Wednesday, January 19, 2005

Why There are So Few Women in the Hard Sciences: Part III

In this post the story of disecting the provoking thoughts of Lawrence Summers continues. As you may remember, he offered three hypotheses for the reasons why there are few women in hard sciences:

1. The eighty-hour workweek expected
2. That women might be biologically less able to do hard sciences than men
3. Discrimination against women.

I have discussed the second one quite extensively in my previous post in this series, and here I wish to tackle the first and second suggestions of our Lawrence.

First, why would the eighty-hour workweek in hard sciences keep women back? What is the average workweek in the soft sciences or in the humanities, where there are more women, relatively speaking (though not by the amounts that the so-called biological studies of female superiority in writing skills would suggest...)? As far as I can tell, the average expected workweek is pretty long for all Ivy League professors and scholars, but this appears to have less effect on women outside the hard sciences.

Second, why would women react differently to a long workweek than men? This is the hidden agenda in this suggestion: it has to do with who is going to take care of the children. So it's very simple, really: according to this hypothesis, women are not in the hard sciences because they need to have more time for family life. Men appear not to want to have a family life to the same extent. Why would this be the case? Here, once again, opinions differ. Some argue that women have a biological imperative to spend more time chauffeuring their children to hockey meets than men do. Others argue that the upbringing we are all subjected to convinces both girls and boys of the necessity for such tasks to belong to women. Or perhaps both of these reasons apply at the same time.

Third, given this sexual division of labor in many families, why are eighty-hour workweeks so necessary? Here the answer appears to be that the quality of research would diminish if the scientists could go home, once in a while, or even take the white coat off for a day every week. I disagree with this. There is no evidence proving that this long workweeks would be good for creativity and productivity in general. In fact, there is a pretty good counterargument, based on the likelihood that we get less productive the longer the time we have already worked. Besides, eighty-hour weeks translate to zero hours with children except during the weekends, and I don't care what sex you are; if you are a parent this is crap.

Talking about the "eighty-hour week" is a shorthand way of pointing out that so many things in the academia assume that the scholars have a well-equipped home base to which they return only to sleep. The scholars are certainly not expected to give birth to children, for example, though change in this has taken place in the last few decades. But hard sciences may not look like very hospitable places to a lot of women for reasons of this sort.

Then consider being the only woman in a department, or one of very few. Consider a lab where people talk about what internet pornsites they visit and so on. And consider that some of your colleagues might very well be thinking all the time that women are by nature less able to do science. It takes a very specific type of person to wade through all this towards the great enjoyment that hard sciences can offer.

Things are not this bad or bad at all in every department, but these departments-from-hell still exist. Most of us don't want to be the symbolic pathbreakers who get beaten to a pulp in the process, and it's not difficult to see why hard sciences might not look that attractive to a woman perfectly capable of practising them.

And this brings me to Dr. Summers' third hypothesis, the one about discrimination against women. Here is the place where Dr. Summers reveals his true beliefs:

Summers' third point was about discrimination. Referencing a well-known concept in economics, he said that if discrimination was the main factor limiting the advancement of women in science and engineering, then a school that does not discriminate would gain an advantage by hiring away the top women who were discriminated against elsewhere.

Because that doesn't seem to be a widespread phenomenon, Summers said, ''the real issue is the overall size of the pool, and it's less clear how much the size of the pool was held down by discrimination."

This little conclusion is from a very early book on discrimination by Gary Becker. If you think about what is being said here for a while, you realize that if Dr. Summers is correct, then there has never been any discrimination at all! All those blacks and women who sued Wal-Mart for discrimination were wrong!

Sadly, this is not the correct conclusion. Becker derived this result in a simple model of the world where he assumed that only employers wanted to discriminate. Both coworkers (colleagues) and consumers were assumed to be totally fair and unbiased. But in reality we cannot assume that neither coworkers nor consumers would be totally fair and unbiased. In fact, when Becker introduced the possibility that discrimination might not be something that only bosses do his conclusion no longer applied.

Also, the conclusion is derived from a model which assumes that there is no ignorance about workers' true abilities, that everybody knows everything relevant. Once we allow for the actual amount of uncertainty and ignorance in the real world, the result would fail to follow even if only bosses were bigots. In fact, some alternative models show just the opposite in this case: that even actually unbiased bosses might discriminate if they expect that this strategy is, on average, rewarded by the overall market.

That Summers served one of the earliest and very partial conclusion from the whole rich field of the economics of discrimination shows what he really thinks. For this is the one piece that is always trotted out by those who never believed in any of that discrimination bullshit in the first place. He could have at least mentioned the many alternative theories of discrimination and the fact that they lead to very different conclusions. Or he could have mentioned the many empirical studies that show gender discrimination exists.

Instead of giving you a long lecture on these theories or evidence, I'm going to assume that you trust me enough to believe that they exist. Thus, Summers' argument fails to say anything very useful about the possible extent of discrimination against women in the hard sciences, even if we define discrimination as only taking place on the level of the relevant university. Then there is the whole issue of steering by teachers and career counselors on school level and the general societal expectations... Really, there is a lot of material here, and Summers' statement is a little bit inane.

Which brings me to my conclusions. Whether women and men have the same average ability to do hard sciences or not, shouldn't we clear out all the other obstacles women in the hard sciences meet before we decide to throw up our hands and agree that it's-just-the-way-things-are? Especially as it might not be the way things are, after all.
If you are interested in real-world descriptions of sex discrimination, you might want to read through this thread in the Democratic Underground.