Numerous studies have found that female scientists publish at slower rates than male scientists. In a classic statement of the problem, Jonathan Cole and Harriet Zuckerman characterize this gender gap in publication rates as "the productivity puzzle."1 A widely held statistic stemming from the study by Cole and Zuckerman is that female scientists are slightly more than half as productive as male scientists. If this sounds like the 59-cent button (the symbol adopted to protest the gender gap in wages), you are on the right track.
It was not until the late 1980s that women's average wage began to improve toward equity relative to men's. A parallel but more dramatic story is true with the gender gap in research productivity. We recently conducted a systematic and detailed analysis of data from four large, nationally representative surveys of postsecondary faculty in 1969, 1973, 1988, and 1993.2 Our research yielded two important and surprising findings. First, the raw gap in productivity between female and male scientists declined significantly, with the female-to-male ratio increasing from about 60 percent in the late 1960s to 75-80 percent in the late 1980s and early 1990s. This held true even when we did not control for any other factors, such as age, rank, field, and employment positions and resources. Second, when we did control for those factors, the gender gap disappeared entirely, suggesting that when all things are equal, female scientists are just as productive as their male colleagues.
There's just one catch: All things aren't equal. In our research, we identified three dimensions along which gender inequality contributes to the gap in productivity rates between female and male scientists: personal characteristics, structural positions, and facilitating resources. We documented, for example, that women are less likely to work in research universities, spend more time in classroom teaching, and are less likely to secure research funding and research assistance than men. Not surprisingly, these gender differences in the distribution of structural positions and facilitating resources are to women's disadvantage. We also identified two personal characteristics that put women at a disadvantage: time between a bachelor's degree and a Ph.D. degree and marital status. Across the multiple data sets, for example, we found that women were 40 to 80 percent more likely than men to have taken more than 10 years between earning a bachelor's degree and a Ph.D. Scientists who took more than 10 years between earning a bachelor's degree and a Ph.D. were 30-40 percent less productive than those who completed their Ph.D. within four years of earning a bachelor's degree. We also found married scientists to have significantly higher (by 7-11 percent) rates of productivity than unmarried scientists and that female scientists were much less likely (by 20-40 percent) than male scientists to be married.
For those who wish to see gender equity in research productivity among scientists, it may be discouraging to know that the remaining 20-25 percent gender gap in productivity was entirely due to the unequal distribution of personal characteristics, structural positions, and facilitating resources between the two genders. However, it is worth noting that a major reason for the large decline in the raw gender differences in research productivity over the last three decades is the trend of narrowing gender gaps along the same dimensions. For example, the two personal characteristics we identified (time between earning a bachelor's degree and a Ph.D. and marital status) became more equally distributed between female and male scientists from the late 1960s to the early 1990s (roughly by 50 percent). With the secular improvement of women's role in science, one can be optimistic that women will gain further ground in science relative to men, and this continuing trend will help reduce the gender gap in research productivity.
Did we solve "the productivity puzzle"? The answer is both yes and no. The answer is yes in the sense that our study successfully identified, for the first time, differences between female and male scientists in personal characteristics, structural positions, and facilitating resources that account for women's lower research productivity. That is, we show there is very little direct effect of gender on research productivity. However, we still do not know why female and male scientists differ along these important dimensions, and in this sense the puzzle remains unsolved: Why do career trajectories of female scientists still differ systematically from those of their male colleagues? In our current work on women in science, we explore the influences of various social and personal factors, including but not limited to the role of family responsibility for women pursuing scientific careers. The results of this study will be reported in a book.
Yu Xie is John Stephensen Perrin professor of sociology and associate director of the Population Studies Center at the University of Michigan, and Kimberlee Shauman is assistant professor of sociology at the University of California, Davis.