When it comes to the huge and persistent gender gap in science and technology jobs, the finger of blame has pointed in many directions: sexist companies, boy-friendly science and math classes, differences in aptitude.
Women make up almost half of today's workforce, yet hold just a fraction of the jobs in certain high-earning, high-qualification fields. They constitute 20 percent of the nation's engineers, fewer than one-third of chemists, and only about a quarter of computer and math professionals.
Now two new studies by economists and social scientists have reached a perhaps startling conclusion: When it comes to certain math- and science-related jobs, substantial numbers of women - highly qualified for the work - stay out of those careers because they would simply rather do something else.
One test, a standard personality-inventory test measured people's preferences for different kinds of work. Personal preference it concluded was the single largest determinative factor in whether women went into IT. They calculated that preference accounted for about two-thirds of the gender imbalance in the field.
A second study found something else intriguing: Women who are mathematically gifted are more likely than men to have strong verbal abilities as well; men who excel in math, by contrast, don't do nearly as well in verbal skills. As a result, the career choices for math-precocious women are wider than for their male counterparts. They can become scientists, but can succeed just as well as lawyers or teachers. With this range of choice, their data show, highly qualified women may opt out of certain technical or scientific jobs simply because they can.
Despite these studies women still seem to make choices throughout their lives that are different from men's, and it is not yet clear why. For example we don’t know about the role of mentors or experience or socialisation.
These findings on self-selection only open new areas of inquiry. The end result may be surprising - and an equal-opportunity workforce may look a lot less equal than some had imagined.
I would be very interested in your views on this.