Frequently Asked Questions

Women in Physics and Astronomy, 2005

June 2005

Perhaps the most controversial and most publicized finding of the women’s report is the finding that in physics and astronomy, there is no “leaky pipeline” at the faculty level. In other words, women are represented on the faculty at about the levels we would expect based on degree production in the past. Below are several frequently asked questions about this conclusion.

Q: Do your conclusions mean that there are no problems for women in physics?

A: No. The situation for women in physics and astronomy still needs improvement. Although women are hired into tenured and tenure-track positions at respectable rates, they are hired into part-time faculty positions at even higher rates. Controlling for sector of employment and for time since degree, women earn significantly lower salaries than men. Very few minority women earn degrees in physics. Finally, the fact that women are represented on the faculty at expected levels does not mean that they got those positions without tremendous personal sacrifices. In many physics departments, women encounter climates that range from chilly to hostile. Nevertheless, the results of our analysis send a positive message to younger women considering a career in physics. Women can, in spite of obstacles, make it to the top faculty positions in physics and astronomy.

Q: I know more than one woman who has left academic physics. How can your data be correct?

A: Not everybody who gets a degree in physics pursues a career in academics. Historically, the majority of men and women with physics PhDs have non-academic careers. In other words, while many women choose not to pursue academic careers, so do many men. The report shows that women are not more likely to leave the academic pipeline than men are.

You may pay particular attention to the women who have left. People tend to remember either what is important to them, or what is different. And women in physics are still “different” because there are so few of them that what they do becomes very noticeable.

Q: What about women at the top physics departments? Perhaps women can get faculty jobs, but not at the elite departments.

A: Women can, and do, get jobs at elite departments. Among the top 20 departments (using NRC rankings), 6% of the full professors are women, 11% of the associate professors are women, and 12% of the assistant professors are women. These percentages are also consistent with degree production in the past, meaning that women are represented on the top 20 faculties at about the levels we would expect.

Q: Are there any other studies that show results similar to yours?

A: Yes. Xie and Shauman (2003) analyzed longitudinal data on women and men entering science careers and found that a major leak in the pipeline happens before college because “ girls are less likely to declare a science major [even though they] have done the work and have what it takes to move into these majors. In college, the persistence rates are just as good between young men and women—in fact, once in the programs, women are slightly more likely to complete science degrees." However, their analysis found that married women with children encounter many obstacles in pursuing a career in science. http://www.mentornet.net/news/2005/junenews2.aspx.

Sonnert and Holton (1995) found that women are not more likely than men to leave science, although they found other significant differences in men’s and women’s careers in science. http://www.aps.org/apsnews/0704/070407.cfm.

Q: Where do you get your data on the percentage of faculty who are women and on the age distribution of the faculty?

A: Data on the percentage of faculty members who are women come from a population survey of physics departments in the US. This survey had a 94% response rate. Data on the age distribution of faculty come from our Membership Survey, which is sent to a randomly selected sample of the US-resident members of the ten AIP Member Societies. The Membership Survey has more than 10,000 respondents and a response rate better than 60%. Of these, nearly 2,000 are PhD physicists working in universities and colleges.

Q: Although you say that the differences on time since PhD for men and women faculty are not significant, it appears that women faculty members are younger than men. Shouldn’t you use the age distribution of women to predict what percentage women on the faculty should be?

A: No, to do so would be incorrect. Let’s take a hypothetical example. Suppose our data showed that the middle 50% of male faculty members earned their degrees between 1965-1975, and that the middle 50% of female faculty members earned their degrees between 1970-1980. If we base our prediction on degrees awarded during 1970-80, we are missing the years before 1970, when women did earn PhDs in physics. These women should be represented on the faulty. Therefore, we use the age distribution of all faculty members, both male and female, to predict the percentage of faculty members who should be women.

Q: I know that many of the PhDs awarded in physics go to foreign citizens. Does your analysis take this into account?

A: Yes. In the US, citizens earn almost all physics bachelor’s degrees. We looked at the percentage of physics bachelor’s degrees earned by women compared to the percentage women among first-year, US-citizen grad students. The percentages are very similar, in that the percentages of women earning a physics bachelor’s degree and then entering graduate school the next year are almost identical. This means that among US citizens, there is almost no leak in the pipeline between earning a bachelor’s degree in physics and entering physics graduate school.

Q: What about faculty who earned their PhDs outside the US?

A: Unfortunately, because of the limitations of our data, we cannot exclude these faculty members in calculating the percentage of faculty who are women. We compare this percentage to the percentage of women earning degrees in the US. This is one example of how an individual-level survey (rather than the aggregate numbers we use) would improve our understanding of how people progress up the academic ladder. We don’t yet have these individual-level data