Despite increased attention to gender disparities in academic promotion over the past decade, women are still substantially underrepresented in the highest academic ranks.1–3 Academic productivity is often used as a criterion for the selection of leaders, and therefore disparities in productivity may help explain why few women are in leadership roles.1,4 As a result, few female academic physicians are available to provide gender-concordant role modeling and mentoring for more junior female colleagues.5,6
Gender disparities in academic productivity have been explained by observations that women are more likely than men (1) to work part-time,7–10 (2) to dedicate a greater proportion of their professional effort to teaching and patient care as opposed to research,11,12 (3) to shoulder a greater proportion of family responsibilities,13–15 (4) to receive gender-discordant role modeling and mentoring,5,6,16,17 (5) to perceive gender-related obstacles to academic advancement in the work environment,17 and (6) to hold personal values incongruent with traditional measures of academic success.12,18,19 However, most studies examining gender differences in academic productivity are cross-sectional and do not examine progress over the full span of academic careers. Whether disparities in academic productivity persist throughout careers is unknown.
Therefore, we aimed to compare the publication records, academic promotion, and leadership appointments of women and men academic physicians over the span of their entire careers using a longitudinal cohort design. These analyses will help determine whether disparities in academic productivity between genders persist throughout careers. On the basis of prior studies showing that women may prioritize family responsibilities over academic goals during the early phases of their careers,13–15 we hypothesized that men would publish more than women early in their careers but that publication rates may be similar for both genders in the later stages of academic careers. Yet, on the basis of data demonstrating consistent gender disparities in leadership roles,1,4 we hypothesized that a reduction in the gap in publication rates between genders would not attenuate inequity in leadership appointments.
In 2007, we conducted a retrospective, longitudinal cohort study of the publication records, rates of academic promotion, and leadership appointments of women and men academic physicians at Mayo Clinic. The Mayo Clinic institutional review board approved this study.
We included all women scholarly clinicians who were, at the time, employees of Mayo Clinic with 20 or more years of service at Mayo Clinic. We defined scholarly clinicians as physicians who spend >25% of their professional effort directly providing patient care. We excluded clinician investigators and physician scientists, defined as physicians who dedicate ≥75% of their professional effort to research. We categorized physicians by career track based on their initial appointments. For each woman participant, we identified two male physicians, matched by appointment date and career track as scholarly clinicians, to serve as controls.
Mayo Clinic is a not-for-profit academic health center with a nontenured system. Continued employment and salary are disassociated from academic rank, and Mayo Clinic does not provide financial incentives for publishing, achieving academic promotion, or being selected for most leadership appointments.
We identified physicians using an institutional database containing years of service of physicians. We used an institutional database containing physicians' curricula vitae to determine the following:
* the number and date (year) of peer-reviewed publications,
* the dates (years) of academic promotion to assistant professor, associate professor, and professor, and
* the total number of appointments to divisional, departmental, institutional, and national leadership roles for each participant.
Divisional and departmental leadership roles included division or department chair, vice chair, or associate chair. We did not include lesser roles such as head of a section within a division. We defined an institutional leadership position as dean, associate dean, or chair of a prominent institutional committee. We defined a national leadership role as president or chair of a national organization such as a scientific society or foundation.
We summarized continuous variables (years of service and total publications) using the mean and standard deviation (SD) for each group, and we made comparisons between groups using linear regression with a random effect for matched sets. In addition, we graphed the mean publications by years of service for men and women. We used a linear regression model to compare the publications per year, with a random effect for matched sets and repeated measures. This model included gender, years of service, an indicator variable for >27 years of service, and an interaction variable between gender and >27 years of service. We selected this model because the graph of mean publications by years of service seemed to show a reversal of the trend by gender after 27 years of service.
We summarized categorical variables (highest academic rank achieved and highest leadership position attained) using the number (percentage) in each group, and we made comparisons between groups with polytomous logistic regression with a random effect for matched sets and with robust standard errors. We summarized time-to-event data (years to assistant professor, associate professor, and full professor) using the mean and SD for those who achieved the academic rank; we made comparisons between groups using Cox regression with a random effect for matched sets. We also calculated Kaplan–Meier curves for time to associate and full professor. We compared the number of leadership positions held at the divisional, departmental, institutional, and national levels between men and women using a Poisson regression (for event data) with robust standard errors and a random effect for matched control. We adjusted the number of leadership positions held for the years of service by including a natural log offset in those models. We considered a two-tailed P < .05 to be statistically significant for all analyses. We analyzed data using SAS 9.1 (SAS Institute, Cary, North Carolina).
We included all 25 women physician scholars then employed at Mayo Clinic who had 20 or more years of experience at Mayo; we identified 50 men who, matched by appointment date and career track as scholarly clinicians, served as controls. Women physicians in this cohort practiced at Mayo Clinic a mean of 26.6 (SD: 4.2, range: 22–37) years, and men practiced at Mayo a mean of 26.6 (SD: 3.2, range 22–34) years. Seven women and 13 men had each, at the time of the study, provided more than 27 years of service at Mayo Clinic.
The women in this cohort published a mean (SD) of 29.5 (28.8) and a median (interquartile range [IQR]) of 18 (8, 52) publications over the span of their careers compared with the men, who published a mean (SD) of 75.8 (60.3) and median (IQR) of 59 (26, 112) publications (P = .001) over the span of their careers (Table 1). The range of publications by individuals was 2 to 106 among women and 1 to 303 among men.
Figure 1 shows the mean publications for women and men according to years of service. In our study, women produced a mean of 1.94 fewer publications than men per year throughout the first 27 years of service (P = .001, Table 2). However, after 27 years of service, mean publications by women increased to 2.72 publications per year compared with a mean of 1.15 publications per year by men (P = .03), such that women produced a mean of 1.57 more publications per year than men in this time period (Table 2). The difference in the mean publications for women compared with men increased by 3.51 publications per year from −1.94 for ≤27 years of service to 1.57 for >27 years of service (P < .001 for the interaction between gender and more than 27 years of service, Table 2). Yet, women produced 1.74 fewer publications per year than men on average over their entire careers (P < .001). We found no associations between academic productivity and total career duration.
In our study, men achieved higher academic rank over the span of their careers compared with women (P = .01, Table 1). Throughout their entire careers, 38 men (76%) achieved the rank of associate professor and 33 men (66%) achieved the rank of professor compared with 10 (40%) and 7 (28%) women, respectively (P = .01). The mean (SD) years to promotion to associate professor was 10.1 (5.6) years for women and 9.3 (4.1) years for men (P = .052). Men achieved the academic rank of professor in a mean of 16.3 (5.8) years compared with 19 (5.1) years among women (P = .090). We compared the time required for women and men to attain the academic ranks of associate professor and professor using Kaplan–Meier curves (Figure 2).
Throughout the span of their academic careers, the women in this cohort held fewer total leadership roles than the men (P < .001, Figure 3). Nearly half (44%) of women physicians (no. = 11) did not attain a divisional, departmental, institutional, or national leadership position during the span of their careers, compared with 30% of men (no. = 15). Just one woman in the sample held a single departmental leadership position compared with 22 departmental positions held among 20 men. Seven national leadership positions were held among 5 women, compared with 15 national positions among 10 men.
This retrospective, longitudinal cohort study examining the academic productivity, promotion, and leadership appointments of women and men academic physicians over the span of their careers demonstrates that women publish less, advance through academic ranks more slowly, and ultimately achieve lower academic rank and fewer leadership appointments compared with men. These findings align with previous cross-sectional studies indicating that publication productivity and academic promotion are lower among women.1,2 These data are unique, however, in that the nontenured system of the Mayo Clinic allows examination of a longitudinal cohort that is not selected on the basis of academic productivity (unlike in systems in which academic productivity is often an important criterion for tenure).
Because Mayo Clinic is a nontenured academic health center where salary is dissociated from academic rank, all physicians in this cohort were equally likely to be retained regardless of academic productivity. Examination of this unique cohort across more than 20 years reveals that the publication productivity of women “catches up” to and even exceeds that of men in the latter stages of academic careers. To our knowledge, this finding has not been previously reported in the literature. Our findings have important implications: academic health centers with tenure systems that include financial incentives for academic rank may lose female physicians with less academic productivity early in their careers even though, as our findings suggest, these women may become more productive in the latter stages of their careers.
Despite increases in publication rates among women toward the ends of their careers, the women in our study did not achieve parity with men in attaining leadership positions. In this study, many leadership appointments occurred at 10 to 20 years of service, a period when, for this sample, the publication productivity of women was half that of men. Because academic productivity is often a criterion for the selection of leaders, otherwise qualified females may not be chosen on this basis. A paucity of qualified women in leadership positions both deprives organizations of the unique skills and perspectives women bring to such roles20,21 and suggests that criteria other than publication record should be considered in the selection of leaders.
This study has several limitations. First, it occurred at a single institution; thus, the findings may not be generalizable to other academic health centers. Second, the duration of careers varied among physicians, so fewer participants represented the latter years of an academic career; however, we noted no association between career duration and productivity. Third, we did not examine other variables (e.g., marital status, number of children, personal values) that may explain discrepancies in academic productivity between genders. Fourth, although we matched physicians by career track (e.g., scholarly clinician) and excluded those who worked part-time, we did not examine the allocation of physicians' professional effort in patient care, teaching, and research, which fluctuates over the span of careers and cannot be accurately obtained retrospectively. Finally, we could not account for the possibility that physicians may have changed career tracks (e.g., from scholarly clinician to career scientist) over time.
These limitations notwithstanding, the findings from this longitudinal cohort study indicate that although the academic productivity of women lags behind men in the early and middle stages of careers, publication rates are similar between genders in the later stages of academic careers. Thus, academic productivity measured in midcareer may not be an appropriate measure of leadership skills, and factors other than publication record and academic rank should be considered in the selection of leaders. Academic health centers should recognize the disparate trends in scholarly productivity and promotion between men and women and develop new mechanisms to identify qualified female leaders.
The Mayo Clinic institutional review board approved this study.
This work was presented at the 31st Annual Meeting of the Society of General Internal Medicine.
2Ash AS, Carr PL, Goldstein R, Friedman RH. Compensation and advancement of women in academic medicine: Is there equity? Ann Intern Med. 2004;141:205–212.
3Nonnemaker L. Women physicians in academic medicine: New insights from cohort studies. N Engl J Med. 2000;342:399–405.
4Hager M, ed. Women and Medicine. New York, NY: Josiah Macy, Jr. Foundation; 2007.
5Mayer AP, Files JA, Ko MG, Blair JE. Academic advancement of women in medicine: Do socialized gender differences have a role in mentoring? Mayo Clin Proc. 2008;83:204–207.
6Levinson W, Kaufman K, Clark B, Tolle SW. Mentors and role models for women in academic medicine. West J Med. 1991;154:423–426.
7Uhlenberg P, Cooney TM. Male and female physicians: Family and career comparisons. Soc Sci Med. 1990;30:373–378.
8Shanafelt TD, West CP, Sloan JA, et al. Career fit and burnout among academic faculty. Arch Intern Med. 2009;169:990–995.
9Levinson W, Kaufman K, Bickel J. Part-time faculty in academic medicine: Present status and future challenges. Ann Intern Med. 1993;119:220–225.
11Shea S, Nickerson KG, Tenenbaum J, et al. Compensation to a department of medicine and its faculty members for the teaching of medical students and house staff. N Engl J Med. 1996;334:162–167.
13Carr PL, Ash AS, Friedman RH, et al. Relation of family responsibilities and gender to the productivity and career satisfaction of medical faculty. Ann Intern Med. 1998;129:532–538.
14Levinson W, Tolle SW, Lewis C. Women in academic medicine. Combining career and family. N Engl J Med. 1989;321:1511–1517.
15Warde C, Allen W, Gelberg L. Physician role conflict and resulting career changes. Gender and generational differences. J Gen Intern Med. 1996;11:729–735.
16Bright CM, Duefield CA, Stone VE. Perceived barriers and biases in the medical education experience by gender and race. J Natl Med Assoc. 1998;90:681–688.
18Pololi L, Kern DE, Carr P, Conrad P, Knight S. The culture of academic medicine: Faculty perceptions of the lack of alignment between individual and institutional values. J Gen Intern Med. 2009;24:1289–1295.
19Wright SM, Beasley BW. Motivating factors for academic physicians within departments of medicine. Mayo Clin Proc. 2004;79:1145–1150.
20Levinson W, Lurie N. When most doctors are women: What lies ahead? Ann Intern Med. 2004;141:471–474.
© 2011 Association of American Medical Colleges
21Eagly AH, Johannesen-Schmidt MC, van Engen ML. Transformational, transactional, and laissez-faire leadership styles: A meta-analysis comparing women and men. Psychol Bull. 2003;129:569–591.