July 10, 2006

A response.

Eugene Volokh responds to my "oh, please."

MORE: In vlog form!

38 comments:

Patrick said...

Is it possible men are more driven? Men often strike me as more narrowly focused than women, or if you prefer, women have more balance between a personal and business life. Look at the upper reaches of the corporate world, dominated by men who put in very long hours.

Ann Althouse said...

The collection of individuals is so small and so minimally skewed toward the male that it's a terrible basis for making generalizations about the sexes! Hey, I'm a girl, but I can still see that's damned unscientific. If we're going to just speculate, well, how about not doing it?

Bruce Hayden said...

I agree with Eugene that 1.2 to the fifth power is nearly 2.5, and that the differences can most likely be attributed to this dynamic. In other words, it isn't one thing, but rather, a combination of several different factors, among which I would suggest may be:
- at the ends of the bell curve for IQ, males predominate. Essentially same mean, but slightly broader standard deviation. Most SC clerks are probably at least 2, and more likely 3 SDs out from the mean.
- women do, on average, try to balance their lives a bit more, and it takes a fairly unbalanced person to make it to the top of their Harvard or Yale law school class.
- lower court judges are still predominently male, and probably prefer, at least by a little bit, male clerks. You may not find that many real chauvenists, but tenure is for life, and there are some out there.

Bruce Hayden said...

Ann, but that is what we do all the time - speculate.

Seriously, I don't see a resolution to this problem for exactly the reason you suggest - the numbers are so small that the noise will defeat any attempt at a valid answer.

ben wallace said...
This comment has been removed by a blog administrator.
ben wallace said...

I agree with Ann. The IQ argument relies on two assumptions: (1) men and women differ in IQ and (2) that the group selecting individuals is choosing over the range of IQ where there are differences between the sexes, i.e. the extreme ends of the IQ distribution. Larry Summers applied this logic to tenure decisions in hard sciences at the top departments in the world where there is reason to believe that the pool of individuals is from the extreme tails of the distribution. Unless we are willing to assume that the pool of applicants to the Supreme Court is similar to the pool of applicants seeking tenure in the physics department at MIT, the second assumption is violated. Thus, the speculation by Summers might make sense in the case of hard sciences but far this hypothesis makes far less sense with respect to clerkships; hence, "oh please." Summers also hypothesized that intrinsic differences are much less important in explaining tenure decisions than social factors, implying there should be even less relative importance of intrinsic factors in SC decisions to hire clerks.

Patrick said...

Oh, you want actual substance eh. There is this from a debate between Steven Pinker and Elizabeth Spelke. Much more at the link, but this is pertinent to what I theorized.

http://www.edge.org/documents/archive/edge160.html

Just one example. In a famous long-term study of mathematically precocious youth, 1,975 youngsters were selected in 7th grade for being in the top 1% of ability in mathematics, and then followed up for more than two decades. These men and women are certainly equally talented. And if anyone has ever been encouraged in math and science, these kids were. Both genders: they are equal in their levels of achievement, and they report being equally satisfied with the course of their lives. Nonetheless there are statistical differences in what they say is important to them. There are some things in life that the females rated higher than males, such as the ability to have a part-time career for a limited time in one's life; living close to parents and relatives; having a meaningful spiritual life; and having strong friendships. And there are some things in life that the males rated higher than the females. They include having lots of money; inventing or creating something; having a full-time career; and being successful in one's line of work. It's worth noting that studies of highly successful people find that single-mindedness and competitiveness are recurring traits in geniuses (of both sexes).

Patrick said...

Oh, you want actual substance eh. There is this from a debate between Steven Pinker and Elizabeth Spelke. Much more at the link, but this is pertinent to what I theorized.

http://www.edge.org/documents/archive/edge160.html

Just one example. In a famous long-term study of mathematically precocious youth, 1,975 youngsters were selected in 7th grade for being in the top 1% of ability in mathematics, and then followed up for more than two decades. These men and women are certainly equally talented. And if anyone has ever been encouraged in math and science, these kids were. Both genders: they are equal in their levels of achievement, and they report being equally satisfied with the course of their lives. Nonetheless there are statistical differences in what they say is important to them. There are some things in life that the females rated higher than males, such as the ability to have a part-time career for a limited time in one's life; living close to parents and relatives; having a meaningful spiritual life; and having strong friendships. And there are some things in life that the males rated higher than the females. They include having lots of money; inventing or creating something; having a full-time career; and being successful in one's line of work. It's worth noting that studies of highly successful people find that single-mindedness and competitiveness are recurring traits in geniuses (of both sexes).

ben wallace said...

Patrick: This data implies that the variation in clerkships is expalined in part by prefernces, as opposed to IQ. These differences in preferences can be explained in part by gender roles. Hence, the data imply a preferce-based or sociological explanation for differences in clerkships. At the same time, I suspect that there are an equal number of men and women who actually prefer clerkships among the brightest law students.

In addition, any such data on people who test off the board in math does not necessarily have any implications for clerkship patterns unles the SC is selecting from a similar sample as the sample you described.

37383938393839383938383 said...

This whole "girls getting better grades" thing annoys me. It isn't as if grades accurately reflect anything but gaming the grading system. Where 10% of your grade is class participation (read: showing up early and answering the easiest questions); and another 5%-10% is attendance (just showing up); and professors slash off points late-filed assignments (or refuse to give partial credit), then it is pretty easy to note that a consistenyl punctual, ass-kissing rote-memorizer without any true understanding of the material can get a baseline of around 25 points higher than the other less anal students. All it takes is a B-minus mind to get an A in such a grading system. So I'm not sure why pointing to a proliferation of A's on someone's transcript is proof of intelligence or knowledge; it may just be proof of meticulous puntcuality, ass-kissing, and rote-memorization. We have all met these kind of people, both male and female. They suck. No one likes them. They are not deserving of praise.

And people also seem to ignore that there is massive cheating throughout the American college system, there is grade inflation all over the place, and college students exit college without basic understanding of math or history...so what exactly are all these As worth, anyway? The gender disparity in and of itself tells us nothing except that girls value having nice GPAs more than boys do; this is not a surprise: if I planned on dropping out of the labor market in 5 years to have a kid, I would want to maximize my earnings potential before I did so also. I mean, duh.

Palladian said...

Love the vlog!

Freeman Hunt said...

Love the vlog also!

Much better than podcasting. Great stuff!

reader_iam said...

Cool! But please don't stop podcasting ... I couldn't walk, or fold laundry, or drive around, or sit at the playground, and watch the vlog!

Eh, how selfish of me, I know.

I know!--how about you devote your entire life to blogging, vlogging and podcasting, meeting our needs in every medium?!

Probably not, huh?

Ann Althouse said...

There's also video podcasting...

Well, I really like doing the little movie clip. Now that I've figured it out, I'm going to use it for variety here. It might be faster than some updates, and it's pretty amusing.

Dave said...

Interesting vlog. It's interesting to see you speak, as opposed to reading your writing or listening to podcasts.

HD_Wanderer said...

Align left: I don't know how much of this you have control of., but your defined "Object" width is 425 while your defined "Embed" width is 600. If you make you embed width 425 as well, you'll be able to align as you choose, left or otherwise.

Jennifer said...

Love it too! You have an expressive face that is great for this medium.

And, I love the blue glasses.

Christy said...

Call me a luddite. I prefer the written word over podcasting or vblogging unless there are 2 or more people involved. Not that you vblog wasn't lovely.


With regard to the subject at hand, I note that there are few references to mentors, or is that what people mean when they refer to feeder judges? It struck me in the publicity around the Robert's hearings that he had been well mentored. That calls up perhaps some of the difficulties of women getting ahead. With the numbers as they are today, the mentors would be largely male. How comfortable are they mentoring young women?

Dawn said...

Ann, have you read every book there behind you?

The blue glasses are the best!

Editor Theorist said...

There seems, perhaps, to be an assumption behind AA's comments that the default position is that men and women intrinsically think and behave the same - and that therefore differentials in career choice and aptitude are most likely to be due to 'cultural' (ie. non-biological) factors.

If this is indeed AA's assumption, I believe it goes against the scientific evidence (this is my field). Biologically, men and women are psychologically different - this is not really a matter for scientific debate - it is the case for virtually all male and female animals. What is at issue is whether the psychological differences are significant, and how big they are, in any specific situation.

For example: The scholarly work of Kingsley Browne, a law Professor at Wayne State University, has plausibly documented the major biological contribution to men-women differences at work in relation to matters such as average salaries, promotion levels, sexual harrassment, combat military careers and many other issues.

http://www.law.wayne.edu/faculty/
profiles/browne_kingsley.html

But my guess is that (rather than being a matter of IQ) the men-women differential in supreme court clerks probably has the same general cause as the differences for all high status jobs that require single-minded committment. For well-established evolutionary reasons, men (on average) value status more than women do, and are prepared to make greater sacrifices to attain it.

Paddy O. said...

Seems to me that the better explanation is that Supreme Court justices, with all the power and freedom they have, can be as chauvinistic as they want to be.

Or maybe it's their wives. They don't want young, ambitious women law clerks working for their husbands.

Or maybe it's a generational thing. It hasn't been so long since there has been an equality of women graduating from law schools.

Or, maybe women make better lawyers and have so much success in the courtroom they don't want to be clerks.

Maybe women have an predisposition to dislike the weather in Washington DC more.

Maybe men have a stronger tendency to brown nose, and feel they have to perform on their resume to make up for their natural lack of talent.

Maybe there is a special code imbedded in the Constitution told only to Supreme Court justices that has been passed on for 200 years which guides these sorts of things.

I think it's more fun to keep speculating rather than stop speculating. Speculation opens up the possibility of a lot of fun, especially when the sure fire response is to demand a study saying why the speculation is wrong.

ben wallace said...

I think Volokh has an implicit hypothesis about the effect of IQ on the size of the pool of qualified women. Volokh assumes that the pool of qualified women is smaller than the pool of qualified men. His hypothesis is that IQ differences explain differences in the number of qualified women. Some of the alternative hypotheses predicting smallers size of the pool of qualified women include differences in the preferences of women; societal bias against women; and any other factors besides IQ that may differ systematically between men and women. The IQ hypothesis is weak because there is no real difference between men and women on the part of the intelligence distribution that would determine qualifications to become a SC clerk; hence, Althouse rightly points out that some other hypotheses are likely to be relevant in explaining the variation in clerkships.

A second issue is that intelligence does not say anything about sexism that may lead to a fewer number of women seleted. Differences in the size of the qualified applicants is not the only predictor of judicial choices over clerks. Under the assumption justices reward intelligence and intelligence varies, any IQ differences will explain some of the variation, but so too will biases or sexism of justices, the non-IQ factors that explain differences in the pool of qualified women, and any differences in preferences of women that may lead to a smaller number of women seeking these clerkships. Hence, there are few reasons to believe theories based on IQ will explain these differences.

In fact, if we assume that being a clerk requires a high level of intelligence but not genius level, then the data on sex-based intelligence imply there are higher numbers of women qualified to be clerks in the population, which implies societal factors are an even stronger factor explaining the disparity in clerkships.

Mariam said...

Awesome! All I could think was "Snap! Go Ann!"

I have often thought about the disparity in the legal profession between the sexes. More specifically, why men seem to dominate the higher level firm positions, clerk positions, etc. I think the difference is much more societal than biological. But not exclusively so. Biologically, women bear children. Socially speaking, traditionally women are also the ones who take time off to care for the children. And that's only one factor that could contribute men being the predominant ones in those positions. Maybe the predominantly male Court simply prefers male clerks? Who knows!

I think your reference to the biological contributions of race as compared to sex in intelligence differences is the most germaine. Ultimately, the lesson learned is you can't make gross generalizations without scientific substantiation.

Art said...

Do people ever tell you that you look like a different person in different photographs...and you look very different in photos than you do in person?

Ann Althouse said...

Art: Actually, no.

proudtobealiberal said...

I would recommend that commentators read the chapter in Sorcerers' apprentices: 100 years of clerks at the United States Supreme Court. It is clear that the proportion of women as law clerks varies by justice, suggesting that there the highly personal hiring decisions of the justices may be the source of the lower number of women.
For example, looking at other factors, Justice Scalia seems to limit his clerks to those who have worked for a very limited number of feeder judges. According to Wikipedia (which may not be entirely reliable), for example, he has not chosen any clerks who had worked with Edith Jones (a very conservative judge from the Fifth Circuit) whereas Clarence Thomas has chosen a couple of law clerks who had clerked with Edith Jones. (Both Thomas & Scalia have typically hired at least one Luttig clerk).
Some justices hire 40% or more women; Scalia in contrast through 2002 had hired 15% women.
Thus, one question may be what criteria are used by the feeder judges to hire their law clerks.
(I would note that the "feeder" system appears to be an old boys' network).
Given that there are 9 separate employers, a few judges who hire only a handful of women law clerks can make a significant difference in the proportion of women law clerks in total.

proudtobealiberal said...

A correction: Justice Scalia has in fact hired one law clerk from Judge Edith Jones.

Interestingly, Eugene Volokh, a UCLA graduate, clerked for Sandra Day O'Connor, who hired more than 40% women and from a very diverse group of law schools and of feeder judges (compared to some of her peers who focus primarily on Harvard & Yale).

Daryl Herbert said...

Oh, please. I know it's in vlog form, but really:

1 - IQ is real and valid. People generally score consistently across different intelligence tests, and it does measure something important. And the LSAT is such a test that measures IQ.

2 - You didn't challenge that IQ is real and valid, you erected a friendly straw-man that some people don't believe it:

"even if you believe IQ tests, and obviously people challenge IQ tests, but even if you believe IQ tests..."

If this is indeed a battle of "wills," "will" you stand behind the arguments you rely on?

3 - The "idiots and geniuses" phenomenon is real and valid. You don't have to accept that IQ scores are meaningful in order to accept that Dr. Volokh's description of the statistics is correct. Even a belief that IQ doesn't matter doesn't change the fact that men have more of the highest scorers.

4 - There's a very, very simple way to take into account your criticism that we need to look at who pursues law: There are more men than women at Harvard Law School. According to HLS, it's got 46% women. 54 divided by 46 is 1.17, which is almost 1.2, so it's almost exactly the number given by Dr. Volokh needed to explain 1/5 of the phenomenon.

5 - The racial disparities can also be explained by two factors, namely how few minorities there are at the elite law schools and how many of those that are, are there because of affirmative action, and would otherwise be at second-tier schools, based on their LSAT scores. Again, you don't have to think it's right or wrong for there to be affirmative action in order to admit that fewer blacks have very high LSAT scores. In fact, if blacks had as many very high-scorers, there obviously wouldn't be any need for affirmative action at the elite level.

6 - You don't have to believe IQ scores are valid to recognize the statistics that there are fewer high-IQ blacks.

7 - You don't have to believe it's because of racial or genetic reasons, that there are fewer high-IQ blacks.

8 - You don't have to believe that women's (current) lesser membership in the highest eschelons of IQ is necessarily biological in nature, but you still must recognize it as a fact.

9 - There is nothing inherently racist about believing that IQ is largely inherited from one's parents, which is completely independent of one's parents' skin color(s).

10 - Feel free to address any of these points. The paragraphs are numbered so it should be easy enough to go down the list and "fisk" me, but I doubt even a single one is rebuttable. Describing them as being among a class of opinions that is "not helpful" or "not cool" does not count.

11 - You did say you wanted comments, right?

Daryl Herbert said...

I meant to source my claim that HLS is only 46% female:

http://www.law-school-admission.com/Harvard/#students

None of the other claims need sourcing because they are simple logical statements based on numbers no one has (yet) disputed here.

Hazy Dave said...

Perhaps a lack of social graces and empathy for other humans is a positive attribute in law clerks? Just a thought.

I'm surprised that the presumption that superior intelligence is a prerequisite for success in any particular endeavor seems to be so widespread among people who should know better. Do "successful" people see their success as proof of their intelligence? Do "intelligent" people consider "success" to be their due?

ben wallace said...

Daryl: Points 1, 6, 7, 8, and 9 are empirical statements about the distribution of IQs. Althouse does not dispute the possbility of such differences. Althouse criticized the appeal to IQ distribution as the first hypothesis offered by Volokh in explaining clerkships.

The data on IQ is not in dispute. You do not provide any evidence that the Supreme Court clerks are selected from the range of the IQ distribution where there are significant differences between men and women. The Idiot-Genius claim would only apply if there are fat tails on the IQ distribution for men relative to women and that this is the range from which clerks are selected.

You also do not describe how to disentangle the effect of IQ differences from all other factors that differ systematically between men and women, including differences in preferences of women and socialization effect that influence what women prefer. Sex captures any intelligence-based, preference-based, and sociological differences between men and women; relating these differences to clerkships requires separating out these effects. Moreover, you do not describe how do separate out the effects of factors that vary systmatically between men and women and the direct effects sexism may play in the appointment process.

Point (4) also demonstrates the problem with selecting observations consistent with your theory. Harvard is consistent with your theory but a lot of lawschools are not. Hence, we cannot draw strong conclusions from that case because Harvard is not randomly selected.

Now, I would not characterize the IQ hypothesis as irresponsible, but the hypothesis is unlikely to be supported when all other relevant factors are considered in the analysis. There are no articles I could find that analyze the relationship betwen clerking and IQ (there is one economics paper that deals with clerkships and beauty). So I guess there could be more work done in this area to test these hypotheses, but I doubt rigorous inquiry will support the hypothesis IQ differences between men and women drive differences in outcomes.

dbp said...

The collection of individuals is so small and so minimally skewed toward the male that it's a terrible basis for making generalizations about the sexes! Hey, I'm a girl, but I can still see that's damned unscientific. If we're going to just speculate, well, how about not doing it?

Hi Ann,

It took me a while to post this since I had to dust off (literally) my old stats book and relearn how to use some complex functions on Excel. What I have calculated is that the probability of 11 or fewer women and 27 or more
men being clerks is just less than 0.7 %. This strikes me as highly unlikey to be due to random chance. In fact this is roughly one Standard Deviation more than is required in scientific journals.

My point, and I do have one, is that there is a some cause (other than random chance) for this disparity in numbers. If one is to speculate about the causes, I see no reason from a scientific standpoint, to hold biological reasons to a higher standard than other reasons.

dbp

Daryl Herbert said...

You also do not describe how to disentangle the effect of IQ differences from all other factors that differ systematically between men and women

Do I have to? I never said the idiot-genius effect explained 100% of the phenomenon. I've made a reasonable showing that it could have some sort of effect. So the burden falls on you to explain why it can't have an effect, if that's what you want to claim.

The data on IQ is not in dispute. You do not provide any evidence that the Supreme Court clerks are selected from the range of the IQ distribution where there are significant differences between men and women. The Idiot-Genius claim would only apply if there are fat tails on the IQ distribution for men relative to women and that this is the range from which clerks are selected.

Let's have some data, then. First, like I said earlier, clerks are chosen from the best law schools, which are most selective re: LSAT scores, which correlate very closely with IQ scores.

Now I just need to demonstrate a "fat tail" on the distribution. Here's a chart I ran into in the comments section on the VC web site. The fatness of the tail cannot be disputed. No sir.

Point (4) also demonstrates the problem with selecting observations consistent with your theory. Harvard is consistent with your theory but a lot of lawschools are not. Hence, we cannot draw strong conclusions from that case because Harvard is not randomly selected.

Harvard was the first to come to mind. But let's do all the top five, shall we (in no particular order):

46%, 46%, 45%, 42%, 46%
Harvard, Yale, Chicago, Stanford, Columbia

http://www.law.yale.edu/fastfacts.asp
http://www.law.uchicago.edu/Life/diversity.html
http://www.law.stanford.edu/about/facts/
http://www.law.columbia.edu/jd_applicants/studentbody
http://www.law-school-admission.com/Columbia_Law_School/

The truly astute researcher would notice how Stanford and Columbia won't list percentages on their web sites. Columbia won't even give numbers, I had to go to a third party source. They say they've got "approximately half" women.

So really, I was selling myself short by choosing HLS.

ben wallace said...

The IQ issue is boring unless you describe how to separate the effects of IQ and other factors. I think that would be an interesting project but I do not think anyone has described how to separate out these effects. You affirm that there may be separate effects, and I want to know how to separate them, so we have some agreement that there is an interesting hypothesis here.

The relationship between LSAT, IQ, and appointments is suggestive. IQ may cause differences in LSAT scores which cause differences in appointments. When we consider this hypothesis against other societal factors that influence appointment decisions, the IQ effect is probably going to be weak. Thus, I agree that some of the variation in in top law schools might be explained by IQ, but I agree with Ann's sentiment that this effect will be small relative to other factors.

I guess if I wanted to say IQ cannot have this effect then I would point to the fact that on average the top law schools now have an equal number of men and women. Hence, the distribution of IQ cannot explain differences in composition of the sexes in law school.

ben wallace said...

Daryl: Just to clarify, I think this IQ issue could make a good journal article, but I think the results would indicate minimal effects of IQ. I would never rule out the benefit of exploring this hypothesis a priori.

Daryl Herbert said...

I guess if I wanted to say IQ cannot have this effect then I would point to the fact that on average the top law schools now have an equal number of men and women.

Did you even read my post? Males outnumber females by about 5 to 4 at those elite institutions. And that's still not with an even playing field--men are losing at least 1-2 of their LSAT points to pay the penis tax, because if the top5 schools didn't do that, they would have even fewer women. Even with this affirmative action in women's favor, even with women dominating the grades curves in undergrad*, even with so many of the smartest men being drawn away to math/science/engineering majors--even in the face of all that, penises don't merely survive at the top5 law schools, they don't merely peek out here and there, they have a rigid, dominating presence. The social factors all weigh against men dominating the top law schools--yet men do. So quite obviously, you have to look beyond social factors.

* the highest-ranked law schools have LSAT scores that are on average, a few points higher than 2nd-tier schools. But the average GPAs are much, much higher, probably as a way to arrive at a higher number of females who will pass a facially sex-neutral admissions criteria.

very suggestive... IQ may cause differences in LSAT scores

Preparation and study is important, but your IQ for all practical purposes sets a score ceiling beyond which you will never achieve. The correlation between IQ and LSAT is very strong.

You are having a hard time with the link, so I will spell it out for you:

1 - A high IQ is practically necessary to achieve a high LSAT score (I doubt many people "luck out" and stumble into 170+).

2 - A high LSAT score is practically necessary to get into a top law school. There are a few exceptions (affirmative action), but I doubt many clerks are chosen from the group of people who got in because of skin color points.

3 - Being from a top law school is practically necessary to be considered for clerkship at the Supreme Court. Again, there are exceptions, but not many.

Without the high IQ, it's very difficult to enter consideration for the position in the first place. It's only high-IQ people competing with other high-IQ people for those jobs.

The idiots-and-geniuses theory goes straight to the question of "who has high IQ," in a far more direct manner than the social theories possibly could, because they completely ignore IQ.

The IQ issue is boring unless you describe how to separate the effects of IQ and other factors.

Then pretty much all social science must be "boring," because this is just about the clearest case you're going to find of one variable being separate from other variables, in terms of how you define it, how you measure it, and what its effect is.

ben wallace said...

Daryl: The possible causal relationship is fairly obvious. What would make this more interesting is a description of (1) a procedure for determining the magnitude of the effect of IQ on LSAT scores holding constant other factors and an explanation how to measure the other variables included in the model of LSAT score and (2) a procedure for estimating the maginitude of the impact of LSAT scores on clerkships, as well as a description of other variables and measures of the other variables that would be included in the model of clerkships (e.g. a probit model). Your argument about IQ is consistent with an outcome but there are ways to test the IQ hypothesis, and until you describe the statistical procedure to do this, the speculation about IQ, LSAT, and the sex gap in clerkships is not very interesting from a social science point of view.

Maria said...

Nice response! I think she is looking for Tuxedos it is?