Foster, Sharon W. PhD; McMurray, Julia E. MD; Linzer, Mark MD; Leavitt, Judith W. PhD; Rosenberg, Marjorie PhD; Carnes, Molly MD
It is easier to overcome a series of known obstacles than tilt at a series of shadowy spectres.—MARY ROTH WALSH, 19771
Dr. Walsh, a medical historian, was commenting on the physician training requirements, formalized in the late 19th century, that offered women clear criteria by which to gain entry into the medical profession. And indeed, since that time, women have entered medicine in vast numbers. But some obstacles remain shadowy spectres: why, at the beginning of the 21st century, have we not seen a proportionate rise in the number of women in positions of leadership in academic medicine? Although women comprise 44% of the students currently entering U.S. medical schools and up to 64% of the physicians in some residency programs,2 they constitute only 27% of U.S. medical school full-time faculty, and the proportion of women in leadership roles has remained quite stable for more than 15 years.2
Women are not alone at this threshold. Academic medicine has sought to involve groups that have not traditionally been incorporated by initiating local, regional, and national efforts to improve career development and leadership opportunities for both women and underrepresented minorities. Such initiatives may be made more effective by taking into account the perceptions of those within the nontraditional groups regarding what changes are needed.
Previous work suggests that several factors may interfere with women medical school faculty members' potential to succeed, including inadequate mentoring,3 isolation and less access to informal networks,4 salary inequity,5 disproportionate care-giving and household responsibilities,6,7 and both subtle and overt gender biases.8,9 The Council on Graduate Medical Education (COGME) concluded that gender bias and harassment are the greatest obstacles preventing women physicians from achieving their full potential.10
In 1996, the University of Wisconsin (UW) Medical School's Faculty Equity and Diversity Committee (FEDC) issued a report showing that the proportion of women in the medical school's faculty (18.4%) was notably lower than it was nationally (26%). This discrepancy suggested that the academic climate at our institution affected women differently than men. To understand the “known obstacles” at the school, we conducted a survey in the spring of 1997 to obtain data on the faculty's perceptions of the work environment and to identify differences in the ways women and men perceived career obstacles. The goal was to establish baseline data for evaluating future interventions. Our hypotheses were that, compared with men, women would be more likely (1) to report awareness of informal networking that, despite the availability of formal mentoring, acts to exclude certain faculty from career advancement, (2) to experience difficulties in the professional environment, such as not being sought for collaborative research, not being included in informal decision making, and being subjected to denigration or harassment, (3) to face obstacles related to family responsibilities, and (4) to consider leaving academic medicine.
Sample and Study Design
We adapted a questionnaire developed by Fried and colleagues for their survey of a single department of medicine.11 We added (1) an item on appointment track, (2) an item on any history of change of track, and (3) several items pertaining to departments and divisions (as we were surveying all UW Medical School faculty members in 25 departments). The instrument consisted of 115 items. The UW Institutional Review Board (IRB) approved the survey with the stipulation that we not analyze data by department, as the small cell sizes would risk violating confidentiality. Accordingly, we categorized the departments into five clusters: basic science, non-surgical subspecialties, primary care, surgical specialties, and non-clinical.
The survey, administered in 1997 by the Wisconsin Survey Research Laboratory, was accompanied by a letter of introduction and a description of provisions for ensuring confidentiality. It was sent to all medical school faculty. Non-respondents received two follow-up mailings. Survey responses were tabulated by gender, track, rank, and departmental cluster.
Using Fisher's exact two-tailed test, we assessed gender differences overall and within track, rank, and departmental cluster. A p value < .05 defined a significant difference in the percentage of men versus the percentage of women who agreed or disagreed (or said yes or no) to a question. For items with Likert scales, we combined “strongly agree” with “agree” and “disagree” with “strongly disagree.” For statistical testing, “don't know/not applicable” responses (except where specifically noted) were considered to be missing; thus, the denominator we used to determine the percentage of respondents who agreed or disagreed with an item excluded the “don't know/not applicable” responses.
Demographics of Respondents
Questionnaires were sent to all 836 UW Medical School faculty (tenure, clinician-educator, and clinical tracks); of those, 507 (61%) responded. The numbers of evaluable questionnaires varied by gender, track, rank, and cluster because some respondents did not respond to one or more of the items: 497 respondents provided gender, while 489 provided both rank and gender.
Table 1 compares respondents and non-respondents by gender, track, and rank. The last column shows the response rates by faculty category. Seventy-one percent of the women responded to the survey, compared with 55% of the men. Only 21% of the clinical faculty responded. The percentages by rank are fairly close to the overall response rate of 61%. The confidential nature of the survey precluded any further analysis of data on the non-respondents, so we base our analyses only on those who filled out the questionnaire; we do not extrapolate results to the whole faculty.
Perceptions of the Academic Climate
Mentoring. Of the respondents who were assistant professors, 75% of the men and 69% of the women reported that they had mentors. Both women and men of all ranks indicated satisfaction with their mentors' facilitation of their professional development and review of their progress toward promotion. There was no significant difference by gender. Women were more likely to be mentored by men than were men by women, with 100% of women full professors, 74% of women associate professors, and 65% of women assistant professors having or having had men as mentors. In contrast, 5% or less of men faculty at all ranks had women as mentors.
Networking. When asked “Are you aware of informal networking which systematically (even if not purposely) excludes faculty members on the basis of gender,” 24% of the women compared with only 6% of men who responded to this item said yes (p < .001).
Professional environment. Sixteen items on the questionnaire addressed components of the professional environment at UW Medical School. The women and men faculty differed significantly on 14 of them (Table 2). Many of these differences were generalizable across rank, track, and departmental clusters. Specifically, more women felt that (1) the professional environment in the medical school as a whole, as well as in departments and divisions, was less supportive of them, (2) men faculty with comparable expertise were more likely to be sought for collaborative research efforts, (3) men had difficulty taking seriously the careers of women faculty, (4) collegial relationships were more difficult between faculty of different genders, (5) there was a greater likelihood of seeking consultations from someone of the same gender, (6) women were more likely to have been used to further their mentors' careers, (7) women felt they were less likely to be put up for promotion early, (8) full and equal participation in problem solving and decision making was not available to them, (9) they were not welcomed members of the scientific community, (10) men faculty had denigrated women faculty based on gender, and (11) gender bias issues were difficult to raise with colleagues. All of the differences were statistically significant at the p < .001 level except (9) (p < .003).
We assessed the denigration of women faculty by asking about perception, actual observation, and direct experience. Thirty percent of the women perceived that men faculty denigrated women based on their gender, while 26% of the women (compared with 9% of the men; p < .001) reported that they had “observed situations in which a woman has been denigrated by male colleagues based on gender.” When asked “have you been harassed sexually during your time at the University of Wisconsin?”, 13% of the women versus 4% of the men answered yes (p = .001). Forty-six percent of the women and 22% of the men indicated they did not feel comfortable raising issues regarding gender bias with colleagues (p < .001).
Obstacles to academic success. Our fourth area of interest was in the faculty's perceptions of obstacles to academic success. They checked all that applied from a list of six potential obstacles (Table 3). We found that (1) women perceived obstacles at a rate two to three times that of men, and (2) both women and men perceived the primary obstacle to be meetings held at 7 AM, after 5 PM, and on weekends (42% of women, 21% of men, p < .001).
Desire to leave or stay in academic medicine. Of the 479 faculty who responded to the item regarding leaving academic medicine, similar proportions of women (45%) and men (43%) indicated that they had seriously considered leaving, but their reasons differed (Table 4). The women most commonly cited conflicts between work and family, while the men cited uncompetitive salary. Women cited work-family conflicts and job stress significantly more frequently (p < .05).
When asked “Do you want to be in academic medicine ten years from now?”, 52% of the faculty responded yes, 14% no, and 34% don't know. Of the 187 faculty who indicated that they had seriously considered leaving academia, 36% stated nonetheless that they wanted to still be in academic medicine in ten years. These faculty represent a group “at risk” for not achieving academic medicine goals. Identifying such faculty groups helps target interventions.
Impacts of Track, Rank, and Departmental Cluster
Gender differences in perceptions of the five components of climate persisted across track, rank, and departmental cluster. Responses of faculty differed most pervasively between women and men by track, with significant differences (p < .01) on 16 of 31 comparisons. By rank, associate professors also showed widespread gender differences, with significant findings in 14 comparisons. By departmental cluster, primary care faculty showed significant gender differences on 15 of 31 comparisons, followed by basic science faculty (nine of 31 comparisons). Faculty who do not feel welcomed as members of the UW scientific community included 57% of the women and 37% of the men primary care faculty (Figure 1) and 54% of the women and 36% of the men clinician-educator faculty (Figure 2). Also of note are the women (53%) and men (46%) associate professors who reported that “conflicting family responsibilities have contributed to… thoughts of leaving academic medicine.” The climate survey thus identified issues for faculty of both genders.
We were not able to cross-tabulate non-respondents, as we were not able to simultaneously describe individuals by track, rank, and departmental cluster, due to the likelihood of small cell sizes and the confidential nature of the survey. We were thus not able to develop a model predicting the non-responders. The percentages of respondents by both rank and track (except clinical faculty) were comparable to those in the faculty as a whole. A higher percentage of women than men responded. The data are based on 71% of women faculty responding. Even if all non-respondent women faculty perceived the climate, for example, as equally supportive of women as of men, the data would suggest that 42% of all (respondent and non-respondent) women faculty perceived the climate as “less supportive of women than of men.” These results are informative, even without comparison to non-responders or men faculty.
In this survey of the medical school faculty at a major midwestern university, women and men differed considerably in their perceptions of several aspects of gender climate: networking, professional environment, obstacles to academic advancement, and reasons for considering leaving academic medicine. While the women and the men had equal access to and similar experiences with mentoring, the women were more likely to perceive the existence of an informal network that excluded faculty on the basis of gender. They were less satisfied with many aspects of the working climate that ranged from feeling less welcomed by the scientific community and less likely to feel an equal participant in informal decision making to perceiving gender-based denigration and harassment. Meetings outside regular working hours were concerns for both men and women, but work-family conflicts were a greater concern for women. Although the medical literature on faculty issues often focuses on mentoring, our survey suggests that having a mentor does not mitigate the effects of institutional structures that lead to isolation, stress, strained collegial relationships, and work-family conflicts.
In addition to gender, other factors that may impact faculty perceptions of the academic climate were considered. Regardless of gender, faculty on the tenure track were more likely to feel welcomed as members of the scientific community, but they also were more likely to feel like outsiders, and had concerns regarding funding for research, uncompetitive salaries, and isolation. Clinical faculty were less likely to feel welcomed by the scientific community, but had fewer child care concerns. Compared with either clinical or tenure-track faculty, clinician-educators appeared to have more concerns about support, research collaboration, and clinical supportiveness of colleagues, as well as issues related to balance of work and family. Issues related to whether or not faculty felt included in research collaboration or promotion equity were important to all departmental cluster groups, with the primary care group demonstrating concerns more than basic scientists and subspecialty clinicians. For surgical specialties, sexual harassment and lack of inclusion in informal networks were salient issues. Thus, there were important faculty concerns about the current climate in academic medicine that varied by track, rank, and department cluster as well as by gender. At the same time, when we looked separately at the impact of gender within track, rank, or department cluster, we found significant gender differences.
One limitation of our research is that we were not able to develop a model of how the non-responders might have responded, given the IRB constraints. However, the data are based on 71% of all women faculty. Even if all non-respondent women faculty perceived no gender climate problem, our findings still apply to the majority of women faculty at the University of Wisconsin Medical School. Some perceptions of adverse gender climate, therefore, are so widely held that the views of non-respondents, even if divergent, would not change the summary findings.
A second possible limitation of our results is that the data reflect faculty perceptions. It is reasonable to question the degree to which these perceptions accurately reflect the actual work environment. Although we did not directly study gender differences in salary equity, space, or staff support, such inequalities in allocation of institutional resources have been documented at other institutions.11,12 Moreover, perceptions can guide actions. Because the rate at which women enter academic medicine continues to be lower than the rate for men, women medical students carefully watch women faculty to gauge their own reasonable expectations for success. As academic medicine seeks to diversify faculty to serve an increasingly diverse society, it is critical that we identify factors that lead current and future women faculty to feel welcome in the profession. Finally, a robust research literature on social cognition has detailed how perceptions may maintain or change the status quo. Fiske noted that individuals selectively pay attention to certain types of information about others, based on their goals. Her research suggests that the situation that changes what information one pays attention to about others, thereby decreasing the potential for stereotyping, is when there is something at stake that requires inter-dependence and a demonstrable outcome.13 Because academic medicine seeks to diversify the physician work-force, specifically by encouraging greater numbers of women and underrepresented minorities to join the profession, increasing the involvement and success of these groups is a shared goal in which we are common stakeholders. To accomplish this goal, everyone's perceptions need to be considered carefully.
Based on this survey, we made a number of recommendations to the dean of the UW Medical School. These were to: (1) increase women's networking opportunities via a “First Fridays” conference series, to which local and national women leaders would be invited to make presentations; (2) develop a UW Medical School Faculty Mentoring Program, with ties to the campus Women Faculty Mentoring Program; (3) hire an ombudsperson to track and address climate issues and other faculty concerns, such as denigration, harassment, and misuse of intellectual property; (4) focus the policy development role of the Faculty Equity and Diversity Committee; (5) develop a gender climate video for use within the medical school; (6) provide professional development seminars for all faculty; (7) name a special assistant to the dean on gender to independently track gender issues and initiate further efforts; (8) hold fewer 7 AM, 5 PM, and weekend meetings; (9) develop resources for both routine and unanticipated child care; (10) assess gender equity in compensation; (11) preserve opportunities for faculty track changes; and (12) require that department chairs report annually to the dean regarding progress in improving gender climate. The dean and other institutional leadership groups accepted these recommendations.
The viability of academic medical centers depends on the survival of their faculty. The rapid drop in the number of physician-scientists, heralded as a national crisis,14 has occurred concurrent with the increasing number of women physicians.2,4 Our study and others suggest reasons for this crisis: many women may choose not to become physician-scientists due to inadequate mentoring, a lack of role models, and an unwelcome academic climate. The Women Physicians' Health Study9 provides evidence of additional obstacles, noting that of 3,655 women physicians, 48% reported having experienced gender-based harassment and 37%, sexual harassment. Indeed, given the current gender distribution of U.S. medical students, it will be crucial to invest in efforts to increase the number of physician-scientists in ways that pay careful attention to the existence of gender bias in academic medicine and its damaging effects on career success.15 Our women faculty reported persistent perceptions of denigration of women and harassment. Eradicating harassment and denigration of faculty will be crucial for the survival of academic medical centers as the number of women in the ranks of medicine increases. It is crucially important that attention be paid to the work environment regarding collegiality, adequate resources, leadership, mentoring, and equitable salary structures to lead to successful and stable faculty.
The process of conducting and discussing the survey itself contributed substantially to the forward momentum of both women faculty's development and the development of a cadre of individuals interested in and committed to women's health. During this period of time, UW earned a designation as a National Center of Excellence in Women's Health, and development of a Women's Health and Women's Health Research Center was named a strategic priority by the medical school. The above recommendations have earned the ongoing support of our dean. What were once shadowy spectres are now known obstacles—obstacles that we hope to continue to overcome.