Ann Althouse (whose work I generally very much like) bristles at one of the several candidate explanations I included in my earlier post:
Eugene Volokh asks why are there so many more male Supreme Court clerks than female? His first guess is:
Is the cause possible differences in innate intelligence at the tail ends of the bell curve (what I'd heard called the idiot-genius syndrome, which leads men to be overrepresented both among the very low-IQ and the very high-IQ)?
Oh, please. I know it's in question form, but really...
Perhaps I misunderstand Ann's reason for the "Oh, please." Still, I'd have thought that when one is seeking the causes of disparities such as this, the possibility of men's biologically caused overrepresentation at the tail ends of the IQ bell curve must be one candidate explanation that one should consider. Even if the effect is relatively small, too small to explain by itself a 2.5 to 1 disparity (just to use as an example the disparity the year I clerked), it may still be important as one of the explanations: Five causes, each of which would cause an extra 1.2 to 1 disparity, can multiply out to 2.5 to 1. And I'd think that any analysis that infers some cause -- or estimates the magnitude of some cause -- from the disparity but fails to consider this other possible cause would itself be unsound.
Of course this potential explanation might ultimately prove to be entirely unsound -- but is there really any a priori reason why we should dismiss it as implausible? Has scientific research, for instance, conclusively dismissed this possibility? I'd be delighted if people who (unlike me) have actually seriously studied sex-based cognitive differences (or the absence of such differences) could speak to that.
(I should note that my listing the possibility of intelligence differences first wasn't meant to suggest that I think this is likely the most important factor, just as my listing the possibility of sex discrimination second wasn't meant to suggest that I think it was the second most important factor. I really don't know which if any of the factors are important, or how important they are.)
Related Posts (on one page):
- Prof. Ben Barres' Response:
- Credentials and Interdisciplinary Work:
- Scientific Debate, Proof, and Conjecture:
- Should Speech About Gender Cognitive Differences "Not Be Tolerated" on Campus, and Instead Treated as "Verbal Violence" Rather Than "Free Speech"?
- Be Careful Trusting Data, Even in Nature:
- Gender and Science:
- More on Sex and Supreme Court Clerks:
- Data on Women in Legal Academia:
- Why So Few Women Supreme Court Clerks?
If you make the claim "hmm, maybe we men are more likely to be geniuses than you ladies", without any good statistical reasons FOR IT, as opposed to a lack of studies definitively rule it out--I perceive people as being positively EAGER to believe this.
Do you understand how bad a track record "maybe it's not discrimination, maybe we're biologically superior" theories have?
It's fair to offer anything, I suppose, but yesterday's NYT had an extensive article documenting how undergrad women are blowing past men and getting a disproportionate number of the honors recognitions. I'm sure that's cultural and not biological too
And while some might point out that a large number of recent presidents were left handed (Ford, Reagan, Bush I, Clinton) only 7 of the 43 have been left handed, or 13% (not counting multiple terms as multiple presidents).
1) you list it first. this usually implies that you see it as the most plausible or otherwise most salient possibility.
2) consider your choice of language in the parenthetical. "idiot-genius syndrome"; "leads men" (emphasis added in both places) suggest that you see substantial innate differences as proven, and the dispute is about whether they're explanatory in this particular context.
so why put it first, and why use that language?
I don't necessarily agree with Althouse's premise -- the SCOTUS clerks I know are pretty darn clever -- but I think that was her point.
This is a no-brainer for math ability at the extreme upper tail of distributions of large nationally representative samples.
Women might perform better than men at the central tendency of a distribution, but men perform much better than women at the extreme upper end of a distribution on more complex math problems (see p. 148, Selectivity of Sample, d’s=~.50)
REF
Benbow and Stanley (see Science article on mathematically precocious youth) rule of a host of sociological explanations and focus on biological ones (and more recently on testosterone diffs). One hypothesis is that women who have higher levels of testosterone, or who are exposed to higher levels of the hormone in utero (compared to other women) should perform better on math problems.
If that were true, though, that would in no way dispel the cultural explanation, though, as there are various reasons why common cultural features emerge. The book, the Evolution of Culture, contains much material here.
According to Arthur Jensen's 1998 book, "The g Factor," there is a "near zero" difference between the sexes in the general cognitive-ability factor "g", but greater variability in males and hence, as mentioned before, more males in the extreme high and extreme low end of the IQ range. More recent studies have suggested that there might in fact be a mean difference -- in addition to a difference in variability -- amounting to 3 to 4 IQ points. I understand (based purely on hearsay) that Jensen now thinks that he may have been premature in his "near zero" conclusion. The reason for the newer numbers is that many studies are based on pre-adult samples, and males complete their maturation process (both mental and physical) later than females.
In a post on the other thread, I noted several good books on the subject of sex differences in cognition or sex differences more generally.
There is a recent large-scale British study showing a greater variance in boys from three-years of age on (as well as a higher male mean from 10-years of age): Rosalind Arden &Robert Plomin, Sex differences in variance of intelligence across childhood. Personality and Individual Differences Vol 41(1) (Jul 2006): 39-48.
Perhaps you'd be willing to provide a cite for this claim, before we go any further?
As the father of a three-year-old, I can tell you that by that age kids have internalized a huge amount of cultural conditioning about how boys and girls are "supposed" to behave. I'd be very hesitant to draw conclusions about innate differences between the genders on the basis of such studies.
Incidentally, one explanation for the disparity is that in hunter-gatherer societies, where men tended to be hunters and women tended to be gatherers, different kinds of mental skills were rewarded with surival. This is used to explain why men have (marginally) higher spatial intelligence than women, and are accordingly over-represented in math and science.
I don't know of anything that states that men have higher verbal intelligence. So the above does not really explain why there are fewer female law clerks, (assuming that law is primarily a function of verbal rather than spatial intelligence).
Perhaps, though that claim seems to be more speculative then the "g" explanations others here have offered. I guess the question is - why?
On a sidenote, I observe that whenever the subject of intelligence comes up, those who are uncomfortable with it tend to attack the methods of social scientists who gather the data. Usually, this involves assumptions about the quality of their work that seem intricate to (those of us who are) laymen, such as, "Oh, but did they control for variable Y?" or "They probably didn't consider test bias." Well, "they" probably did. In fact, "they" thought of much tighter controls on the studies than I could think of sitting here. Anyone who thinks this is junk science should go out and read "Bell Curve," and take the time to read the technical stuff and uderstand it. It's worthwhile.
Supplemental points:
1. Standardized mathematics tests are hardly pure measurements of analytic skill; a lot of questions require (or are at least greatly eased by) familiarity with their subject areas.
2. A much better analysis would look at the AP or higher-level IB mathematics exams, which (IIRC) have free-form analysis sections.
3. If anything, legal analysis skills ought to be better measured by "comprehension"-style questions from the verbal/analytic sections.
4. On the other hand, verbal/analytic tests stress advanced vocabulary more than anything else.
5. Performance on the GRE's analytical writing section would be a decent measurement, except that it's very tightly targeted towards a particular demographic.
6. Couldn't a wider distribution be most obviously accounted for by an increased willingness to take risks? I'd be very surprised if men weren't innately prone to riskier behavior — culture has a large effect here, too, but so does biology.
7. This list was much more attractively formatted before the forum software forced me to strip the <ol> tag from my post.
But since you ask, I'll step up. As someone who believes that essentialist explanations for gender differences are generally ill-founded and misguided, I'm happy to concede the "possibility of intrinsic differences in intelligence between the genders."
I concede the possibility. Now what?
http://www.edge.org/3rd_culture/debate05/debate05_index.html
Now, thank you. That's a reasonable start, (some apparent vitriol aside).
In Excel, the percent of men above 4SD:and for women:
which gives a ratio of about 2.2. Actual individual LSAT scores are cut off at 3 SDs (180), but that doesn't significantly affect this aggregate calculation.
I was amused that the NY Times' otherwise rather PC agitprop front page article on women's high college admissions and GPAs yesterday mentioned this tendency ("Professors interviewed on several campuses say that in their experience men seem to cluster in a disproportionate share at both ends of the spectrum...") in a favorable authoritative way but without mentioning its connection to the dangerous ideas of the Summers affair.
frankcross: True, though, as I note above, the NYT article actually acknowledged the distribution in question. Also... Just to pick one point in a lengthy article, the article seemed pointedly disinterested in why women choose different fields than men, even when their choices predictably pay significantly less. It nodded to it, but only to lump today's hot subfield of engineering in with physics as "male-dominated." It did not deign to make the connection when it mentioned men's higher wages on graduation. Nor did it deign to address the popular stereotype of different grade inflation in such field. (Possible GPA connection? Inconceivable!) And evidently no cognitive dissonance was expected among the better class of reader when men afflicted with "lassitude" who "don't seem to hustle as much" and so forth grind partial differential equations and becoming chemical engineers (at least in those years when it looks like the job market will be good), while higher wages seem insufficient to motivate the women "who feel more pressure to achieve" to do so. So one man's "extensive documenting" becomes another man's "PC agitprop."
anon: The track record of "genetics has nothing to do with it" advocates is rather poor as well. Identical twin studies on heritability, anyone? 20/20 hindsight doesn't make it shameful to've guessed wrong in that controversy, but, e.g., it does make it at least very embarrassing to've been so certain of one's wrong guess that one designed one's studies to simply ignore the possibility instead of controlling for it.
(I'm also more interested in what happens to normal people rather than what is going on in the extremes of the distribution, but that's another matter.)
There are obviously biological differences between men and women. But people are too quick to leap to this explanation. The cultural differences are not so crude as sex discrimination. They are different societal expectations on matter such as risk taking.
Mike [2], my "now what" wasn't rhetorical. Now what? What does conceding the possibility lead us to in terms of policy, in your view?
LPC, I don't have an opinion as to whether genius-level IQ can be cultivated by educational advantages. But it does seem likely to me that whether a person realizes his or her intellectual potential is at least somewhat dependent on the education --- both formal and informal --- that he or she receives.
Now, would you say that there is some genetic feature that predisposes both to family crisis and genius? Or would you think that perhaps early life crisis alters a person in such a way to make genius more likely.
This sentiment, I suppose, sums up aptly enough why people shouldn't be afraid of this topic. The people who fear this discussion are afraid, not without some justification, that accepting group cognitive differences will pave the way for moral and political subjugation of those groups. That does not have to be the case at all. We are not building a master race here. It would just be very nice to know some of these things to adjust social policy to the real world, instead of addressing it to perceived wrongs that may be explained away by genetics, rather than anything more nefarious.
One concrete policy example would be as follows. If we concede the possibility of intrinsic differences, and then have an honest public discourse about it, what we MAY find is that there is a strong correlation between gender and certain "types" of intelligence. If we can properly assess the relationship between genes and environment, instead of just putting all of our chips into the environment camp, we can change affirmative action policy accordingly.
It's a touchy subject. And just to underline my point: though you have to be very smart to make it to Supreme Court clerk, the behavioral element is big and the biggest geniuses probably don't go into law.
Is there an IQ threshhold that might correlate to super law school achievement? What's your guess?
As I understand it, there are statistically significant gender differences for certain types of cognition, especially at 3+ standard deviations. Is that true for IQ? Can any of these cognitive abilities be correlated to super law school achievement? The LSAT data hints that there might be.
This is incorrect. Average IQ is 100 by definition. It wouldn't matter if everyone's intelligence suddenly doubled, the average IQ would still be 100.
What does this tell us about gender differences? It makes it possible that they are genetic, but unlikely. Examine recent history. Consider the gender differences that existed in 1900 or 1950. Now consider how foolish one would look if one ascribed them to genetics rather than environment, given the vast changes that have occurred since that time.
We may discover, as science advances, that there are provable intrinsic differences in cognition between the sexes, or we may not. These differences, if they exist, may be substantial, or they may not. They may translate into measurable differences between populations, or they may not. They may be relevant to educational or professional aptitude, or they may not.
And even if they do exist, and they are substantial, and they are measurable on the population level, and they are relevant to real-world aptitude, we may decide that they are relevant to public policy, or we may not.
Given all that, it is not obvious to me that any great social benefit would accrue from a high-profile push to uncover genetic differences between the sexes, at least at the present time. And the social costs of such an effort --- particularly if it uncovered spurious "differences" that were prematurely incorporated into public policy --- could be immense, and catastrophic.
I don't think anyone argues that IQ is determined solely by genetic factors. Most estimates of heritability of IQ range between .5 and .8 (meaning that 50 to 80% of variability in intelligence is accounted for by genetic differences). So, everyone concedes that environment plays an important role. Therefore, demonstrating that environment plays a role does not disprove a genetic contribution (any more than showing that environment plays a role in determining adult stature means that genes play no role in causing individual differences in stature).
anonyomousss wrote:
That is why one can make no strong prediction about the sex compostion of the bottom of the law school class (or even the bottom of the LSAT distribution) based upon variability. The entire distribution of a law school class (at least we hope) comes from the right side of the ability distribution. That is not to say that what is being called the "idiot-genius effect" would not show up at the very far right of the distribution. The geniuses may be in the distribution, but the idiots would not be.
Brooklynite wrote:
In fairness to Eugene, his original post asked whether a whole series of things might be "the cause" and also whether it might be a combination of multiple factors.
Brooklynite (mail) (www):
I'd be interested in knowing what the mechanism might be for cultural expectations creating greater variability among male toddlers than female toddlers and what support there is for that explanation.
Bear in mind that children are unreliable at even identifying people's sex until they are about three, and even after they can they tend to stereotype themselves less than they stereotype others.
For example, 3- and 4-year-olds will justify their selection of toys for others on the basis of sex stereotypes, but they justify their own preferences in terms of the toys' characteristics. Eisenberg, N., Murray, E., and Hite, T. 1982. Children’s reasoning regarding sex-typed toy choices. Child Development, 53:81-86.
Moreover, many behavioral sex differences show up before children can even identify their own sex. Lisa A. Serbin, Diane Poulin-Dubois, Karen A. Colburne, Maya G. Sen, &Julie A. Eichstedt, Gender Stereotyping in Infancy: Visual Preferences for and Knowledge of Gender-Stereotyped Toys in the Second Year, International Journal of Behavioral Development, 25:7-15 (2001).
As for cognitive differences, psychologist Diane Halpern reports that “the male advantage in transforming information in visual-spatial short-term memory is seen as early as it can be tested – perhaps at age 3 – and in mathematical giftedness as early as preschool.” Diane F. Halpern, Sex Differences in Intelligence: Implications for Education, American Psychologist, 52:1091-1102, at 1093 (1997).
One of the difficulties in attributing differences in cognitive performance to "cultural expectations" is that those expectations have to explain not only why males tend to do better in certain tests of ability but also why girls get better grades throughout their school careers. That is, why is it that girls suppress their performance on anonymous standardized tests but not on more visible measures such as school grades?
Another thing to bear in mind in this discussion is that one of the alternative hypotheses on the table is that the apparent sex disparity among Supreme Court clerks is a consequence of discrimination by Supreme Court justices in their hiring. So, even if sex differences in variability of intellectual performance can plausibly be blamed on some pre-school social experience that children have, that is still not something that can be laid at the door of the justices. That is, the range of explanations goes beyond "Is it biology or are the Justices discriminating?"
Hank, if the mean is the same and the variance is larger, it's a necessary consquence that males will be over-represented at both ends of the distribution. In fact, saying it that way is logically equivalent to saying "greater variance".
i think that depends on your theory of what consitutes discrimination. i think it's extremely plausible that justice scalia, at least, is discriminating, given how overtly sexist his opinion in the vmi case was.
to be pedantic, this isnt true, unless you are assuming a curve of a certain shape. to take a simple example, compare a normal distribution centered on zero with standard deviation s and a distribution where equal numbers of data points are at 1.5*s and -1.5*s, and no other values are represented (i.e. the distribution is two vertical straight lines). obviously the normal curve has infinitely more data points at, say, +/-3*s
I don't know what the mechanism might be. In my view, the default assumption is that any two human populations are of equal intelligence, and any contention otherwise bears the burden of proof.
Awareness of gender difference isn't a precondition of gender-based behavior. If adults treat children differently on the basis of gender from birth --- and they do --- then that difference in experience is likely, it seems to me, to translate into different behaviors and aptitudes at a very early age.
Again, I'd say this is irrelevant. If a child is discouraged from certain behaviors and encouraged in others because of his or her gender, that child is going to be shaped by that experience regardless of his or her own understanding of gender difference.
By the age of three, a typical American child has been playing with gendered toys --- to pick just one small example --- for three years.
That's a difficulty that both cultural and biological explanations encounter, it seems to me, and I don't see why it's a bigger difficulty for one than the other.
I'd certainly agree with this, and whole-heartedly.
No, it's not. The scientific knowledge however is largely already there. Public discourse is needed to get our social policy on par.
dont all people stereotype themselves less than they stereotype others? i dont know what the research says about this, but i'd be surprised if three-year-olds were in some way unusual.
Sorry, I thought it was obvious that we were talking about a large population and a gaussian (normal), approximately symmetrical distribution ... since we were talking about the distribution of IQ across the whole of humankind, I guess I imagined this was clear from context.
As for the biological explanation, IMHO, the reason it is so strongly resisted is that many of us who believe in women's equality see that reliance on such arguments is often used as an excuse to maintain the status quo. Oh well, no need to reconsider social or work structures for example. After all, women just don't tend to be as smart. So, I agree with the question "now what?" Even if it were true that there is some minor biological difference causing men to be overrepresented at the two ends of the bell curve, I can't see how that would change a single policy reform that I would advocate, and I would consider putting a lot of resources into finding the answer to the question to be a significant waste.
Brooklynite wrote:
I have no problem with that as a default assumption, but what I do have a problem with is the imposition of a "beyond a reasonable doubt" standard of proof. (I don't mean this addressed to any particular commenter, but it is my experience that any old conjectural sociological explanation is taken seriously and is taken to trump even well-founded biological arguments).
It's true in concept that awareness of a difference between sexes is not logically necessary in order to result in sex differences in behavior. However, the differences in treatment tend to be pretty subtle, so I don't think one can make a strong argument for the effect of differential socialization without conceding that children are actually predisposed to learning "gender roles" very easily.
Also, bear in mind that much of the treatment of children, even infants, is responsive to the child's own characteristics. I hate to quote myself, but here is an excerpt from an article of mine from the Cardozo Women's Law Journal article that I referred to on the other thread:
Discussing Joyce F. Benenson, Maria Philippoussis, &Rebecca Leeb, Sex Differences in Neonates’ Cuddliness, Journal of Genetic Psychology, 160:332-342 (1999) &Christine L. Williams, Gender Differences at Work: Women and Men in Nontraditional Occupations, at 11 (University of California Press, 1989).
Brooklynite wrote:
Again, there is substantial evidence that toy choice is responsive to child preference. Fathers are certainly less likely to give dolls to their sons than their daughters, but sons are also less likely to play with them if they are given.
In many ways, children are more sex-typed than adults. For example, parents are much less likely to give children sex-typed toys when they choose the toys themselves as opposed to when they give children the toys they ask for. See Campenni, C.E. 1999. Gender stereotyping of children’s toys: A comparison of parents and nonparents. Sex Roles, 40:121-138; Fisher-Thompson, D. 1993. Adult toy purchases for children: Factors affecting sex-typed toy selection. Journal of Applied Developmental Psychology, 14:385-406.
Even young monkeys exhibit the same kind of sex-typed toy preferences found in humans, and as far as I know no one argues that they have been indoctrinated into sexist societal expectations. See Gerianne M. Alexander &Melissa Hines, Sex Differences in Response to Children’s Toys in Nonhuman Primates (Cercopithecus aethiops sabaeus), Evolution and Human Behavior, 23:467-479 (2002).
Brooklynite wrote:
I think it is a bigger difficulty for the "cultural" side of it, because the argument for sex differences in math performance are usually based upon such things as expectations that "math is for boys" or that girls are "playing dumb" or that teachers aren't serious about teaching math to girls, etc. If that is true, then I would have expected girls to do relatively better on anonymous measures of ability and relatively worse on public ones.
Instead, it is the other way around. What you see is that on tests of mathematical reasoning, boys tend to do better, both on average and at the high end, even though they get worse grades. Part of the reason for that is that at least in primary and secondary school, grades are given in part for "compliance," and girls tend to be more likely to give teachers what they want than boys. Also, in general, girls do relatively better on curriculum-based tests of knowledge compared with tests of reasoning ability, which is thought to account for the lesser sex difference on portions of the ACT (which is heavily curriculum-based) compared with the SAT (which is less so).
I have not outlined any of the substantial evidence for hormonal contributions to sex differences in cognitive abilities (and in temperament), which provides a biological mechanism for some of the sex differences to develop.
Finally, "Another Clerk's" response should remind us that there are in fact lots of causal factors involved and that no one is arguing that cognitive differences (if any) are solely (or even mostly) responsible for existing patterns. If I had to guess (and this is just a guess), I would guess that the kinds of factors she refers to (and that were discussed in the other thread) are more important than the ones that are the subject of this thread. They, of course, have some biological roots, as well, but that is another story and one I will leave for others.
Kingsely Brown: You missed this article: LINK
One possibility is that competitive pressures, probably mediated by testosterone, drive men to higher levels of achievement (and, indirectly, to higher levels of performance on IQ tests) – although this is a theory. If true, then individual (and perhaps sex differences) in testosterone should correlate with IQ. (Hmmmm: LINK)
Ed (Re: IQ scores going up for decades): “This is incorrect. Average IQ is 100 by definition. It wouldn't matter if everyone's intelligence suddenly doubled, the average IQ would still be 100.”
You’re right, and that’s why raw scores are used. Although average raw scores on g-loaded tests (Ravens) have been increasing over historical time (the so-called Flynn effect), average differences between groups (White-Black) have remained stubbornly persistent. I would expect this to be true for male-female differences, too, to the extent that such differences are mediated by biological/genetic differences (hormones, brain size) that are relatively immune to cultural influence.
Is the cause possible differences in innate intelligence at the tail ends of the bell curve (what I'd heard called the idiot-genius syndrome, which leads men to be overrepresented both among the very low-IQ and the very high-IQ)?
Oh, please. I know it's in question form, but really..."
A fair question dserves a fair answer. It is really pretty simple to anyone who has extensively researched the structure of the legal professin and standardized testing for more than 16 years. Men ARE cognitively different than women, and one of those differences is the ability of men to have higher success rates on standardized tests. Then there is the structure of the legal profession where grades earned at a top tier law school are not the equivelant of grades earned at a second or third tier law school, and the "percentiles" (e.g., top 10%) screening criteria that screens out everyone who lacks this standardized test taking ability from all meaningful opportunities in the law -- udicial clerkships, big firm employment offers, law review in the schools that do not utilize write-on competitions.
The upshot is that the current state of affairs not only perpetuates the exclusion of the disabled and African Americans, but serves to perpetuate the male dominated heirarchy at the highest levels of the legal professon. Sorry women, but that is how it is, cognitively speaking (supported by research).
EV, no watering down of your accomplishments intended. You are a class act. But a fair questions deserves a fair answer.
P.S. Please name for me each and every mentally disabled, learning disabled, and autistic clerk to Supreme Court Justices since the institution's inception. Bet there are none.
“The study made it understandable how an exponentially growing male/female ratio at the high end of the g distribution—with the exact numerical ratio being a function of the size of the average sex difference in mean g and of the dispersion of g—provides part of an explanation of the male dominance in high society.”
you were sounding very convincing until you got to this. when did we develop these curriculum-independent tests of "reasoning ability"?
Wow. An admission the LSAT violates the Americans With Disabilities Act by design to exclude the mentally disabled.
"See the chart above; the bell curves for men and women are nearly identical in shape (women do slightly worse than men on the LSAT, but men aren't noticeably more variable)."
Significantly absent from the statistical study is the category of "disabled," as well as the agenda of the outfit that paid for the statistical research. Why do people just take statistical studies at face value without critical thought? the results can be slanted just as any version of one side of a case by that side's legal counsel.
Another reason for requiring joint J.D./M.B.A. degrees.
If this were true, why are women outperforming men in college today?
I would say the reason you need very strong evidence to depart from presumptions of equality is because history is littered with countless examples of claims of genetic differences that are plainly wrong. We've had a lot of false positives.
Do women genetically underestimate themselves, or is that more likely to be cultural?
Duh ... learning disabilities and autistic savants are predisposed to genius. However, they are also disabled, and heck if the legal profession is apt to embrace opportunities for such individuals.
As is asked ... name one Supreme Court clerk who is autistic.
I'm curious as to what policy reforms would rectify this.
It finds that men are at the top of the class in some subjects and not others but attributes this to culture. The interesting data, though, is that men and women in law do equally well. As many women in England get top scores in law as do men. Also true at the bottom.
And ... maybe we should completely change the structure and how we test and provide opportunities to those seeking to be a part of the legel profession?
Generally, boys get worse grades but score higher on tests. IQ is tested, not graded. QED.
Ann, review the six decade debacle in Florida of Virgil Hawkins. The Florida Bar advocated Mr. Hawkins not be granted bar admission 'until he turns white' or the KKK rides out of Gainsville. 'Until he turns white' is the phrase describing drapetomania, a psychiatric disability made up about African Americans, that is the inability to turn white. And this was occurring as late as the mid-1970s!! This is pretty close to a "racial disparity."
This is critical. Before you can say that this "more men at the extremes" idea has an important impact you have to show that clerks in fact come from that far out on the curve. I find this dubious. Looking pure IQ scores, for whatever they are worth, I suspect that the top law students do not match the top students in mathematics, say, or the sciences. So this may not matter at all.
The statistics may even cut the other way. If there are more men way out on the right of the curve, then there are more women in some range that stretches from the median to well above average. There is some point at which the curves cross. Are we sure that top lawyers come from a region where there are more men than women? Maybe, or maybe not, or maybe even if they do the difference is so small that other factors like social pressures are far more important. I'm dubious, but then I'm not a lawyer, so I have no investment in the assumption that the law is an especially demanding profession intellectually.
I'm all for equality of opportunity and access. But that's quite different from endorsing equality of condition. Put another way, everyone should have a chance to go to a decent school and to live in a safe home - without regard to gender or race or any other such intrinsic factor. NOT every group should be assured of equal representation.
Will you also have us change the standards of being allowed to play in the NBA, on the ground that the present ones disproportionately exclude the short, (and whites?)
College performance is largely based on GPA; GPA and mental ability tests have different validity coefficients.
In particular, GPA doesn’t come close to matching IQ in predictive utility (i.e., predicting later outcomes such as job or academic performance): LINK (cf. grade point averages to mental ability tests).
These differences in predictive utility can probably be attributed to GPA being loaded with measurement error (differences in computing GPA across institutions) and unique variance that has nothing to do with mental ability (e.g., a teacher liking a student based on characteristics that do not generalize outside of the classroom). IQ, in contrast, has low measurement error, high test-retest reliability (your IQ is more reliable than your weight, measured at 12 mo intervals), and excellent predictive validity beyond the classroom. Based on everything we know about IQ and GPA and their predictive utility, the sex differences in mental ability (as measured by IQ) should account for more variance in real-outcomes (including clerking at the SC) than the sex differences in GPA. This pattern has been confirmed to my knowledge for almost all (all?) jobs that have been studied to date. (IQ validity coefficients are actually stronger for jobs that involve higher levels of complexity, but you still find correlations in the predicted direction for jobs that ostensibly involve low complexity, like janitorial work.)
The classic article on the predictive utility of mental ability tests (which include IQ tests) is: Hunter, J.E. and R.F. Hunter (1984) The validity and utility of alternative selection predictors of job performance, Psychological Bulletin, 96, 72-98.
Frank: “I would say the reason you need very strong evidence to depart from presumptions of equality is because history is littered with countless examples of claims of genetic differences that are plainly wrong. We've had a lot of false positives.”
I don’t disagree with the sentiment. But if you don’t think the findings concerning mental ability are important in explaining sex differences at the highest levels of achievement, what (evidence) would satisfy you that mental ability *is* important?
It's just fairly frustrating to have the entire discussion focus on whether or not these differences exist, with little or no attention paid to the actual obstacles in the way of talented female law students who want to clerk on the Supreme Court. (I really don't thinkthere is very much intentional discrimination going on - indeed, I believe there are one or two justices who practice a degree of gender-based affirmative action - but there are difficulties for most married women of childbearing age in clerking on the Supreme Court that do not necessarily exist for the equivalent pool of men.)
In math, gender intelligence differences at the highest levels of performance seem reasonably well documented, but most of the discussion of whether the same is true in law is based on wholly unfounded speculation not backed up by any data whatsoever. Further, even if such differences do exist and are a cause of the Supreme Court disparity (and, in this regard, it's important to note that attaining a Supreme Court clerkship depends on many factors in addition to raw intelligence), unless they are the only cause, so what? It is just as intellectually dishonest to ignore other possible causes of the disparity as it is to ignore the possibility of cognitive differences, yet I don't sense the same rush to explore and document those other potential causes here.
Blaming the higher moments of the distribution (well, higher than the standard deviation anyway) is an intuitively attractive model for two reasons. First of all, you can construct a leptokurtotic distribution from two normal distributions with the same mean but different standard deviations. If there is something going on with the X chromasome, and there are several traits with differing standard deviations, then women have more chances to end up with mixed smart/dumb than men do. If one variation is not fully dominant over the other, or there are more than two variations, the statistics get very interesting, and start looking like the sort of distributions we see in real life.
The other effect is when you look at the distributions from the point of view of college admittence, naming to SCOTUS clerks, placement in regular classes rather than special ed, you expect that there is not some hard and fast cutoff, but some function where as you move from the smartest to dumbest ends of the distribution, you exclude a larger and larger fraction. If you think of how this works from the point of view of a fat-tails-fat-mean distribution, vs a tall-shouldered distribution, then you would expect a characteristic shape, which would be a function of how "exclusive" the grouping is. So SCOTUS clerk might be basically populated only by people on the genius tail of the distribution. While looking at say the undergraduate population of a good liberal-arts college your population is centered around the smart shoulder of the distribution, where the women outnumber the men. And you would expect such a college to have a small number of very smart kids, disproportionately male, and then of the 95% not in that group, it would be 60-40 female male with a higher proportion of women on the smarter side. If you compare to a not-very-selective college, there you would expect very few members of either tail (geniuses got attracted to better places; idiots weren't allowed in). But you would reach down somewhat into the "dumber" shoulder. So there you would expect men to cluster at the middle, while the women would have a far flatter distribution, and on average, smarter than the men.
Having observed close up an excellent college, a very good college, and a pretty lousy college, what I saw was pretty much what I would expect to see if the kurtosis of the male distribution were significantly higher than the female distribution, and then you applied the 3 different college admissions criteria.
cathy :-)
Oh, my. Wasn't it Chief Justice Roberts who expressed concern whether his law clerk might be able to sue him over employment? See, the Chief Justice is really so intelligent, even I would not sue a Justice over not hiring an autistic Supreme Court clerk, but rather would try to use my very best skills to persuade how capable an autistic clerk can be. Isn't that what diversity is about?
Another thing, I have to say the Chief Justice has many supporters, and during his nomination when I asked his views on disability discrimination, one of his very good supporters won me over by citing me to a Circuit Judge Roberts Rehabilitation Act decision to show me why he deserves to be the Chief Justice. Like EV, the Chief Justice and his supporters are also class acts, by trying to persuade my point of view rather than attacking the disabled.
How has this turned into Justice bashing? For every VMI there is a Goodman v. Georgia (unanimous), upholding the rights of the disabled.
If what some people are saying is somehow the Supreme Court Justices are responsible for setting the policies for criteria and testing format to enter the legal profession, including selection for clerkships, then they should say so. And if THIS is the perceived problem, maybe it is high time for the Supreme Court to revisit bar admission.
Even if we take these claims as proven, in other words --- at least for the sake of argument --- they don't add up to a demonstration that men and women possess broad differences in cognitive capacity. At most, they suggest that men or women may, on average, be more adept at solving certain problems in specific ways. They seem to make no credible case for any biological difference in aptitude in a field like law, and no credible argument for revision of public policy on questions such as affirmative action.
I have not outlined any of the substantial evidence for hormonal contributions to sex differences in cognitive abilities (and in temperament), which provides a biological mechanism for some of the sex differences to develop."
Men have such a propensity to oversimplify. No doubt men were the creators of the faux multiple choice testing format. Now I am wondering in light of the statistics, correlations, causations, and gender, what happens to all of these assumptions in light of 1 in 166 children born being autistic, fact that more men than women are autistic, and the new research indicating hormonal balances in the womb affect autistic outcome.
I am just saying there are many potentially significant factors other than simple conjecture ...
Maybe because they take classes with papers and assignments or group projects instead of standardized tests? Has anyone tracked this?
Now that’s interesting. A great deal of effort has been spent trying to determine why men fall at the extremes of a distribution. Explanations for males at the lower tail have focused on sex-linked genetic disorders. Explanations for males at the higher tail are more varied (recent theories have targeted testosterone). I haven’t heard anyone proffer the XX dilution theory, though, which seems very plausible. Got a citation?
One point Professor Volokh made that doesn't seem to be getting much notice is that there may be multiple causes. The intelligence/gender angle is just one, and I think he mentioned it first so it wouldn't get lost or easily dismissed.
I commend Adam on placing a wide range of left-handers in the population. His point was that they are overrepresented at the extremes like males, but include both genders. As a left-hander, I made a point of checking how many other lefties were in my school classes, from the first grade on. There were as few as 2 in grammar school classes of 40-50 (we baby boomers had LOTS of classmates) and as many as 6 in my college classes of 28-36. My point is that there's so little accurate data that it's impossible to determine how many lefties there are in a population, much less draw any conclusions.
It's anecdotal, but I noticed most lefties were boys in grammar school, gender-equal in high school, and female-dominant in college - and the overall gender ratio in college was nearly 1:1 four decades ago.
As for Ann Althouse's comment that Eugene's getting all the comments, that's easy to figure. Anyone can comment here, but at Ann's site you have to register. Ann can get more comments if she opens the comments, but if she wants a better class of commenters, she should rename her blog "Ann Althouse's Law Blog" - and stop posting about her favorite contestants on American Idol.
As for IQ, it is a stew of nature and nurture. I don't doubt that and I recognize the possibility that gender may have an effect. But I doubt it, simply because of all the past false positives. They don't prove that today is not a true positive but I'd want strong evidence that it is. I think there's a human tendency to think that current conditions in the locale are the "normal" or "natural" state of the world, but they inevitably are not.
If we link male dominance at the top in US law schools with intelligence, then a failure of this trend in Britain could speak not to the domestic trend's validity, but to British testing methods. For example, are their tests blindly administered? Highly g-loaded? Do men and women go into very different types of law, the result being that the former take test-based classes, while the latter opt for the essay-based?
...Or perhaps you are right - necessarily conjectural as the above paragraph is.
The suggestion that "innate intelligence," whatever that is, was at the root of the situation could only have been made by a former male Sup. Ct. clerk with a fair amount of hubris among his own characteristics, at least that's what I would bet.
Many supporters of the concept of "innate intelligence" will tell you that things you call common sense, or good judgment, or linguistic ability etc., all comprise what the psychometricians would call "g." Volokh's claim may seem facially ridiculous to you, but I assure you that there are many reasonable people to whom it does not. Also, way to end it with an ad hominem.
Essential functions to participate in the NBA are mostly physical and visual-spatial skill based (i.e. hands-on performance skills). no different than the essential functions to ride and jump a horse in show competition or (analogizing to the NBA coach, teaching someone else to do so).
Essential functions to be a Supreme Court clerk are highly language-oriented, written and spoken (one has to relate to their boss and the other clerks), not physical visual-spatial hands-onperformance based.
I can give you a classic example of why your coparison does not hold. Job Candidate A applies for equestrian riding and instructor position. Employer conducts office-type interview based on written and spoken language criteria, screens person out of employment for dictating responses to inaccessible employment application form by Dragon Dictate softeare on attachment in essay format, implies person is "illiterate," and begins quizzing person as to any "crimes" committed. Employer never permits job applicant to ride a horse or give a sample riding lesson. Result: Employer violates Title I of the Americans With Disabilities Act for using prohibited criteria (office style interview) that tends to screen out the disabled qualified to perform the essential functions of the position (riding a horse, giving a riding lesson), and for using prohibited non-essential functions to deny employment.
Ahh, but you must be a reader of Justice Scalia's and Thomas' dissenting opinion in PGA Tours, whereby they described the California Bar Exam as an employment test for entry into the competition to be a licensed lawyer not different than that for which Martin sought accessibillity to perform the hands-on performance requirements to play in the PGA Tour. This definitely complicates everything, because now it would appear Title II (incorporating Title I) employment provisions of the ADA apply to the legal profession, bar admission, bar examinations, and attorney discipline.
That still does not convert prohibited non-essential functions into essential functions, only adds additional protections to bar applicants (no physical or mental medical examinations in the character and fitness stage until after a conditional offer of bar admission has been made.
"I'm all for equality of opportunity and access. But that's quite different from endorsing equality of condition." Uh, no ... see US Airways v. Barnett (US Sup.), accommodations preferences are required to level the playing field because the disabling conditions create the inequality of opportunity vis-a-vis the non-disabled. Where did you go to law school? There is no distinction between "equality of opportunity and access" vs. "equality of condition" written into the language of the ADA. Another example of why your argument is nonsense: person who is autistic cannot place information into written format by manually keyboarding due to both cognitive limitations and physical conditions. Autistic person is entitled to "equipment and devices" to equalize her condition in comparison to a non-dsiabled person's condition, i.e., speech recognition software and extra time. Somehow you just don't get it.
"Put another way, everyone should have a chance to go to a decent school and to live in a safe home - without regard to gender or race or any other such intrinsic factor. NOT every group should be assured of equal representation." And this is ... more of the same of the last 100 years of oppression of the disabled? You need to read the Solicitor General's briefs in Medical Board of California v. Hason (US Sup. dismissed) and Tennessee v. Lane.
In sum, there are differences that cause some people (largely non-disabled white men) to suceed with high scores on standardized tests at the expense of many women, African Americans, and the disabled, thereby perpetuating the heirarchy of such men over the rest.
- (1) I wasn't even talking about the disabled. I'd be out of my element to discuss the ADA - I don't know anything about it.
- (2) To answer your question, I'm actually still in law school, that law school being Boston University. I was hoping my name here made that kind of obvious, but I guess not to you.
- (3) I respect your position in that I find it well-meaning and sympathetic. You apparently do not respect mine. I urge you