Affirmative action for males? (part 1)

Over the last few months, there have been plenty of stories in the United States press about affirmative action for males — colleges giving preferential admission to males so as to avoid a severely gender-imbalanced student population (currently, the mix is 3:2 for girls to boys — administrators fear it could go as far as 2:1 without their affirmative action policies).

Some links:

• A New York Times piece (March 23, 2006) titled “To All The Girls I Have Rejected” by the Kenyon College Dean of Admissions.

• Inside Higher Ed (November 2, 2009) covers a probe by the U.S. Government into charges of affirmative action for males.

• National Public Radio (November 11, 2009) has a discussion on whether colleges favor male applicants (Transcript included).

• A mention of the NPR story and other observations (November 11, 2009) on Reason’s Hit & Run blog by Ronald Bailey.

• Inside Higher Ed (December 22, 2009) has a discussion titled “Soon-to-be open secret” about the issue of preferential admissions for men and the dominance of women in higher education.

• Minding the Campus (December 23, 2009) publishes a response by Mark Bauerlein to the Inside Higher Ed piece, titled “the problem with the boy problem”.

Some other blog posts discussing female dominance in higher education:

• Women’s role in the economy (June 10, 2006) by Gary Becker on the Becker-Posner blog.

• Is the growing gender gap in college enrollments a cause for concern? (July 17, 2006) by Gary Becker on the Becker-Posner blog.

• Response on gender gap in enrollments (August 5, 2006) by Gary Becker on the Becker-Posner blog.

• The new gender gap in education (March 2, 2008) by Gary Becker on the Becker-Posner blog.

• The new gender gap in education — Posner’s comment (March 2, 2008) by Richard Posner on the Becker-Posner blog.

There are three issues here: (i) Why do women dominate in higher education? (ii) Given that they do, is this a problem? (iii) If it is a problem, is the solution to be found in affirmative action, or preferential admission policies for males?

The variance explanation for female dominance

One explanation is offered by a general statement that on most traits relevant to academic success, women and men have roughly the same average (mean) but women exhibit lower variance. See this, for instance. In other words, there are fewer females at the upper and lower tails of the distributions, and more females are concentrated near the middle of the distribution. It has been claimed that lower variance for women is true for many traits, and both biological and cultural explanations have been sought.

Biological explanations include explanations based on genes (men have a larger variable portion of chromosome than women), hormones, and evolutionary explanations (e.g., women need to be more steady and dependable because of the burdens of childbirth). I don’t know enough about the topic to judge the veracity or general applicability of these theories.

Cultural explanations would have the general flavor of saying that men are encouraged to take more risks (which would make them more likely to both slog hard for success and become petty street thugs) and women are encouraged to conform more to expectations. This actually borders on another explanation, which we discuss later.

(There have also been some recent studies claiming that males have a slightly higher average IQ than females, and that most of this IQ difference develops during the teenage years. This conclusion, if correct, would mean that IQ is like height — the divergence between male and female heights happens in the teenage years where male height keeps increasing and female height stabilizes. Again, I do not know of the extent to which these findings have been validated by further studies).

Here’s what lower variance would imply. It would imply that in the very upper echelons (say, the top 10%) men would be dominant. Thus, if college-going were a rarity reserved for the smartest, men would be the dominant sex in college. On the other hand, if, say, the top 70% of people are going to college, women would dominate because there are more women concentrated near the middle, and people at the very low end, who are left out, are male. This is broadly consistent with empirical findings: in the most high-IQ majors and professions (such as mathematics, physics, economics, and engineering) male dominance is greater, while females dominate in political science, English, and history. The variance argument is also borne out by looking at distributions of things such as SAT scores and IQ scores.

Note that the ability of lower variance to explain a phenomenon does not mean that it is the correct explanation, or that it explains all of the difference. Further, this explanation needs to be tested against facts for each individual trait. For instance, a study on why men are better chess players than women, discussed here at Science Blogs and here at Marginal Revolution finds that there is no significant difference in the variance for men and women when it comes to chess.

Also, simply saying “lower variance” is not in itself a complete explanation — a full explanation would explain which trait is causing the difference in educational attainment, and further show that for trait, women exhibited lower variance than men. It might further seek an explanation for the difference in variance.

Early-age language learning

Another possibility that has been suggested is that women are better at learning lanuage in their early, formative years. At young ages, males have an advantage in visuo-spatial abilities and women in verbal abilities, the cliche goes. Since much of education and learning crucially hinges on the ability to understand and process language, girls have an advantage over boys in the younger ages and this advantage carries over to good performance in school, which encourages them to work harder and learn more, and so on.

The “verbal ability” argument, though a cliche at some level, is hotly contested, with different studies showing different results. See, for instance, this meta-analysis of the male-female differences in verbal ability.

Attitude differences

Another explanation for the differential in high school performance is attitude: females have a better attitude when it comes to listening to teachers, following instructions. They have, on average, somewhat better work discipline, which allows them to sit still and learn a topic to a reasonable degree of competence.

This is somewhat related to the “variance” argument in that the argument could be modified to claim that females have lower variance in attitude — with some men being overly dedicated and some totally unenthusiastic about work, while women coming closer to average.

It is also somewhat related to the “verbal ability” argument in that it may be argued that the ability of girls to sit still and work is that their better verbal abilities makes this more pleasant for them.