Minority Faculty and Academic Rank in Medicine

Context.— Previous studies have found that fewer minority medical school faculty hold senior professorial ranks than do majority faculty and may not be promoted as rapidly.

Objective.— To determine whether minority faculty were as likely as majority faculty to have attained senior rank (associate professor or full professor) after adjusting for other factors that typically influence promotion.

Design.— A self-administered mailed survey of US medical school faculty using the Association of American Medical Colleges database. The sample was stratified by department, graduation cohort, and sex.

Participants.— A stratified random sample of 3013 full-time faculty at 24 representative US medical schools. All underrepresented minority faculty at these schools were sampled.

Main Outcome Measure.— Attainment of senior academic rank (associate professor or full professor).

Results.— Of 3013 faculty surveyed, 1807 (60.0%) responded, including 1463 white (81.0%), 154 black (8.5%), 136 Asian (7.5%), and 54 Hispanic (3.0%). Overall, 980 faculty (54%) had attained senior academic rank, including 47 (30.5%) of 154 black faculty, 59 (43.4%) of 136 Asian faculty, 22 (40.8%) of 54 Hispanic faculty, and 852 (58.3%) of 1463 white faculty. White faculty had significantly more first-authored and total peer-reviewed publications than the other groups. After adjusting for the medical school, department, years as medical school faculty, number of peer-reviewed publications, receipt of research grant funding, proportion of time in clinical activities, sex, and tenure status, we found that the odds ratios of holding senior rank relative to white faculty were 0.33 (95% confidence interval [CI], 0.17-0.63) for black faculty, 0.36 (95% CI, 0.12-1.08) for Hispanic faculty, and 0.58 (95% CI, 0.30-1.12) for Asian faculty.

Conclusions.— Minority faculty were less likely than white faculty to hold senior academic rank. This finding was not explained by potential confounders such as years as a faculty member or measures of academic productivity.

RECENT EFFORTS to improve the representation of minority faculty in academic medicine have focused on increasing the number of minority physicians who pursue academic careers.^1,2 However, the number of minority students entering medical schools has plateaued, despite efforts to achieve racial and ethnic diversity in US medical schools, such as the Association of American Medical Colleges (AAMC) Project 3000 by 2000.^1-3 Minority medical students and resident physicians in academic medical centers have few role models.⁴ Data derived from the AAMC Faculty Roster System show that only 3.9% of all faculty identify themselves as black, Native American, Mexican American, or Puerto Rican.⁵ The AAMC classifies these groups as "underrepresented" in medicine because the proportion of physicians in these groups is less than in the general population.⁶ Other minority groups, such as Asian Americans, are not classified as underrepresented.

In 1989, Petersdorf et al⁶ reported that underrepresented minority faculty were far less likely to have achieved the rank of associate or full professor than majority faculty. This difference may have been partly due to minority faculty being, on average, younger. They also noted that minorities were not promoted to associate professor rank as quickly as majority faculty.

Literature on promotion and tenure includes opinion pieces regarding the merit of tenure,^7-9 articles that explore the process of promotion for medical school faculty,^10,11 and reports on the promotion of women faculty in academic medicine.^12-14 The AAMC has reported on the promotion experience of minority medical school faculty⁶ but, due to limitations of the AAMC database, was not able to control for other important factors that affect promotion, such as academic productivity. In the current study, we examined whether minority faculty were as likely as majority faculty to have attained senior rank (associate or full professor) after controlling for such factors.

In 1995 we used a 2-stage sampling plan to draw a stratified random sample of full-time salaried faculty of US medical schools. Of the 126 medical schools in the AAMC's 1994 lists, we excluded the 6 schools outside the contiguous United States because the AAMC considers them to be significantly different from mainland schools. Also, to make it possible to select about 150 faculty (including similar numbers of women and men and adequate numbers of minorities) from each sampled school, we excluded 14 small schools (ie, having fewer than 200 faculty members). Retained schools had faculties of at least 200, containing at least 50 women and 10 minority faculty. Altogether, only 5% of medical school faculty are affiliated with the small schools. We randomly selected 24 schools from the remaining 106 eligible schools. Numbers of faculty at the sampled schools were similar to those in this population of eligible schools. This sample also achieved regional diversity across the 4 AAMC regions and a balance of public and private institutions.

In the second stage we created a sampling frame from the full-time salaried faculty listed by the AAMC Faculty Roster System at the 24 selected medical schools exclusive of 720 faculty (4%) in unique departments (such as veterinary medicine) that did not exist at other medical schools. We identified our survey sample from the 16714 faculty members in the sampling frame, first through a 2 × 24 × 4 × 3 factorial design that sought 6 randomly selected male and female faculty members from each sampled school in each of 4 academic department groupings (generalists vs surgical specialists vs medical and other specialists vs basic scientists) in each graduation cohort (those who completed their first doctorate before 1970 vs between 1970 and 1980 vs after 1980). The oldest graduation cohort cells were filled first and back-filled, if necessary, with younger faculty. We expanded the sample to include all underrepresented minority, generalist, and senior women faculty at these 24 schools. The total sample included 4051 faculty members.

The questionnaires were mailed to faculty at their professional addresses; 1038 were ineligible because they had left their institution, were not full-time faculty, or had died. Nonrespondents among the remaining 3013 subjects were mailed a reminder postcard and, if necessary, had a follow-up telephone call and were sent a second questionnaire. We assigned a final project disposition to every participant through contact records, which contained dates of the mailings and telephone reminder calls and the final disposition.

The survey instrument examined various aspects of academic life among medical school faculty. The survey was pretested by 45 medical school faculty at 3 institutions to ensure that the respondents understood the meaning of the questions and could answer them appropriately. Many questions were taken from previously published studies.^15,16 Thirty questions used in the survey were developed by Linda Fried, MD, MPH, and Clair Francomano, MD, Johns Hopkins Medical Institutions, Baltimore, Md, and were used in the original form or modified for use in our instrument with permission. The self-administered survey included 177 questions about faculty demographics; professional goals and work situation; current academic environment and rank; mentoring relationships; experiences with bias, discrimination, and harassment; academic productivity; family responsibilities; faculty compensation; and career satisfaction. Items examining attitudes used Likert scales.¹⁷ Other items required circling 1 item within a group of items or recording a number or percentage. This study was approved by the Boston University School of Medicine Institutional Review Board.

We coded departments as follows: generalists (general internal medicine, general pediatrics, family medicine, and geriatrics); surgical specialists (general surgery and its subspecialties); medical and other specialists (internal medicine subspecialties, pediatric subspecialties, neurology, physical medicine, radiology, emergency medicine, anesthesia, and psychiatry); and basic scientists. For analysis, we created 4 racial and ethnic groups from the self-reported ethnic and racial categories in the questionnaire: white, not of Hispanic origin; black, not of Hispanic origin; Asian; and Hispanic.

We used descriptive statistics to characterize the majority and minority survey respondents because the distribution of characteristics of surveyed white faculty is a function of our factorial sampling design, whereas the proportion of the other ethnic groups reflects the results of full census sampling in the 24 study schools.

We used conditional logistic regression¹⁸ to determine the likelihood of minority faculty holding senior rank (associate or full professor) compared with white faculty within the strata defined by departments within medical schools. We adjusted for the following potential confounders: the number of years as a medical school faculty member (continuous) and years as medical school faculty, squared, to capture the declining influence of additional years on the likelihood of senior rank; the number of first-authored, peer-reviewed publications (<5 vs 5-9 vs ≥10); current research grant support (yes or no); sex; percentage of clinical time; and tenure status (tenured or tenure track vs nontenured).

We also classified respondents as white, underrepresented minority (black, Puerto Rican, and Mexican American), and nonunderrepresented minority (Asian and other Hispanic, such as Cuban)⁶ and replicated the previous analysis for underrepresented and nonunderrepresented minority groups relative to white faculty, adjusting for the same covariates. We tested for the presence of interaction terms for ethnicity by sex, tenure status, school, region of school, and private vs public school. Analyses were performed using SAS statistical software, Version 6.11 (SAS Institute Inc, Cary, NC).

Of 3013 eligible faculty members who received the survey, 1807 returned it for a response rate of 60%. Due to confidentiality concerns of the AAMC, we are not able to calculate response rates within racial and ethnic subgroups. Eighty-one percent identified themselves as white, not of Hispanic origin; 8.5% as black, not of Hispanic origin; 7.5% as Asian; and 3.0% as Hispanic. There were fewer women among black and Hispanic faculty in our sample (Table 1). Fewer black faculty reported having board certification and postgraduate degrees. Black faculty were less likely to be in a basic science department than other faculty. At the time of the survey, 980 respondents (54%) reported senior faculty positions (associate professor or full professor), including 47 (30.5%) of 154 black faculty, 59 (43.4%) of 136 Asian faculty, 22 (40.8%) of 54 Hispanic faculty, and 852 (58.3%) of 1463 white faculty. More white faculty than others were tenured or on a tenure track, and they were also more likely than other faculty to have attained senior rank regardless of their tenure status.

Many of the minority faculty were not US born, particularly Asians, for whom 80% were born outside the United States (Table 2). For Asian and Hispanic faculty, being foreign-born was associated with a primary language other than English. All groups rated their aspirations to become full professors similarly. Only 20% of the faculty respondents felt their rank was lower than other faculty with equivalent accomplishment and experience at their institution. The reasons these individuals attributed for their lower rank were similar across groups, except that more black and Asian faculty reported ethnic or racial bias and more Hispanic and Asian faculty reported inadequate negotiation skills.

All groups reported similar hours worked during a typical workweek (Table 3). Black faculty spent more time in clinical activities and less time in research relative to other faculty. More black and Hispanic faculty felt pressure to serve on committees due to their race or ethnicity, but they did not spend more time in hours per month on committee-related activities than the other groups. More white faculty held research grants during the previous 2 years than Hispanic faculty, but there were no differences in the types of grant funding (governmental vs private vs industry) or the median number of research grants held between the groups. White faculty also had more first-authored and total peer-reviewed publications than the other groups.

After adjustment for measures of academic productivity (publications and grants) and other factors that affect attaining senior rank, minority faculty were less likely to have been promoted to associate or full professor compared with white faculty (Table 4). Compared with white faculty, the adjusted odds ratios (ORs) of having senior rank for black faculty were 0.33 (95% confidence interval [CI], 0.17-0.63); for Hispanic faculty, 0.36 (95% CI, 0.12-1.08); and for Asian faculty, 0.58 (95% CI, 0.30-1.12). These results were unchanged whether first-authored or total publications were used, either continuously or categorically defined, or whether years as a faculty member was coded as a continuous or categorical variable. Adjusting for the same variables, we found that underrepresented minority faculty were significantly less likely to have been promoted to senior academic rank compared with white faculty. The adjusted OR for having attained senior rank for these underrepresented minority faculty was 0.29 (95% CI, 0.16-0.54), and for nonunderrepresented minority faculty, the OR was 0.64 (95% CI, 0.34-1.20). In a similar multivariate model with rank of full professor as the outcome variable (data not shown), the estimated effect for each racial or ethnic group continued to show a lower likelihood of holding senior rank compared with white faculty. These effects were not as large as they were in the main analysis and the statistically significant effect for black faculty was less (P=.05 vs P=.007, primary analysis). The influence of each racial or ethnic group on the likelihood of attaining senior rank was similar by sex, tenure status, location of the medical school, or public or private status of the medical school.

Our study found large racial and ethnic disparities in the attainment of senior rank among US medical school faculty. Black faculty were significantly less likely than white faculty to hold senior rank. Moreover, this disparity persisted after adjustment for potential confounders. Hispanic and Asian faculty were also much less likely than white faculty to have attained senior rank, although these differences were smaller and not statistically significant (P=.07 and P=.10, respectively). Underrepresented minority faculty, who were predominantly black and Mexican American in our sample, were less than a third as likely than majority faculty to have attained the rank of associate or full professor.

Difficulties with the advancement of underrepresented minority faculty have been noted by Petersdorf et al.⁶ Their study also relied on the AAMC database but did not have the ability to adjust for potentially confounding factors. To our knowledge, our study is the first to comprehensively address the issue of promotion among minority faculty in all medical school departments in a representative, national sample of US medical schools. The underrepresentation of minority faculty at senior ranks is not explained by their younger age. After controlling for the number of years of medical school faculty appointment, these faculty remained substantially less likely to have attained senior rank than white faculty.

Underrepresented minority faculty have been reported to have a greater debt burden than other faculty,^1,6,19 which may partly explain the greater clinical activity and lower levels of research time among black and Hispanic faculty in our study. Economic factors may make it more difficult for these faculty to participate in research activities and publish in peer-reviewed journals. In addition, more black and Hispanic faculty may be on clinical tracks, in which promotion to senior levels is slower. However, after controlling for the percentage of time spent in clinical work, black faculty were still less likely to hold senior rank.

All of the groups rated their aspiration to be a full professor and the likelihood of becoming a full professor similarly, so that the observed shortfall in attaining senior rank cannot be attributed to lesser ambition. Negotiation skills have been recognized as important for career development and advancement.²⁰ Yet, we found little overall differences in the self-reported assessments of negotiation skills by racial and ethnic group. Nevertheless, among the one fifth of faculty who thought they were at a lower rank relative to similarly prepared and accomplished peers, more Hispanic and Asian faculty attributed their lower rank to inadequate negotiation skills. This may be due to differences in communication styles and cultural differences, since higher proportions of faculty from these 2 groups were born outside the United States.

Reasons to explain the underrepresentation of minority faculty in senior ranks are unclear. Discrimination, which may be less obvious than in the past, is a possibility.²¹ The stereotypes of minority groups that permeate society at large may carry over into academic medicine. Cultural differences may contribute to some minority faculty being excluded from certain opportunities or informal information sharing that could be helpful to their career development. Cultural differences and historical factors also may make some minority faculty reluctant to network at the divisional or departmental level, thus reducing their chance to forge personal and professional relationships with majority colleagues.

A potential limitation of our study is that the faculty productivity measures were self-reported. However, we see no reason to suspect systematic discrepancies between actual and reported numbers of publication by race or ethnicity. We also could not evaluate the quality of the publications or other academic activities of our respondents, but there is no evidence that minority faculty publish articles of lesser quality. Our measure of academic performance was research based and, thus, did not capture the achievements of faculty in other areas, including teaching and administration. Nevertheless, on the basis of time commitment to these latter 2 activities, white faculty were similar to the largest minority group, black faculty. Another potential concern about our study is response bias. Because of confidentiality concerns of the AAMC, we cannot compare the characteristics of respondents and nonrespondents by race and ethnicity. However, the proportion of minority faculty with senior rank in our sample (30% for black faculty, 43% for Asian faculty, and 41% for Hispanic faculty) is similar to that recorded in the AAMC Faculty Roster (28% for black faculty, 38% for Asian faculty, and 39% for Hispanic faculty).⁵ Thus, black faculty at lower ranks did not disproportionately respond to our questionnaire. We found that 58% of white faculty had senior rank in contrast with the 52% recorded by the AAMC, but this difference most likely is due to our oversampling of senior faculty and should not affect the findings of our multivariate analysis. We could not adjust for all factors that affect attaining senior rank that may differ among racial and ethnic groups, such as the different tenure tracks that are in place at various medical schools. Also, our results cannot be generalized to historically black medical schools because they were not included in our study. Finally, given that this was a cross-sectional survey and not an inception cohort, we have no data from individuals who have left academic medicine. It is likely that data from those who remain at academic centers underestimate the negative effect of being in a minority group on the chance of attaining senior rank.²²

Our findings suggest a number of possible interventions. First, the research careers of minority faculty, particularly black and Hispanic faculty, should be encouraged and supported financially because more research activities should increase the likelihood of promotion.²³ Second, the promotion criteria at each institution should be evaluated to ensure that teaching and administrative activities are sufficiently rewarded. This would include greater value placed on scholarly activities other than publication, such as excellence in teaching, the performance of administrative responsibilities, and service to the community served by the medical school. Third, external reviewers blinded to the race and ethnicity of the faculty member should evaluate the candidate's curriculum vitae for promotion to limit the potential for ethnic and racial bias. Fourth, medical schools should disseminate their promotion and tenure procedures, offer formal career counseling and faculty development programs, and establish formal monitoring processes for their faculty as recommended by the American College of Physicians.¹⁹ In addition, qualitative studies may have a role in identifying the factors minority faculty perceive as barriers to their advancement and in suggesting effective interventions. More minority faculty in senior positions would help provide role models and mentors to minority junior faculty, resident physicians, and medical students.

Armed with the results of our study and the reality of an increasingly diverse US population, medical school deans and department heads need to foster and provide greater support for the careers of minority faculty to ensure their equitable representation at all levels in academic medicine.