Gender Differences in Publication among University Professors in Canada *
Nakhaie, M. Reza, The Canadian Review of Sociology and Anthropology
VEBLEN (1918) LONG AGO ARGUED that persistence and success in higher education involves two lines of work, scientific research and teaching. He insisted that the former of these is primary and indispensable. In fact, the former is a necessary activity for one to be an effective teacher. Scholarly publication is also a key to career success and recognition in the scientific community as well as a visible criterion that determines institutional efficiency in academia. Despite its importance, there is little research on the publication activities of Canadian professors. In particular, we know little about the publication patterns of Canadian professors across ranks, disciplines, type of university, by ethnicity, gender and/or family responsibilities. In this paper, I evaluate publications of Canadian university professors, using a large 1987 study and paying special attention to gender differentiation. The main guiding question is whether or not women and men scholars produce the same research output. If not, wh y?
Review of the Literature
One of the most persistent differences in research productivity found in American and European studies is between academic men and women. Cole and Cole (1973), Cole (1979), Astin and Bayer (1979), Cole and Zuckerman (1987), Franklin (1988), Zuckerman (1991), Bailey (1992), and Long and Fox (1995) all found that female professors are less productive with respect to research and publication than male professors. Blackburn and Fulton (1975: 266) showed that in both the United States and the United Kingdom, in all subjects men publish more articles and more books than women. In some subjects, such as the humanities, the difference is substantial, with men publishing two or three times as much; while in others, such as the social sciences, the sexes are much closer together and the difference is in the range of 20%. Similarly, Lie (1990) showed that in Norway, as in the United States, academic men are more productive than women. The difference in the Norwegian male and female publication rates is smallest in the n atural sciences (20%). Women in other areas publish 30% to 35% less than males. In the social sciences, however, women are actually more productive.
Long (1992: 160) argued that although attempts have been made to explain publication differences between men and women, "unfortunately none of these explanations have been successful in accounting for sex differences in productivity." Such findings of less productive women than men are often referred to as "the productivity puzzle" (Cole and Zuckerman, 1984). There are two broad explanations of gender differences in publication rates.
One focusses on various gender-specific personal and social responsibilities that negatively affect professional lives. It is argued that male professors publish more than female professors because having a spouse who takes care of household responsibilities frees men to devote larger amounts of time and energy to publication. However, studies that examine the effects of women's and men's domestic responsibilities and private-life identities on their publishing records do not seem to support such a hypothesis (see Cole, 1979; Cole and Zuckerman, 1987; Menges and Exum, 1983; Simone, 1987; Long, 1990; Davis and Astin, 1990; Grant and Ward, 1991). The results of these studies show that married women and women with children publish more than women who are single and childless. Despite the fact that child care is traditionally one of a married woman's responsibilities, research shows that women with children are more productive than women with no children (Lie, 1990: 115). Moreover, Fox and Faver (1985) found tha t having young children is positively related to women's productivity. The evidence suggests that it is not marriage and children per se that explain women's lower productivity. However, one problem with such a conclusion is that most of the research that evaluates gender differences in publication has not employed multivariate models (Bailey, 1992). …