Overcoming Barriers to Implementation
The analysis draws substantially from the focus group report, by Tasseli McKay and Dr. Christine Lindquist of RTI International, which was commissioned for this study. The full research paper appears at: www.nap.edu/catalog/25585.
As Chapters 3 and 4 demonstrate, there are a variety of evidence-based and promising practices and strategies to advance the participation and advancement of women in science, technology, engineering, mathematics, and medicine (STEMM). Some institutions have adopted such policies and practices and have seen improvements in the representation and experiences of women in STEMM education and careers. However, in most institutions and organizations there are particular entrenched patterns of underrepresentation across disciplines that still exist; namely, that women remain underrepresented at all levels of education and career in disciplines like computer science, physics, and engineering—and are underrepresented among more senior leadership roles in disciplines like medicine, biology, and chemistry (see Chapter 2). The fact that some institutions, departments, or schools are doing better than others in improving the recruitment, retention, and advancement of women in STEMM raises questions about why this is the case. Which factors serve as barriers or facilitators to institutional adoption and sustainability of effective policies and practices? To quote the statement of task for this report, why is it that “effective interventions have not been scaled up or adopted by more institutions?”
In this chapter, the committee provides an overview of the characteristics of successful programs and describes the common institutional barriers to sustainably implementing these practices. This analysis is supported by the research literature, as well as by the findings from a series of focus groups with faculty and administrators carried out by RTI International on behalf of the National Academies of Sciences, Engineering, and Medicine (see www.nap.edu/catalog/25585 for the full results from the focus group research). At a high level, both the
research literature and the focus group findings point to a common set of conditions that support positive institutional change, including:
- Committed leadership at all levels
- Dedicated financial and human resources
- An understanding of institutional context
- Accountability and data collection
- Adoption of an intersectional approach
The sections below elaborate on these five key points.
COMMITTED, SUSTAINED LEADERSHIP
Organizational transformation requires changing institutional culture (Bilimoria, 2008; Eckel and Kezar, 2003), which in turn requires leadership. Research demonstrates that leadership is a major factor in organizational transformation and is critical to successful equity and diversity efforts (Bilimoria, 2006; Eckel and Kezar, 2003; Garvin, 2000; Plummer, 2006). Eckel and Kezar (2003) describe four core strategies common to institutions undergoing transformation, including senior administrative support, collaborative leadership, flexible vision, and visible action. In particular, senior administrative involvement is a prerequisite for successful organizational change (Bilimoria, 2008; Eckel and Kezar, 2003; Garvin, 2000). Collaborative leadership is also critical in institutional transformation because it shapes organizational vision, sends institutional messages and signals, and has authority to implement change (Bilimoria and Liang, 2011; Eckel and Kezar, 2003).
Bilimoria and Liang (2011) note that universities can increase women’s representation in science by creating and supporting a transformation team composed of senior faculty leaders and administrators to comprehensively address the issues of women’s underrepresentation. Similarly, Plummer (2006) notes that communication and leadership strategies are key to the successful implementation of policies, processes, and programs designed to achieve institutional transformation. In addition to being a strategy for implementing lasting change, senior administrative and faculty leadership can serve as a preliminary indicator of lasting institutional change (Plummer, 2006).
In order for transformative change to be sustainable, leaders should be alert to institutionalizing successful features promoting cultural and structural change, and mobilize adequate resources to support change in the long term (Bilimoria and Liang, 2011). Also, direct or indirect access to and support from the highest levels in university administration are cited as being critical in bringing about changes in institutional policies, infrastructure, and climate to address the recruitment, advancement, and retention of women and minority faculty, as well as to create new positions and offices for the implementation of future changes (Bilimoria, 2006; Plummer, 2006).
Based on an evaluation of National Science Foundation (NSF) ADVANCE programs, Plummer (2006) found that:
Presidents, chancellors, provosts, vice presidents, and deans created an environment that supported the goals of ADVANCE by communicating frequently about the educational value of diversity and the productivity possible in supportive college and department climates and by expressing support for the goals of ADVANCE. Senior administrators communicated and modeled institutional values and norms by articulating their commitment verbally in formal and informal settings and by underscoring the importance of ADVANCE endeavors.
A co-principal investigator from one NSF ADVANCE program summarized what several other focus group participants expressed about the important role university leaders, namely: “The leadership of the administration matters. Central leadership from the top is crucial. It’s amazing how much difference this makes—what the president says and does” (Plummer, 2006).
The results of the focus group research carried out for this report also highlighted the important role of leadership. In particular, a lack of strong leadership support from university presidents, provosts, deans, and others is a major barrier to equity and diversity efforts. Even if academic leaders are personally supportive of gender-equity practices, the lack of willingness to risk controversy on equity-related initiatives can be an additional barrier. Some leaders fear backlash from vocal opponents and may see little incentive to implement changes, given the risks to them personally. In the words of one participant:
I think a lot of times people know what the best practices are, and would personally be supportive of them, but they feel like they’re going to incur too much backlash . . . if they’re not secure in their base of power, they feel like rocking the boat too much isn’t something that they want to push for . . . ‘Why am I going to go out on a limb to do this? There’s no real incentive for me to do it, for me personally as the leader.’ And so, they’re just unwilling to go up against the very strong faculty members who are loud, and don’t want to make the changes.
To be successful, equity work needs to actively involve those who have power within their institutions. Yet such work is frequently delegated to university diversity and inclusion officers, who are often marginalized within their institutions, are women and minority faculty tapped by virtue of their service on relevant committees, and who have limited power to bring real change (in addition to risking being overburdened and harming their individual careers). Delegation of equity work to nonacademic staff, such as human resources personnel, was also reported to be a concern, given the perception that human resources is focused foremost on protecting the institution from legal liability. As one participant noted:
Human resources. . . . they’re so bogged down in following the letter of the law that they lose track of what the spirit of the law often is. And so, they’re not willing to be flexible about a lot of things.
Leadership transitions are also a point of vulnerability for equity and diversity efforts. Inadequate planning to implement new procedures and the failure to identify a new “champion” when turnover occurs can undermine any progress that has been made under prior leadership. Although new leaders may be committed to change, they often bring new agendas to advance and may give less attention to existing policies or practices. Even when such policies have been formalized, the extent to which they are communicated and encouraged to the campus depends on implementation by new and existing leaders alike, which makes them vulnerable when administrative transitions take place. To quote one participant:
There used to be a feminist statement to married women, ‘Most women are only one man away from welfare’. . . I feel like a lot of these programs are only one man away from existing . . . I hope every day [that the provost] is not out looking for jobs, because I don’t know what will happen to a lot of these programs. Even if you think it’s institutionalized, it’s really not institutionalized . . . it’s all very vulnerable, it’s still peripheral.
The focus groups also discussed the role of leaders in facilitating the implementation of research-based policies and practices. First, they cited as a facilitator equity-related initiatives from boards of trustees, with some participants noting that this push from the top could be particularly effective, given trustees’ roles in allocating resources and concern for public relations. Equity-related initiatives are also part of the governing boards’ responsibilities. The Association of Governing Boards (AGB) notes the following responsibilities as part of a Board’s duty:
Accordingly, higher education governing bodies must ensure institutional compliance with applicable federal, state, and local laws, including those that prohibit discrimination based on age, race, gender, sexual orientation, religion, disability, and other characteristics, and those that protect freedom of speech and academic freedom (Association of Governing Boards, 2016).
The role of governing boards with respect to campus climate was addressed in 2016 in the AGB Board of Directors’Statement on Governing Board Accountability for Campus Climate, Inclusion, and Civility (Association of Governing Boards, 2016). The statement defines diversity across a number of demographic characteristics, including gender, and notes the importance of campus climate, culture, and norms in terms of building diversity. Building from diversity, the statement recognizes the role of inclusion: “Inclusion recognizes and embraces the need for all members of the institutional community to have a sense of
ownership in the institution and a place of belonging. It requires sustained and intentional institutional commitment and action.” Further into the statement, the AGB lists recommendations to boards to develop a safe and inclusive campus climate, with significant mentions between the board and the campus leader. As part of their governing function, boards select campus leaders, delegate the management of the institution to those leaders, and provide oversight to the leader’s performance (Eckel and Kezar, 2016). For supporting a diverse, equitable, and inclusive campus climate, the AGB recommends the following actions between the board and campus leader: collaboration and transparency, periodic updates to review policies and ensure compliance, ensuring appropriate allocation of resources to address needs, and a communication plan with regular updates on the implementation of campus climate activities. While board members may not spend the majority of time on campus, they have the right to hold leadership accountable to the school’s mission, to determine the progress of institutional change, and to adopt new policies to address issues.
Second, the focus groups discussed strong alignment among academic leaders and academic staff at all levels as another enabler of research-based policies. Participants emphasized the benefits of equity work receiving simultaneous effort and priority among both leadership and faculty. As one participant suggested, “senior-level support coupled with policies and coupled with leadership development at all levels of the institution” produces the largest impact on “institutional traction and progress.”
Alignment between the formal and informal emphasis placed by deans and chairs on equity efforts was also identified as a facilitator. The formal actions of a dean or chair, such as allocating hiring resources and enforcing policies, are necessary for successful implementation. Their informal actions, particularly the communication of motives behind the formal policies or requirements, are equally important for motivating faculty who are involved in carrying them out. One participant noted:
It’s up to a chair to energize a [search committee] and not just hear it from the office of advancement or equal opportunity . . . why we’re doing this, why it’s important. Just kind of that motivation of why these policies came into being in the first place, and [that] it’s not a check box.
In industry settings, research points to a need to intentionally connect the diversity and inclusion strategy with the business strategy. Although most business leaders voice support for diversity and inclusion initiatives, the show of support tends to be insincere. For example, one study noted that when private sector leaders rated a number of potential business priorities, nearly all of them ranked diversity and inclusion initiatives in last place. Chief diversity officers intimated that business strategy was accordingly the weakest driver of diversity and inclusion. Recommendations from this study noted that success-
ful diversity and inclusion programs should—similarly to academia—focus on defining the problem, provide funding support, and require data, accountability, and buy-in from leaders to gauge effectiveness and drive progress (Paikday et al., 2019).
Leaders must also work to ensure that they effectively communicate and enforce policies (see Chapter 4 for additional discussion). Focus group participants reported that a lack of standardization and communication at the university level often results in individual departments implementing policies on their own, despite inadequate department-level familiarity with and communication about equity-related policies. It was noted that some department chairs and search committee members who are in a position to implement the university’s equity-related policies (e.g., family leave entitlements, search committee processes) are simply not familiar enough with them to ensure they are uniformly available and implemented. Relatedly, the “soft adoption” of equity-related policies, where implementation is left to department judgment, was perceived to be a barrier to successful implementation. For example, some universities may automatically implement tenure clock stoppage when family leave is taken, whereas at others, this stoppage is recommended to departmental leadership as a “best practice” but essentially left up to them to implement. The latter approach has much greater potential for uneven application. Similarly, the lack of department-level accountability for university-wide efforts was identified as a related barrier, with several participants noting that it is necessary to move from department-idiosyncratic policies toward more standard practices at the university level.
Finally, both the research and the focus group participants noted that leaders in all sectors should work to embody the respectful, inclusive behavior they expect from members of their organization since the behavior of leaders sends powerful cues about organizational expectations. Individuals in an organization take notice of the behaviors of leaders and model this behavior—and sometimes attitudes—accordingly (NASEM, 2018b).
DEDICATED FINANCIAL AND HUMAN RESOURCES
Both the literature and the focus group research emphasize the need for institutional leaders to set aside adequate resources to support equity and diversity efforts. As described above, the vast majority of equity work is assumed, in an uncompensated manner, by existing women faculty and faculty of color. The predominant model for accomplishing such work was perceived to rely on those who are passionate about equity to assume it voluntarily—on top of their regular workload—and with no teaching release or other compensation.
It’s perceived as an extra workload kind of thing . . . you’re willing to do the extra work if it’s your passion . . . if you have one person with the passion to do it they figure out a way to do it, but then it’s extra work for them. They don’t get any kind of release . . . and then it goes away [if they leave].
In addition to the burden imposed on the individuals carrying out the work (described by one participant as “putting a burden on the oppressed”), this approach can be ineffective, making equity issues easier to sideline or compartmentalize. Participants suggested the need for greater involvement from male and/or majority culture faculty:
All too often, the ethnic minority or the gender minority population that is most affected is being asked to solve these kinds of issues. But women can’t solve the fact that they are paid less than men. They need men on board to solve that problem. And ethnic minorities can’t solve their pay inequalities either. They need people from the majority culture to also say, ‘hey, let’s solve these kinds of issues.’ So, I think will is absolutely one of those things where if the university doesn’t have that, everything else is just going to be lip service. And a lot of extra service work to hide the fact that it’s just lip service.
The research literature points to similar findings regarding general lack of resources to support equity and diversity work. A recent study of 234 chief diversity officers at Standard & Poor’s 500 companies found that many in this position were in fact business leaders selected to lead internal diversity and inclusion efforts in addition to their other job responsibilities. The multiple layers of responsibilities unrelated to diversity and inclusion frequently hindered their ability to perform their jobs fully. The study also indicated that chief diversity officers often lacked data and analytics, such as employee demographic data, requisite to their positions (see below for a discussion of the importance of data collection for accountability and targeted intervention).
Similar findings have been reported in the field of pediatrics (where White women and women of color are underrepresented among senior leadership roles). In a recent publication, Spector et al. (2019) used the example of hospital safety as a comparison to illustrate the point:
Institutional hospital safety leaders, not patients (those most affected), are held responsible for identifying and prioritizing inadequacies, eliciting solutions, assigning institutional funds and resources, and collecting, analyzing, and compiling outcome data into reports distributed both internally and to regulatory agencies. In contrast, gender-equity initiatives have been largely driven from a grassroots level (with little or variable institutional recognition or support) by those most affected (women who are underpaid and underrecognized) with few resources (volunteering their spare time and often underwriting the initiatives themselves) (Spector et al., 2019).
After conducting a comprehensive evaluation of the state of White women and women with intersecting identities in pediatrics, Spector et al. concluded that achieving gender equity in medicine needs to include a number of specific scientific principles, namely: leadership accountability; dedicated financial and human resources; and an evidence-based, data-backed, and transparent approach
to evaluation and reporting (see discussion below) (Spector et al., 2019). These conclusions are consistent with the research presented throughout this chapter.
Beyond compensating those individuals in an organization that take on responsibility for equity and diversity efforts, it is important to ensure that such support is sustained over time. Focus group participants noted that even when money is allocated to equity work, its vulnerability to budget cuts or re-appropriation is a barrier to sustained implementation. True institutionalization of policies and practices was reported to be expensive and labor intensive, and the vulnerability of efforts is particularly pronounced when resources are dispersed across budgets in different departments and offices within the university. One participant offered an example illustrating the potential for funds to be misused in such arrangements:
We had funds that were put aside for opportunity hires for underrepresented women and underrepresented minority men and women. And over time they were used for all kinds of things. Whoever the provost was who came in or whatever the president wanted, the funds got used, and they eventually disappeared.
Focus group participants additionally observed that the most effective approaches for addressing issues with equity and diversity are also often very resource- and labor-intensive to implement and sustain (e.g., in-person instead of online trainings). Other resource-related barriers identified in the focus groups included the lack of investment in equity-related initiatives from major donors and/or alumni, who wield strong influence in the institution, and the unwillingness or inability of departments to allocate the additional resources that may be required to successfully recruit an under-represented candidate (e.g., associated with many women faculty needing jobs for partners).
Participants viewed grant-funded efforts as playing an especially important role in facilitating the implementation and sustainability of gender-equity efforts. They emphasized the receipt of NSF ADVANCE institutional transformation grants as a major facilitator of implementation success within their institutions (and others).
I really think NSF, between the ADVANCE program and broader impacts, made it sort of important for institutions to pay attention to this. You know, there’s probably some effects from the changing environment at large. But, I think the first big push at my institution came from more ADVANCE institutional transformation grants.
Participants viewed the NSF ADVANCE program as drawing initial attention to gender-equity issues as well as serving as a catalyst for sustained equity efforts. Stakeholders at some institutions were able to secure ongoing, university-level funding to continue initiatives that were implemented with the grants, which was identified as a facilitator of sustained implementation. One participant com-
mented that institutionalizing efforts initiated with NSF ADVANCE grant funding enabled the institution to continue to innovate and be a continued leader in the field, and that “institutional resources have to undergird that initial external funding.”
Participants also identified school- or department-level funding support was also identified as a facilitator of implementing and sustaining research-based policies and practices. For example, some STEMM departments or schools allocated funding for search committee training that was formerly covered by outside grants.
UNDERSTANDING INSTITUTIONAL CONTEXT
While there are shared characteristics of successful programs and common institutional barriers and facilitators to sustainably implementing these practices, the committee recognizes that institutions have different goals and missions, values, cultures, and resources and this institutional context can impact the efficacy of any program, regardless of that program’s designation as “successful” and “evidence-based.” There is no one-size-fits-all solution, policy, or practice that will perfectly fit the needs of all institutions. As Hardcastle et al. (2019) state: “Greater participation of women and faculty of color in STEMM fields is complicated and dependent on complex and multi-layered interactions between activities and actors.” In addition, because researcher and institutional goals vary as a function of target population and context, “generalizable models can struggle in the face of larger broadening participation efforts.”
To further explore these issues, Hardcastle et al. (2019) conducted a social network analysis, an exit survey of departed faculty, longitudinal analysis of career trajectories and research productivity, and a survey on the interaction between values and climate to assess the barriers for women in STEMM across institutions. The authors found that a “dynamic, multi-scaled and organizational level approach is required to reflect the reciprocal dialogue among research questions, best practices, tailored applications and quantifiable goals” (Hardcastle et al., 2019).
The authors identified three strategies to better retain women in STEMM across institutional contexts, including (1) improving women’s professional networks; (2) re-aligning policy documents and departmental practices to better reflect faculty values; and (3) improving departmental climate. Regarding the need to improve women’s professional networks, the authors found that helping women proactively develop professionally-oriented connections, while also working with department heads to assist with this process, should help improve a sense of fit and belongingness, which should in turn decrease attrition (Hardcastle et al., 2019).
Hardcastle et al. (2019) noted that changing explicit policies has a much greater and immediate impact than trying to change hearts and minds. “Claiming institutionally that we value diversity and diverse forms of scholarship is one thing, but formally recognizing diverse scholarship and having policy to point to in promotion and tenure
cases are more convincing. Moreover, policy can outlast any set of ‘hearts and minds,’ since they stay when people leave” (Hardcastle et al., 2019).
Regarding improving departmental climate, the authors state that a strong and continuous organization can promote sustained change by holding the institution accountable for achieving diversity and inclusion (see discussion of accountability in the section below). “True institutional transformation will not come from the work of select individuals across campus; instead, it must be driven by organized groups over a significant period of time who connect both with leadership and faculty, who also can leverage each other’s successes, and who can ensure that institutional leaders enforce policies, standards, and expectations” (Hardcastle et al., 2019).
Similarly, the results of the focus group research carried out for this report highlighted the issue of institutional context and how it can impact whether a policy will have a positive impact. For example, focus group participants noted that there is a need to recognize and engage the specific strengths and challenges of different institutional contexts. Participants noted that cross-context adaptation is challenging and can be poorly guided. Another theme from the focus groups included that adapting research-driven policies and practices to different institutional contexts is critical to achieving large-scale equity across STEMM disciplines, since most such initiatives had been developed and tested in a single type of university only. To quote one participant:
We know a reasonable amount about these kinds of initiatives; what we don’t know is how to do them in all the different contexts. And so, I think that’s an enormously hard and important problem . . . there’s not a magic bullet process or procedure to use. It has to be adapted, and how do we analyze the examples that exist and say, ‘This would work here, but it would have to change in this way’?
Also, focus group participants highlighted gaps in cross-contextual and translational research that could inform such adaptation efforts, noting that the lack of research constrained efforts to scale or adapt “evidence based” policies and practices to their own institutional contexts. In the words of a focus group participant:
‘What about this program works that can then be applied at other institutions?’ A lot of times [what is] presented as research is actually a single institution implemented a practice and it worked really well for them.
DATA COLLECTION FOR ACCOUNTABILITY AND TARGETED INTERVENTION
As it can be difficult to predict the interventions that will be most successful within particular institutional contexts, it is necessary for individual institutions and organizations to collect data and monitor trends in the recruitment, retention, and advancement of women in STEMM to better adopt or adapt targeted interventions and to monitor their efficacy. By collecting and monitoring data and
evaluating it over time, employers and admissions officers can better understand whether recruitment, retention, or advancement—or some combination—is the major issue affecting the representation of women (see Figure 5-1 for an example of an online dashboard at Stanford University). If recruitment is an apparent issue, institutions can evaluate and monitor the diversity of applicant pools at every stage of the recruitment process and keep track of who decides to enroll or accept the job once an offer is made. By tracking trends in the recruitment process, employers or admissions officers will be better able to determine whether the underrepresentation of women is related to particular stages in the recruitment process (e.g., a limited pool of candidates, shortlist, interview group, or final choice). Data on who is completing a given program or leaving an organization is useful in diagnosing whether retention of women is an issue, and examining data on the representation and rate at which White women and women of color are advancing, as compared with other groups, is helpful in identifying problems of advancement. In addition, examining patterns of advancement longitudinally, rather than only at key transition points, will provide a better evaluation of issues related to advancement. It is only by tracking attrition and delay (e.g., of promotions) at each stage, that organizations can gain a greater understanding of how apparent parity at the beginning of a process (e.g., admissions or hiring) can result in large disparities at the end (graduation or degree completion, and advancement to the highest positions). Several organizations and higher education institutions already make use of publicly available dashboards that include data on gender representation. For example, Google,1 Stanford,2 and the University of Michigan3 have such interactive dashboards (see Figure 5-1 for an example).
The importance of data collection to “diagnose” and “treat” equity issues was also discussed extensively during the focus group sessions. Participants noted that when data were available or data collection systems were in place, they made implementation of equity-related policies and practices more likely, and they could get university stakeholders and professional associations representing specific disciplines interested in solving equity-related problems, tracking progress toward solutions, and establishing organizational priorities.
Beyond the numbers, it is also important for institutions and organizations to understand the experiences of White women and women of color through periodic climate research carried out by an evaluator outside the relevant unit. Enlisting the services of an evaluator external to the unit is important in that it will permit assessment of the climate in a school, company, or department in a manner that is methodologically sound, independent, objective, and free from bias and conflict of interest. Such climate research can take the form of surveys, focus groups, and/or interviews. That being said, given the extremely low representation of women of
1 See https://diversity.google/annual-report/.
2 See https://ourvision.stanford.edu/design-teams/ideal-homepage/ideal-dashboard.
color in most STEMM fields, it is important to adopt a methodological approach that can protect the anonymity of such individuals and accurately capture their experiences. In some instances, interviews may serve as the most appropriate means to gather this information; in others, conducting such research within a single institution may be deemed unsafe for the women of color who make up an extreme minority in certain fields. In instances where there is a small sample size, specifically in the case of women of color in most STEMM fields, qualitative research may be extremely valuable, particularly as it can provide richly-textured information, relevant to the phenomenon under investigation. For example, focus
group methods may be uniquely well suited to exploring issues with existing processes or strategies and gaining insight into the functioning of institutions (Stewart et al., 2009), as is the case in for many of the issues addressed in this report.
Additionally, data collection, monitoring, and evaluation, if done transparently, can increase accountability, which can, in turn, serve as a driver of positive change. The social science and business literature offers many examples of the positive impact of greater accountability on equity and diversity efforts in education and business (Dobbin and Kalev, 2016). Take, for example, Emilio Castilla’s field study of the Massachusetts Institute of Technology’s Sloan School of Management, where African Americans were consistently given smaller raises than White employees, despite identical job titles and performance ratings (Castilla, 2015). To address this pervasive issue, Sloan began posting the average performance reviews and associated raises for each unit by demographic characteristics (i.e., race and gender). As soon as managers realized that bias in compensation by race and gender would become public knowledge within the school, they developed an increased sense of accountability and the discrepancies in compensation disappeared.
Deloitte offers another compelling example. In 1992, chief executive officer (CEO) Mike Cook realized that despite gender parity in hiring, the company was struggling to retain and advance talented women (Gaventa and McGee, 2013). He assembled a high-profile task force to address the issues with retention. Adopting a strategy that relied on accountability, the task force got each office within the company to monitor the career progress of its women and set goals to address the problem within the context of the specific unit. To quote a Harvard Business Review article on this case:
When it became clear that the CEO and other managing partners were closely watching . . . “Women started getting their share of premier client assignments and informal mentoring.” And unit heads all over the country began getting questions from partners and associates about why things weren’t changing faster. An external advisory council issued annual progress reports, and individual managers chose change metrics to add to their own performance ratings. In eight years turnover among women dropped to the same level as turnover among men, and the proportion of female partners increased from 5 percent to 14 percent—the highest percentage among the big accounting firms. By 2015, 21 percent of Deloitte’s global partners were women, and in March of that year, Deloitte LLP appointed Cathy Engelbert as its CEO—making her the first woman to head a major accountancy.
Similar examples are found in educational settings. When teachers realize that they will have to explain their evaluations, they rely less on their biases. For instance, studies have shown that when teachers are told that they will have to discuss and justify the grades they give students on papers, racial bias in grading disappears (Kruglanski and Freund, 1983). Equally, when departments are expected to present short lists of potential candidates to the dean’s office for review, those lists include more diverse candidates (Bilimoria and Buch, 2010).
In addition to improved accountability, data collection allows an organization to gain a more complete understanding of its specific issues with recruitment, retention, and advancement and develop targeted strategies to address these issues. In the example offered above, it would have made little sense for Deloitte to invest additional resources in the recruitment of women candidates since women were being hired at the same rate as men—the specific issue they were facing was one of poor retention. To address a problem and make best use of the (often) limited resources available to address the issue, we must first understand the nature of the specific problem.
Institutions and organizations that have collected, monitored, and reported data over time to assess the recruitment, retention, and advancement of White women and women of color have been able to implement targeted interventions and seen marked improvements in equity and diversity. For example, within some schools and colleges at the University of Michigan, every department is provided annual data about the rate of Ph.D. attainment by women and minorities in the
relevant field at Ph.D.-granting institutions, at institutions that the University of Michigan considers “peers,” and within the department at Michigan itself. This practice allows for departments and deans to notice when the Ph.D. production in a department is not meeting that of the national production and when this needs to be addressed. Additionally, it helps remedy the often incorrect assumptions about the potential availability of applicants so that realistic goals can be developed (Stewart and Valian, 2018). When these data were first distributed, many were surprised by how few women and minority Ph.D.s they had graduated and that the pool of applicants was larger than they expected. These data are now expected among the departments to achieve the diverse applicant pool they desire to fill their faculty positions. Other efforts that place a significant emphasis on data collection in support of driving greater accountability and the adoption of evidence-based, targeted practices by institutions include the Athena SWAN (Scientific Women’s Academic Network) Charter in the United Kingdom and the STEMM Equity Achievement (SEA) Change effort in the United States (see Box 5-2).
One challenge, however, is that the extremely small numbers of women of intersectional identities in a department or field can serve to hinder organizational accountability and quality improvement with regard to their representation. Focus group participants noted that, within a given discipline or university, tracking how women of color and women with other intersectional identities benefited (or not) from policies and practices intended to increase women’s representation or advancement is difficult.
We are necessarily going to overlook those individuals with intersecting identities . . . because the sample sizes are so small they often get collapsed into broader categories . . . The data systems don’t allow for really looking at [this].
Without the ability to rigorously assess the status of women of intersectional identities or examine whether the representation of particular groups of women expanded with the introduction of a certain initiative or policy, institutions lack accountability for promoting the inclusion and advancement of all women. As one participant put it: “The lack of data . . . has tended to perpetuate the underrepresentation.” This suggests the need for careful consideration of how to balance the importance of collecting and monitoring data on the experiences and participation of women of color and women with other intersecting identities, while also ensuring sufficient protections and anonymity for such a small group of people.
ENSURING THAT ALL WOMEN BENEFIT FROM EQUITY AND DIVERSITY EFFORTS BY ACCOUNTING FOR INTERSECTIONALITY4
The literature and the focus group research indicate that not all women in STEMM benefit equally from policies and practices designed to support their representation, advancement, and academic contributions. Instead, such efforts tend to be unevenly successful depending on women’s life experiences (particularly racialized life experiences), their career stages, and the institutional contexts in which they work (such as distinctions between public and private universities, historically minority- and majority-serving institutions, and academia and industry). As discussed in Chapter 2, most research on women in STEMM has focused almost exclusively on middle-class White women and very little empirical attention has been paid to the intersection of ethnicity, race, gender, and the scientific culture. Further, well-intentioned efforts to support women in STEMM have historically failed to account for the intersectional experiences of
4 The concept of intersectionality considers the complex, cumulative ways in which the effects of multiple forms of discrimination (e.g. racism, sexism, homophobia, classism, etc.) intersect in the experiences of women of multiple marginalized identities (e.g. women of color, women with disabilities, sexual minorities, etc.). For a discussion of intersectionality, see Chapter 2.
women of color and women of other intersecting identities, such that “programs intended to serve women disproportionately benefit White women, and programs intended to serve minorities mainly benefit minority males (Ong et al., 2011).” This was further emphasized in a recent paper by Corneille et al. (2019), which synthesized the available literature barriers to the advancement of women of color faculty in STEMM, and found that “there is limited research that examines STEM women of color faculty experiences at minority-serving institutions and in leadership roles. Further research is needed to examine the long-term efficacy of mentoring strategies and institutional transformation efforts for women of color” (Corneille et al., 2019).
The results of the focus group research are consistent with the findings in the published literature. Focus group participants agreed that existing policies and practices have been inadequate or ineffective for supporting the representation, advancement, and contributions of women of color. They noted that active institutional recognition of distinct challenges relevant to women of color in STEMM lagged considerably behind rhetoric and policies regarding gender or the needs of academic “women” as an undifferentiated group.
Despite some positive change in gender composition in many institutions and fields represented in the focus groups, participants observed that efforts to address the inclusion of women and their representation in academic leadership roles have not brought corresponding shifts in the presence of women of color in their fields.
There’s definitely a sense of full inclusion based on gender in my department, but that doesn’t carry forward or I don’t think the same can be said when we consider race and ethnicity and women of color, as we have very low representation . . . there’s not a sense of that we’re really achieving all that we could achieve on that.
Participants suggested that this persistent underrepresentation of women of color in many STEMM fields, even those in which gender composition had shifted in recent decades, is symptomatic of greater discomfort or intolerance regarding efforts at inclusion and advancement for women of color.
Most colleagues, I think, are reluctant to engage it . . . the indifference or resistance to hiring underrepresented minorities. It’s quite astonishing that in some schools they’ve just hired their first African American. In fact, if you look at African American or Latino women, in some cases there are whole segments of higher education, STEMM fields, that haven’t hired any.
The failure to address issues of underlying racial bias or the specific needs of women of color with regard to representation and advancement meant that White women had tended to benefit more heavily from efforts to address gender composition than their colleagues of color.
The one exception noted by focus group participants was the inclusion of nonnative-born women, whom a few participants observed tended to benefit disproportionately from university initiatives to increase the representation of women and faculty of color. Participants felt very positively about the presence of immigrant scholars in STEMM fields and about efforts to encourage their full contributions. However, they noted that immigrant faculty might not have the same capacity to relate to, and steward the contributions of, women students of color from racialized and marginalized American communities. For this reason, it is critical that their presence is not seen as replacing ongoing efforts to recruit and promote native-born faculty of color who shared those experiences of racialization and marginalization.
Many of the people who are being promoted are not ethnic minority Americans, but instead recent immigrant ethnic minorities . . . Somebody who just got here from Nigeria has a very interesting and worthwhile perspective [but] it doesn’t advance the civil rights movement in this country because they’re not tied to it at all . . . [and] what we’ve always asked people as any underrepresented group, whether that’s women or African Americans or Latinx populations, is that, as they attain positions of responsibility and authority, they reach back and help people who are like them, who are similarly disadvantaged. But that kind of system breaks down if they don’t have any relationship with the natural communities that [faculty of color] should be helping to bring up.
Participants noted that the introduction of requirements to address intersectionality in applications for the NSF ADVANCE grants helped to bring attention to intersectionality generally and issues faced by women of color at their institutions. Still, this nascent attention was perceived as “barely scratching the surface,” unaccompanied by a well-developed understanding of issues faced by women of color or how to address them.
That being said, those institutions that have taken an intersectional approach in their efforts to improve the representation of women in STEMM offer some important lessons learned. For example, a 2014 study of NSF ADVANCE5 institutional transformation (IT) grants evaluated the programs’ approaches to, and strategies for, addressing issues faced by women of color in STEMM fields (Armstrong and Jovanovic, 2017).
The study identified five “intersectional facilitators” for institutional leaders that can help drive new strategies for supporting and improving the diversity of women of color in STEMM. These included:
- Creating accountable leadership that participates actively and cooperatively in efforts aimed at supporting women of color in STEM. This includes proactive institutional leaders who are supportive in more than word and share responsibility for outcomes. An example includes “senior level administrators who
5 See https://www.nsf.gov/funding/pgm_summ.jsp?pims_id=5383.
understand significance of issues and proactively create consequences for actors who do (not) attend to those issues.”
- Recognizing and engaging with multiple “institutional climates” and adopting strategies for intervention and change in a locally intersectional context. The authors cited the example of developing an array of programming that reflects an understanding that faculty work within multiple structural climates within an institution.
- Understanding the implications of the “small N” problem and leveraging it as an opportunity to name and intervene in the dynamics of majority privilege while learning how to be effective allies to women of color. An example includes majority faculty consistently listening to underrepresented minority women and becoming responsive to their needs.
- Becoming knowledgeable about common obstacles and solutions, as well as key scholarship and research findings, on issues commonly affecting women of color in STEMM, in order to close the knowledge gap between current research and the agents of change at any given institution and among change agent team members themselves.
- Promoting “counterspaces,” or community structures that provide women of color opportunities to find others with whom they share a particular identity, allowing for collaboration or mentorship. Specifically, underrepresented minority women in STEMM should benefit directly from structures that “bring them together, increase their investment in organizational change, and allow them to define their own needs” (Armstrong and Jovanovic, 2017) (see Box 5-3).
While approaches that incorporate intersectionality in their design may have promise in addressing representation of women of color, programs that address the needs of the most marginalized populations may similarly prove to have a positive impact on all groups. The concept of universal design may be applicable in these cases.
Mace et al. (1988) coined the term universal design as “the design of products and environments to be usable by all people, to the greatest extent possible, without the need for adaptation or specialized design.” The initial conception of universal design was focused on the design of buildings and roadways to ensure accessibility to the entire public, including those with disabilities. A familiar example of universal design is the sidewalk “curb cut,” which allows individuals with mobility disabilities to more easily transition from a sidewalk in to the roadway. While designed to meet the needs of individuals with disabilities, these curb cuts also improve the experience of individuals pushing strollers, bicyclers, and many others (Schreffler et al., 2019).
There is evidence that applying a universal design approach may address equity for marginalized populations. The goal of universal design extends beyond eliminating discrimination toward people with disabilities. “A universal design benefits everyone or, at least, a large majority . . . Universal design demands creative thinking and a change in perspective. It is not sufficient merely to apply design criteria in accessibility regulations in a mechanistic way. Often a change in perspective is needed” (Steinfeld and Maisel, 2012).
Pliner and Johnson (2004) note that universal design “not only serves students with disabilities, it also serves the increasingly diverse student population at large (diversity in terms of race, class, gender identification, religion).” Newman et al., 2011 add that “universal design for learning is one way to make every lesson accessible to every student. By making STEMM content accessible to all students, colleges and universities may see an increase in STEMM enrollment by underrepresented populations. Universal design is “a goal that puts a high value on both diversity and inclusiveness. It is also a process” (Burgstahler, 2013).
By incorporating the concept of intersectionality and universal design as key components in programs, strategies, and policies to address the underrepresentation of women of color in STEMM, particularly ensuring that the most marginalized groups are at the forefront of the design, the impact will likely be felt more broadly across the STEMM enterprise.
Nevertheless, there is a clear need for additional research on the experiences of women of color in STEMM and on the impact of specific strategies and practices intended to support the improved recruitment, retention, and advancement of women in STEMM (such as those reviewed in Chapter 4), and on women of color, specifically. We should be cautious about assuming that an intervention that has benefited middle-class White women will benefit all women. The only way to determine whether this is in fact true is to carry out rigorous comparative studies. This committee urges additional research on whether and how the strategies and practices outlined in this report benefit women of color and women of other intersecting identities.
FINDINGS: CHAPTER 5
FINDING 5-1: Organizational transformation requires changing institutional culture, which in turn requires committed leadership. A lack of strong leadership support from university presidents, provosts, deans, and others is a major barrier to equity and diversity efforts. Leadership transitions are a point of vulnerability for equity and diversity efforts. Inadequate planning to implement new procedures and the failure to identify a new “champion” when turnover occurs can undermine any progress that has been made under prior leadership.
FINDING 5-2: Governing boards are an effective way to hold institutional leadership accountable for creating diverse and inclusive environments, given trustees’ roles in allocating resources and concern for public relations.
FINDING 5-3: Strong alignment among academic leaders and academic staff at all levels facilitates the implementation of research-based policies and practices.
FINDING 5-3: For equity efforts to succeed, leaders in all sectors should work to embody the respectful behavior, including meaningful communication, and equity-related policies, and ensure that institutional departments and diversity and inclusion officers receive adequate resources.
FINDING 5-4: Women faculty and male faculty of color assume, in an uncompensated manner, the vast majority of equity work. In addition, a general lack of resource allocation exists in gender-equity work, including:
- Lack of funding or teaching relief for equity work,
- Vulnerability to budget cuts or re-appropriation of funds allocated for equity work, and
- Lack of investment in equity-related initiatives from major donors and/or alumni.
Thus, there is a need for institutional leaders to set aside adequate resources to support equity and diversity efforts. School- or department-level funding support, as well as support from the federal government, can facilitate implementing and sustaining research-based policies and practices.
FINDING 5-5: To be successful, equity work needs to actively involve those who have power within their institutions. Such work is frequently delegated to university diversity and inclusion officers, who are often marginalized within their institutions, are women and minority faculty tapped by virtue of their service on relevant committees, and have limited power to bring real change.
FINDING 5-6: While there are shared characteristics of successful programs, and common institutional barriers and facilitators to sustainably implementing these practices, institutions have different goals and missions, values, culture, and resources and this institutional context can impact the efficacy of any program, regardless of that program’s designation as “successful” and “evidence-based.” There is no one-size-fits-all solution, policy, or practice that will perfectly fit the needs of all institutions.
FINDING 5-7: Given that it can be difficult to predict which interventions will be successful in which institutional contexts, it is necessary for individual institutions and organizations to collect data and monitor trends in the recruitment, retention, and advancement of women in STEMM to better adopt targeted interventions and monitor their efficacy. Collecting and monitoring data and evaluating it over time, employers and admissions officers can increase utility in ascertaining whether recruitment, retention, or advancement (or some combination) is the major issue affecting low representation of women.
FINDING 5-8: Not all women in STEMM benefit equally from policies and practices designed to support their representation, advancement, and academic contributions. Participants suggested that such efforts tended to be unevenly successful depending on women’s identitites (particularly race and ethnicity), their career stages, and the institutional contexts in which they worked.
FINDING 5-9: By incorporating the concept of intersectionality and universal design as key components in programs, strategies, and policies to address the underrepresentation of women of color in STEMM, particularly
ensuring that the most marginalized groups are at the forefront of the design, the positive impact will likely be felt more broadly across the STEMM enterprise.
FINDING 5-10: A clear need for additional research exists, specifically on the experiences of women of color in STEMM and on the impact of specific strategies and practices intended to support the improved recruitment, retention, and advancement of women in STEMM on women of color and women with other intersecting identities.
This page intentionally left blank.