Authors' reply

In our recent Correspondence1 we presented a study based on the career trajectories of PhDs trained in the United States in the discipline of physical oceanography, focussing on the differences in the rate at which men and women attain tenure-track positions, and changes in these rates over time.

We speculated that the changes in success rates for women over time could reflect different family choices that women have made in recent years, but we did not mean to imply that having children explains all differences between men's and women's careers. At least in the USA, recent literature shows that women's careers are negatively affected by being married as well as by having children, whereas men's careers are not2. The tenure system in the USA requires individuals to be most productive in the first five years after securing their positions, and this period often coincides with the time when women choose to have children. A longitudinal study in the astronomy community of 800 PhDs who will be followed for 15 years is designed to assess whether family status directly impacts career decisions3. By contrast, our study was retrospective and we had no direct contact with individuals.

Hiring of female faculty in science, engineering, technology and mathematics fields has been observed to tail off after a critical mass of women in a particular discipline has been reached, but before the gender distribution of faculty hired represents that of the potential applicants4. Issues surrounding affirmative action may therefore provide most of the explanation for our results.