I happened upon a list of figures that display lots of information about who majors in science and engineering (S&E), all available at the NSF page on Women, Minorities, and Persons with Disabilities in Science and Engineering. All are for institutions in the U.S. only, as far as I can tell.
Here we see the number of bachelor’s degrees in S&E and non-S&E fields, by sex:
Percent of S&E and non-S&E bachelor’s degrees earned by racial minorities:
Number of doctoral degrees earned by sex:
Percent of female workers in selected occupations in 2007:
Percent of S&E Ph.D.-holding employees at 4-year colleges and universities that are women:
Notice that while women are pretty well represented among “full-time tenured or tenure-track faculty,” they make up barely 15% of “full-time full professors with children,” “married full-time full professors,” and “full-time full professors at research institutions.” Some of this may be a matter of time, in that because being promoted to full professor takes years, an increase in the number of women getting Ph.D.s in a particular field will take a while to show up as a similar increase in number of women as professors in that field, everything else being equal.
But everything else isn’t equal. Women, even highly-educated ones, still do the majority of childcare and housework, and are more likely than men to consider how potential jobs could conflict with future family responsibilities when they are deciding what type of career to choose. So the lower proportion of female full professors at research institutions is likely a combination of some old-school unfriendliness to women in some departments, but also of women opting out of those positions–that is, deciding that the time and energy required to get tenure in a science/engineering department at a research school would conflict too much with family life, so they pursue other career options (for instance, women make up a higher proportion of faculty at community colleges, which are often perceived as being easier to get tenure at, since you don’t necessarily have to do lots of publishing–though anyone who has ever taught 5 courses a semester may question the assumption that it takes less time and work to teach than to publish).
And of course some women try to combine a tenure-track job with their family responsibilities and find it difficult, and either leave academia voluntarily or are turned down for tenure because they have not published enough or are not see as adequately involved and committed. Because of inequalities in the family, men are less likely to face those situations, though certainly some do.
I wish the report had a graph for non-S&E faculty, although it might just depress me.
UPDATE: Commenter Sarah says,
Check out the report “Women in Science: Career Processes and Outcomes” by Yu Xie and Kimberlee A. Shauman. They find no evidence to support the hypothesis that women are under-represented at the tenure level due to a lag between getting PhDs and acquiring tenure track positions. In the life sciences, women have been getting PhDs at the same rate as men for many years now, but inequality persists at the tenure-track level. Xie and Shauman find that a variety of social factors are responsible for actively hindering women at the highest levels of academia.