3. Careers

In the Preface to Ahern and Scott (1981), Hornig states that, ``generally speaking, women are far more likely than men to be involuntarily unemployed and underemployed; they were much less likely than men to attain senior ranks or move to management levels, and their earnings not only reflected these differences but were persistently lower even at equal ranks. Women who appeared equal to men in all respects at receipt of the doctorate had less assured careers than men with slower progress and lower ceilings.'' Six years later, Vetter (1987) draws the same conclusions and suggests further there has been little change since 1977. On the other hand, Zuckerman (1987) is more encouraging, noting that while gender differences are present in all categories of academic institutions, most especially at the top ranked ones, trend data suggested that the gap is narrowing, especially at the lower ranks.

By far, the most important contribution to an analysis of career attainments is the superb study by Ahern and Scott (1981). This study (to be referred to as the matched triads study, originated by Dorothy Gilford) is based on triads consisting of one women and two men matched by year and field of the doctoral degree, the institution at which the degree was earned, and race. For some analyses, there is further matching by years of full-time equivalent experience and employment sector. Taking into account perceived levels of quality proved too complicated to handle in the study. Given the difficulties encountered by women in publication rates (see below), the fact that quality could not be addressed adequately is perhaps unsurprising. The data were compiled as to the observed outcomes of 1979 of approximately 50,000 individuals for four cohorts, viz., those who received the doctorate in the period 1940-1959, 1960-1969, 1970-1974, and 1975-1978, respectively. While the entire study was concerned with academic careers over the sciences, engineering and humanities, our analysis here will be restricted to the results for mathematics (including statistics) unless otherwise stated. We shall consider the issues of employment/unemployment, rank and tenure, including progress in promotion. Issues surrounding salaries will be addressed in the next section. More complete details can be found in Ahern and Scott (1981). As we look at the study, we should bear in mind that it was not until 1972 that the legislation prohibiting discrimination in federally assisted programs (Title IX of the Education Amendment) and prohibiting discrimination in employment in educational institutions (Amendment of Title VII of the 1964 Civil Rights Act), was enacted.

Employment

In 1979, 82.4% of men and 84% of women with doctorates in mathematics were employed in educational institutions (Vetter (1981)). For the matched triads, Table 2 shows that consistently fewer women are employed full-time than are men and that, excluding the 1940-1959 cohort (where some are now retired), roughly three times as many women are unemployed, though the numbers are reasonably low. More importantly, the percent of men and women unemployed for the most recent 1975-1978 cohort is just one. For all mathematics doctorates in 1985, Vetter (1987) reports the unemployment rates have decreased further to 0.4% for men and 0.9% for women; by 1987, these rates decreased slightly for men to 0.3% but increased substantially to 1.8% for women (NSF (1990)). Approximately six times as many women as men in the matched triad study are in part-time positions. Zuckerman (1987) reports that women are more likely to be underemployed, being either in involuntarily part-time positions or in jobs outside their training. Finally, men (9%) and women (11%) in the 1970-1974 cohort were approximately equally likely to be employed in the top ranking universities (primarily due to overrepresentation of women at the lower ranks), which shows an improvement over the corresponding figures (11% for men and 6% for women) in the 1960-1969 cohort.

Rank and Tenure

For those identified in the matched triad study, the percent of the men and of the women occupying the traditional academic ranks of professor, associate professor, assistant professor, and instructor/nonfaculty positions is shown in Table 2 for each of the four cohorts. Also shown is the percent who are tenured. It is immediately apparent that, despite the fact that the data represent men and women starting out with equivalent matched credentials, women do not fare as well as men, with the women being ranked consistently lower than their male peers. Those in the first (1940-1959) cohort do eventually progress up the ranks, although twice as many women (12%) as men (5%) appear to be frozen in rank below the full professor level. For the 1960-1969 cohort, men were 1.4 times as likely to be full professors, and 97% of men but only 80% of women were tenured by 1979. Of those (in all fields) in this cohort who were tenured, the men took an estimated average of 5.9 years to attain tenure while the women took 6.3 years.

The 1970-1974 cohort spans the 1972 date of the nondiscrimination statutes, and by 1979 such individuals should have reasonably expected to be promoted and tenured. Therefore, the fact that still considerably fewer women (45%) than men (70%) have been promoted to full or associate professor suggests further analysis is warranted. The matched triad study investigated three possible explanations. Ahern and Scott (1981) found that a lower commitment to research with a corresponding heavier teaching load did not affect women differently than men in the same position. In fact, a higher percent of those whose primary activity was teaching rather than research was promoted by 1979 from assistant to associate professor (64% teaching versus 60% research for men and 48% and 32%, respectively, for women). Another commonly believed explanation, that women lost time due to childbearing, did not stand up to scrutiny. Indeed, married women with children (with 51% being promoted from assistant to associate professor) fared better than married women with no children (41%) who in turn fared better than unmarried (including widowed and divorced) women with no children (37%). Of those unmarried with children, 33% were so promoted. The corresponding figures for men were 66%, 51%, 53% and 80%, respectively. Thirdly, the perception that women are less mobile than men was not substantiated by the matched triad study. To the contrary, more women (28%) than men (19%) changed jobs between 1975 and 1979. While men who moved improved their status significantly, women who moved did not improve their status, possibly because the women moved due to a failure to receive promotion and tenure. Few faculty with tenure in 1975 changed jobs between 1975 and 1979.

Comparison of Ranks

For all four cohorts, although individuals in the matched triads started our together, differences in rank and tenure status persisted over the years. Accordingly, Ahern and Scott (1981) developed prediction equations (using stepwise regression analysis) to estimate rank and salary and to estimate the salary women should receive if paid like a man with similar characteristics. The data used here were confined to those triads for which the doctorate was earned since 1958. This analysis also included matches that had equivalent full-time work experiences. A prediction equation was established for men and was then used to predict the ranks (and salary) women in the same field and institution category should receive if rewarded the same as a man with comparable characteristics. Taking into account the possible biases of this study, Ahern and Scott opined that their method should provide good estimates of the differences for younger faculty but underestimate the differences for older faculty.

The standardized regression coefficients in the prediction equations for women and men in the MPE (mathematics, physical sciences and engineering) field are displayed in Table 3. (In this equation rank was coded as 4 for full professor, 3 for associate professor, 2 for assistant professor and 1 for instructor). The most important predictor was a weighted measure of time. For men, this was the time since receiving the doctorate, while for women, the number of full-time equivalent years experience dominated. The sum of the coefficients of these two (0.747 for men and .638 for women) estimated the average yearly increase in rank. For men, the second most important predictor was the variable indicating the person had children under 18. The effect was positive. This predictor also had a positive effect for women (in MPE, but it did not enter in the equation for other fields). For women, the prestige of the department from which the doctorate was received had a positive effect. Since the two variables -- being married at the time of the doctorate, and being married at the time the data were collected in 1979 -- are collinear, only one of these variables usually entered the prediction equation, producing a slightly negative effect when present. Prediction equations which included sex as a predictor variable showed the effect for female to be always negative and significant, with women predicted at a rank approximately one-third lower than their matched men.

However, while these studies show that women are promoted more slowly and gain tenure more slowly than their male counterparts, Zuckerman (1987), using the CEEWISE 1983 data, reports that over science and engineering fields, generally 67% of men and 40% of women in tenure track positions had received tenure by 1983. This suggests the gap is narrowing. More recent data on the distribution of all doctoral men and women mathematical scientists employed in 1987 at four-year colleges and universities compiled in NSF (1990) suggests this trend is continuing. From Table 4, we observe that 58% of the women but 75% of the men are either at the Full or Associate Professor rank; and 75% of the women but 85% of the men are in tenured or tenure-track positions. Of all those in the Full (Associate) Professor rank, 5% (13%) are women with 7% of both these ranks combined being women; and 7% (15%) of those who are tenured (tenure track, but untenured) are women with 8% of all those in tenure track positions being women.

Next: 4. Salaries Up: The Past, Present, and Previous: 2. Education

Copyright ©1991 American Mathematical Society. Reprinted with permission.
Brought to you by the Association for Women in Mathematics.
Comments: awm-webmaster@awm-math.org