Investigators studying issues of gender equity in medical and other academic settings have found that, compared with male faculty members, female faculty members hold fewer senior-level, tenured positions,1–3 have a slower rate of advancement3–5 and a higher rate of attrition (a reflection of job satisfaction),2 and receive lower pay.1,6,7 According to the Association of American Medical Colleges (AAMC), although nearly 50% of medical school applicants and graduates are female, women constitute less than 30% of medical school faculty, less than 20% of full professors, and less than 15% of department chairs.2,8,9 Further, the percentage of women among first and senior authors of original research and editorials in leading medical journals is significantly lower than that of men10,11; similarly, women are less likely than men to receive major career funding grants.12 The potential effects of gender inequity also extend to national and international health workforces.13
Although the Department of Pediatrics at the University of Colorado School of Medicine (UCSOM) has grown substantially in size over the past decade and our overall male-to-female ratio among faculty has approached 50/50, a disproportionately high percentage of senior-level and tenured faculty are male. In this study, we sought to define the extent of that disparity and to identify its causes. We tested two potential hypotheses: (1) The department has treated women and men unequally and has failed to adequately advance the careers of female faculty, and (2) our female faculty entered academic medicine more recently than our male faculty and have just begun to catch up in seniority, tenure, and leadership roles. We also considered the likelihood that these two hypotheses were not mutually exclusive. We believe that the approach we used to study gender equity in our department can serve as a template for similar efforts by other departments in academic medicine. After detailing our findings, we share the preliminary results of interventions in our department that target the inequities we identified.
In 2009, we evaluated five parameters to assess gender equity in the UCSOM Department of Pediatrics: promotion, tenure, leadership roles, faculty retention, and salary. One of us (D.M.) gathered data from two departmental faculty databases: the Faculty Information Database Online (FIDO), a Web-based software application designed and maintained exclusively by our department that contains data reflecting each faculty member's career since joining the faculty and is used for faculty assessment, mentoring, and promotion purposes; and PeopleSoft (Oracle Corporation; Redwood Shores, Calif), a software program for employment and salary records for which data from 2000 to 2009 were available. D.M. coded and deidentified the data by assigning random-number identifiers to each faculty member. The rest of us used the deidentified data in the analyses described below. The Colorado Multiple Institutional Review Board determined that federal regulations governing human subject research did not apply to this project.
In 2009, our department's faculty included 543 members, 263 of whom were on the “promotion track” (i.e., assistant, associate, or full professors); the remaining 280 faculty (instructors or senior instructors) were not included in this study. Whereas most of our promotion track faculty are located on the UCSOM's Anschutz Medical Campus, a smaller but significant number are based at one of two affiliated institutions—National Jewish Health or Denver Health. We did not include the 34 faculty members at these institutions in our evaluations of tenure or salary because UCSOM rules specify that they are not tenure eligible, and their salaries are determined by their institutions rather than by the department.
For all of our analyses, we only included faculty at ≥0.5 full-time equivalent (FTE) employment in 2009. For purposes of calculating time to promotion and time to tenure, we “corrected” to 1.0 FTE those faculty who worked part-time (≥0.5 FTE) at any time during 2000–2009 (the period for which PeopleSoft records were available). For example, if a faculty member was at 0.6 FTE for 3 of 6 years since her appointment as an assistant professor, we recorded her duration at that rank as 4.8 years (3 full years plus 3 years at 0.6 FTE). We compared men and women on time to promotion, time at rank, time to tenure, and time with tenure using two-tailed t tests, and we compared rates of tenured faculty among eligible professors by gender using chi-square analyses, using PASW Statistics 18, version 18.0.0 (SPSS, Inc., Chicago, Illinois). We considered a P value <.05 to be significant.
We defined leadership positions as medical directors, section heads, and vice chairs. We determined faculty attrition by assessing all departing faculty, by gender and rank, from 2000 to 2009. Because the department did not conduct exit interviews during this period, the exact reasons for the departures could not be identified; we could only determine whether the departing faculty member resigned/relocated or retired. We conducted a separate analysis of attrition for the two years from January 1, 2008 through December 31, 2009 for purposes of comparison with available national data.
We performed the salary equity analysis using salaries for academic year 2008–2009. As noted above, for this portion of the study, we excluded the 34 faculty at affiliated institutions because their salaries are set by their institution rather than by the department. We also excluded the 71 non-MD faculty members in the promotion track for our salary analysis because of their disparate job descriptions and the lack of national salary benchmarks. We corrected the salaries of MD faculty at <1.0 FTE employment to what they would have been had the faculty members been working full-time. For example, if an MD faculty member earned $100,000 at 0.75 FTE, we calculated his or her salary as $133,333 ($100,000 divided by 0.75). We then matched each MD faculty member's “base plus stipend” (excluding incentives and bonuses) salary to the national median base plus stipend salary (excluding incentives and bonuses), as reported annually by the Association of Administrators in Academic Pediatrics (AAAP),14 using the three demographic categories used by the AAAP (rank, years at rank [0–5 years, 6–10 years, and 11+ years], and subspecialty). We then compared each individual faculty member's salary with the median salaries of individuals with the same demographic characteristics. For example, an assistant professor at 4 years in infectious diseases would be compared with the national median salary for assistant professors of 0 to 5 years with infectious diseases subspecialties. We expressed each faculty member's salary as a percentage of the national median and determined the percentage of our faculty who fell below, within, and above a range of 95% to 105% of the national median.
To increase precision, we performed a second salary assessment using multiple regression analyses, with gender, leadership role (section head and/or medical director), and years at rank calculated on a continuous scale (rather than the AAAP bracketing) as predictor variables. The continuous scale reflected the actual years at rank, calculated to a single decimal point and corrected (as above) for any years at <1.0 FTE. This allowed us to distinguish, for example, an associate professor at 6.2 years from another at 10.7 years at rank, although we then compared both faculty members' salaries with the AAAP national median for associate professors at that rank for 6 to 11 years because those are the only national benchmarking data available. We did not separately compare salaries of male and female medical directors, section heads, or vice chairs because of the wide range of job responsibilities within each category.
To assess the actual dollar value discrepancies in faculty salaries by gender compared with national medians, we chose representative subspecialties and determined median salaries at all academic ranks and all years at rank. We then calculated the dollar value of a variation from those representative salaries. For example, if an assistant professor in infectious diseases who had three years at rank earned 98% of the national median of $113,500 for that faculty member's profile, the 2% shortfall was $2,270 in annual income.
Of 543 faculty in the pediatrics department in 2009, 263 (48%) were on the promotion track and were included in our study. Of these 263 promotion track faculty, whom we will refer to as “current” faculty, 118 (45%) were women: 60 of 112 (54%) assistant professors, 39 of 70 (56%) associate professors, and 19 of 81 (23%) professors.
Table 1 shows the time to promotion and years spent at each rank. We found no significant differences between male and female faculty in average time to promotion from assistant professor to associate professor for all current associate professors; years at rank for current associate professors; or time to promotion from associate professor to professor for all current professors. However, among current professors who had also been promoted to associate professor at our institution, the average time for that earlier promotion was one year longer for women than for men. Current male professors had held that rank nearly twice as long as female professors, reflecting the relatively recent ascent of more women to senior rank at our institution.
Of the 50 tenured faculty members in our sample, 9 (18%) were women, a reflection in part of the disparity between the numbers of female and male full professors. Among professors eligible for tenure (i.e., all promotion track faculty not at an affiliated institution), 9 of 18 (50%) women were tenured whereas 39 of 52 (75%) men were tenured (χ2(1) = 3.88, P < .05).
In 1997, a change in UCSOM tenure rules increased the stringency of criteria; since then, tenure has been awarded mostly to full professors. Male faculty received 25 of 28 (89%) pre-1997 awards of tenure; only 2 of those 28 faculty, both men, received tenure after promotion to professor. Since 1997, tenure has been awarded to 22 faculty members (16 men, 6 women). More of these faculty received tenure following promotion to professor—6 men and 3 women—and 3 declined to be considered for tenure (1 man, 2 women). Since 1997, UCSOM has promoted 10 women and 15 men in the pediatrics department to professor; 2 of these women (20%) and 5 of these men (33%) received tenure with that promotion. Overall, two-thirds of the tenured male faculty members (25/41; 61%) received tenure prior to 1997, whereas two-thirds of the tenured female faculty members (6/9; 67%) received tenure after 1997.
In 2009, women held 13 of 21 (62%) medical directorships, whereas men held 18 of 24 (75%) section head positions and 6 of 7 (86%) vice chair positions. Medical directors are clinical administrative positions, whereas section heads and vice chairs are academic administrative posts. The 6 female section heads had served in that role for an average of 5.5 years—2 (33%) of them for more than 10 years (10.3 and 15.4 years) and the others for less than 5 years. The 18 male section heads had served an average 12.3 years—10 (56%) of them for longer than 10 years (range: 10.3–31.3 years) and 3 of them for less than 5 years. All 7 vice chairs had held their positions for at least 10 years (range: 10–18 years).
Between 2000 and 2009, 34 women and 33 men left the department (Table 2). All 31 of the departing assistant professors (17 women, 14 men) resigned, leaving academia or relocating to other academic institutions. Of the 10 departing female associate professors, 3 retired and 7 resigned/relocated; all 8 departing male associate professors resigned/relocated. Among departing full professors, 4 of the 7 female professors resigned/relocated compared with 5 of the 11 male professors; the other 3 female and 6 male professors retired. Because the data regarding departing faculty are aggregate for the nine-year time period, we could not determine yearly totals.
For comparison with national faculty retention data, we looked separately at faculty departures during January 1, 2008 through December 31, 2009: 10 faculty members left the department (3 women [30% of departing faculty], 7 men [70% of departing faculty]). Our data are comparable to national faculty retention data—36% of departing faculty at U.S. medical schools nationally during the same time period were women.2 During this two-year period, 5 departing faculty members were assistant professors (4 men, 1 woman), 2 were associate professors (1 man, 1 woman), and 3 were professors (2 men, 1 woman).
Of the 263 promotion track faculty in our overall sample, 105 were excluded from the salary analysis because they were either at an affiliated institution (34) or were non-MD faculty (71). The 158 MD promotion track faculty included in our salary analyses consisted of 67 women and 91 men. At all faculty ranks, more women (48/67 [72%]) than men (46/91 [51%]) received pay below the AAAP national median, when matched for the three AAAP demographic categories—rank, years at rank (0–5 years, 6–10 years, and 11+ years), and subspecialty (Table 3). A higher percentage of women than men received pay at less than 95% of the national median (28/67 [42%] versus 32/91 [35%]), and a higher percentage of men than women were paid more than 105% of the national median (31/91 [34%] versus 12/67 [18%]).
In a secondary assessment using multivariate analysis, controlling for years at rank on a continuous scale and leadership role, our female MD faculty received on average 98% of the national median salary compared with male MD faculty, who received 105% (P < .05). In a separate, subset analysis of the salaries of MD faculty members in leadership roles, female leaders received pay at an average of 93% of the national median for all faculty of similar rank, years at rank, and subspecialty (the AAAP does not separately report salaries for faculty in leadership roles) compared with male leaders at 112%.
Using AAAP actual dollar values for salaries, depending on subspecialty, the potential gap between the salaries of the women (98% of the national median using our multivariate analysis noted above) and men (105% of the median) in our department ranged from $8,000 for “low-earning specialties” to $11,000 for “high-earning specialties” for assistant professors; $9,500 to $14,500 for associate professors; and $12,500 to $16,500 for professors.
Gender equity studies in numerous academic settings, including medical schools,6,15,16 have examined specific aspects of career achievements. To the best of our knowledge, previous studies have not specifically examined academic pediatrics departments, wherein the common perception is that women have traditionally had more substantial opportunities for advancement than in other clinical departments. Furthermore, we are not aware of any previous studies in any academic medical department that examined all five metrics of equity—promotion, tenure, leadership, salary, and retention—that we evaluated here.
We found that both of the hypotheses we tested were true, at least in part: Women in the UCSOM Department of Pediatrics belatedly entered academics and have just begun to catch up to men in seniority and tenure. However, our department failed to treat men and women equally in certain areas, thereby possibly failing to adequately advance the careers of female faculty. Whereas current male and female associate professors spent the same number of years as assistant professors, and current male and female professors spent the same number of years as associate professors, male professors had been at their current rank for an average of six years longer than their female counterparts had at the time of the study, and men had held tenure an average of six years longer than women had. Both of these observations speak to the later start women had in joining our faculty. Statistics kept by the AAMC confirm the late entry of women into academic medicine nationally.8,9 Additionally, current male professors in our department were promoted from assistant professor to associate professor (the promotion prior to their most recent promotion to professor) almost a year sooner than were current female professors during that earlier promotion stage.
In 1998–1999, women constituted 44% of accepted applicants to U.S. medical schools, a percentage that rose to 48% by 2008–2009.2 In contrast, only 34% of assistant professors, 23% of associate professors, and 11% of professors in U.S. medical schools were female in 1998–1999; by 2008–2009, the percentage of women at each rank had increased only modestly to 41%, 30%, and 18%, respectively.2 Pediatrics has fared somewhat better nationally, with women constituting 56% of assistant professors, 42% of associate professors, and 27% of professors at U.S. medical schools in 2009.2
In this study, only 18% of our tenured faculty members were women, a disparity attributable in large part, but not entirely, to pre-1997 UCSOM practices. Tenure awards to our department's male faculty prior to 1997 exceeded those to women by a ratio of nearly 8 to 1; since 1997, more men than women have received tenure, but the ratio by 2009 was less than 3 to 1. This disparity also results from post-1997 UCSOM rules that preferentially award tenure either with or following a faculty member's promotion to professor, and there were fewer female than male professors in our department. However, female associate professors now outnumber male associate professors in our department, and the time to promotion now is equal between the genders; the combination of those two factors should result in increasing numbers of female professors and a resultant increase in tenured female faculty.
In 2008–2009, only 20% of tenured faculty at U.S. medical schools were women,2 which is comparable to the 18% among our department's tenured faculty. In academic settings outside medicine, the American Association of University Professors reports that women constitute nearly 50% of tenured faculty at community colleges but only 33% at colleges that award master's and baccalaureate degrees, and 25% at universities that award doctoral degrees.17
Our pediatrics department has a disparity among academic leadership roles held by men and women, with most section head and vice chair positions being held by men. At the time of our study, our vice chairs had held their titles for 10 to 18 years and our section heads had held their titles for an average of 10.6 years, reflecting appointments that date back to a time when there were many fewer female faculty members in contention for those positions. Our department's statistics parallel national demographics in that only 21% of vice chairs and 21% of section head/division chief positions in U.S. medical schools are held by women.2 The literature suggests that causes of the gender disparity in leadership positions include a lack of mentorship, sexism, unconscious bias, and the “constraints of traditional gender roles,” including personal choices in priority setting.3,18–20 Among the faculty appointed as section heads in our department within the five years leading up to and including our 2009 study, 4 of 7 were women. Thirteen of 21 medical directorships were held by women. It is unclear, however, whether women in our department are disproportionately offered these clinical administrative positions or if they are more likely than men to accept them.
We observed no gender differences in the numbers or percentages of departing assistant professors; this is similar to national observations in which attrition rates are 43% for both male and female assistant professors across 10-year study periods.21 However, more women than men in our department retired as associate professors, and more female than male professors resigned or relocated. The gender imbalance among professors on our faculty (62 men versus 19 women) in 2009 implies an even greater imbalance in prior years when there were fewer female professors. As a result, the resignation/relocation of four female professors from 2000 to 2009 is disproportionate compared with that of five male professors over the same period. The retirement of three women, but no men, at the associate professor level raises additional concerns regarding the department's promotion and retention of female faculty. Because neither the pediatrics department nor UCSOM conducted faculty exit interviews, we cannot determine whether inequitable efforts at faculty retention or support for career advancement played a role in the retirement of female associate professors or the resignation/relocation of female professors. In national surveys, female faculty express lower job satisfaction than do their male counterparts,22,23 which may explain the poor retention rates. During 2008–2009, 30% of our departing faculty members were women. This statistic compares with national data indicating that 36% of departing faculty at U.S. medical schools were women2 during the same time period.
Our findings show that our department remunerated women less than their male colleagues at all ranks in 2008–2009. Nationally, similar disparities have been identified in medical schools1,6,7 and in colleges and universities.7,17 The U.S. Census Bureau reported that in the general workforce in 2008, women earned 77 cents for every dollar earned by men.24 Our analysis was limited somewhat by the AAAP's groupings of years at rank14; however, our multiple approaches to assessing our faculty salaries make it clear that corrections for underpaid female faculty are necessary.
Our study identified numerous gender inequities in our large department of pediatrics and described a method by which other academic medicine departments can assess such characteristics. However, an important limitation of our study is that we did not identify why these inequities developed or persisted. The underlying reasons for disparities in awards of tenure, leadership positions, attrition rates, and salary remain to be determined. Possibilities include unconscious bias in recruitment,19 acceptance of publications,10,11 and awards of funding,12 as well as differences in productivity due to work–life balance issues,7 constraints of traditional gender roles, sexism in the medical environment, and lack of effective mentors.18 Future studies should methodically evaluate these potential root causes of gender inequity in academic medicine. One recently published study found that productivity, as measured by publications, of female faculty lagged behind that of men at midcareer when leadership positions are typically determined,25 but that, later in their careers, women published at a higher rate than men did. The authors suggest that using midcareer productivity to recruit academic leaders may hamper women from achieving these leadership positions.
Moving forward: Using our findings to target gender inequity
On the basis of the results of our assessment of gender equity in the pediatrics department, we designed an intervention strategy, summarized in Table 4, for which preliminary results are available. First, the promotions process required no specific intervention. Second, although most tenure disparities predate 1997, fewer women than men have been hired with tenure or awarded tenure simultaneously with promotion to professor in recent years. The department has now instituted proactive efforts to recruit senior female faculty with tenure and will refer all faculty being considered for promotion to professor to the tenure committee for simultaneous consideration. Next, men hold most vice chair and section head positions because of appointments that occurred many years ago. The department will undertake a review of vice chair roles and will emphasize searches for senior female subspecialists as section head positions become available. Furthermore, the department will explore the reasons why more women than men hold medical directorships and the comparative leadership and professional growth opportunities of each leadership position. Finally, at the first opportunity for salary adjustments following this study, the department used the complete dataset of individual faculty salaries compared with their national medians to correct inequitable salaries, and the department will use the same process annually.
Our department is joining with Children's Hospital Colorado to design exit interviews for all departing faculty. Additionally, the same FIDO database that we used extensively in this study facilitates annual career development interactions with each faculty member and may help with real-time intervention in the event that gender equity concerns arise for an individual. This system was not in place before the retirements of the three female associate professors and the resignation/relocation of the four female professors.
Since the implementation of our intervention strategy in 2010, we have made progress on several fronts. At the two annual departmental promotions committee meetings that have occurred since we completed this study, the pediatrics department has nominated 13 associate professors for promotion to professor (10 women, 3 men) and 5 professors for tenure consideration (4 women, 1 man). Three newly recruited senior faculty at the professor level have joined our department (1 woman, 2 men). We also have hired 2 new section heads (1 woman, 1 man). A female vice chair was recruited to replace a retiring female vice chair.
The salary adjustments that followed the completion of this study decreased the percentage of our faculty members who are paid below the national median; among female faculty, salary improvements were the most marked: The percentages of female faculty below the national median fell by double digits at all academic ranks (23% for assistant professors, 26% for associate professors, and 13% for professors). Following salary adjustments, we repeated our multivariate analysis of the same MD faculty members included in our initial analysis and found that the average female faculty salary increased from 98% to 105% of the national median, and the average male faculty salary increased from an average of 105% to 109% of the national median. The difference in salaries between men and women were no longer statistically significant after the salary adjustments.
In this study, we identified the need for change in departmental policies for advancing the careers of women and described interventional strategies that we have undertaken to date to counter the disparities that we found. These are comparable to the perceived needs and recommendations made by others.26,27 We will continue to closely monitor the effectiveness of our intervention strategies. And we recommend that our methodology serve as a template for other academic medicine departments in their assessment of gender equity.
The authors gratefully acknowledge Dr. Mimi Glode, Dr. Georgeanna Klingensmith, Dr. Amy Brooks-Kayal, and Dr. Nancy Krebs for helpful advice in study design and analysis. Dr. Peg Christon designed the FIDO system and provided invaluable access to data. The authors also thank Jennifer Ackerman for data support, Dr. Edward Garrido for statistical back-up, and Martha Huckaby for secretarial assistance.
The Colorado Multiple Institutional Review Board determined that federal regulations governing human subject research did not apply to this project.