Summers and the Economic Way of Talking
Editors don't always inflict painful cuts on my articles. Sometimes I inflict them myself. My latest NYT column, arguing that the reaction to Larry Summers's remarks about women in science is an attack on economists' approach to social problems, is a great example. By the time I explained the controversy and quoted Summers there simply wasn't much room for some of the most interesting reflections (mine and others') on how economists' habits of speech and mind got him into trouble. The article concentrates mostly on what happens when you dispassionately examine hypotheses about human life and human behavior: People with an emotional stake and without the disciplinary habits of separating "is" from "ought" get pissed. But there's more to the story.
Take Summers's use of the word "marginal," a concept so central to modern economics that economists can hardly think without it. Even as a journalist who tries hard to avoid jargon, I know from personal experience that if you slip and say "marginal" rather than "additional" or "incremental"--the economic meaning--people will think you mean "unimportant," "wasteful," "worthless," or just plain bad. If misunderstandings can happen in a speech on the economic importance of aesthetics as the absolutely critical "marginal value" that determines whether a good or service succeeds, imagine what happens when you're talking about affirmative action.
Then there's the concept of "path dependence." Biology and discrimination aren't the only possible explanations for why a group would be underrepresented in a particular profession. History plays a major role. Especially to a liberal economist like Summers, whose scholarly work looked for examples of market failure, history can lead to bad results that market interractions won't correct. So when he looks at workplaces that reward employees who devote nearly every waking hour to thinking about their work, he sees a bad equilibrium that unfairly penalizes women--a system stuck in a dysfunctional groove because it developed under different historical conditions.
I, on the other hand, think the system is not only efficient but pretty darned fair, regardless of how it may differentially affect the sexes. "Time on task" makes a huge difference in productivity and problem solving. People who spend all their time thinking about work are more likely to be better at what they do than people who have other interests and priorities, just as people who actually spend time with their kids are probably better parents than parents who are never around. The general culture stigmatizes that single-minded dedication to work as weird or morally suspect. The workplace rewards the added productivity, and also pays a "compensating differential" (as the economists say) to make up for the sacrifices elsewhere. That seems reasonable to me. A research university is an odd institution to join the campaign for "balanced lives."
Finally, I find it quite possible that more men than women are really, really great at math and science, just as I find it not only possible but likely that on average women have poorer spatial abilities than men. (That goes quadruple for me, which only proves the importance of motivation and social assumptions. If I can drive a car, a lot more kids can learn calculus.) I doubt, however, that "fat tails" make a huge difference in the run-of-the-mill Ivy League job. In a country of nearly 300 million (not to mention the rest of the world) even a skinny tail includes a lot of women. My guess is that, aside from family issues, temperament and interests matter as much as raw ability.