Is John Tierney pulling a Larry Summers?

NY Times columnist John Tierney wrote a pair of columns on the much-discussed question of why women are underrepresented in math and the physical sciences.  I didn’t see these columns until someone pointed out this response to them on the PBS Inside NOVA blog:

Why aren’t there more women in the upper echelons of science? It’s a question with many answers, but John Tierney at the New York Times is only interested in one: Maybe women just aren’t smart enough.

This is such an inaccurate description of Tierney’s position that the authors would seem either not to have read his columns or to be deliberately misrepresenting them.  Tierney:

So why are women still such a minority in math-oriented sciences? The most balanced answer I've seen comes from two psychologists at Cornell, Stephen J. Ceci and Wendy M. Williams €” who, by the way, are married and have a daughter with a graduate degree in engineering. After reviewing hundreds of studies in their new book, "The Mathematics of Sex" (Oxford), they conclude that discrimination is no longer an important factor in keeping out women.

They find consistent evidence for biological differences in math aptitude, particularly in males' advantage in spatial ability and in their disproportionate presence at the extreme ends of the distribution curve on math tests (the topic of last week's column). But given all the progress made in math by girls, who now take more math and science classes than boys and get better grades, Dr. Ceci and Dr. Williams say that differences in aptitude are not the primary cause of the gender gap in academic science.

Instead, they point to different personal preferences and choices of men and women, including the much-analyzed difference in the reaction to parenthood. When researchers at Vanderbilt University tracked the aspirations and values of mathematically gifted people in their 20s and 30s, they found a gender gap that widened after children arrived, with fathers focusing more on personal careers and mothers focusing more on the community and the family.

Dr. Ceci and Dr. Williams urge universities to make it easier for a young scientist to start a family and still compete for tenure, but they don't expect such reforms to eliminate the gender gap in academic science. After all, the difficulty of balancing family and career is hardly unique to science, and academia already offers parents more flexible working arrangements than do other industries with smaller gender gaps.

The gap in science seems due mainly to another difference between the sexes: men are more interested in working with things, while women are more interested in working with people. There's ample evidence €” most recently in an analysis of surveys of more than 500,000 people €” that boys and men, on average, are more interested in inanimate objects and "inorganic" subjects like math and physics and engineering, while girls and women are more drawn to life sciences, social sciences and other "organic" careers that involve people and seem to have direct social usefulness.

Ceci and Williams (and hence Tierney) may be right or they may be wrong.  But to take this position and replace it with “women just aren’t smart enough” is shamefully dishonest.

For what it’s worth, my best guess is that the key factors explaining the gender gap are:

  1. Girls being discouraged from doing math and science in school, starting from a very young age, by teachers, parents, and peers.
  2. The family-unfriendly nature of the tenure-track job process.
  3. Discrimination (almost 100% unconscious but no less harmful as a result) against women scientists by their peers.

Tierney does give more credence than I do to the Larry Summers hypothesis, that the gender gap is partly explained by the fact that more men than women lie at the extreme high-end tail in the distribution of math ability (as well as the other tail).  The NOVA blog post does a good job at laying out some of the reasons why this seems like an unlikely explanation.  Among many other reasons:

  1. If this were the explanation, the effect would be uniform across cultures, but it isn’t.
  2. In my experience, lying in the extreme tail in the distribution of scores on math tests is not all that highly correlated with being a successful scientist.  On the contrary, many of those extreme outliers lack the other skills needed for success in science.

So to the extent that Tierney is advocating this as the explanation for the gender gap, I think he’s probably wrong.  But it’s not fair to replace his actual position with an inflammatory and inaccurate straw man.

Published by

Ted Bunn

I am chair of the physics department at the University of Richmond. In addition to teaching a variety of undergraduate physics courses, I work on a variety of research projects in cosmology, the study of the origin, structure, and evolution of the Universe. University of Richmond undergraduates are involved in all aspects of this research. If you want to know more about my research, ask me!

15 thoughts on “Is John Tierney pulling a Larry Summers?”

  1. The Summers position – greater spread on the male intelligence curve – is not wrong. You’re right that high intelligence does not correlate with the social requirements of thriving in academia or anywhere else, for that matter. However, I wonder if the brute realities of right-tail intelligence distribution don’t make a big difference in three-sigma plus positions, however.

    I grant little credence to your points 1 and 3, which seem like PC boilerplate, but you’re absolutely right on point 2: women have babies and nurse, and are thus ill-suited to the careerist grind of academic science.

  2. Regarding point 1: PC boilerplate is sometimes true. One thing I’m 100% sure of: my female students very often think that they can’t do math, all evidence that they can notwithstanding; my male students are just the reverse. Pretty much everyone I know who’s taught math or physics in the US reports the same thing. That fact needs an explanation. It could be nature, but nurture (or rather anti-nurture) sounds more plausible to me (especially since this effect does not appear to be uniform across cultures).

    On #3: Some experiments have been done in which scientists reviewed papers with authors with clearly female names, while others reviewed the same papers with just initials for the authors. The ratings were significantly different. Tierney cites similar studies that yield different results, though, so maybe the research is more ambiguous than I thought.

    Explicit gender discrimination is not socially acceptable anymore (fortunately!), but the time when it was is not that far behind us. Jenny Harrison, roughly 20 years ago, had to fight incredibly overt sexism to be granted tenure at Berkeley. 20 years seems like a long time, but some of her most vocal opponents are still active. I find the hypothesis that those attitudes are still present far more plausible than the hypothesis that they are not.

  3. Oh, and I’m not disputing the fact that men populate the tails of intelligence tests more than women. I think it’s unlikely to be a major explanation of the gender gap, though.

  4. Whew, where to start? First, check out the discussion today over at
    http://telescoper.wordpress.com/2010/06/21/sexism-of-course/ (Peter Coles’s blog).


    For what it's worth, my best guess is that the key factors explaining the gender gap are:

    I think that, in the last few years, it is a good thing that it has become possible to discuss this intelligently in some circles (automatically disqualifying Harvard as an intelligent circle, but there you go) without immediately being labeled a male chauvinist pig.


    1. Girls being discouraged from doing math and science in school, starting from a very young age, by teachers, parents, and peers.

    This might happen, often unconsciously and unintentionally, but I don’t think it is a big effect. There are parents who do everything they can to reverse the traditional gender roles, and the effect is minimal. Probably, parents and others tend to encourage people in things they show an aptitude for, so while we might have a vicious circle, thinking of it as positive feedback might be more helpful.


    2. The family-unfriendly nature of the tenure-track job process.

    Irrelevant. Yes, it is family-unfriendly for many people. But then again, the fact that universities do anything at all to address the needs of dual-career couples goes far beyond what happens outside the academic world. Also, there are differences in many places outside of academia (see below).


    3. Discrimination (almost 100% unconscious but no less harmful as a result) against women scientists by their peers.

    This certainly happened in the past (like when Lise Meitner was told “no, you can’t work here because we don’t have a women’s toilet”), but I think, today, it plays almost no role. If anything, some women benefit from positive discrimination (no, I’m not naming any names).


    Tierney does give more credence than I do to the Larry Summers hypothesis, that the gender gap is partly explained by the fact that more men than women lie at the extreme high-end tail in the distribution of math ability (as well as the other tail). The NOVA blog post does a good job at laying out some of the reasons why this seems like an unlikely explanation.

    I find the outrage over Summers to be the most puzzling thing. The gender gap and the fact that men are more often in the high-end tail (and in the low-end tail, but that doesn’t get as much press) and that this probably has a genetic origin is obvious to anyone who has taken an unprejudiced look at the data. It might not be the only effect, but it is a significant effect.


    1. If this were the explanation, the effect would be uniform across cultures, but it isn't.

    Non sequitur. Perhaps the effect is there, but perhaps there are other effects which compensate it. For example, there are many female astronomers in Italy (a country not especially known for women’s liberation). I was told by many astronomers working in Italy that one reason for this is that the jobs are badly paid. As such, a man can’t support a family as an astronomer, so there are less men then there otherwise would be, and it is a job taken up by women when they don’t have to support themselves and their family from it (because they have husbands with higher-paying jobs elsewhere). Many people say “women’s jobs are badly paid” but there is also the effect that women who take up a job because it’s something they want to do, and not for the pay, are attracted to badly paid jobs since there is less competition for those. I’ve read that there are many women in academia in Turkey, again not a country known for women’s lib. Maybe the situation there is similar.


    2. In my experience, lying in the extreme tail in the distribution of scores on math tests is not all that highly correlated with being a successful scientist. On the contrary, many of those extreme outliers lack the other skills needed for success in science.

    It’s not just math tests. Probably, most astronomers lie in the extreme tail of the “astronomer test”, it’s just that there is no standardised “astronomer test”.

    Such discussion tends to concentrate on (what are perceived to be) good jobs in which women are underrepresented. The additional assumption is that this must be due to discrimination. If we broaden our perspective, we see that women earn much more than men when working as models and in pornographic films. Is this because there is discrimination against men here? Yes, most chess grand masters, 3-star chefs and Nobel Prize winners are men, but so are most garbage men, most people who work in dangerous jobs and most mentally retarded people. This really looks like the result of a wider distribution (at both tails).

    What about rock music? Can you name one reasonably well-known female rock musician who is a) not a singer, b) not romantically involved with nor related to some man in the band and c) not a member of a group which is a “girl group” as a matter of principle? I can think of one, and even that is borderline. (I’ll mention her name in a later comment.) I’m sure that such musicians exist, but they are not reasonably known rock musicians.

    In other types of music, there are as many or more men than women. However, most composers are men.

  5. Just to be clear, the comments about rock music are just to point out that the gender gap is by far from being exclusive to academia, and is more pronounced in some other areas (and less pronounced in still others), so focusing on academia-specific putative causes (such as the long path to a permanent job) is probably not a good place to start (which doesn’t mean that they are completely non-existent, of course).

  6. “Perhaps the effect is there, but perhaps there are other effects which compensate it.”

    Oh, Phillip, you know better than that! If the claim is that A is an explanation for B, and in order to compensate for the fact that that claim doesn’t fit the data you have to invoke other unspecified mechanisms, then A has little or no explanatory power. A may or may not be true, but it’s not the explanation.

    (In this case, I concede that A probably is true — men disproportionately populate the tails of the distributions of certain tests. I just don’t believe it’s an explanation.)

    Despite some (modest) efforts to accommodate dual-career couples, I don’t think that the family-unfriendliness of the academic job market (at least in the US) can be seriously disputed. What other career requires the combination of low income, frequent moves, and enormous uncertainty about future career prospects, timed perfectly to go along with the years when people typically want to start a family?

    Just to be explicit, this, as an explanation of the gender gap, requires one more ingredient: that women are more apt than men to be unwilling to accept this sort of career-family tradeoff. That of course also requires an explanation, so one can ask the usual nature-nurture question again here. Innate (biological) sex differences are far more likely to play a role here.

    Not sure what point you’re making about other career options. Differences between men and women in the pornography industry have an obvious market explanation. Are you seriously saying that garbage collectors tend to be male, because men are more likely to be stupid? I doubt very much indeed that the preponderance of male garbage collectors is because men overpopulate the far tails of IQ scores.

    I don’t know enough about the culture of haute cuisine to speculate on why top chefs tend to be male, but I don’t see any evidence for the hypothesis that it’s because of sex differences in the distribution of … what exactly? Similarly (but even more so) for rock musicians. There’s no doubt that the culture treats men very differently from women in a bunch of ways that are relevant to success in rock music. That seems to me much more fruitful ground to search for explanations than hypothetical innate cognitive differences, which are bound to be small.

    Of all your examples, chess players seem to me the most likely to involve innate biological cognitive differences.

  7. One thing Phillip and I surely do agree on: I don’t believe we should demonize people who believe in the Larry Summers hypothesis, and I don’t think we should rule it out of polite discussion. It’s perfectly reasonable and indeed quite interesting to study innate sex differences and to ask whether they plausibly explain various observed phenomena. I just think that this particular explanation is unlikely to be correct, for the reasons I’ve tried to outline.

    In fact, the thing that prompted me to write this post was annoyance over what the NOVA people wrote: they seem to have skimmed a few lines of Tierney’s piece and lumped him into the Larry Summers category. As a result, they apparently felt no obligation to actually read or engage with what he wrote.

  8. Oh, Phillip, you know better than that! If the claim is that A is an explanation for B, and in order to compensate for the fact that that claim doesn't fit the data you have to invoke other unspecified mechanisms, then A has little or no explanatory power. A may or may not be true, but it's not the explanation.

    Note that I gave a concrete example, the fact that astronomy jobs in Italy are badly paid. If men and women working in astronomy in Italy tell me that this is why they think there are more women astronomers in Italy, then I shouldn’t ignore that datum just because I have another prejudice (which is not necessarily right or wrong, per se). This was a direct reply to your “if that ‘s the case, it must be the same in all cultures”, but I still think that is a non-sequitur.

    What other career requires the combination of low income, frequent moves, and enormous uncertainty about future career prospects, timed perfectly to go along with the years when people typically want to start a family?

    It would apply essentially to anyone who is a) unemployed or b) employed by a temporary work agency or c) working in any number of jobs (lorry driver, for example) where companies can go bust and he might have to look elsewhere, even if he is not on a fixed-term contract. The number of such people is huge compared to the number of people in academia. And there aren’t that many female lorry drivers, though physical strength, in these days of power steering and power brakes, is hardly needed. My point is that the gender gap exists in many places outside academia. While there might be academia-specific causes as well, I think it more likely that, at least in part, there is a common cause for many different walks of life.

    Just to be explicit, this, as an explanation of the gender gap, requires one more ingredient: that women are more apt than men to be unwilling to accept this sort of career-family tradeoff. That of course also requires an explanation, so one can ask the usual nature-nurture question again here. Innate (biological) sex differences are far more likely to play a role here.

    Of course. I think we all agree here. While there are many factors, even if there were just one—women need to have children when they are young, and men don’t—it would seriously skew the distribution.

    Not sure what point you're making about other career options. Differences between men and women in the pornography industry have an obvious market explanation. Are you seriously saying that garbage collectors tend to be male, because men are more likely to be stupid? I doubt very much indeed that the preponderance of male garbage collectors is because men overpopulate the far tails of IQ scores..

    My point here is only the following: the gender gap exists in many areas besides academia, and it is not always the case that, as in academia, women are underrepresented and/or paid less. Yes, there might be individual explanations for some areas, but I don’t think this rules out a cause which applies to more than one area, including academia. (With regard to porn, the obvious market explanation begs the question. Following up the garbage-collector example, I haven’t heard anyone say that there are so few female garbage collectors because of discrimination, but if we’re talking Harvard professors then obviously discrimination is the cause. In other words, the discrimination card is played when women are perceived to be disadvantaged, but not otherwise. This doesn’t mean that discrimination doesn’t exist, but a more balanced and objective assessment than that which one usually finds in the popular media is called for.)

    One thing Phillip and I surely do agree on: I don't believe we should demonize people who believe in the Larry Summers hypothesis, and I don't think we should rule it out of polite discussion.

    I agree that we agree here. Not only is it going too far to lump more reflected discussions into the Larry Summers camp, it is also going too far to lump Larry Summers into the male-chauvinist-pig camp.

    Getting back to the subject of a recent thread, Steven Pinker’s books How the Mind Works and The Blank Slate contain extensive and what I think are good and balanced discussions of the gender gap.

  9. I’m really not getting your point. Are there more male lorry drivers [truck drivers over here in the US] because men populate the tails of lorry-driver-aptitude distributions more than women? You’re the one advocating a Grand Unified Theory with a single explanation for all gender differences, so if that’s not what you mean, I can’t work out what you do mean.

    Personally, I don’t see a unified theory as likely in this case. Human society is complicated, and sex differences are woven through it in a bewildering variety of ways. I think that the many different gender gaps probably have many different explanations. For the specific case of academic science, I’ve tried to indicate why I think that certain explanations are more likely than others.

    The Larry Summers hypothesis is a lousy fit to the observations, for a number of reasons. The main one is that presence in the tail of a distribution of intellectual ability is, in my experience, a weak predictor of academic success. Scores on intelligence tests (of any sort) are surely correlated with academic success, but with an incredible amount of scatter. If we were talking about a small gender difference, then this could be a plausible explanation, but I don’t think that such a weak correlation can do the work of explaining a large effect.

  10. “I'm really not getting your point. Are there more male lorry drivers [truck drivers over here in the US] because men populate the tails of lorry-driver-aptitude distributions more than women? You're the one advocating a Grand Unified Theory with a single explanation for all gender differences, so if that's not what you mean, I can't work out what you do mean.”

    There are two aspects: a) is the gender gap caused by longer tails for males (you know what I mean) and b) is the gender gap academia specific. The lorry-driver example addresses the second aspect. The chef-de-cuisine example addresses the second, but perhaps also the first.

    When discussing the gender gap in academia, and pointing out that soft-money positions are not good for families and might prevent some women from remaining in the field, the discussion seems biased against a longer-tails explanation. I think one needs to take a broader perspective and a) look at the gender gap elsewhere and b) weigh the relative merits of longer tails and other explanations in all areas.

    I think the rock-musician example is a good one. Yes, female groupies are probably more welcome to the male rockstar than vice versa, and famous rock stars are rich so there is the whole issue of money making women more attractive but not men, but the example is interesting because all famous rock stars used to be not famous, and we don’t find more women in the less famous ranks (with the exceptions I mentioned: singers, husbands/boyfriends or relatives of boys in the band and members of girls groups). A low-level rock musician doesn’t earn much and probably isn’t in it for the groupies (though they do enter at a lower level than most people expect). There is no reason physical strength etc should play a role, and in other branches of music, women are at least as numerous as men, in many even more so. Yet the gender gap is much greater than among astronomy professors. I can think of several well known female astronomy professors. Even though there are more rock musicians than astronomy professors, and more famous rock musicians than famous astronomers, I can think of only the one example (which is not completely straightforward, as I mentioned) of a famous female rock musician. I don’t think there is huge discrimination against female rock musicians. So what is the explanation.

    Can anyone provide an example of a famous female rock musician (or even a female musician in a well known rock group, even if she herself is not famous)? No point in searching the web: if you have to search, then the “well known” criterion is not fulfilled.

  11. Yes you are right. Being that there are so many more famous female rock singers than astronomy professors they should come rolling off the tongue but they don’t.

Comments are closed.