Please visit the new home of Majikthise at

« "Out" heterosexual | Main | Why whippets never win Best in Show »

February 17, 2005

What Summers said

Looks like Summers caved and released a transcript of his remarks to the NBER Conference on Diversifying the Science & Engineering Workforce.

In other Summers news the Harvard Crimson reports that Summers Faces Crisis of Confidence from the Harvard faculty. Apparently, "I'm gonna provoke you" is not turning out to be a wildly popular leadership style.

Here's some vintage Larry Summers provocation, a 1991 memo in which he claims that less developed countries are underpolluted and that economists should take a leadership role in filling the toxic waste gap.

For a decade, Lawrence H. Summers has been dogged by a memo, bearing his name, purporting to advocate exporting polluting industries to poor nations ("underpopulated countries in Africa are vastly underpolluted") and dumping toxic wastes there ("the economic impeccable"). The memo, dated December 12, 1991, during his service as vice president and chief economist of the World Bank, found its way to the press, and has since circulated widely on the Internet. It was the subject of a question at his March 11 news conference at Loeb House, and of student protests then and the next day. Summers responded to the reporter, "I think the best that can be said is to quote La Guardia and say, "When I make a mistake, it's a whopper.


TrackBack URL for this entry:

Listed below are links to weblogs that reference What Summers said :


“[We must] show the public that we are not cowards, we are not spineless, and we are not with you,” said Arthur Kleinman, chair of the anthropology department, addressing Summers in the early minutes of the meeting.

He seems to be trying to ride it out. Apologize, release the transcript (which really does him no credit), and lobby behind the scenes. The question is how many people want to be associated with him in people's minds.

"Apparently, "I'm gonna provoke you" is not turning out to be a wildly popular leadership style."

Should it be? That's seems like a serious question. Should people in leadership roles be allowed to say wildly unpopular things? Should they be allowed to offend their own workforce?

Economic logic: I've always thought his Africa memo was an example of perfect logic given questionable premises. It's like the sylogism "All black people are lazy, Ezra is a black man, therefore Ezra is lazy." The logic is impecabble, but the opening premise is questionable. Likewise, if you accept the premise that the market can never make a mistake and that all things should be comodified and sold, then it follows that African countries have under-pollution to sell to advanced countries. As Summers says, "the economic logic... is impecable."

But isn't the fact that he is prepared to take his thinking to a point where it outrages people give what he says more credibility when it supports liberal/social democratic causes. When Summers said that primary education for girls was the development investment with the highest benefit/cost ratio, it had greater meaning because he was the kind of person who might have picked building hazardous waste sites instead if that was what his investigation led to.

Personally, I think Summers is wasted as a college president. I just hope that we don't see some feminist-anti-Clinton wingnut lineup to keep him from heading up the World Bank or IMF.

Gareth, by that logic, the most effective critiques of racism and Jim Crow came from Stalin since he was prepared to massacre entire minority populations like Volga Germans and the Don Cossaks.

On another note, is Summers' intro to the three hypotheses with its reference to the dearth of Catholics is investment banking, etc. meant to suggest that there's some gene for usury?!?!?


Your analogy would be valid if I was saying that Summers is a bigot and therefore when he says something progressive it has more force. But Summers isn't a bigot, and the transcript makes that obvious. He's just someone who is willing to say certain empirical relationships hold which do in fact hold.

The value of a statistical life, defined on the basis of conventional economics, is much less in Africa than in Europe. There is a difference in variance of mathematical aptitude between men and women. And there are not a lot of Jewish farmers in the United States.

The first point is true, but reflects a situation that is clearly unfair. But Summers has tried to do more about it than many people in public life.

The second point is also true, but Summers is probably right that it is less important than the disproportionate share of domestic and childrearing work women take on. For this, he is David Duke?

I'm sorry, but that transcript consisted of reasonable points, put forward tentatively for discussion. I thought that the idea that feminism depends on an exact identity between male and female minds was long dead, but I guess not.

That World Bank memo is a perfect example of why economists have no special claim to deference in public-policy debates.

Thanks for the link.


First, my apologies for the snarky comment; I did regret it after the pushin the save button. And I don't have a general problem with the research program of evolutionary psychology, though the tendency of its proponents to find socio-biological answers to everything and offer a bunch of just-so stories is a bit irritating--think of Steven Pinker's suggestion that suburban lawns somehow are residuals, genetic memory from our optimization to the savanna's of Africa.

And I don't think Summers is a bigot, but I do think that he has had quite a justified reputation for putting his foot in his mouth and discounting and attacking ad hominem arguments that don't fit his calcified framework on some issues, even in economics. In addition to the Catholics, whites and Jews comment (also note that the first isn't tied to any sort of phenotype and is a belief category), the reference to Gary Becker's piece on discrimination I think is telling. Becker's piece is insightful, but as a sort of benchmark. The question of why is the racial premium exploited by non-discriminating firms exploit to reap larger returns in a competitive sector is tossed out to suggest what? that discriminatory premises are true? the marginal productive of black labor is smaller? the marginal return of female scientific inquiry is smaller?

And then there's the function of the reference to differential abilities. Is it well, here's an interesting scientific question? (If that's the case, why not also point out differences in the variance of test scores in the sciences between boys and girls, with boys having a much, much larger variance) Is it to suggest, well, here's a biological source and given that we've decided for a host of reasons to seek gender diversity, we should have policies aimed at increasing diversity take these factors into account? Is it to suggest that the goal of diversity should be revised in the face of differential ability, and Harvard to place less stress on greater representation of women in the sciences, which would also make Summers job easier? The first two are quite reasonable, assuming that we have a full grasp on the complete sources of differential performance and the implications thereof; but if he's pushing for the third, then he should state so. The Catholics in banking, Jews in farming, and whites in basketball intro doesn't really suggest that he's taking it as a problem.

I admit I haven't actually followed what Summers said very closely, so let me preventively declare ignorance as a defense for the moment.

My concern is that as an educator myself (both corporate and collegiate), I often deliberately "provoke" my students by challening their beliefs. Granted, I teach rather benign telecom and infosec courses, but still, there's dogma in every human endeavor and I generally pose questions in class rather than provide answers. Those questions are often deliberately staked on an apparent position that mainstream thought is wrong, and I force my students to defend what they take for granted.

So if Summers had been taking a similar approach in this instance, should he really be excoriated?

Is this another example of "it ain't the crime, it's the coverup?" Do I need to actually expend mental energy to go through the transcript to pinpoint his sin?

When I hear the words "the economic impeccable," I reach for my Excedrin.


Summers is not a teacher. He's an administrator. And he's not presenting a research paper in psychology, or even engaging in good-faith debate by playing Devils Advocate.

The reason he brought all of this stuff up is that the number of women hired and and offered tenure has dropped precipitously since Summers took over at Harvard, and he's trying to let himself off the hook for that. And the reason he will almost certainly face a vote of non-confidence from the Harvard faculty on Tuesday is not merely because of this one incident -- it's a pattern of behavior that started long before he came to Harvard. This was just the last straw.

And yes, you should read the transcript. This isn't a "witch-hunt" -- Summers hanged himself.

think of Steven Pinker's suggestion that suburban lawns somehow are residuals, genetic memory from our optimization to the savanna's of Africa.

I don't recall Pinker ever mentioning lawns, but there is extremely good empirical evidence that human beings all over the world prefer two kinds of vistas above all others (in photographs, paintings, and actual real estate): the landscape they grew up with, and the African savannah.

Thad - fair enough--I was just doing "provoking" in the thread! Thanks for disabusing me of any notion of a witch hunt, or that the cover up was the issue, rather than the crime. I'll read the transcript after I've had some dinner (or maybe later--I'm enjoying a nice Shiraz right now).

there's this kind of pained, "i hate to do it, believe me, this hurts me more than it does you, but you know, my dedication to liberal ideals of free inquiry forces me" ethos behind the defense of those who castigate critics of summers. this is nothing but hoo-ha.

how do we know? well, obviously, there is more than simply "provoking" going on, or else we'd see more provoking along the lines of, "maybe blue eyed white men really are devils?" and subsequent defenses from eminently thoughtful liberals who dare to challenge the status quo by demanding to know, "aren't we even allowed to ask hard questions any more? isn't it the height of illiberalism to shut down questioning before we even begin?"

but of course we don't and we won't. as much as the empiricists hate to admit it, the post-modernists were on to something - the discursive totality of science, despite the claims of objectivity that ground its method, has been shaped by deep biases that persistently channel the focus of questions in one direction or another. the summers affair has helped highlight this truth better than any other science/pop culture event in recent memory. for that alone, i'm glad he said what he did. let him now reap his just rewards


I appreciate the reasonableness of your response.

The first point, which I think Summers was precisely right on, is that the main issue of gender equality in modern North America is the second shift. Part of the answer to this is changes in male behaviour at home. But part of the responsibility lies on institutions. Summers, being an economist, points to costs of accommodating career interruption and more reasonable work weeks, but this is the debate we need to have.

The second point, on innate abilities and interests, leads to interesting normative questions if you entertain it. Merit is defined, at least in part, by incumbents: part of the value of diversity is that it redefines what a good academic or lawyer, for example, is.

However, I wonder if it is really normatively important that each discipline and sub-discipline have 50/50 representation.

I also wondered about the Becker point. For one thing, oddly for an economist, he presumes that universities have the same incentives in hiring as firms in a competitive market. That doesn't seem like a reasonable assumption to me.

But I did think that Summers provided an honest contribution to the conversation, and I think the backlash is unfortunate.

(What you think of Pinker depends on your toleration for non-peer reviewed speculation in popular science. Mine is pretty high, as long as it is disclosed as such.)

FWIW, I do think the second shift is an issue, but not necessarily for the reasons stated by Summers,(just about as pejoratively as possible) that women aren't "willing" work long hours. Many women start out being willing to work long hours until they conclude what I did in my own profession: it won't make any difference. Sure, you have to put in enough time to be credible but when you look around you see that people are often rewarded for plus factors that are only marginally related to what we normally think of as merit. Some of these plus factors really are plus factors -- the ability to achieve consensus -- and others are not -- one's social similarity to one's boss.

Other issues: women may want to make more of a "difference" in their career choice. Hence my scary bright Harvard math major former co-worker chose to be a lawyer, and she now crunches numbers in high profile commodities cases. Believe me, she works 80 hour weeks.

And again: women who do go into hard sciences are excruciatingly isolated. Their colleagues are generally older married men, and social opportunities can be very limited.

And another one: Hard sciences aren't exactly booming, federal funding has been cut, universities are increasingly parsimonious about tenure track positions -- this exacerbates the phenomenon of underrepresentation, although it is not responsible for it.

I also think it's rich to convey that college level professors are working 80 hours a week. Not where I went to college, I guaranteed you that (yes, it was a major university with graduate level departments in every major field you can think of). Or that all scientists are two or three or even four or five standard deviations above the norm. Oh sure, those people exist and I can accept that there may be some innate differences that tend to make males predominate among them, but there are a lot of universities out there and they aren't all staffed by baby-Einsteins. Far from it. Women are still underrepresented.

Finally, my biggest beef with Summers is his apparent outright blindness to the subtle and not so subtle socialization of girls out of science. When I took advanced algebra I usually did the extra credit proofs. One day the teacher asked if anybody had attempted it and I said yes. He spent the next 30 minutes deliberately misrepresenting my work in an effort to humiliate me in front of the class. Stupid chick. You're not supposed to be good at math. No, I realize that probably won't happen today in quite such an overt way, and even then, if I had complained I probably would have gotten a very popular teacher reprimanded. But there are many women my age who might have made a career in science who thought better about it.

First, I did not mean to disparage Pinker generally--I'm a fan. The lawns comment was in what I think is his weakest book, How the Mind Works, and used to suggest that these explanations seem to attract these kinds of quick and easy just so appeals, even among people who are generally careful. And while the preference for the savanna is not surprising at all (and is referenced), the tendency to use these observations to blithely explain, e.g., home landscape preferences is endemic. Remember the rape as an ESS thesis--which had such little work on things like female counterstrategies, such as infanticide of offspring of rapes. These explanations need a lot more than a mix of survey data and a demonstration of subgame perfect equilibria--the latter aren't that hard to come up with for any direction. With that level of evidentiary minima, why not a preference for hot tubs as genetic memory of the cambrian oceans?

Barbara's comment does raise something interesting on the socialization side. Every discounting of socialization as an explantion, at least relative to the role it played in past attempts to explain the phenomenon, offers up as evidence the relative ineffectiveness of alternative parenting with no measure of the role of parents vs. the rest of world in socialization.

But to get to the issue: as near as I understand it, the puzzle, so to speak, is not the underrepresentation of women or why it isn't 50/50 yet. There have been improvements and continuing improvements as indicated by different measurements--enrollment in each discipline, labor force composition by job category, etc. And the momemtum is not declining so that we can expect continued increases in women's representation in sphere's they've been traditionally excluded from. The Summers' point I think is something else: why has the rate of increase of women in the faculty of the human sciences and arts been larger than the rate of increase in math and the natural sciences. Second shift arguments won't work since second shift and child rearing constraints are a constant across disciplines. (Or maybe not. Maybe women who go into the humanities are less likely to have children than one who go into the sciences.) In either case, the way the issue is structured, appeals to second shift and other institutional factors won't work since it is "why this differential representation across the disciplines"? They're offered as arguments to be quickly rejected by all else being equal considerations, and the structure of the question leads almost immediately to innate abilities. (Socialization, the other possible answer is very quickly dismissed with the mama truck, baby truck example . . . although I always thought that there are a lot of interesting scientific issues involved in biological relations.) And that's why I question the function of Summers' hypotheses.

I think this is an interesting, and useful, discussion. Summers should be rationally criticized, not ritually denounced.


Summers does consider the possibility that the difference in male/female ratios in physics compared to medicine (for example) is due to differences in innate interest rather than (or as well as) ability.

I'm sure discrimination is more rampant in male-dominated professions and industries than in ones with more equal representation. But this does not (by itself) explain why some professions have lower ratios of XXs. Fifty years ago, discrimination against women existed about equally in medicine, law and physics, and was greater than it is in any of those areas now. But women's representation in medicine and law has risen much faster than in physics.

There may well be a self-reinforcing effect: if there are fewer women in an area, other women will tend to be deterred both because of overt sexism and cultural differences. But something else must be driving the difference.


Summers rated the second shift as a bigger reason for disparity in the physical sciences than innate ability/interest. I take your point to be that this cannot be right, since the second shift is equally (if not more) an issue in medicine and law. I think that's right; however, the second shift helps explain why there are fewer female senior law partners than their proportion of the law school graduating classes of twenty years ago would warrant.

It's true that socialization means more than parents. But I just read Summers to be saying that we can't exclude innate explanations, not that they are exclusive.

Finding ways to encourage girls to be more numerate is an important social objective (as is finding ways to deal with a lot of primarily male learning disabilities). I'm not sure that a 50/50 representation in the theoretical physics department of Harvard is.

The comments to this entry are closed.