News your connection to The Boston Globe

Grading Larry Summers


Here are some comments on your seminar presentation dealing with the underrepresentation of women in math and the sciences. Your grade, I regret, is a C-plus.

As you know, this is a low grade at Harvard. I do hope that when you develop this into your term paper and as you mature as a scholar, you will take these comments to heart:

Pages 1 and 5 of the transcript: You offer three possible explanations for the underrepresentation of women scientists: Women have a harder time succeeding in ''high-powered jobs" because of the demands of intense work schedules and their own preferences. Second, women have lesser innate aptitudes. Third, women suffer the effects of discrimination and prior role expectations.

What's missing is an empirical discussion of the immense gains women have in fact made in the past three decades as barriers have fallen and expectations changed. Women PhDs went from 0.6 percent to 17.3 percent in engineering, 2.9 to 15.5 percent in physics, 2.3 to 22.8 percent in computer science, 7.6 to 29.0 percent in math, and so on.

This trend suggests that institutional factors are far more an influence than innate ability. If there are serious innate differences, how do you explain this remarkable shift?

Page 1. In pointing to other cases of underrepresentation, you say, ''The data will, I am confident, reveal that Catholics are substantially underrepresented in investment banking." Larry, there's no place in academia for this kind of casual conjecture. Imagine submitting that for publication! Back up your assertions with evidence. You should have learned this as a freshman.

You also note the lack of white men in the NBA. Do you really mean this as a serious analogy to underrepresentation of women in the sciences?

On page 4, in dismissing discrimination as a principal factor, you approvingly cite professor Gary Becker, who asserted that if discrimination were prevalent, institutions that discriminated would suffer competitive disadvantages and there would be ''very substantial opportunities for a limited number of people who do not discriminate."

This is a widely discredited argument that I'm surprised you would make. As many critics of Becker have noted, whole classes of people benefit from the privileges that discrimination confers (i.e., whites, males). This reflects structural power. There was a time when there were no white women professors or university students at all, hence no opportunity for institutions to gain by nondiscrimination. When discrimination is pervasive, Becker's hypothesis doesn't apply. Historically, when barriers have fallen, it has been due to political and institutional struggle and reform, not because nondiscriminating institutions edged out discriminating ones.

Your presentation, overall, is heavy on pop sociology and light on data. On page 4, in suggesting that girls may be hard-wired to put child-rearing ahead of career goals, you quote your own daughters playing with trucks and saying, ''The daddy truck is carrying the baby truck." Larry, the word anecdote is not the singular form of the word data. Or as our colleague Daniel Bell memorably put it, ''For example is not an argument."

I do credit you with observing that the social institutions we have inherited and their outcomes are not necessarily efficient or just. I commend your suggestion of further study of the impact of child care, familial arrangements, tenure rules, and career interruptions on women scholars. This brief section saves the presentation from being a D. Please expand it.

But as an aspiring economist, you need more rigor in your methodology. It doesn't do to claim you are just being ''provocative." We economists have extended our influence into other scholarly disciplines, including law and political science. If we are going to practice sociology as well, we need to rely on logic and evidence; otherwise this is the sort of thing that gives economists a bad name.

Larry, you are a student of great promise and the best mathematician in the class. Your problem is certainly not lack of intellect. But you need to prepare for class, document your assertions, and not just rely on your brilliance to spout glib generalizations. This is especially problematic when you get into qualitative and institutional territory where you're not expert.

I'm concerned that you are displaying a self-defeating combination of overachiever and underachiever. With more self-awareness, empirical curiosity, and a little humility, you could go far. You could become a tenured professor at a young age, and, if this doesn't sound too improbable, you might even become president of Harvard someday.

Larry, please see me after class.

Robert Kuttner is co-editor of The American Prospect. His column appears regularly in the Globe.

Today (free)
Yesterday (free)
Past 30 days
Last 12 months
 Advanced search / Historic Archives