Underrepresented minority faculty in the USA face a double standard in promotion and tenure decisions

The academic community in the USA significantly underrepresents Black and Hispanic faculty compared to their proportion in the general population. This underrepresentation negatively impacts equitable access to academic careers and hinders the quality and breadth of research. Diverse teams produce more innovative research, and a lack of scholars from underrepresented groups leads to less focus on issues affecting their communities. While various barriers contribute to this underrepresentation, this paper focuses on the potential for bias in promotion and tenure (P&T) decisions. Past research has shown bias in admissions, peer review, grant funding, and hiring processes. P&T decisions are critical advancement steps, affecting a scholar's career trajectory, yet there is a lack of quantitative research on the role of bias in these decisions. This study hypothesizes that racial disparities extend to the P&T process, particularly for those with intersectional identities (e.g., URM women), and investigates how external review letters (ERLs) can mitigate these disparities.

Existing literature demonstrates the pervasive nature of bias against underrepresented minorities (URMs) in academia, affecting various stages from admissions to grant applications and peer review. Studies highlight the impact of gender bias on women in academia, but fewer have examined the interaction of race and gender in P&T decisions. The concept of shifting standards suggests that minority group members are held to different standards than majority group members, often being judged more harshly for similar levels of performance. In the P&T process, research productivity, often measured by metrics like the h-index, is a crucial factor. Previous research indicates that scholarly productivity is weighted heavily in P&T decisions, potentially leading to biased evaluations of URM faculty. The study also considers intersectionality, acknowledging how race and gender intersect to create unique challenges, particularly for URM women, who may face a “double bind.” The role of external review letters (ERLs) in mitigating bias is explored, building on existing suggestions for writing identity-affirming letters.

This multi-institutional study analyzed data from five US research-intensive universities, encompassing 1,571 P&T decisions (2015-2022). The data included voting outcomes at the department and college levels, candidate characteristics (race/ethnicity, gender, discipline, rank, years in current rank, number of external grants), and linguistic analyses of external review letters (ERLs) using LIWC software. The key dependent variables were the percentage of negative votes received and whether a candidate received a unanimous positive vote. Independent variables included URM status, gender, h-index (a measure of scholarly productivity), and the presence of scholarship-related language in ERLs. Ordinary least squares (OLS) regression and logistic regression were used to analyze the relationships between variables, examining direct and interaction effects. Control variables included institution, discipline, university ranking, rank sought, number of grants, and STEM status. Multilevel modeling was used for analyzing ERL data, accounting for nested data structure (ERLs within candidates).

The analysis revealed that URM faculty received 7% more negative votes at the college level and were 44% less likely to receive unanimous votes than non-URM faculty. This disparity was particularly pronounced at the promotion to associate professor level. There was evidence of a double standard in evaluating scholarly productivity. URM faculty, especially URM women, with below-average h-indexes received more negative votes and fewer unanimous votes than non-URM faculty with similar h-indexes. The interaction between URM status and h-index was significant for both negative vote percentage (β = -0.34, p = 0.044) and unanimous votes (OR = 4.24, p = 0.003). The negative impact of lower h-index was significantly greater for URM women than other groups. The inclusion of scholarship-related language in ERLs significantly mitigated the negative impact on URM women's voting outcomes, reducing the negative vote percentage (β = -0.22, p = 0.007) and increasing the likelihood of unanimous votes (OR = 15.67, p = 0.005). This effect held even when controlling for other linguistic features of ERLs and candidate productivity metrics.

The findings demonstrate that URM faculty, particularly women, face a double standard in the P&T process, where research productivity is evaluated more harshly. This double standard contributes to the underrepresentation of URM faculty in tenured positions. The study highlights the significance of biased evaluations even when accounting for differences in resources like grant funding. The results also underscore the importance of ERLs in mitigating bias. Emphasizing scholarship in ERLs, especially for URM women, can positively influence P&T outcomes. This suggests that targeted interventions, such as training for ERL writers, could help address racial and gender disparities in academia.

This study provides compelling evidence of a double standard in the P&T process disadvantaging URM faculty, particularly URM women. The findings highlight the need for reforms to P&T processes, including reconsidering reliance on the h-index and implementing strategies to improve ERLs. Future research should explore interventions to address these issues and investigate the potential for a similar double standard in other academic evaluation processes such as annual reviews. Further research should also investigate the generalizability of these findings to a wider range of institutions and disciplines.

The study's generalizability may be limited by its focus on US research-intensive universities and the reliance on self-reported racial/ethnic data. The h-index, while widely used, is not a perfect measure of scholarly productivity. The study's analysis of ERLs focused on the presence of scholarship-related language, rather than the nuances of how this language was used. Future studies should address these limitations by expanding the sample to include a more diverse range of institutions and employing more sophisticated linguistic analyses.