Two Decades of Gains For Female Faculty?

There is conflicting evidence about the degree of success women have had in academic careers.1 On the positive side, the record is clear that both the number and the percentage of women earning doctorates in all fields has been increasing since 1965.2 Particularly, and importantly, women have become a growing segment of the science and engineering work force.3 In addition, there has been a steady gain in the number of women employed at the 3,300 colleges and universities in the United States.4 On the negative side, one reads that women are concentrated in the lower professorial ranks (and underrepresented in the full professor category) and have lower salaries even when occupying the same rank as their male counterparts.5

Most of these reports regarding female faculty are snapshots at one point in time and fail to reveal the ways in which important characteristics of academic women have been changing over time. Our investigation differs from these reports in a significant way: Our work is based on national surveys of faculty extending over twenty years (in 1969, 1975, 1980, and 1988). We examine five status changes for female faculty in the highly stratified U.S. system of higher education. In each instance, we pose this question: How have matters changed over the past two decades? The five areas we report on are:

1. Place of work: While not everyone wants to be on the faculty of a leading research university, these are the institutions with the richest resources. Nearly all faculty holding key national disciplinary and association positions work in Research-I universities.6 In 1969, women held 6.7 percent of the faculty positions in these institutions.

2. Academic discipline: While people ought to study what interests them most, the underrepresentation of women in the sciences—and especially in the physical sciences—has received national attention.7 Many believe women are driven out by the structure of the discipline, the criteria set by the universities, and the way science is taught.8 In 1969, on average there was but one woman in every three Research-I chemistry faculties.9

3. Rank and tenure: Obviously, becoming a full professor and possessing tenure signify success. In 1969, women represented 9.6 percent of the faculty work force and 8.1 percent of the tenured faculty. Nearly 15 (14.9) percent were assistant professors while only 5.6 percent were full professors.

4. Grants: Receiving a research grant is a status symbol in and of itself, but there are other incentives for faculty to engage in grant competition. Securing grants is often directly rewarded with merit raises and promotion. Furthermore, the additional monies for research lead to publishing, the heart of the scholarly enterprise (see number 5 below). In 1969, women were underrepresented in grant awards according to their representation in the academic work force. Women constituted 9.6 percent of the work force yet they received 7.5 percent of the institutional (small-dollar) grants, 6.0 percent of the federal (largest number of dollars) grants, 2.4 percent of the industrial (mixed number of dollars) grants, and 5.7 percent of other (state, foundation, etc.—on the dollar side) grants.

5. Research publications: Publishing scholarly work is the sine qua non of the academy. Publications determine how reputations are earned, grants acquired, promotions awarded, salaries allocated. In 1969, women’s two-year publication rate was about three fourths that of their male counterparts.

Our objectives-are to weigh the best existing evidence from its many sources and to display the record of changes in women’s numbers and status in higher education over the past two decades as they inform these five questions. Movement toward status equity is what our study explores.

FRAMEWORK

Success in the academy goes beyond symbolic equity—equal salaries, equal numbers, and the like. Success includes the possession of power—women’s being able to influence the rules by which the game is played. Success means being able to alter the current rules, to create an alternative set that are accepted as fully legitimate. Success is more than white, male standards. Behaving like men is dysfunctional for most women. Male success relegates a private life to an unacceptably low status and, at worst, rejects a human commitment to another and to children.10

Our investigation is set in Tetreault’s evolutionary or stage theory of women.11 Tetreault’s five stages are (1) womanless, (2) compensatory, (3) bifocal, (4) feminist, and (5) multifocal. Our study goes beyond the womanless first stage, past Logan Wilson’s Academic Man and the 1969 Carnegie national survey of faculty that uses nothing but male pronouns.12 We are also past the second stage, the compensatory one that glorifies a handful of women who have exceptional records. We operate principally in the third stage—bifocal—where women and men are compared and contrasted, occasionally touching the feminist phase (stage 4), mostly in some observations in the epilogue. We do not intersect with Tetreault’s fifth stage, the multifocal.13

As for our methodological approach, we draw on Epstein’s cogent set of categories for examining research conducted on women: (1) biological/sociobiological, (2) psychological/developmental, (3) sociological, and (4) cultural/symbolic.14 Our study falls within her sociological category (although the inferences we draw from what we have learned call for psychological inquiries). Sociological analyses consider the linking of behaviors with the sexes (e.g., men with research, women with teaching), socialization processes, stratification of institutional types, structures for holding power, and the like. The studies we cite most often in connection with our findings have sociological variables of these kinds.

THE SAMPLES

This study is based on data drawn from four national surveys of U.S. faculty, conducted in 1969, 1975, 1980, and 1988. The first two surveys were sponsored by the Carnegie Foundation and the American Council on Education. The 1980 survey was conducted by faculty at UCLA, and the 1988 survey was carried out by faculty at the University of Michigan.15

One would expect research findings from the same and interconnected databases to agree rather than conflict. However, corroboration is not always the case. One reason is that each of the national faculty surveys suffers from sampling errors. For example, the 1969 and the 1980 surveys greatly over-sampled research universities and had a poor response rate from community college faculty. Since the percentage of women is lowest in the former and highest in the latter, comparisons of women in higher education will be in error unless corrections are made. Consequently, it is not surprising that researchers may report disagreeing and even conflicting outcomes.16

Our data represent full, associate, and assistant professors in eight disciplines—humanities (English, history), natural sciences (biology, chemistry, mathematics), and social sciences (political science, psychology, sociology). These eight disciplines are large and common to most colleges and universities. The subsamples target faculty from Research, Doctoral, and Comprehensive-I colleges and universities.17

RESULTS

Our analyses resulted in a number of findings regarding the changes in the status of women in academe in the past 20 years.18 We report the findings in the order of the questions raised at the outset, namely, (1) where women work; (2) the academic disciplines women pursue (or are hired in); (3) the rank women have achieved; (4) research support women are awarded; and (5) their publication rates. In addition, we report some salary data comparing men and women, but without 1988 figures. When relevant, we proffer explanations for the data.

PLACE OF WORK—COLLEGE AND UNIV73RSITY TYPE

While women academics still remain underrepresented in proportion to their fraction of society, they have made gains in entering faculty positions over the past two decades.19 The percentage of women working at Research, Doctoral, and Comprehensive-I institutions increased from 9.6 percent in 1969 to 19.5 percent in 1988.

Most notable was the increased participation of women in more selective universities. Whereas the earlier surveys (especially 1969 and 1975) revealed a clear maldistribution of women by institution type—with women more concentrated in Comprehensive-I and Doctoral-II institutions-this was no longer the case in 1988, as the percentage of women hired by Research and Doctoral-I institutions increased about 12 percent. Unlike the earlier three survey years, the 1988 data show no significant difference in the distribution of sex by institutional type.20 Still, women held less than 15 percent of the ranked positions in Research-I universities, the smallest percentage of the five types.

PLACE OF WORK-ACADEMIC DISCIPLINE

Women continue to be concentrated in certain disciplines. For example, while women represented 19.5 percent of the total faculty work force in 1988, they constituted 11 percent of the political scientists (7.3 percent in 1969). By contrast, women represented 33.5 percent of all English faculty (vs. 16.8 percent in 1969) and 22.9 percent of the sociologists (vs. 4.8 percent in 1969). Although women continue to be less prevalent in the natural sciences, here too there have been gains over the four time periods. The percentage of female chemists rose from 2.6 percent in 1969 to 9.9 percent in 1988. The proportion of female mathematicians also increased during the same period, from 6.1 to 14.1 percent.

STATUS/RANK

The data show that women are still concentrated in positions of lower academic rank, namely, assistant professors.21 For example, women constituted 35.2 percent of all assistant professor positions in 1988 at the same time that they were 19.5 percent of all faculty.

In order to see how the proportions have changed over the last two decades, one needs to compare the ratios of rank to percentages of women in the work force. In 1969, women comprised 9.6 percent of the work force and 14.9 percent of the assistant professors. That is, women were 1.6 (14.91 9.6) times as frequently found in the lowest rank. The number for full professors is .6 (5.6, the percent of all faculty who were full professors, divided by 9.6). That is, they were but 60 percent of an equal share. The numbers for 1988 are 1.8 (35.2—the percent of all assistant professors who are women—divided by 19.5, the percent of women in the work force) for assistant professors and .6 (11.8/19.5) for full professors. There has been a loss at the introductory rank and no move toward equity at the top. These data suggest that the revolving door is still spinning.

There are, however, other possible reasons for women’s being overrepresented in the lower ranks. One is that their numerical increase reflects recent hiring from the larger pool of women Ph.D.‘s. If that is the case, one would expect the average age of women to be less than that of men. The average age for men increases almost linearly, but at a number that is less than the time intervals when the samples were drawn. In the twenty years, the average age of men has increased about eight years. With retirements being very low over these two decades, the hiring must have been accomplished with younger people, presumably most often newly minted Ph.D.‘s.

The picture for women is different. They are older than men at the beginning of the time span and younger at the end, even dropping slightly over the last eight years. This picture suggests a greater proportionate influx of women and partly explains the increased proportion of assistant professors.

The survey data also show that the percentage of both men and women who have tenure is larger at almost every point in time. In 1969 the figures for women and men were 49 and 58 percent, respectively; in 1988 the numbers were 74 and 85 percent, an equal gain, and a constant difference, between sexes. These data support equal treatment of men and women and suggest that the success rate is the same for both sexes. That is, despite instances of clear discrimination and claims that hiring women is a travesty, the surveys indicate otherwise. To inform the debate, what is needed is a longitudinal study of a cohort of men and women being hired in the same institutional type and discipline as assistant professors immediately after graduate school.

As to whether women are disproportionately located in the lower-status nontenure-track positions such as lecturer, adjunct, and the like—annual appointments with little likelihood of conversion to a tenure-track position- the data are not adequate to render a confident conclusion. Each survey year has more women than men identifying themselves with other than a professorial title (e.g., as lecturers). The numbers are small, however. The largest is in 1988, where 8 percent of the women and 3 percent of the men responded to one of the “no rank” categories. However, most surveys were delivered only to ranked faculty and must have fallen accidently into the hands of these nontenure-track people. Consequently, sampling errors can be expected.

SUPPORT FOR RESEARCH

With respect to grant support, the data indicate that gender inequities have diminished over time. The percentage of women receiving institutional, federal, industrial, and other grant support rose over the four time periods. The data show that differences in the distribution of institutional grants between men and women found during the first two survey periods (1969 and 1975) ended in 1980 and 1988. Likewise, significant sex differences in the distribution of federal and industrial grants disappeared by 1988. For example, using the same mathematical technique employed in looking at changes over time for rank, in 1969 women had 62 percent as many federal grants as men did; in 1988 the number was 89 percent, no longer a significant difference. Gender differences for grants from foundation and state government sources remained significant -men received more, although here too the gap narrowed.

PUBLICATION

To begin with, past research shows that, on average, women publish less than do men.22 To account for the discrepancy, researchers claim organization and social-psychological differences. Women tend to work at other than research universities, receive fewer grants, are not in the natural sciences. Also, women have greater family responsibilities, less interest in research than in teaching, and less competitive personalities. In addition, the evidence for women’s publishing less is not infrequently contradictory. For example, Hamovitch and Morgenstern as well as Astin found no evidence that child rearing is related to the number of publications of academic women.23 Hargens, McCann, and Reskin found a negative relationship between fertility (number of children) and productivity, one that was true for both sexes.24

Turning to scholarly publication, our data show that two-year publication differences between men and women have narrowed considerably since 1969. As a result, by 1988 there was no significant difference in two-year publication rates between men and women, although men still published at a slightly higher rate—3.1 compared to 2.8 publications for women.

A narrowing trend in publication output by sex is evident in biology, psychology, and English. The largest publication gap remains in biology, where women averaged 2.6 publications in the past two years as compared with 3.1 for men. This difference, however, is not significant.

Past research has been mixed as to whether gender publication differences persist when other relevant correlates of productivity are taken into account. Blackburn, Behymer, and Hall reported that sex differences were not significant when factors such as institutional affiliation, preference for research, rank, and age were controlled via multivariate analysis. Yet other studies have found gender differences even after controlling for other productivity correlates.25

We conducted multivariate analyses to examine the relative importance of sex on faculty productivity while controlling other correlates. These analyses showed that gender is consistently the least important predictor of two-year publication rate when other variables such as age, rank, institutional affiliation by Carnegie type, and types of grant support were taken into account.26 Moreover, a comparison of partial beta coefficients, which provide a useful rank order importance of predictors, shows sex to be the least important predictor.

In addition, the total variance explained by our independent variables increased over the four time periods—from 24 percent in 1969 to 33 percent in 1988. Two reasons account for this increase. First, multivariate analyses showed a stronger negative relationship between age and publication output. Younger faculty became the most productive group at successive points in time. Second, a strong, positive relationship between Carnegie type and two-year publication rate is in operation. Productivity rose more at research and doctoral universities than at comprehensive institutions. Again, gender differences accounted for very little of the explained publication variance.

Last, we ran three additional, separate multivariate analyses for biology, psychology, and English faculty. Sex was less important than age, rank, Carnegie type, and most of the grant sources in predicting two-year publication output. Biology had the largest sex difference. A comparison of the adjusted two-year publication rates for women by discipline showed that female biologists averaged 2.47 publications in 1988 compared with 3.01 for male biologists. By contrast, in English men and women published at nearly the same rate in 1988.

In all, then, statistical analyses demonstrate that sex differences in two-year publication rates have narrowed to such a point that by 1988 they were no longer significantly different. Furthermore, when other correlates of faculty productivity are controlled, minimal sex differences are the result over the four time periods, including 1969, 1975, and 1980, when sex publication differences existed.

The recent higher publication rates for women are probably linked to the rise in the number of women at Research and Doctoral-I institutions where scholarly output is a high priority. The increased female productivity also can be tied to the more equitable distribution of grants over the past two decades.

SALARY

Salary equity has received appreciable attention in the press as individual and class-action suits are brought to court. However, the extent of female and male equivalences on a national basis and how changes over time have occurred are more difficult to come by. Since the 1988 survey did not collect income information, we are unable to present parallel data for salaries. This inability is especially unfortunate in light of the publication findings just presented. As Tuckman and Leahey have demonstrated, there is a high correlation between salary and number of articles published.27 On that basis, one predicts that the reported salary differences would now be eliminated—if everything were indeed equitable and fair.

Barbezat reports female/male salary differences for 1969, 1977, and 1984.28 Her 1969 data are from the same survey as ours. Her 1977 survey data are a follow-up to our 1975 data—the same instrument by the same people two years later. Her 1984 data are from the Carnegie Foundation for the Advancement of Teaching survey, in which the salary items are identical to the 1969 and 1977 surveys.

The news is bad. Controlling for key variables (productivity, rank, age, institution type) and using the Oaxaca procedure for estimating the proportion of salary difference that can be attributed to discimination, Barbezat finds “that roughly three quarters (74-76%) of the observed [23%] salary difference in 1968 [in favor of men] was the result of discrimination.” The situation in 1977 is still adverse, but somewhat improved. “The most conservative estimate for 1977 suggests that 29-38% of the 19% observed salary difference was the result of discrimination.” However, by 1984 the female/male salary differences had increased 2.5 to 21.47 percent. In addition, the difference that can be attributed to discrimination has risen to at least 36-43 percent.29 The prognosis of equity in 1988 is still a dream.

SUMMARY

The most evident sex difference continues in the differential status of men and women by rank.30 Despite gains over time in publication output, access to more faculty positions at elite research universities, and successful competition for grant support, women are still more likely to hold assistant professor positions. They are also less likely to have tenure than are their male colleagues. This overrepresentation of women as junior faculty underscores a concern that women may be especially susceptible to the increased publication pressures facing junior faculty. They are coming in, but their success rate is less than that of men.

Returning to our five questions, we see marked progress in terms of women working in Research-I universities, although that is still where the percentage is lowest. These universities are the most competitive and also have the reputation of being the most sexist. These are also the institutions that most strongly demand an uncompromised commitment to the advancement of knowledge in the disciplines, a time and career commitment women are less likely than men to make.

Disciplinary imbalances are still great despite increases in females Ph.D.‘s in mathematics and the sciences. The proportion of female faculty is not increasing in this area as rapidly is the number of Ph.D.‘s earned by women. The chilly climate in the sciences apparently shows few signs of defrosting.

The best explanation for the large proportion of women in the assistant professor rank remains their more recent arrival in the work force. As for women being overrepresented in nontenure-track positions, as indicated above, the data are not adequate to allow a definitive answer. There appear to be more women in these nonprofessorial positions. However, we have no data on whether this is a consequence of this being their only opportunity or whether for personal reasons they do not want to commit to a career path that places more demands on them than they are willing to make.

The notable achievement in grant equity is probably linked to affirmative action at the institutional level, as well as at foundations and government funding agencies. The sources can control the distribution of funds. Institutional grants, where women have had the most success, almost invariably carry with them the lowest number of dollars. Still, they provide seed money for generating grants, gathering data, and publishing. The strategy seems to be working.

Women’s accomplishments in publication show they do not lack the necessary knowledge and skills. Since promotions, and ultimately salary, are intimately linked to scholarly research, one would predict the disproportionalities and gap to end. However, a long track record of reward difference suggests they will not.31

Finally, while it also is unknown how much of the female gains can be attributed to affirmative action regulations, changes in laws and practices now going on in Washington require political attention. Harassment, tokenism, and sexism are still experienced. in academe. It is too soon for women to claim victory in their effort to attain equity and equal status in U.S. colleges and universities.

EPILOGUE

Unfortunately, monitoring women vis-à-vis men will need to continue—Tetreault’s bifocal phase cannot be completely abandoned. Retrogressions have occurred in the past. Some feminist progress has been made in both scholarship and curriculum. The gains, however, have been almost exclusively in the humanities and in research university settings. Twombly found little support for feminist activity in her community college literature survey.32 Also, one sees little activity in liberal arts colleges and comprehensive colleges and universities, the places where larger numbers of women are on the faculty.

Still, it is time for creative energy to be directed to the feminist phase. Succeeding at playing the man’s game by his rules is not enough. Furthermore, doing what men have been doing ad infinitum limits our understanding of faculty life. It also fails to enrich and deepen our disciplinary knowledge. Both require a woman’s perspective.

By way of illustration—in the sociological mode but also introducing psychological variables—Trautvetter and Blackburn compared female and male academic scientists’ article production and the factors that predicted their output.33 A key difference was self-competency, the belief that one possesses the necessary research skills. Women who expressed self-competence were publishers; those who did not published little.

Competency was not only a significantly stronger predictor for women than it was for men; it also was but one of two predictors for women whereas many variables accounted for publication by males. One infers from these findings that the data collected to account for publication output failed to capture critical variables that can explain female research publication. Trautvetter and Blackburn call for a study of female academic scientists using more of the variables that fall in Epstein’s psychological category—intellectual abilities, emotions, goals, intrinsic and extrinsic motivations, personal support, intention.34

Gumport has combined psychological and sociological factors to account for women’s surviving in academe and being productive scholars.35 Aisenberg and Harrington’s interviews with female academics, successful and otherwise, also combine sociological factors (the structure of the university and the location of power, for example) with the very intense psychological pull of power (“doing” vs. “are,” “ mind” vs. “body”) on private life, especially vis-à-vis marriage and children.36 Women face competing claims that are unresolvable for many. Feminine counter values call for an alternative system for the university, if not a revision. Neither change has a realistic future until the number of women grows large enough so they can exercise political power.

Meanwhile, some men are having their lives enriched by insights of women, insights they otherwise never would have had. Many may decry the extravagances that engage some Marxist theorists, deconstructionists, and feminists. Serious scholars, however, know that their views of phenomena on which they are “experts” no longer are the same after reading alternative perspectives. Feminists have produced insights and raised questions that cannot be casually dismissed or put aside.

Multicolored glasses enrich understanding for all of us. The bright side of a sometimes gloomy short-term prognosis for women (and other underrepresented groups) is that a foothold has been established, especially for deconstructionists and feminists, one that has to continue to expand as it illuminates the human condition for everyone.

We are indebted to Susan Twombly and Estela Bensimon (chair and critic respectively for the American Educational Research Association session at which we presented the interior part of this paper, Chicago 1991). They provided the challenging questions that led to parts of the Introduction and the Epilogue. We thank Lisa Mets for her critical reading of and editorial assistance on the revised paper.