Gender differences in the careers of academic scientists and engineers: a literature review

There is substantial evidence that women, as a group, ies and include measures of human capital, measures of are underrepresented in senior academic ranks. The mod- productivity, personal characteristics, and academic field.
eling issues discussed below should be considered wheninterpreting the results of empirical research on advance- The rationale for including human capital variables as controls in studies of academic rank (and tenure sta- tus) is similar to the rationale for their inclusion in sal- Many of the studies on academic rank that we re- ary studies. Other things being the same, one would ex- viewed attempted to determine the effects of gender on pect that individuals who have accumulated more hu- academic rank after controlling for the effects of other man capital are more likely to receive tenure and to be factors that might affect promotions (e.g., experience and scholarly production). In most cases, these studies em-ployed one of two kinds of analyses: discrete outcome Experience
The number of years elapsed since earning the doc- torate is perhaps the most commonly used measure of Discrete outcome models permit multivariate analy- experience in academic rank studies. McDowell and ses of outcomes that are observed as discrete events. This Smith (1992), however, included a variable measuring kind of model is appropriate for analyses of discrete ca- years of academic experience in their study. Several au- reer outcomes, such as academic rank or tenure (e.g., the thors, including Ransom and Megdal (1993) and individual is either tenured or not tenured). Two kinds of Raymond et al. (1993), included years of service at the commonly used discrete outcome models are logit analy- employing university as an institution-specific measure sis and probit analysis.23 Long (2001), Olson (1999), and Raymond et al. (1993) all used logit analysis in their stud-ies of academic rank. Ransom and Megdal (1993), Education
McDowell and Smith (1992), and Farber (1977) used Some studies of academic rank include measures of probit analysis. Logit and probit analyses allow research- educational quality as controls. For example, Long ers to estimate, for example, the effect that gender has (2001) controlled for the prestige of the doctorate-grant- on the probability of being promoted to the rank of full ing institution in his study of tenure and promotions.
professor after controlling for other factors that might Olson (1999) included as controls post-doctoral appoint- affect rank, such as experience or scholarly productivity.
ments and the Carnegie classification and departmentalrankings of the doctorate-granting institution. Broder Hazard analysis is a useful tool for analyzing factors (1993) also controlled for the quality of the department that affect the length of time required to achieve a given from which individuals earned doctorates. When data academic rank.24 Both Weiss and Lillard (1982) and Kahn included faculty who had not earned doctorates, some (1993) used hazard analysis in their studies of academic studies included control variables for the highest degree rank. Hazard analysis allows the researcher to estimate, for example, the effect that gender has on the time re-quired to reach the rank of full professor after control- Characteristics of Employing Institution
ling for other variables affecting promotions.
Several studies, including Long (2001), Olson (1999), Broder (1993), Kahn (1993), and McDowell and The kinds of control variables used in the literature Smith (1992) controlled for the characteristics of the on academic rank are similar to those used in salary stud- employing institution. These controls could be inter-preted as measures of human capital, given that individu- 23Logit and probit analyses are similar statistical tools but differ in als who have accumulated the most human capital are assumptions about the distributions of random modeling error.
most likely to be employed at the most prestigious uni- 24Hazard analysis, sometimes referred to as duration analysis, is versities. In studies of academic rank, however, employer superior to ordinary least squares regression analysis in that it can deal characteristics are probably better interpreted as proxies with censured observations. Observations on the length of time between for variations in tenure and promotion requirements.
promotions are censored in that individuals who have not yet beenpromoted are still observed in lower ranks.
Because promotion requirements are likely to be most stringent at the most prestigious institutions, institutional column in Table 4-1 identifies the years covered by each quality is likely to be negatively related to the probabil- study. The second column briefly summarizes the find- Taken as a whole, the findings from the literature suggest that, other things being the same, female faculty Many of the studies of academic rank we reviewed find it more difficult than male faculty to achieve tenure controlled for scholarly productivity, but few controlled and to be promoted to senior academic ranks. Of the stud- for teaching output and those that did used relatively simple ies that we have reviewed, only two found no statisti- controls. Only one of the studies reviewed included any cally significant gender differences in promotion rates.
controls for service to the academic community.
Raymond et al. (1993) found no evidence of gender hav-ing an effect on academic rank, but this study used data Scholarly Productivity
for a single institution. A study by McDowell and Smith As in the salary studies, most of the academic-rank (1992), who used data for only the field of economics, studies we reviewed used simple counts of the number found no statistical difference in promotion rates between of articles published as measures of scholarship. Olson men and women after allowing for gender differences in (1999) controlled for the number of papers presented at the effect of experience on academic rank. They did find conferences as well as the number of publications.
that women receive less credit for experience than men Raymond et al. (1993) included research grant money do. Interpreting gender differences in returns received awarded. Studies by Olson (1999) and Farber (1977) in- from experience has raised controversy in the literature.
cluded indicators that research was the primary work Gender differences in credit for experience could be due either to gender differences in human capital accumula-tion (caused by family responsibilities and workforce Teaching
As noted above, controls for teaching output are rela- tively rare and are simple in the academic-rank studies The findings from some of the studies we reviewed we reviewed. Two studies, Olson (1999) and Farber (1977), suggest that women faculty are placed at a particular dis- controlled for teaching as a primary work activity.
advantage by family responsibilities during child-rear-ing years. For example, Farber (1977) found that women receive significantly fewer promotions when they are Generally, fewer academic-rank studies than salary young but found no significant differences in promotion studies controlled for personal characteristics. A few rates for older women. McDowell and Smith (1992) con- studies controlled for such factors as age, age at the time cluded that promotion rates for women are lower than of earning the doctorate, and race/ethnicity. Unfortu- those for men because women receive less credit for years nately, only three studies, Long (2001), Olson (1999), of experience. Gender differences in family responsibili- and Winkler et al. (1996), included marital and parental ties may be responsible for this finding.
Kahn (1993) found that women are less likely than men to receive tenure but found no gender effect for the time between promotion from associate to full profes- Table 4-1 summarizes the findings of multivariate sor. The tenure decision, which usually coincides with studies of the effects of gender on academic rank. Each promotion from assistant to associate professor, often of the studies listed in this table controls for at least some occurs during early child-rearing years.
measure of experience and academic field.25 The first Long (2001) and Olson (1999) estimated separate promotion models for women and men and included con- We adopted two criteria for including in Table 4-1 studies that we reviewed. First, the studies must include original empirical The results of the academic rank studies are more difficult to research on the relationship between gender and tenure or academic summarize quantitatively than are the salary studies. This is due in rank. Second, the studies must attempt to control for factors other part to differences in modeling approaches across studies and the than gender that might affect tenure and promotion.
kinds of quantitative results reported by the authors.
trol variables reflecting the number of children at home.
Long’s results do not show consistent, statistically sig- Olson found that having children significantly reduces nificant gender differences in relations between promo- the chances of promotions for women but not for men.
Table 4-1. Estimates of gender differences in rank and tenure
promotions when under age 40; promotions comparable at ages 40–50 Women wait twice as long as men to be promoted Several S&E fields Women less likely than men to be in senior ranks; All academic fields promotion rates of women about the same as for men with 1–2 years less experience Women’s experience counts less for promotions Women less represented at full professor level Descriptive statistics Everett et al. (1996) Women less likely to be full professor, tenured, or All S&E fields Gender does not affect likelihood of promotion Women with more than 6 years of postdoctoral experience more likely than men to be in lower ranks Women less likely to be tenured; no gender effect All S&E fields for time between tenure and full professor rank Women make up 51% of instructors, 38% of Descriptive statistics Carnegie Foundation (1990) assistant professors, 28% of associate professors, fieldsand 13% of full professors; also less likely to be tenured or on a tenure track Women disadvantaged with respect to rank and Regression analysis Sonnert and Holton (1995) Small number of women associate professors Atmospheric sciences Descriptive statistics Winkler et al. (1996) About 21% of women employed at full professor Descriptive statistics Everett et al. (1996) Women less likely to be full professor, in senior Women more likely to be employed as instructors Geosciences Descriptive statistics Ongley et al. (1998) 1Indicates years covered by data used in study. 2Study conducted for a single academic institution.
3Senior ranks include associate- and full-professor ranks.
KEY: S&E = science and engineering
27Neither Long nor Olson standardized the timing of when children are observed during the postdoctoral career. The timing of fertility mightaffect the influence that children have on academic careers (e.g., havingchildren before or after the tenure decision).

Source: http://www.dletemplates.com.ua/statistics/nsf03322/pdf/sect4.pdf