to such abuse. Because I think, based on the example given in the end, such "parenting" is abuse. What "success" can motivate that?

Hyena writes:

One thing to note, though: it's difficult to argue that character and performance are primarily hereditary while placing any weight on negative experiences. If the rest of your thesis holds, then that event, however unpleasant, is probably completely unimportant to the child's development.

Otherwise, you'd have to concede that there are ways to condition performance and character. Rather than heredity being dominant, it's merely that parenting is actually weak from psychological standpoint. Perhaps parents should be more severe with their children.

Drive them forward under the lash! Literally!

Jim Ancona writes:

I suppose it's possible that Asian parenting "works" by denying kids the opportunity for the type of peer-group socialization that Judith Harris postulates is responsible for much of the non-genetic variation in outcomes, leaving parents to fill the void.

But first Chua would have to prove that there is something to explain; i.e. that children of Asian parents actually do better than they would be expected to from genetics alone.

Tracy W writes:

Judith Harris suggests that there might be parent-group to peer-group effects that drive children's personalities. Eg kids don't learn or create a whole new language unless they have to, if they get to school/pre-school and discover that most of the other kids there speak English with the same accent as them, said kids keep on speaking the same way, just adding on some slang and in-jokes.

So groups of Asian parenting like that, along with Asian kids being sent to schools where there's enough other Asian students for them to self-identify as Asian kids, could result in higher Asian outcomes. Or perhaps not. It hasn't really been tested yet.

Brian Clendinen writes:

I am with Hyena. I just have seen way to many cases were adopted kids are just like their parents. One would not know if they where adopted if one were blind and had never been told.

I just have way to much real life experience to believe this nonsense. I am not saying genetics does not plays a role but “almost all” defies common sense. This contradicts almost every single piece of evidence I have ever seen in my life that parenting does matter and a lot.

Bryan, the logical conclusion is if almost everything is based on genes is to then let children raise themselves. Feed and house them and that is it. In almost every single case in which parents did this, the kids I knew were screwed up and had major issues. Granted some of them got their life together latter in life. But most ended being royally screw-ups. Don’t get me wrong a few turned out fine and never made any major mistakes. On the opposite end, the parents which I knew quite well and were good parents. Most of their kids turned out well. A few of the kids are screwed up and the parents could do nothing about it. So I am not saying there is any guaranties.

I attribute the studies erroneous conclusions to that it is really hard to judge in an objective way good parents. I don’t know how many people I know that appeared to be good parents but when I got to know them they really were nothing special. Or how many people which it did not appear they were that great (not bad just not great) parents and one looked closer and saw they were really good parents. Also I know parents who became good parents but sucked when their kids were younger, and it ended doing a lot of damage. There is just to many aspects of good parenting and each parent and child are different. I would have no idea how to categorize let alone validate that the parents really lived that specific categories attributes on a daily bases. This would be need to validate if good pareting matters.

Now if your point is that quite often good parenting is no guarantees of good outcomes and bad parenting is not a guarantee of bad outcomes. Then I absolutely agree. We are talking principals here not theories or laws.

ER writes:

I wonder if the observation of asian success is skewed by immigration effects. It takes some amount of drive/IQ to come to America for a shot at better life and asian immigration was heavily restricted until the 60s. To know whether or not asian success is genetic, we would need to look at asians who do not live in America. And what about other instances of selection bias? Are they counting those asian shopkeepers in chinatown or the workers at the illegal sweat shops upstairs as asian success stories?

Greg Rehmke writes:

A key question for those who call this kind of upbringing abuse is: "compared to what?" Lots of kids get yelled at and ordered around arbitrarily. When parents lives are a mess, they visit this turmoil on their children. But even stable families with mixed up views of reality, education and human nature can cause turmoil in their children's lives. As an example, John Stossel has a great segment interviewing parents who tried and tried to suppress gender roles in their children.

Leigh Bortins (author of "The Core") argues that young children enjoy memorization because they are good at it, at that age. They don't enjoy being pushed to try to answer questions they can't quite understand (though they can memorize the expected responses). Advocates of Classical Education say the Trivium models fits the natural development of children.

Children of sports coaches and enthusiasts often spend two or three hours a day in training. From swimming practice to weight lifting or basketball, hours a day are invested in a combination of training, work, and play. It is both hard and fun. Jr. High football was an hour or two a day of running, and we had another hour in P.E. (with the same drill-seargent teacher).

So, all that said, why are similar hours spent with music practice and study so terrible? Millions of kids spend that amount of time training each day in the complex and intellectually stimulating skills needed to succeed at various video games. Maybe there is a stimulating "violin-hero" game (but I doubt it).

Long hours each day of training in music and memorization seem the norm for Christian homeschool students I meet at workshops around the country. The kids seem fine. Though with wide variations for genetics and nurturing, these students are generally thoughtful, polite, competitive, and curious.

Chinese parent's child writes:

"To my mind, the mere memory of this experience is lasting damage of a heinous kind."

It's one thing for you to criticize the argument that the parenting style makes a lot of difference - it's a whole other to judge the child's experience in such a family, with which you appear to be totally unfamiliar.

The household described in the article looks very warm and full of love, and the bond between the mother and her children appears very strong. I wouldn't mind being a child in this family - in fact, I was a child in a similar family and I loved it. You simply don't understand how this looks from the inside, so please stop pontificating in that direction.

sourcreamus writes:

Traditionally one way of advancement in China that was open to all people was the civil sevice. They had an exam to take and each person was competing against everyone else taking the test. In this case a 1 percent advantage could mean the difference between a life of power and prestige and being a rice farmer. Authoritarian parenting may make sense in that context. In present day America it seems pointlessly cruel.

Curt writes:

Years ago, I remember examining some data that suggested the economic success of Asians in the U.S. was due largely explainable in terms of a cultural preference for deferring child-bearing to later years and having fewer children -- a preference they shared with American Jews.

Any comments from those who know something about this?

Another chinese parent's child (no really, would I lie to you) writes:

[Comment removed for supplying false email address. Email the to request restoring this comment. A valid email address is required to post comments on EconLog.--Econlib Ed.]

Hyena writes:

Mr. Cleninden,

One thing you should keep in mind, however, is that the studies focus primarily on economic outcomes (i.e., earnings). In the OECD this is strongly tied to educational performance as measured by credentialing. But, perversely, the credentials are awarded based largely on performance on an IQ test.

We should always expect economic outcomes to be inherited if we expect IQ to be inherited since our system is largely built on monetizing IQ. In short, the studies could always just be measuring the extent to which we rely on IQ proxies for deciding who gets credentials, etc.

Father of 9 writes:

Although Mrs. Chua may not like drama, methinks I've seen this play before. A culture or country, destroyed by war and/or violent political despotism, tries to cram 5 generations of development into one and a half (or less). It works splendidly for a while, and foreign visitors trumpet its successes and predict it will come to rule the world. The crash follows. This type of drama plays itself out in miniature in overly strict families all over the world-people get tired of working like dogs after a while. By the third generation they're painting their hair purple or joining communes.

Finch writes:

> Years ago, I remember examining some data that
> suggested the economic success of Asians in the
> U.S. was due largely explainable in terms of a
> cultural preference for deferring child-bearing
> to later years and having fewer children -- a
> preference they shared with American Jews.

My understanding is that the literature Bryan cites strongly suggests that this is basically an illusion, and such minor differences in parenting style as starting later and being richer have little effect on outcomes.

I do have a hard time squaring Bryan's references with my personal observation that only-children are messed up, for example, or that kids who have a parent die suffer long term effects. I wonder how powerful the tests in the literature are... What level of out-performance would a parenting style need to have before the literature would detect it?

Jew with an Asian fetish writes:

Jews have been more successful because unlike the Chinese, Jews are white. This allows them to blend in with the supermajority and avoid discrimination, unlike other ethnicities with skin of color.

Steve Sailer writes:

Ms. Chua has, in her own life, grabbed the brass ring of her field, a professorship at the top law school in America. Perhaps she knows thing about how to claw your way to the top in 21st Century America that less successful people don't know.

(Also, at times, she's trying to be funny. There's a little bit of Dave Barry in her writing style.)

tom writes:

Bryan, you admit having no evidence directly about this. Then you go off on Chua. I suggest this model instead:

1) Many, many people should not bother to try hard at extremely challenging tasks. They will not learn them and will not get better.

2) Some people have the ability to excel at some extremely challenging tasks, but only with extensive practice and focus. And some types of the extremely challenging tasks reward practice more than others do.

3) Some people have the ability to excel at extremely challenging tasks even if they do not practice extensively.

Bryan criticizes Chinese parenting by comparing Jewish parenting. But what if Chinese kids fall more in group 2, and Jewish kids more in group 3? Then you have an answer that shows different optimal strategies for each group of parents.

(I also think there are a lot of Jewish guys who under-perform. I can think of no way on earth to test that next to Asians, and I'm overgeneralizing wildly, but I'd guess that children of Chinese moms have less variance in their success than children of Jewish parents.)

One final Harris-idea: if an immigrant group wanted to design an activity that would reduce peer influence on their kids, they would pick piano, cello, violin, etc..., practice. It takes every bit of after school time and reduces opportunities to socialize with academic losers. Did Harris study anything like that?

Are you really that certain that the Harris studies do not miss successful strategies of parents of Group 2 and 3 kids, in the vast mix of group 1 kids?

JamieNYC writes:

"To my mind, the mere memory of this experience is lasting damage of a heinous kind."

Bryan, did you read the part that describes how mother and daughter felt after achieving success? You may be right to attribute most of the success of East Asians to heredity, but Ms. Chua makes some good points when comparing the true self-confidence that results from hard-won successes to empty praises that leave the child feeling like "garbage". Too many US parents encourage their children to "pursue their dreams", and then tacitly approve of giving up at the first obstacle.

I also agree with Chinese parent's child's comment.

Tracy W writes:

Hyena - just because an event isn't completely unimportant to a child's development doesn't mean that it doesn't matter more generally. Children are thinking feeling human beings, a miserable time for them is miserable and that matters even if it doesn't influence their eventual adult personalities and successes in life.

Brian Clendinen - I can think of cases where adopted children were not like their parents. For example, one of my mum's cousins was adopted at birth, and we only found out about her when she turned 21 and traced down her birth father. She was adopted by a small-town Pakeha family that placed a very low value on education and basically had the view that the best a girl could hope for was to leave school at 15 and get a job in the bank (as a cashier, not a currency trader). My family places a high value on education. She had been desperate to get to university, but unable to persuade her parents to support that, persuaded them to support her going to teacher's college, because it would have a stready job at the end of it. There are a lot of teachers in my family.

Feed and house them and that is it. In almost every single case in which parents did this, the kids I knew were screwed up and had major issues.

There's a conflicting problem. The standard in our society is that parents are more actively involved in their children's lives. A person who breaks social standards is unusual. The person either doesn't care about social standards, perhaps because they're focused on other things (eg Asperger's syndrome), or care but because of other problems in their lives can't manage to meet them (eg the alcoholic). We know that both Aspergers and alcoholism have genetic components, and so it's not surprising that the biological kids of those parents who can't or won't conform will also sometimes inherit similar behaviours genetically, and these of course can show up in milder or stronger forms. (This is not to say that all parents who don't conform are that way for genetic reasons, but if a lot of parents who don't conform are non-conformists because of genetic reasons then they will bring down average results). And then there are indirect genetic effects on peer relationships. So if a family diverges substantially from the norm of parenting we need to control for the genetic relationships before we can form any view about the causality.

it is really hard to judge in an objective way good parents.

There should be a logical test here. If parents affect how kids turn out, then you should be able to connect adoptive adults to their adoptive parents, given sufficiently divergent parenting styles, at a rate higher than pure chance.

Finch - with my personal observation that only-children are messed up, for example, or that kids who have a parent die suffer long term effects.

Funny, I know several only-children, and my observation is that they aren't messed up, at least not any more than anyone else I know. There should be a test for this too, introduce you to a bunch of adults, some of whom are single-child, some of whom come from matched socio-economic backgrounds to the single-childs but from multiple-child families, but you don't know which, and see if you can do better than random at working out which is which.

As for kids who have a parent die suffering long-term effects, there's a lot going on here. Losing one parent affects the family's income (stay-at-home parents contribute a lot of household production), often leading to moving home, disrupting children's peer group. The behaviours that give an increased risk of death for a parent in our time with good vaccines, sewage systems, etc, can be passed down genetically (eg risk-taking behaviour, alcoholism.)
And, also, what do you mean by long-term effects? Grief hurts, and normally children love their parents, even if they learn their accents and behaviours from their peers.

Finch writes:

Tracy W,

I didn't mean to strongly argue for those observations. I don't know if they are really true. I meant them as indications that parenting can have some effect - remove a parent, and you get an effect. But I don't know how large a parenting effect would have to be for it to show up in the literature as significant.

I'm reminded of the literature on active management of equity funds. The initial papers in the field declared active management a waste with tests that couldn't detect 3 percent annual out-performance (I'm remembering this from grad school, I may have the number wrong). Well, 3 percent is a lot, you'd pay for that. It took serious work to get the error bounds down, but the popularization of the story missed that.

If 20 percent of parents were out-performing by 10 percent in some sense (child income, number of grandchildren, "happiness"), would the literature Bryan cites to be able to detect that? On average the effect is low, but we're not asking about the average here.

Tracy W writes:

remove a parent, and you get an effect

No argument there, losing someone you love is a very painful experience.

If 20 percent of parents were out-performing by 10 percent in some sense (child income, number of grandchildren, "happiness"), would the literature Bryan cites to be able to detect that

The level of detected effect would depend on the study design, which isn't consistent, and the natural variation in the measured outcome (eg if the measured outcome has a normal distribution with a standard deviation of 10 percent, I'd expect 17% of parents to produce a kid that was 10% or more above the norm by pure chance, darn close to your 20%).

The "parents have little or not impact" is a null hypothesis. All we can say is that as far as we know no one has repeatedly detected such an effect after properly controlling for genetics yet, despite developmental psychologists looking for it for some 50 years. Effects show up, but they're not consistent. That doesn't prove that such an effect doesn't exist, though it does provide some reason for being doubtful about its existence.

A thing Judith Harris does explore is the possibility that a parent with enough children of about the same age to let the children form a peer group might be able to hothouse their kids and change the values of the peer group, she discusses a case in her book of a poor man with five daughters who managed to turn his daughters into far higher achievers than would be expected by their socio-economic group, part of it was by stopping them from playing with their classmates and only playing with each other. But that's only anecdotal, it needs more study.

parviziyi writes:

Jews were a cultural group that formed from several different gene pools, in places as far separated as the western Mediterranean and Central Asia. The reason why Jews as a group have been successful is purely cultural, and not at all genetic.

An immigrant group that US statistics shows to be approximately as successful as Jews, and mabye a little more successful than Jews, is immigrants from Arab countries. Most of this group are Christians whose Arabic forebearers arrived in the US in the earlier part of the 20th century from Syria, Lebanon, and Egypt. Christians from Arabic-speaking countries who arrived in the later 20th century have also been successful. Not so much as the earlier arriving families. But they are clearly very upwardly mobile in terms of rising intergenerational education levels. The rate of Ph.D's among Arabic-ancestry immigrants is higher than it is among Jewish-ancestry immigrants. In the data I've seen about this at the US census bureau, the classification is by country of ancestral origin, and not by religion, and of course some percentage of these Arabic-ancestry immigrants are not Christians but muslims. I don't know what percent, but I know the Christians are at least 51%. If you go to today's Aleppo in Syria, and take a sample of people from among the hundreds of thousands of Christians living there, and compare them against Arab-ancestry Americans, you find huge differences in education standards, IQ tests, incomes. I hold it as a certainty that these big differences arise from cultural factors not genetically hereditary factors. The hereditary-associated differences found in the Twin Studies are very minor compared to them.

Kelly McMenamin writes:

This was a great piece. I think Ms. Chua's need to prove her parenting style is not only okay, but superior, has kept her from seeing the evidence proving her wrong and/or she's suffering from Stockholm Syndrome from her own childhood.

Here's my two cents on Ms Chua's WSJ article.

parviziyi writes:

When I was a boy in the 1970s, all the heavyweight boxing champions were black-skinned Americans. I was told, at the time, that that class of people had heritary (genetic) endowments that made their dominance in heavyweight boxing inevitable -- such as having long arms. But nowadays there isn't a single black-skinned American among the various heavyweight boxing champions. Most of the recent champions have been Russians, Ukranians, Uzbeks, and Belarussians. See "List of heavyweight boxing champions" at Wikipedia. I regard this as one of the most compelling examples I know of, of the supreme importance of cultural factors for human achievement.

Bryan Caplan's contention that genetic variations are more powerful than culture variations shows one and only one thing to me: something is very seriously wrong with the interpretation of the Twin and Adoption data.

Tracy W writes:

parviziyi - I don't think that Caplan is arguing that genetic variations are always more powerful than cultural variations, I think what he is arguing is that parenting behaviours don't affect adult children's outcomes much once you control for genetic links. Parenting behaviours are only a subset of cultural influences.

For example, the children of immigrants will normally learn the accent of their peers, not their parents, if the two are different. What accent you speak is normally culturally-determined, but not parent-determined.

parviziyi writes:

In response to Tracy W, I'll say a bit more carefully what I meant to say: Bryan Caplan's contention that genetic variations are more powerful than parenting culture variations shows one and only one thing to me: something is very seriously wrong with the interpretation of the Twin and Adoption data.

gcochran writes:

Bryan Caplan is right about the weak influence of parenting style. You can't trump statistically solid studies with a few anecdotes - and of course there are plenty of anecdotes that point in the opposite direction. Reminds me of the time that a high-school teacher commiserated with my (valedictorian) little sister - saying it must be difficult to be under such parental pressure. My sister said "What pressure?" She was just smart.

Tracy W writes:

parviziyi - shows one and only one thing to me: something is very seriously wrong with the interpretation of the Twin and Adoption data.

Is there any evidence that could convince you that you might possibly be wrong, and the interpretation of the twin and adoption data be right?

parviziyi writes:

In 2005 the journal American Political Science Review published an article that used the twin study method of behavioral genetics to argue that much of the variance in political ideology in USA is attributable to genetic inheritance from parents (as opposed to cultural inheritance from parents). In response to that, some other social scientists wrote "Why Twin Studies Are Problematic for the Study of Political Ideology", an article that attacks the interpretation of Twin Studies data in general, while using the domain of political attitudes as a concrete context for the attack. You can find the article free online by searching for its title. It's not the last word on the subject; it's one place where one can start thinking about what's wrong with the interpretation of Twin Studies data.

ChineseChild writes:

For whatever it's worth...

I was raised by a Chinese single mother until the age of 15, when she died and I ended up in foster care. I was then raised by several white families. Since she was a single mother, she was never around, I did whatever I wanted so long as I didn't make trouble like most latch-key kids.

My biological mother's family were all musically gifted, each person playing at least one instrument (piano, guitar, mandolin, flute, etc). Nobody went to college... except me. I play the piano decently, although I've never taken lessons before, play completely by ear. I played the violin (with free lessons at school) from grades 2-college in the orchestra and competitions. Music comes naturally to me. I've been composing my own music since the age of 12. And go figure, I am excellent at math.

In college I studied chemical engineering and took up ballet. I got sick of all the socially awkward engineers and graduated with a finance & accounting degree. I speak 4 languages fluently. I am now a 26 year old Sr analyst for an international consulting firm. I still play the violin, am still immersed in math.

My Chinese mother was the complete opposite of Ms. Chua, with no education, no success, no money, a totally hands off parent, but I have a feeling her kids will turn out somewhat similarly. Could high achievement be in the genes, regardless of how dysfunctional the upbringing and background? Perhaps I'm an outlier? I'm only a sample of 1. But I personally think Bryan is onto something about genes trumping nurture. Lord knows I got royally screwed on the nurture scale.

parviziyi writes:

A readable attack against Bryan Caplan's interpreation of twin studies data, written to be accessible to first-year college students, is: "The Genetics of Political Behavior: Claims and Refutations", by Jay Joseph, published year 2010 at Jay Joseph's own website JayJoseph.Net:

The above article cites many other articles. Here are two from among the articles it cites that are good ones:

  • "Twin Studies of Political Behavior: Untenable Assumptions?", by Jon Beckwith and Corey A. Morris (2008)

  • "Twin studies in behavioral research: A skeptical view.", by Leon J. Kamin and Arthur S. Goldberger (2001)

Rob writes:

Amy Chua has written elsewhere about the fear and resentment that is commonly found when a minority group becomes more economically successful than the majority. The most infamous example is the vilification of European Jews at various points in history, but there are plenty of other examples and it's not entirely unreasonable to wonder if Asians in the US might become the subject of resentment at some point during the lives of her children.

My guess is that such resentment would be greater if the basis for Asian success were to be seen as hereditary and thus "unfair". One striking thing about her article on parenting is that she explicitly frames her parenting as "hard work" designed to earn a particular outcome. In other words, if her kids are successful it's because both they and she worked very hard, an option which is open to everyone, rather than having the good fortune of superior genes.

Now, if one happened to have a particularly high awareness of the risks of being seen as a part of an innately-gifted minority, one might have a particular interest in spreading the belief that this minority's advantages are down to cultural practices that can be adopted by anyone, rather than innate and insurmountable advantages of a hereditary nature. Not saying that's necessarily what Amy Chua is doing, but it might provide a rationalisation for it.

Tracy W writes:

parviziyi - as far as I can tell, the papers you link are arguing that twin studies overestimate the genetic links for politics, because twins grow up in similar environments. For example from the Jay Joseph paper, page 202:

"Several studies have found that the social environments of MZ twins are more similar than those of DZ twins. For example, habits, activities, personal preferences, parental treatment, and self-image tend to be more similar between MZ twins.

Fine, but this is irrelevant to the question of whether twins are influenced by parenting behaviours or by cultural variations. And there are some suggestions in the paper itself that cultural variations might be important. For example:

One of many such examples we fi nd in the 85-year history of the
twin method is that, when studied, same-sex DZ pairs (DZSS) usually correlate higher than
opposite-sex DZ pairs (DZOS) for psychiatric disorders and behavioral traits.

Now DZSS pairs who go to a normal Western school will have plenty of opportunities to socialise with other boys or girls, and thus be socialised into the social norms for that culture, while DZOS will be socialised into different norms to the extent they are affected by gender norms.

The Jay Joseph paper also doesn't discuss adoptive studies, which are an obvious solution to the problem that family studies can't tell the difference between genetics and family influence (basically, look at adopted children).

I note that you didn't answer my question of what evidence could possibly convince you that you might be wrong.

Incidentally, the link to the Jay Joseph paper didn't work for me, I suggest

Anz writes:


"You can't have environmental contributions to a child's development without genetics. And you can't have genetic contributions without environment".

Finch writes:

It turns out that parental death is not a very big influence on outcomes:

I find this very surprising, but data trumps anecdote. It's recent and I didn't know about it when I posted above.

Trey writes:

For what it's worth, I've posted my thoughts here:

[blog url fixed--Econlib Ed.]

Ming Ling writes:

There is indeed research (from developmental psychology, cognitive science and education) showing that specific aspects of the parenting methods Chua describes produce greater achievement, while other aspects hinder success and risk undermining motivation. Because these studies utilized random assignment to condition, they averaged out the effects of genetic variation.

I've summarized some of that research in my blog.

Comments for this entry have been closed
Return to top