« SEVEN DAUGHTERS | Gene Expression Front Page | Is Thabo Mbeki a moron? Or are we seeing what we want to see? »
July 09, 2003

Individual impact of affirmative action

Here is an excerpt of a testimony toward the reality of the negative stain that affirmative action has left on a Stanford law professor (Marcus Cole) over at Volokh Conspiracy:


Let me illustrate my point. I am willing to bet that I am the only member of this list who feels compelled to put his standardized test scores and National Merit award on his CV. Why do I do this? For those of you who do not know me personally, it is not a matter of braggadocio. Every September I have to deal with nearly 60 prima donna first year law students whose first and only (initial) reaction to my skin color is that they have been cheated out of a "real" Contracts professor, and are stuck with an "Affirmative Action" instructor. Many of them come around when, as some "gunners" often do, they look up my CV and find that I have outscored virtually every single one of them on the test around which they have centered their lives, the LSAT. Others usually come around by mid semester when they have had an opportunity to compare my teaching to that of their other instructors. If numbers (standardized test scores and teaching evaluations) could obscure my skin color, my life would be heavenly.

I am not complaining. I live a truly blessed life. But to insinuate that my life is nothing more than an Affirmative Action storyline is the insult that I endure year in, year out. I have done everything I can do to distinguish myself. I've worked hard (scrubbed toilets as a janitor to put myself through college at Cornell, among other things). Yet nothing I do is enough to satisfy anyone on the left that I am their equal without need of their help, or anyone on the right that I am not where I am because of affirmative action.


For something similar, check out this page on Jon David Farley, a black mathematician. It states, "He graduated summa cum laude from Harvard University in 1991, where earned 29 A's and 3 A-'s." (not my bold-faced, they are present in the original!) Also, check out this page dedicated to black mathematicians.

This illustrates the trade-off between individual attainment and group achievement-I bold-faced that professor Cole teaches at Stanford because I wanted to highlight that he is not a typical individual, let alone a typical black American. Men like Marcus Cole will and do succeed no matter their race. On the other hand, I think we can admit that if affirmative action ended today, there would be far fewer blacks in the commanding heights of academe. That is the price that we would have to pay for Marcus Cole's self-esteem.

I say this not to belittle the pain Marcus Cole feels, but to remind ourselves that in trading in their individual self-respect black Americans have created a decent sized middle class, for only 1 out of 4 blacks lives the stereotype of the underclass lifestyle today. It might be harsh to state that black Americans have sacrificed their self-respect, but I believe Marcus Cole's assessment is an accurate reflection of reality in how others perceive high achieving blacks, insofar as they will always be high achieving blacks, rather than just high achievers. I myself live at the other end of the antipode of group evaluation, the moderate success of my group (South Asians) often gives people the immediate perception that I am intellectually nimble & technically oriented (or that I drive a cab!). I have known Asian Americans who aren't good at math or straight A students who would complain that they have to live up to group expectations and deal with daily disappointment from acquaintances who quickly realize they aren't geniuses.

The evaluation and legal organization of our citizenry along group lines has individual effects. There are costs and benefits to any given decision-those of us who disagree with affirmative action must admit those costs, and those who revel in the diversity that America's current status quo allows should take time out to reflect on the individual costs that they might never have to bear.

Posted by razib at 10:15 AM




This is why I think all affirmative action and quota systems should end. And there should be a one-off reparations payment for descendants of slaves - which can only be spent on education and training.

After that - no more affirmative action, no more white resentment - just a greatly expanded black middle class, and far less racial division in America. And cheaper in the long run.

Posted by: fredrik at July 9, 2003 12:34 PM


I think I missed one of the logical arguments in your post. Your group (South Asians) is not legally distinguished in the fields you mention. Nor is it handicapped.

So why do you compare your predicament with Marcus Cole. Because of government policy based on groups, Cole feels his acheivement to be suspect in the eyes of others.

Any taint your acheivement might have in the eyes of others has no relationship to government policy. I don't see how the two are comparable.

(And, as I've said before, stereotypes of South Asians is strongly regional. In my area, a smatterring of professionals are brown, but all the coffee-jockeys at the nearest Tim Horton's are from South Asia -- except the manager, who is Eastern European. And in BC, South Asians have more of a criminal-gangsta image. Your local stereotype will also change in time.)

Posted by: Ikram Saeed at July 9, 2003 09:50 PM


"Any taint your acheivement might have in the eyes of others has no relationship to government policy. I don't see how the two are comparable."

Have you really read Razib's post?

Posted by: friedrich braun at July 10, 2003 02:02 AM


Ikram:
"Any taint your acheivement might have in the eyes of others has no relationship to government policy. I don't see how the two are comparable."

Friedrich:
"Have you really read Razib's post?"

I'll second that and ask if Ikram spends as much time on the blogs he actually links to as on GNXP?

Posted by: R at July 10, 2003 02:17 AM


No I still don't get it. Cole is seens as an affirmative action baby. This hurts him. Affirmative action is a gvt program.

Razib is seen as not meeting the cab-driving standards of South Asians (or other 'group' perceptions).

One is a gvt. enhanced group evaluation. The other is not. Are they really comparable?

I understand the main point of the costs of group evaluation on the individual. I don't understand the subsidiary comparison.

I don't mind admitting my er -- 'density' on this one.

R -- what are you talking about?

Posted by: Ikram Saeed at July 10, 2003 06:44 AM


Razib,

I have to say well done! A fair-minded, balanced take on affirmative action, that doesn't simply whitewash the beneficiaries or castigate all blacks as genetic inferiors who would be unable to bring their spoons to their own mouths without white support.

Here is an interesting question to ask: from a Bayesian viewpoint, given the existence of black mathematicians like Wilfrid Gangbo and Adebisi Agboola, just how likely is it that Richard Lynn's claims about African IQ have any truth to them?

Agboola is a genuine, born and raised Nigerian of purely Yoruba extraction, and I'd wager that you can't do the sort of work he's doing without an IQ of at least 150 (any of you who have studied algebraic geometry will certainly agree with this statement). Katherine Okikiolu's father, George Okikiolu, was also a distinguished mathematician, and if anything, he is even more brilliant than Agboola. How likely is it that a continent of supposedly drooling morons, with average IQs of 70, should have produced two such individuals in mathematics alone? And we haven't even touched on the numerous African students on Ivy League campuses with (pre-centering) SAT scores of 1400 and above ...

Finally, I'd be interested in hearing your comments on Leon Kamin's review of Richard Lynn's "work"; laying aside Kamin's attack on the Pioneer Fund, what is your response to his criticism of Lynn's alleged misuse of test scores? For instance:

Lynn chose to ignore the substance of Crawford-Nutt's paper, which reported that 228 black high school students in Soweto scored an average of 45 correct responses on the Matrices--HIGHER than the mean of 44 achieved by the same-age white sample on whom the test's norms had been established and well above the mean of Owen's coloured pupils.

I'd be grateful if you could give a fact-based critique of Kamin's arguments, rather than stray into a discussion of Kamin's politics or motivations.

Posted by: Juvenal at July 10, 2003 08:25 AM


just how likely is it that Richard Lynn's claims about African IQ have any truth to them?

i don't buy it-neither does godless, and i suspect, no one else who has an account on this blog. steve sailer has talked about how there is probably something off in the result of 70 as well. i tend to focus on american/first world results because we can assured people have some level of nutrition and there are MANY MANY tests taken over MANY MANY years.

i don't have time to look up kamin's point at this point, but i will put that on my TODO list and respond, inshalla ;)

i will also get to ikram's point-but the short end of it is is that i was trying to relate on some level to marcus cole's anger-the gov. puts the force of law behind a common human tendency, to make quick judgements of people upon first meeting them by group characteristics.

Posted by: razib at July 10, 2003 09:28 AM


i don't buy it

What is there not to buy? A) That extreme deprivation depresses IQ? B) That testing varies in reliability and validity with conditions? C) That environment and nature work together in a complex way?

or castigate all blacks as genetic inferiors

Cute. And do "we castigate all whites as genetic inferiors" because they score lower than East-Asians? Try to tone down on the hysteria, hmm? It doesn't improve the exchange of ideas.

As for Kamin, do you have an Amazon link for the quote?

Posted by: Jason M. at July 10, 2003 10:33 AM


"And do "we castigate all whites as genetic inferiors" because they score lower than East-Asians? Try to tone down on the hysteria, hmm? It doesn't improve the exchange of ideas."

Hold your fire!! Note that I've been careful not to say that you or anyone contributing to this site has made such a claim. What is true though is that there are many on the web who certainly have been eager to do so.

"As for Kamin, do you have an Amazon link for the quote?"

It's from an old copy of Scientific American that doesn't seem to be online (February 1995, 272(2)); the following link provides a copy of the relevant article.

Here you go.

NB - I'll also go on record as saying that I don't think that whites are actually "inferior" (or "superior") to Asians intelligence-wise. My view on this matter is that while average IQ in various populations may indeed differ, there simply isn't enough genetic distance between most human populations to expect any such differences to be on the scale we see in testing.

Add in the fact that, evolutionarily speaking, intelligence has been a reproductive plus for the last 2.5 million years, and it seems odd to imagine that any alleles for superior IQ wouldn't have spread rapidly through all other populations, as population genetics tells us that we only need a minute amount of intermarriage between neighboring groups for this gene flow to occur globally (n-island model).

Even if the industrial revolution has somehow erased the selective advantage of intelligence (of which I am very much in doubt), that would still mean that, for most populations, no more than 5-10 generations have passed since higher intelligence ceased to be a reproductive advantage. This makes substantial between-group genetic differences in IQ extremely unlikely, in my view.

Posted by: Juvenal at July 10, 2003 11:47 AM


Add in the fact that, evolutionarily speaking, intelligence has been a reproductive plus for the last 2.5 million years

this is a tenuous assertion that needs to qualified and clarified. if you are a multiregionalist, a rather minority camp, you might buy into the idea that intelligence is always advantageous and that all subpopulations of a given species react to the same selective pressure and allele "x" will spread through via recombination. but in our own species lineage we have not seem a gradual jump in biological capacity for intelligence if you draw a line between hominid species. h erectus was fine for a million years or so, until early modern humanity exploded upon the scene, there was some stimulus that selected for high intelligence, and this allele did not spread throughout the rest of "humanity" (hominids)-rather, this early population went through an intense bottleneck and seems to have radiated and replaced other hominid species-though biological evolution seems to have preceded the cultural explosion that allowed the replacement (unless you buy the idea that there was a "great leap" that we can not see in the bones-ie; language, etc.).

even within the context of homo sapiens, intelligence seems to have negative effects, as the correlation with schizophrenia and other mental diseases would indicate (or g. cochran's point about the strength of heterozygotes that have low penetrance of deleterious phenotypes but have a correlated response of high IQ, while the homozygotes are lethal).

there was some population movement and genetic exchange, especially in the past few centuries with transoceanic travel becoming routine-but prior to the modern age, much of humanity has seen a series of population bottlenecks that radiate into localized populations. for instance, "non-africans" are a radiation of a subset of africans, those of the northeast, while amerindians come from a central siberian population, etc. etc.

even you point about the spread of an advantageous allele(s) for intelligence is not totally applicable, you could make a case on the "world island" of eurasia. but what about australia & the new world which have had little population migration in the past few thousand years (aside from the inuit in the new world).

Posted by: razib at July 10, 2003 12:45 PM


"if you are a multiregionalist, a rather minority camp, you might buy into the idea that intelligence is always advantageous and that all subpopulations of a given species react to the same selective pressure and allele "x" will spread through via recombination."

Moi, a multiregionalist? Heavens no! Seriously though, it strikes me as unlikely that any mutation that conferred greater intelligence, without imposing a major accompanying drawback, would have failed to permeate a single breeding population quickly - assuming we are dealing with a single population.

The difference between Homo Sapiens populations and those of older forms of Homo is that we are a lot more mobile than any of the older forms ever were, and this has been true throughout the period since we originally radiated out of Africa; it seems to have taken barely 2,000 years for humans to have spread from the Red Sea coast all the way to the Australian mainland, which is nothing in evolutionary terms. In that sense I think it legitimate to treat our species as effectively a single breeding pool, in a way that wouldn't be justified with our predecessors.

The way I see the difference between us and the species of Homo before our own is in terms of thresholds: the extra intelligence that allows you to see a finished blade in a mere rock may be enough to give you the edge over your companions in your local breeding pool, but it won't suffice to give you the werewithal to migrate far enough to encounter other subpopulations and breed with them.

"but what about australia & the new world which have had little population migration in the past few thousand years (aside from the inuit in the new world)."

Let me take the easier case first - Australia. I don't think that there's been a complete lack of gene flow between Australia and mainland Asia, even though any exchange would have been on a next-neighbor model, i.e, A swaps with B, who swaps with C, etc. So it isn't necessarily true that advantageous alleles would have failed to reach the Aborigines. The data available certainly doesn't support the notion that they are very different from, say, the Andamanese, in a genetic sense.

The more difficult case is North America. All the evidence I've seen suggests that the last stream of migrants crossed the Bering Straits about 6,000 years ago, or about 240 generations. Given a selection differential of 0.3 percent, which is actually rather large for a single gene, that would have been enough time for any putative intelligence gene to have reached a frequency of 49%.

On the other hand, there's been a tremendous amount of European-Amerasian mixing since the conquest of the Americas, and we're talking about two populations that wouldn't have been all that different to begin with. I don't know that there are many Amerindians alive today who have no European ancestors somewhere in the family tree, and it certainly would have taken a keen intelligence to survive the pressures of the colonizers on indigenous populations.

NB - I believe that Hartl and Clark's Chapter 5 covers the issue of gene flow in some detail. In particular, in the island model, given that the number of migrants per generation between two neighboring populations, or N_m, is greater than 1, we ought to expect hardly any genetic divergence between the two populations, regardless of total population size. For all populations in Eurasia and Africa, N_m has typically been much higher, so expecting any purported intelligence gene to have remained in, say, East Asia, without quickly spreading from Normany to the South African veldt, strikes me as unrealistic.

Posted by: Juvenal at July 10, 2003 01:52 PM


Hold your fire!!

I would really like to apologize, I read that as hard irony (i.e 'finally razib, a piece where you don't advocate genocide'.) Now corrected I can see that you have expressed your opinion with nothing but courtesy and intelligence. :)

on to your points:

there simply isn't enough genetic distance between most human populations to expect any such differences to be on the scale we see in testing.

I would say that this assumption is unlikely. Most of the diversity of humanity is already there in all populations, and every population runs the measured curve just with different frequencies. Anything from genetic drift to extreme periods of selection could very plausibly change any sort of frequency in a population. It certainly has with height, and I have a feeling that variation in g operates very similar to variation in height.

Add in the fact that, evolutionarily speaking, intelligence has been a reproductive plus for the last 2.5 million years, and it seems odd to imagine that any alleles for superior IQ wouldn't have spread rapidly through all other populations

Again, the intelligence is already there in all populations, distributed on a curve. Nothing has convinced me, knowing that differences in individual intelligence are substantially genetic, that the distribution on this curve should look the same as it does for Europeans in every population. And the thing is, is that It doesn't: Some populations have some more overall, some have some less overall. I think the fair assumption is that *some* of that is genetic, and the only real debate is how much.

Posted by: Jason M. at July 10, 2003 03:01 PM


"I think the fair assumption is that *some* of that is genetic, and the only real debate is how much."

I am entirely in agreement with you with respect to the difference being populations between a matter of distribution between groups. The problem is that between-group variance only gives us 5-7 percent of total variation. Seeing as a purported intelligence gene, unlike CFTR (cystic fibrosis transmembrane conductance regulator) or MC1R (melanocortin 1 receptor), would be at an adaptive advantage in virtually any environment in which birth control did not exist, I would expect such an allele to eventually sweep the entire population; eventually, all populations would converge to the same value, i.e 100 percent.

The only ways in which I can see this not happening are as follows:


  1. If there were some epistatic interaction which would make the hypothetical gene maladaptive in certain scenarios; but what would such a scenario look like? I have no idea, and I don't know that there's any data out there to support this option.

  2. If the gene existed in the context of some evolutionarily stable strategy, in which having too many highly-intelligent individuals is an unstable state of affairs. But why should such a strategy converge to different proportions in the different populations of hunter-gatherers that were the norm until very recently?

To summarize, I don't rule out a priori that there are indeed major genetic differences in intellectual potential, once we account fully for environmental factors, but I don't see the evidence to support it. If we can account for only 5 percent of human variation on "racial" grounds, what are the odds that the remaining 5% are likely to be the most significant, where intellectual ability is concerned?

Posted by: Juvenal at July 10, 2003 04:00 PM


Razib said,

"i will also get to ikram's point"

No need. I think I get it now.

I would also concur with Juvenal. This was a very good post (on the individual costs and benefits of group evaluations). That did not come across in my initial comments.

Posted by: Ikram Saeed at July 10, 2003 10:03 PM


I am entirely in agreement with you with respect to the difference being populations between a matter of distribution between groups.

Are you saying that you agree there are probably mean differences in g between populations, but just a relatively insignificant amount? If so, that's a lot of common ground!

The problem is that between-group variance only gives us 5-7 percent of total variation.

First of all I think this is a fallacy - nothing tells us what that 5-7 is or does. Humans and Chimpazees share about 99% of our DNA, but it wouldn't make sense to claim that that difference "isn't enough" to account for the differences we see. Also, how much genetic difference is there between a man and a women? Between breeds of dogs (with all of their variety in behavior and ability)? The answer in both cases is a trivial amount. Even if the number were as low as a fraction of a percent, that would tell us very little, if not nothing, about what that could mean.

But again, don't take that the wrong way - what I'm more interested in is what we *do* see in every population, and the variation in most things (IQ included) is already there, so like much of the variation between men and women, most of the differences between world populations would be a matter of degree rather than of kind.

Seeing as a purported intelligence gene ... would be at an adaptive advantage in virtually any environment

I see this a lot, and simply cannot agree. The major problem I see with this is in the assumption that IQ must be all benefits, with no attendent costs. But there are many known things that make higher IQ costly and plausibly unknown ones as well. For instance people with higher intelligence are more likely to have specific physical problems such as myopia and asthma, are more likely to suffer from shizophrenia and depression, and have a greater head circumfrance at birth which means a higher mortality rate. Also illustrative of what I'm saying, is that researchers have found (mostly in Ashkenazi) some candidate high IQ genes that may also shoot up the chances of getting some pretty ugly disabilities.
So genes for intelligence, where ever they might come from, could certainly be selected for or against at differing rates in a variety of environments. The fact that low IQ still exists, at a relatively high rate, in normal individual variety means that it isn't as worthless or "inferior" in some kind of absolute sense in comparison to high IQ as many like to suggest. If so it would have been "selected away" in favor of high IQ a long time ago.

And to answer your other concern, a gene that might contribute to intelligence could remain exclusive to certain populations for the same reasons, whatever they may be, other genes do. Simply having a relationship with those other exclusive genes, could make the intelligene genes exclusive too.

But again, I'm not really interested in differences that come from our differences, so to speak. I don't need to look in the 5%, or whatever, to explain differences. The same genes that account for the great variety of individual differences in intelligence, would be the ones most likely to account for the group differences as well. The key-word to me has always been frequencies.

Posted by: Jason M. at July 10, 2003 11:13 PM


i agree with jason m & godless-but one more thing, if you put an average jew and an average australian aboriginal in a grassland environment and gave both of them a "HOWTO" guide to surviving-i would put my bets on the latter. i think that the average jew is maladaptive for a hunter-gatherer lifestyle, most in my experience seem to be myopic, which would be a big problem when it came to seeing signs of wildlife or cereal food sources. in contrast, the average jew has a lot of generalized abstraction skills that can be brought to bear in the context of a modern industrialized economy. i think it plausible that various subpopulations are a pretty good fit to the niche that most strongly reflected their recent evolutionary history.

maladaptive traits like myopia can only exist in societies where there is enough surplus production of basic resources that a portion of the population does not need to engage in basic food production and/or foraging. ergo, high rates of myopia among askhenazi jews and east asians, low rates among australian aboriginals....

fitness is contextual, and to some extent, "intelligence" is too.

Posted by: razib at July 11, 2003 03:53 AM