Academic medicine is committed to advancing diversity at all levels in response to compelling evidence that inclusion of talented individuals from different backgrounds benefits medical education, patient care, population health, and scientific discovery.1 Since Title IX2 became law in 1972, medical schools are near gender parity at the student level, but women physicians remain woefully underrepresented in certain specialties and subspecialties,3 and their underrepresentation increases in all fields at successive academic career stages.4 Diversity goals for underrepresented ethnic and racial minority5,6 and lower-socioeconomic-status7 groups remain unfulfilled among students and faculty. In addition, members of diverse groups (women; members of ethnic, racial, and religious minority groups; people with lower socioeconomic status, with disabilities, or who identify as LGBTQ; and others) typically experience a more negative educational and workplace environment than do members of majority groups.1
As divisions, departments, and academic health centers strive to increase diversity and inclusion, creating an organizational diversity statement is often a first step. The promise of such statements is that they will help attract and retain diverse faculty, staff, and students; establish a basis for developing policies and practices that promote a welcoming, inclusive environment; and provide a rationale for considering applicants’ ability to foster diversity or work with diverse populations when hiring faculty and staff, or selecting residents and fellows. Although as physicians we value, teach, and attempt to engage in evidence-based practice, well-intended individuals are likely unaware of relevant research to draw on in developing a diversity statement. To our knowledge, there has been no research studying diversity statements in academic medical settings. However, several studies with randomized controlled designs have investigated the impact of diversity statements on participants’ perceptions of various organizations, including some in the health care industry. This research indicates that to achieve desired outcomes and minimize the risk of undesirable and unintended consequences, we must take considerable care in the language used and the messages conveyed when crafting diversity statements. With the caveats that diversity statements alone will not expand diversity but are just one part of a larger strategic plan, and that any research study is limited in generalizability to other settings and conditions, we present recommendations for developing organizational diversity statements that are based on experimental evidence and related research.
Declarative vs. Aspirational Statements
Pitfall: Organization members believe diversity goals have been achieved
If our organizations strive to treat everyone fairly and equally and rely solely on merit or performance in decisions about hiring, promotion, salaries, and bonuses, it seems intuitive to declare these ideals and our commitment to them in our diversity statements. However, several research studies demonstrate that diversity statements that promise to not discriminate, to treat all people equally, and to base evaluations solely on merit can have counterproductive effects. Statements such as “Medical School X does not discriminate” create an impression that the institution has achieved equity and fairness, when in fact, nondiscrimination is an ideal state that may not yet be realized. In addition to feeling disingenuous to those whose experiences are otherwise, research suggests that such statements may promote discriminatory behavior. In a series of randomized experiments with national samples of white adults, Kaiser and colleagues8 found that research participants were significantly less likely to detect and more likely to excuse discriminatory employment practices when a company had a “Diversity Statement” claiming not to discriminate rather than a “Mission Statement” without such language. In one of these experiments, researchers randomly assigned participants to read about a company with either the “Diversity Statement” or the “Mission Statement.” The statements had minor differences in wording. For example, the diversity statement encouraged “collaboration among employees from diverse backgrounds, cultures, and ethnicities,” while the mission statement encouraged “collaboration among employees with different work and learning styles.” The major difference between the two statements was that the diversity statement included an additional paragraph stating that the company “does not discriminate against any employee because of race, religion, gender, sexual orientation, national origin, disability, age, or covered veteran status.” When Kaiser and colleagues provided research participants with clear evidence of discrimination in promotion rates (28% of white and 10% of racial minority employees received promotions), participants who read the diversity statement rated the company as more procedurally fair toward racial minority employees than did participants who read the mission statement. In other words, participants who read the diversity statement did not recognize the discrepancy in promotion rates between white and minority employees as inequitable, while those who read the mission statement did. In another experiment, Kaiser and colleagues found that compared with participants who read the mission statement, participants who read the diversity statement were less sympathetic to a black employee who sued the company for discriminatory promotion practices and less likely to regard his suit as valid. Similarly, in a third experiment using an actual case of sex discrimination, participants read about the class action lawsuit of Velez et al v. Novartis Pharmaceuticals Corporation in which plaintiffs claimed that Novartis paid women less than men and denied promotions to women—especially those with children. Participants who were told that Working Mother magazine cited Novartis as “one of the 100 best companies in the nation for 10 years in a row” regarded the company as significantly more equitable to women and judged the plaintiffs’ case as less valid than did participants who did not receive this information. Taken together, Kaiser and colleagues8 concluded that these studies provide strong evidence that when organizations’ diversity statements make claims about not discriminating, they create an “illusion of fairness” that prevents recognition of bias and unwittingly subverts sincere intentions to avoid race and gender bias in hiring, salary, and promotion.
In related research, Castilla and Benard9 found that emphasizing meritocratic values at the organizational level introduced gender bias in awarding annual bonuses to men and women with equal job performance. In their studies, Castilla and Benard asked research participants, all of whom had managerial experience, to assume the role of manager at a large corporation and to assign bonuses to employees on the basis of their performance evaluations. They randomly assigned participants to read a list of “core company values” that either emphasized meritocracy (e.g., “all employees are to be rewarded fairly”) or did not emphasize meritocracy (e.g., “all employees are evaluated regularly”) before reviewing performance evaluations and assigning bonuses. Participants who read company statements endorsing meritocracy believed the company would be more fair in its performance–reward decisions than did those who did not read meritocratic statements. Yet, despite the fact that the annual performance reviews showed equivalent performance for male and female employees, participants who read the meritocratic statements gave male employees larger bonuses than female employees with equivalent performance ratings. Castilla and Benard9 refer to this as the “paradox of meritocracy”—the finding that emphasizing meritocracy can increase biased outcomes and workplace disparities. They suggest that meritocratic statements promoted by a corporation may increase its members’ confidence in their ability to be fair and objective or may prime them to see themselves as nonsexist or nonracist. Recent research suggests that these conditions make individuals more vulnerable to the influence of cultural stereotypes because they are less likely to question their judgments.10,11
Promise: Create aspirational statements
If messages proclaiming “we do not discriminate” can backfire, how can an organization express its commitment to being equitable, fair, and meritocratic? We recommend framing these values as aspirations. Emphasize that the organization and its members recognize and are working hard to overcome stereotype-based bias and that the institution is striving to provide a nondiscriminatory, fair, and equitable work and learning environment for all its members. Duguid and Thomas-Hunt12 showed the power of aspirational messages in a study comparing the effectiveness of various messages aimed at reducing stereotype-based bias. In a series of experiments, they demonstrated that conveying messages about the high prevalence of stereotyping increased reliance on stereotypes, but that conveying the message that most people try to overcome the influence of stereotypic preconceptions effectively minimized biased responses. In one experiment, for example, research participants who held MBA degrees and had managerial experience were less willing to work with a female job candidate who aggressively negotiated for a higher salary and rated her as less warm than a comparable male candidate. This finding was not surprising because research demonstrates that women who behave counter to gender stereotypes (e.g., that women are timid, passive, nurturing) face social penalties.13,14 When research participants received messages about the high prevalence of such stereotyping, negative reactions to this counterstereotypic woman increased. However, participants who received the message that “the vast majority of people try to overcome their stereotypic preconceptions” rated her as being more warm and were more willing to work with her.12
Another way to provide aspirational messages in a diversity statement is to rely on Dweck’s15 research on “growth mind-sets”—the concept that attributes such as intelligence and personality are malleable and subject to change rather than fixed and immutable. Dweck and colleagues investigated whether a growth mind-set approach could alter stereotype-based bias. In a series of experiments, they demonstrated that when people are informed that racial bias is a malleable (rather than fixed) trait, they were more willing to engage in interracial interactions and more comfortable doing so.16 This work on growth mind-sets together with Duguid and Thomas-Hunt’s findings lead us to recommend a diversity statement incorporating some version of a statement that “the vast majority of people in our organization are working hard to achieve a welcoming, diverse, and inclusive environment.”
Controlling vs. Autonomous Messages
Pitfall: Controlling messages provoke backlash
Diversity statements are crafted, in part, to promote an organizational culture that values diversity and inclusion. Explicitly stating this goal in the statement, however, can have counterintuitive effects if individuals perceive that the institution is trying to “force” this behavior. Legault and colleagues17 randomly assigned nonblack undergraduate students to read brochures framed as part of a new campus initiative to reduce race bias that emphasized either personal choice (autonomous) or complying with social norms (controlling). Students who read the autonomous message showed less bias against blacks and greater motivation to be nonbiased than those who read the controlling message or no message. In a second study, Legault and colleagues primed students with autonomous or controlling messages by asking them to complete surveys about their level of agreement with eight prodiversity statements that were either autonomous (e.g., “I enjoy relating to people of different groups,” “It’s fun to meet people from other cultures”) or controlling (e.g., “It is socially unacceptable to discriminate based on cultural background,” “I should avoid being racist”). They reinforced this priming by asking students in the “autonomous” condition to write three sentences about why it was “personally satisfying,” “enjoyable,” and “important” to be nonbiased, while asking students in the “controlling” condition to describe the “obligation” or “social expectations” to be nonbiased. Compared with a control group that responded only to filler questions, students primed with the autonomous messages showed greater motivation to be nonbiased and lower explicit bias, as well as lower implicit prowhite/antiblack bias on an Implicit Association Test18 requiring rapid pairing of pictures of black or white faces with positive or negative words. Those primed with controlling messages showed the opposite results on all three measures.17
Promise: Emphasize personal autonomy when promoting diversity
Legault and colleagues’17 research indicates the potential pitfalls of controlling and authoritarian phrases such as “zero-tolerance” or “stop racism” in diversity statements. Instead, we recommend using phrases that emphasize personal choice and autonomy in acting without bias. Incorporating messages such as “Our faculty, staff, and students say they value diversity, enjoy relating to people from different groups, have fun meeting people from other cultures, and think issues of diversity are interesting” into a statement or webpage can promote positive personal choices of individuals in the organization, without coercive messages indicating that the institution will force prodiversity actions on its members.
Multicultural vs. Colorblind Messages
Pitfall: Colorblind messages appeal to some groups and repel others
Diversity statements often emphasize either that the organization values differences (multicultural approach) or that it values equality (colorblind approach). Wilton and colleagues19 randomly assigned a diverse group of undergraduate students to examine brochures describing a university with either a colorblind (“we encourage our student body to embrace their similarities”) or multicultural (“we believe that embracing our diversity enriches our campus”) message. When asked to imagine themselves as students at this university, participants who read the brochure espousing the colorblind message expected less student diversity and more instances of bias than those who read the multicultural message. Women of color responded most negatively to the colorblind message. They expected less diversity and performed less well on a math exam than their counterparts who read the multicultural statement.19 Several other research studies similarly find colorblind messaging to be counterproductive.20,21 For example, in a study of workplace climate in a large U.S. health care organization, Plaut and colleagues22 showed that racial minority employees were more engaged in the workplace and perceived less bias when white employees valued multiculturalism rather than colorblindness. However, multicultural messaging can also have unintended consequences. One such consequence is that multicultural messages may cause white people to feel excluded. In the study by Wilton and colleagues,19 for example, white women did worse on the math test under the multicultural condition than they did in the colorblind condition. Multicultural messaging also risks implying that people are only valued for their group identity, which can have negative effects for all groups.22 Emphasizing the need for further scrutiny of the impact of valuing difference or valuing equality in organizational diversity statements, Apfelbaum and colleagues23 found contradictory effects of these messages depending on the context and audience.
Promise: Define diversity broadly
Despite the somewhat mixed findings regarding multicultural vs. colorblind messaging, our recommendation is to use multicultural approaches in diversity statements and augment them with an inclusive definition of what constitutes diversity. This enables all members of the community to see themselves as diverse members of the organization. We recommend statements that include both a message such as “we believe that achieving greater diversity will enhance our organization” and one stating that “we strive to embrace diversity in all its forms—identity, culture, background, experience, status, abilities, and opinions.”
Issuing Diversity Statements vs. Taking Action
Pitfall: Nonwhite members or potential members of an organization are disadvantaged
A general pitfall of publishing any organizational diversity statement is the degree to which it does or does not reflect the experiences of current and potential members of the organization. In a series of studies, Kang and colleagues24 examined black and Asian students’ engagement in “resume whitening”—the practice of concealing or downplaying their racial/ethnic identity. Interviews with black and Asian undergraduate students actively searching for jobs or internships in a range of fields including science and medicine revealed that 31% of black and 40% of Asian participants reported engaging in resume whitening, and 67% reported knowing others who had done so. Qualitative analysis of the interview text revealed two major types of “whitening”: presentation of one’s name (e.g., a Chinese student using an English nickname and a black student using a middle name if it was less racially identifying than the first name), and presentation of one’s experience (e.g., omitting the experience or leaving out the racial/ethnic identifier in an organization’s title). The students’ main motivation was to avoid anticipated discrimination at the initial stage of the application process. In a follow-up experiment, Kang and colleagues demonstrated that 50% fewer job-seeking students whitened their resumes when applying for a job at a company that presented itself as prodiversity by using diversity-affirming messages (e.g., “we strongly value fairness, diversity, and justice”) and images (e.g., a photograph of four people of different race and gender). To assess the consequences of resume whitening, the authors conducted an audit study to examine the probability of an applicant being contacted for a job interview. Using two of the largest online national job-search websites in 16 geographically dispersed U.S. metropolitan areas, the researchers randomly sent whitened and nonwhitened resumes to 1,600 job postings, 800 of which included explicit prodiversity messages. Whitened resumes led to significantly more callbacks than nonwhitened resumes, with no difference in callback rates between organizations with or without explicit prodiversity messages in their postings, regardless of location, job type, or industry. Taken together, Kang and colleagues’24 studies indicate that nonwhite applicants are more likely to abandon the practice of resume whitening when reading statements that potential employers value diversity and that this false assurance disadvantages them in the labor market.
Promise: Take action to implement practices and procedure that align with diversity goals and statements
If underrepresented applicants are more likely to represent themselves authentically when applying to an organization because that organization espouses egalitarian values in a public diversity statement, then the institution must ensure that the statement is backed up by real action to prevent bias and discrimination within the organization. Demonstrating the importance of going beyond statements about diversity and meritocracy to actually implementing processes that can mitigate gender and race bias in organizational decision making, Castilla25 conducted a field study with an actual company. In this study the company successfully reduced gender and race bias in annual performance rewards by introducing transparency and accountability into the process. A growing body of research that includes academic medicine suggests additional evidence-based strategies that institutions can adopt to align their practices with their stated commitments to diversity and equity.1,26–29
Current research supports the following recommendations for developing institutional diversity messages:
- Create aspirational statements, rather than declarative statements implying that the organization is already equitable and diverse;
- Emphasize personal autonomy to promote diversity, rather than promoting controlling messages;
- Use multicultural messages, rather than colorblind statements, combined with broad definitions of diversity.
Although developing a diversity statement is an important exercise for an organization on its journey toward a diverse and inclusive workforce and student body, it is only a beginning. If the goals of an organizational diversity statement are to be realized, this statement is merely one small part of what must be a multilevel strategic plan that attends to nonverbal messaging, hiring practices, evaluative and performance procedures, and fostering evidence-based strategies to improve interpersonal interactions and department climate.
1. Coalition of Urban Serving Universities, Association of Public & Land-Grant Universities, Association of American Medical Colleges. Increasing diversity in the biomedical research workforce: Actions for improving evidence. http://urbanuniversitiesforhealth.org/media/documents/Increasing_Diversity_in_the_Biomedical_Research_Workforce.pdf
. Published 2016. Accessed June 20, 2018.
2. United States Department of Justice. Title IX of the Education Amendments of 1972. https://www.justice.gov/crt/title-ix-education-amendments-1972
. Accessed June 20, 2018.
3. Association of American Medical Colleges. Diversity in the Physician Workforce: Facts & Figures 2014. Tables 9 and 10. http://www.aamcdiversityfactsandfigures.org/section-v-additional-detailed-tables
. Accessed November 16, 2018.
4. Association of American Medical Colleges. Diversity in Medical Education: Facts & Figures 2016. Tables 38B and 38C. http://www.aamcdiversityfactsandfigures2016.org/report-section/u-s-medical-school-faculty
. Accessed November 16, 2018.
5. Association of American Medical Colleges. Diversity in Medical Education: Facts & Figures 2016. Tables 8, 9A, and 9B. http://www.aamcdiversityfactsandfigures2016.org/report-section/applicants-enrollment
. Accessed November 16, 2018.
6. Association of American Medical Colleges. Diversity in Medical Education: Facts & Figures 2016. Table 38A. http://www.aamcdiversityfactsandfigures2016.org/report-section/u-s-medical-school-faculty
. Accessed November 16, 2018.
7. Grbic D, Jones DJ, Case ST. Effective Practices for Using the AAMC Socioeconomic Status Indicators in Medical School Admissions. March 1, 2013. Washington, DC: Association of American Medical Colleges; https://www.aamc.org/download/330166/data/seseffectivepractices.pdf
. Accessed June 24, 2018.
8. Kaiser CR, Major B, Jurcevic I, Dover TL, Brady LM, Shapiro JR. Presumed fair: Ironic effects of organizational diversity structures. J Pers Soc Psychol. 2013;104:504519.
9. Castilla EJ, Benard S. The paradox of meritocracy in organizations. Adm Sci Q. 2010;55:543576.
10. Uhlmann EL, Cohen GL. “I think it, therefore it’s true”: Effects of self-perceived objectivity on hiring discrimination. Organ Behav Hum Decis Process. 2007;104:207223.
11. Monin B, Miller DT. Moral credentials and the expression of prejudice. J Pers Soc Psychol. 2001;81:3343.
12. Duguid MM, Thomas-Hunt MC. Condoning stereotyping? How awareness of stereotyping prevalence impacts expression of stereotypes. J Appl Psychol. 2015;100:343359.
13. Heilman ME, Wallen AS, Fuchs D, Tamkins MM. Penalties for success: Reactions to women who succeed at male gender-typed tasks. J Appl Psychol. 2004;89:416427.
14. Phelan JE, Moss-Racusin CA, Rudman LA. Competent yet out in the cold: Shifting criteria for hiring reflect backlash toward agentic women. Psychol Women Q. 2008;32:406413.
15. Dweck CS. Mindset: The New Psychology of Success. 2006.New York, NY: Ballantine Books.
16. Carr PB, Dweck CS, Pauker K. “Prejudiced” behavior without prejudice? Beliefs about the malleability of prejudice affect interracial interactions. J Pers Soc Psychol. 2012;103:452471.
17. Legault L, Gutsell JN, Inzlicht M. Ironic effects of antiprejudice messages: How motivational interventions can reduce (but also increase) prejudice. Psychol Sci. 2011;22:14721477.
18. Greenwald AG, Nosek BA, Banaji MR. Understanding and using the Implicit Association Test: I. An improved scoring algorithm. J Pers Soc Psychol. 2003;85:197216.
19. Wilton LS, Good JJ, Moss-Racusin CA, Sanchez DT. Communicating more than diversity: The effect of institutional diversity statements on expectations and performance as a function of race and gender. Cultur Divers Ethnic Minor Psychol. 2015;21:315325.
20. Purdie-Vaughns V, Steele CM, Davies PG, Ditlmann R, Crosby JR. Social identity contingencies: How diversity cues signal threat or safety for African Americans in mainstream institutions. J Pers Soc Psychol. 2008;94:615630.
21. Holoien DS, Shelton JN. You deplete me: The cognitive costs of colorblindness on ethnic minorities. J Exp Soc Psychol. 2012;48:562565.
22. Plaut VC, Thomas KM, Goren MJ. Is multiculturalism or color blindness better for minorities? Psychol Sci. 2009;20:444446.
23. Apfelbaum EP, Stephens NM, Reagans RE. Beyond one-size-fits-all: Tailoring diversity approaches to the representation of social groups. J Pers Soc Psychol. 2016;111:547566.
24. Kang SK, DeCelles KA, Tilcsik A, Jun S. Whitened résumés: Race and self-presentation in the labor market. Adm Sci Q. 2016;61:469502.
25. Castilla EJ. Accounting for the gap: A firm study manipulating organizational accountability and transparency in pay decisions. Org Sci. 2015;26:311333.
26. Carnes M, Devine PG, Baier Manwell L, et al. The effect of an intervention to break the gender bias habit for faculty at one institution: A cluster randomized, controlled trial. Acad Med. 2015;90:221230.
27. Sheridan JT, Fine E, Pribbenow CM, Handelsman J, Carnes M. Searching for excellence & diversity: Increasing the hiring of women faculty at one academic medical center. Acad Med. 2010;85:9991007.
28. Moss-Racusin CA, van der Toorn J, Dovidio JF, Brescoll VL, Graham MJ, Handelsman J. Scientific diversity interventions. Science. 2014;343:615616.
29. Isaac C, Lee B, Carnes M. Interventions that affect gender bias in hiring: A systematic review. Acad Med. 2009;84:14401446.