Gender Differences in Academic Medicine Faculty Careers

Abstract

The researchers tracked the careers of U.S. medical school faculty who participated in the 1995 National Faculty Survey through 2012–13. Women were less likely than men to achieve full professor rank, remain in academia, and attain senior leadership roles. After adjusting for 1995 publication records, gender differences in promotion and retention were not statistically significant. However, the gap in senior leadership roles persisted even after this adjustment. The authors propose changing leadership selection processes and organizational culture alongside providing early career support.

Background

Reports that women’s career advancement in U.S. academic medicine lags behind men’s have recurred over the 20 years since Kaplan’s 1996 study of pediatrics faculty. Many prior studies used cross-sectional or retrospective designs or were limited to single institutions. These designs limited accounting for differences in time to promotion. Cross-sectional surveys excluded faculty who had already left academia.

The authors tracked gender differences in career outcomes over 17 years, examining promotion to full professor, academic retention, and attainment of senior leadership roles. Supported by the National Institutes of Health (NIH), the study examined the career status of participants in the 1995 National Faculty Survey in 2012–13.

Method

The researchers randomly selected 24 medical schools in the continental U.S. that had at least 200 faculty members, 50 female faculty, and 10 minority faculty in 1995. They balanced public and private institutions and included the four regions of the Association of American Medical Colleges (AAMC): Northeast, South, Midwest, and West.

From each school, they randomly sampled six faculty members from each of 24 strata defined by three graduation cohorts (pre-1970, 1970–1980, and post-1980), gender, and four specialty areas. They sampled all female faculty who graduated before 1970 and all underrepresented minority faculty. The mail survey received 1,801 responses, yielding a 60% response rate, and 74% of respondents agreed to follow-up contact.

Of the 1,335 respondents who agreed to follow-up, the researchers excluded 60 who had died and two for whom gender information was unavailable, yielding an analytic sample of 1,273. The follow-up group and the original sample were similar in gender, race, specialty, and publication counts by gender. In the 2012–13 academic year, the researchers contacted participants by email, phone, and mail and collected responses through online or paper surveys. Respondents received a small incentive.

The researchers supplemented missing survey responses by reviewing publicly available web data to identify affiliations, academic ranks, and leadership roles. They searched Google and institutional websites using names, past affiliations, and research interests. They checked NIH RePORT for federal grant funding over the previous two years. The institutional review boards of Boston University, Tufts Medical Center, and Massachusetts General Hospital (through an agreement with Tufts) approved the study.

Academic rank was categorized as full professor or other. Faculty working at or retired from universities, foundations, or government agencies were classified as retained in academia. Faculty who moved to private practice, industry, or other fields were classified as having left academia. Two researchers coded positions such as dean, associate dean, provost, and department chair as senior leadership roles.

The primary independent variable was gender. The covariates examined included race, specialty, years since first faculty appointment, the proportion of time spent on administrative, research, clinical, and teaching duties in 1995, marital/partner status, and the presence of children. The models for rank and senior leadership also considered affiliation with an academic institution in 2012–13. Productivity was measured as the cumulative number of peer-reviewed publications by 1995.

The researchers retained gender and race in the logistic regression models. They applied backward selection to variables with P < .10 in bivariate analyses, retaining those with P < .05. Model 1 excluded publication count, and Model 2 added publication count to Model 1. Analyses were conducted using SAS 9.4.

The NIH National Institute of General Medical Sciences and the Office of Research on Women’s Health (R01 GM088470) funded the study. The funders had no role in study design, data collection, analysis, interpretation, or the decision to approve publication.

Results

Survey responses were received from 607/1,273 participants (48%). By gender, 311/632 women (49%) and 296/641 men (46%) responded. Information was reportedly obtained from publicly available web data for 321/632 women (51%) and 345/641 men (54%).1The original Results section states that web data was obtained for 668 individuals, but the sum by gender is 666. The denominators and percentages for rank, retention, and senior leadership in the text also do not entirely match Table 3; therefore, the results from Table 3 are presented here.

Table 1. Characteristics of male and female faculty

Characteristic	Men (of 641): Number of respondents	Men: Value	Women (of 632): Number of respondents	Women: Value	P-value
Race, n (%)
White	639	499 (78.1)	631	520 (82.4)	.0533
Specialty, n (%)
Basic science	613	147 (24.0)	605	134 (22.1)	.1396
Primary care	613	156 (25.4)	605	176 (29.1)
Medical subspecialty	612	188 (30.7)	605	200 (33.1)
Surgical subspecialty	613	122 (19.9)	605	95 (15.7)
Years since first appointment, mean (SD)	633	29.4 (9.2)	616	28.3 (8.6)	.0196
Currently affiliated with an academic institution, n (%)	614	445 (72.5)	601	418 (69.6)	.2612
Marital status in 1995, n (%)
Has spouse or partner	635	557 (87.7)	626	460 (73.5)	<.0001
Presence of children in 1995, n (%)
1 or more children	637	534 (83.8)	628	432 (68.8)	<.0001
Workload proportion in 1995 (%), mean (SD)
Administration	637	18.6 (18.0)	628	18.8 (17.6)	.8307
Research	637	30.4 (29.9)	629	28.7 (29.3)	.3290
Clinical care	639	31.9 (28.9)	631	31.5 (28.7)	.7966
Teaching	638	19.1 (13.4)	630	21.0 (15.0)	.0166
Number of peer-reviewed publications by 1995, mean (SD)	602	33.5 (37.6)	604	22.2 (31.6)	<.0001

SD denotes standard deviation. “Currently affiliated with an academic institution” includes medical schools, government agencies, and foundations.

In Table 3, the proportion of full professors was 398/557 (71.5%) for men and 312/522 (59.8%) for women. In adjusted Model 1, the odds ratio for women attaining the rank of full professor was 0.57. In Model 2, which added publication count, the odds ratio changed to 0.77 and was not statistically significant.

Academic retention was 524/614 (85.3%) for men and 485/601 (80.7%) for women. After adjustment for race and specialty, the odds ratio for women remaining in academia was 0.68. Adding publication count changed it to 0.86, and the gender difference was no longer statistically significant.

Senior leadership roles were held by 137/640 men (21.4%) and 65/631 women (10.3%). After adjustment for publication records, the odds ratio for women attaining senior leadership was 0.49, indicating a persistent gender difference.

Table 2. Unadjusted and adjusted models

Outcome	Number of respondents	Unadjusted model: Odds ratio (95% CI)	Model 1: Odds ratio (95% CI)	Model 2: Odds ratio (95% CI)
Rank	998	0.59 (0.46, 0.77)	0.57 (0.43, 0.78)	0.77 (0.56, 1.08)
Academic retention	1,138	0.72 (0.53, 0.97)	0.68 (0.49, 0.94)	0.86 (0.61, 1.19)
Senior leadership	1,201	0.42 (0.31, 0.58)	0.44 (0.32, 0.61)	0.49 (0.35, 0.69)

CI denotes confidence interval. Odds ratios compare women with men. According to the original table notes, the final covariates for Model 1 were race, specialty, and years since appointment for rank; race and specialty for retention; and race for senior leadership. Model 2 added the 1995 publication count.

Table 3. Univariate results for male and female faculty

Univariate outcome	Value	Men n (%) (of 641)	Women n (%) (of 632)	P-value
Rank	Full professor	398/557 (71.5)	312/522 (59.8)	<.0001
Senior leadership	Senior role	137/640 (21.4)	65/631 (10.3)	<.0001
Academic retention	Retained	524/614 (85.3)	485/601 (80.7)	.03

Discussion

The authors interpreted early publication records as a factor explaining subsequent promotion and retention. Adjusting for the proportion of time spent on research, administration, teaching, and clinical care did not reduce the gender gap. When the 1995 publication count was included, the gender differences in promotion and retention were not statistically significant.

These findings differ from those of a 2015 study by Jena et al., in which rank differences remained after adjustment for productivity. A 2017 study of cardiology faculty found that women’s odds of becoming full professors were 37% lower than men’s. A prior single-institution study that included institutional committee chairs and national organization roles found that 56% of women and 70% of men held leadership positions. The current study limited its scope to deans, associate deans, provosts, and department chairs.

In interpreting the gender difference in early productivity, the authors discussed disparities in funding and administrative support, as well as household and caregiving burdens reported in prior research. They suggested that publication records might themselves represent a pathway through which career gaps develop. The retention measure grouped those who left voluntarily with those who left due to tenure policies. It could not distinguish between these pathways.

The pipeline theory, which posits that increasing the number of women will lead to more women in senior leadership, did not adequately explain the results for this cohort. The authors suggest that perceptions of women as collaborative and nonhierarchical leaders might disadvantage them in the hierarchical organizations of academic medicine. The discussion addressed cultures that recognize diverse leadership styles, multi-level interventions to address stereotypes, and support for women’s leadership self-efficacy. The authors also reviewed previous findings that female deans often served at less research-intensive schools and had shorter tenures.

Specific strategies for career support included equitable support for faculty establishing their careers, mentoring, and promotion criteria that recognize teaching and clinical academic activities. A prior report analyzing the 1995 data showed that women received less mentoring. The authors propose sharing educational outcomes through platforms such as the AAMC’s MedEd PORTAL and recognizing academic outputs such as curricula.

Because this study tracked faculty who were active before 1995, it does not capture the experiences of early-career faculty who entered academia later. Limitations included a low survey response rate and incomplete coverage of institutions. By supplementing survey responses with publicly available web data, the researchers obtained information for 98% of eligible participants. Agreement between self-reports and online data did not differ by gender. The achievements of women who changed their surnames might have been undercounted. The authors identified the national sample and the 17-year follow-up period as strengths of the study.

Conclusions

The authors propose strengthening mentoring and academic support during women’s early careers. To address the gap in senior leadership, they recommend changes in organizational culture and selection processes to recognize women’s leadership competencies and academic achievements.

Reading notes

Because productivity was measured by the number of peer-reviewed articles, the adjusted results are best interpreted within this scope of academic achievement. The authors’ proposal to recognize educational outputs such as curriculum development in promotion decisions raises a question about which achievements evaluations capture. The findings on senior leadership should be interpreted within the 17-year careers of faculty appointed before 1995 and the roles included in the study: dean, associate dean, provost, and department chair.

References

Carr, P. L., Raj, A., Kaplan, S. E., Terrin, N., Breeze, J. L., & Freund, K. M. (2018). Gender differences in academic medicine: Retention, rank, and leadership comparisons from the National Faculty Survey. Academic Medicine, 93(11), 1694–1699. https://doi.org/10.1097/ACM.0000000000002146