Factors that Drive the Underrepresentation of Women in Scientific, Engineering, and Medical Disciplines1
The analysis draws substantially from the research paper by Drs. Michelle Rodrigues and Kathryn Clancy, which was commissioned for this study. The full research paper can be found online at: www.nap.edu/catalog/25585.
For decades, sustained investments from foundations, nonprofits, government agencies, and others have supported efforts to improve the representation of girls and women2 in science, technology, engineering, mathematics, and medical (STEMM) fields.3 Why these efforts to improve recruitment, retention, and advancement have effected little improvement in gender representation in many STEMM fields remains an open question. While it is true that in many STEMM fields the situation has gotten better, many are concerned that the rate of improvement has been too slow and that progress has plateaued, or even moved backward in some cases. These fears are not unfounded. For example, even though the percentage of women earning bachelor’s degrees in engineering doubled from 2001 to 2010, their numbers in 2010 were still extremely low at 16 percent, and even slightly declined by 2015 (Armstrong and Jovanovic, 2015; Nassar-McMillan et al., 2011; NSF, 2002; NSF, 2017). Similarly, even though women’s
1 This chapter offers a comparative examination of research on why women are more underrepresented in some scientific, engineering, and medical disciplines than others, with a particular focus on computer science, engineering, physics, mathematics, medicine, chemistry, and biology. The authors wish to acknowledge the significant contribution of the Committee on Understanding and Addressing the Underrepresentation of Women in Particular Science and Engineering Disciplines to the content of this chapter.
2 In the context of this study, “women” are defined as those who identify as women or are viewed by others as women (or female in the case of girls).
3 Given that medicine is within the scope of this report, the acronym “STEMM” is used when referring to ideas or comments originating with this report; however, when referring to research, evidence, or programs that exist independent of this report, we use the terminology of the material cited, whether it is STEM (science, technology, engineering, and mathematics), STEAM (science, technology, engineering, art, and mathematics), or STEMM.
representation in the physical sciences improved during this time period, women still accounted for only 22 percent of these disciplines in 2010 (Armstrong and Jovanovic, 2015; NSF, 2013). Even in medical disciplines, where, as of 2018, the number of women enrolled in medical schools exceeded men for the first time, there is a persistent underrepresentation of women at senior academic or leadership positions that cannot be explained by a time lag between degree completion and career trajectory (see Figure 2-1).
In this chapter, the committee reviews research on the shared experiences of women across a range of STEMM disciplines, explores the patterns of representation of women across seven specific disciplines—computer science, engineering, physics, mathematics, medicine, chemistry, and biology—and highlights the importance of considering the intersectional experiences of women of multiple marginalized identities (race, class, sexual orientation, disability status) when considering the biases and barriers facing women in STEMM.
BARRIERS TO WOMEN’S PROGRESS
It is not true that women are underrepresented in STEMM because of innate weakness in these fields (NASEM, 2007; Pawley, 2011). Rather, substantial research demonstrates that implicit and explicit biases discourage women from entering STEMM careers (Cheryan et al., 2015; Lehmann et al., 2006; Master et al., 2016) or influence their decision to leave STEMM after beginning their careers (Hunt, 2016). These factors include a spectrum of explicit and implicit biases, as well as structural and interpersonal interactions that impede women’s progress (Grogan, 2018; Urry, 2015). These are factors across the career life cycle:
- Obtaining a Position—bias in recruitment (Milkman et al., 2015; Moss-Racusin et al., 2012); obstacles to accommodating family needs (Urry, 2015; Wolfinger et al., 2008)
- Internal Opportunities and Rewards—unequal allocation of resources (Bronstein and Farnsworth, 1998; Green et al., 2000); mentoring and performance evaluation (e.g., teaching) (MacNell et al., 2014; Reid, 2010a); mentoring access (Chanderbhan-Forde et al., 2012; Moss-Racusin et al., 2012)
- Work Expectations—higher teaching loads (Bronstein and Farnsworth, 1998; Carrigan et al., 2017; Xu, 2008); higher expectations of service without compensation (Hermanowicz, 2012; Kulis et al., 2002; Madge and Bee, 1999)
- External Opportunities and Rewards—lower frequency of speaking invitations (Nittrouer et al., 2018); inequities in access to external funds (Pohlhaus et al., 2011; Witteman et al., 2017), less likely to be on editorial boards and in editor positions (see, for example, Amrein et al., 2011; Cho et al., 2014; Clark and Horton, 2019; Ioannidou and Rosania, 2015)
- Cross-cutting Barriers—harassment and assault (Clancy et al., 2014, 2017; NASEM, 2018b; Nelson et al., 2017)
While women in all fields face bias and discrimination, the way women experience these behaviors differs by discipline and career stage, leading to similarities or differences that create unique climates for women across the STEMM enterprise.
COMMON DYNAMICS ACROSS STEMM FIELDS
Across the STEMM fields, women may experience implicit bias and structural barriers at every career stage, including at critical junctures such as consideration for graduate school admission, recruitment into a laboratory for graduate research, consideration for postdoctoral positions, recruitment to fill tenure-track faculty positions, and evaluation for promotion in rank (Bronstein and Farnsworth, 1998; MacNell et al., 2014; Milkman et al., 2015; Moss-Racusin et al., 2012; Settles et al., 2006; Urry, 2015). These biases are often intensified for women of color, who encounter the double bind of race- and gender-based bias (see “Intersectionality and the Double Bind” section below).
Biases have cumulative effects leading to outsized disparities at more advanced career levels. For example, Li et al. (2019) found that, when junior scientists had the opportunity to coauthor a manuscript with a well-known scientist during the first few years of their career, they experienced a “persistent competitive advantage” throughout their careers compared with those who did not have the same authorship opportunities (Li et al., 2019). When biases result in identification of male students as more promising candidates for initial research experiences, the effect of that bias reverberates, continuing to provide additional opportunities for career advancement for that student. Men are more likely to be
evaluated in ways that lead to opportunities for better pay and mentorship (Moss-Racusin et al., 2012). In a randomized, double-blind study, 127 science faculty at research universities rated a male applicant for a laboratory manager position as “significantly more competent and hirable” than a female applicant, even though the application materials were identical except for one factor: the name of the applicant appearing either as “John” or “Jennifer.” Faculty also offered “John” a higher starting salary and more career mentoring than they offered “Jennifer” (Moss-Racusin et al., 2012). Additionally, culturally engrained biases especially favor White men over men of color and women (Milkman et al., 2015). In a study in which more than 6,500 professors at top U.S. universities, drawn from 89 disciplines and 259 institutions, were contacted by fictional prospective students wishing to discuss research opportunities with names that suggested their gender and race (White, Black, Indian, Hispanic, and Chinese), faculty were significantly more responsive to White males than to all other categories of students. These biases in favor of White males are rooted in deeply seated cultural associations between masculinity and STEMM (see “Cultural Associations Between Masculinity and STEMM” section below).
In academic positions within STEMM, women are more likely to be appointed to teaching-focused positions, where they have less access to external funding or resources and to graduate students (Hermanowicz, 2012; NASEM, 2018b). For faculty positions that focus primarily on scholarship, disparities in teaching evaluations, often rooted in implicit bias, disadvantage women, especially women of color, when being considered for tenure (Jones et al., 2015a; Pittman, 2010; Reid, 2010b). Women are also less frequently invited to be colloquium speakers than men, particularly at prestigious universities (Klein and Briggs, 2017).
With respect to publishing, women are less likely to receive authorship credit and more likely to experience harsher peer review; moreover, manuscripts with women listed as first or last authors are cited less frequently (Bendels et al., 2018; Chawla, 2018; Murray et al., 2018; West et al., 2013). In contrast, men are more likely to receive first authorship or last authorship, and are more likely to be invited by journal editors to serve as reviewers (Chawla, 2018; Lariviere et al., 2013; Murray et al., 2018). Compounding the disparity, all-male reviewing teams are more likely to reject papers from women (Chawla, 2018; Murray et al., 2018).
There are additional gender disparities in receiving grant funding (Pohlhaus et al., 2011; Witteman et al., 2017). Because women are fewer in number among biomedical research faculty, the rate of application by women for National Institutes of Health funding is therefore lower, but women are also less likely to have their funding renewed after the award has been made (Pohlhaus et al., 2011). Even in cases when women’s research is evaluated favorably for funding, their performance and research accomplishments as principal investigator are likely to be evaluated more harshly than that of their male peers (Witteman et al., 2017). As a result, men continue to be funded at higher rates.
Even within fields where women are well represented or overrepresented at lower ranks, they do not have equivalent representation at higher ranks (Addessi et al., 2012; Carnes, 2008; Carr et al., 2017; Isbell et al., 2012; Sheltzer and Smith, 2014). The gendered divisions of labor that exist within academia may be responsible for this disparity. Women shoulder the burden of teaching, mentoring, and service (Armstrong and Jovanovic, 2015; Hermanowicz, 2012; Kulis et al., 2002; Madge and Bee, 1999; Urry, 2015), particularly White women in male-dominated fields and minority women in all fields, who, as Johnson and Lucero (2003) describe, pay a “cultural tax,” whereby they are expected to perform additional service work related to their identity (Armstrong and Jovanovic, 2015). Furthermore, women are often marginalized in low-status jobs such as nontenure-track positions or unstable research associate positions dependent on soft money (Kulis et al., 2002). Across a variety of fields, as women increase in representation, the status and compensation associated with these fields decreases (Kulis et al., 2002; Reskin, 1988). Even as women rise to higher ranks, they themselves often contribute to the perpetuation of culturally ingrained biases—women are just as likely as men to evaluate female candidates negatively, and high proportions of White women and minorities exhibit gender biases in evaluating prospective students (Milkman et al., 2015; Moss-Racusin et al., 2012).
Women also disproportionately deal with the impact of being a member of a “dual-career couple.” Such couples face the challenge of finding employment at the same institution or city, as well as the impact of inadequate maternal/paternal leave and childcare policies (Urry, 2015). For example, Rivera (2017) found that hiring committees for junior faculty positions considered women’s relationship status but not men’s relationship status when making hiring decisions. Hiring committees excluded heterosexual women with partners who held academic or other high-status jobs that were not easily movable when there were male or single female alternatives. Rivera (2017) also found that committees rarely discussed the relationship status of male faculty and saw their female partners as movable. While the exit of women from STEMM has been framed as a choice based on prioritizing relationships and/or motherhood by some (Ceci and Williams, 2011), gender inequities in cultural expectations combined with bias against women who have children, or may potentially have children, often heavily influence these decisions (Wolfinger et al., 2008).
Women faculty at Historically Black Colleges and Universities (HBCUs), Tribal Colleges, and other Minority-Serving Institutions also experience the gender bias-related issues outlined in this report. With the exception of HBCUs, where roughly half of faculty are Black, the faculties in STEMM disciplines of Minority-Serving Institutions, including Tribal Colleges, Hispanic-Serving Institutions, and Asian American and Native American Pacific Islander-Serving Institutions, have demographics that are similar to demographics of faculty in majority institutions (NASEM, 2019a), with a majority of White faculty. Moreover, African American women faculty experience the same issues of bias, discrimination, and uncivil climate at HBCUs as they do at majority institutions (Bonner, 2001).
Although negative biases are more frequently reported, absences of hiring biases and biases favoring women in academic sciences have been reported (NRC, 2010; Williams and Ceci, 2015). According to the 2010 National Academies report Gender Differences at Critical Transitions in the Careers of Science, Engineering, and Mathematics Faculty, female applicants generally fared better than their male counterparts in tenure-track applications to research-intensive universities (NRC, 2010). In all six STEMM fields studied, the percentage of women increased—often substantially—from the applicant pool to interviews to job offers. In electrical engineering, for example, women comprised 11 percent of applicants for tenure-track positions, but 32 percent of those who received job offers. Additionally, a rigorous quantitative synthesis of the experimental literature on gender bias in job-related decisions such as evaluations of competence and hirability showed that, in general, men were preferred for male-dominated jobs, whereas no strong preference for either gender was found for female-dominated jobs (Koch et al., 2015).
However, biases against hiring women still exist in some contexts, such as in hiring laboratory managers (Moss-Racusin et al., 2012) and postdoctoral associates (Eaton et al., 2019). These two contexts are notable because such hiring decisions are often made by individual principal investigators with little administrative oversight and formal monitoring.
Preference for hiring women in tenure-track positions may be explained by the competitiveness and accountability of such positions by them. Because dozens or even hundreds of candidates may apply for certain types of tenure-track positions, search committees typically select only the most outstanding candidates to create a short list of individuals to invite for interviews (Carpenter and O’Neal, 2013). Competition has become more intense, particularly since 2000, as the number of STEMM Ph.D. earners has continued to rise, while the availability of tenure-track positions has remained unchanged (Gould, 2015; NASEM, 2018a). The competition for laboratory manager and postdoctoral positions is likely less intense, and competency of short-listed candidates may be less easily defined than in tenure-track hiring contexts. In a meta-analysis of gender stereotypes and bias in experimental simulations of employment decision-making, Koch et al. (2015) found applicant competence is a critical moderator of gender bias.
Another consideration is that tenure-track hiring at academic institutions involves more accountability than does hiring into other positions. Koch et al. (2015) reported that for male-dominated jobs, no significant difference in average gender bias could be detected when raters “felt accountable for their decisions, believed their decision had real-life consequences, or were reminded of equity norms.” Accountability may arise from pressure from deans and administrators to hire more female faculty in STEMM (NASEM, 2007). In other words, professors are often held accountable to work toward diversity goals in tenure-track hiring, but likely encounter less administrative oversight in other contexts, such as laboratory manager hiring (NASEM, 2007). Goals for diversifying the professoriate
might even create a preference for outstanding women applying to tenure-track positions over equally qualified men, creating a favorable hiring bias for women (NRC, 2010). However, it is likely that these studies underestimate the processes by which many women are excluded from participation in the field at early stages and the ways that small biases in treatment accumulate over time (Ginther, 2006; Mason and Goulden, 2004; Rigney, 2010; Valian, 1998).
Cultural Association Between Masculinity and STEMM
There is a long-standing cultural association between masculinity and objectivity in most segments of American society, which in turn, underlies the associations of masculinity with STEMM (Bejerano and Bartosh, 2015). When Carli et al. (2016) asked study participants to list traits they associate with scientists and with men and women irrespective of profession, the traits identified for scientists and men overlapped to a greater extent that did the traits identified for scientists and women. Moreover, nonscientists are less likely to believe a woman is a scientist if she has a feminine (rather than masculine) appearance (Banchefsky et al., 2016). This expectation that STEMM professionals are White and male is implicitly conveyed in cultural portrayals of STEMM and STEMM education (Banchefsky et al., 2016). These stereotypical associations shape the social and educational environments of children, as well as structural patterns that occur in STEMM professions (Banchefsky et al., 2016).
Traits such as assertiveness, confidence, boldness, risk-taking, independence, and self-promotion are valued, rewarded, and seen as standards in STEMM (Diekman and Steinberg, 2013). Stereotypically “masculine traits” (e.g., assertiveness, ambition, and competitiveness) and “feminine traits” (e.g., warmth, supportiveness, and collegiality) are exhibited by both women and men and, importantly, individual men and women exhibit these traits on a spectrum (Diekman and Goodfriend, 2006). However, many women have less experience with these masculine traits because they are often socialized to be more “other-focused” than their male counterparts (Eagly and Mladinic, 1994; Eagly and Riger, 2014). When women do display these traits, they often encounter backlash in the form of social and economic sanctions (Rudman, 1998; Williams and Tiedens, 2016). In addition to undermining the advancement of women in STEMM to positions of leadership, masculine values can signal to women that they do not belong in these fields in the first place (Bian et al., 2017a).
In reality, a range of traits and competencies, independent of associations with stereotypes, can be differentially advantageous or disadvantageous depending on situations. Negotiation styles of women, for example, tend to be more relationship driven (or more focused on the quality of relationships) than the more stereotypically male outcome-driven style (focused on specific outcomes). Relationship-driven negotiation has been associated with better outcomes in business negotiations, dispute resolution, social movements, marriage reconcili-
ation, crisis resolution, or peacekeeping. For this reason, scholars have increasingly noted the tendency for women to conceptualize issues such as security and the use of military force in different and more productive ways than their male counterparts (Babcock and Laschever, 2003; Boyer et al., 2009). Also, women perform as well as, or better than, men in leadership competency (Folkman, 2015), not only with respect to characteristics typically associated with women (i.e., “nurturing competencies such as relationship building and developing others”), but also for characteristics typically associated with men (i.e. “takes initiative,” “practices self-development,” “displays high integrity and honesty,” and “drives for results” (Zenger et al., 2012). Finally, Tsugawa (2017) noted that, when treated by female physicians, more than 1 million elderly hospitalized patients were less likely to die within 30 days of admission or to be readmitted within 30 days of discharge than those cared for by male physicians. The author estimated that, if male physicians could achieve the same outcomes as their female colleagues, there would be 32,000 fewer deaths each year among Medicare patients and offered as a conclusion the following statement:
There was ample evidence that male and female physicians practice medicine differently. Our findings suggest that those differences matter and are important to patient health. We need to understand why female physicians have lower mortality so that all patients can have the best possible outcomes, irrespective of the gender of their physician.
Further, cultural expectations and biases about which jobs and careers are held, or should be held, by women and men can limit men’s opportunities in certain STEMM fields. Gender discrimination, biases, stereotypes, and microaggressions against men have been well documented, primarily in health fields, such as nursing and other health sciences, where men are not well represented. According to the Kaiser Family Foundation, as of April 2017, only 333,530 of the 4 million nurses in the United States identified as male (KFF, 2017). In a survey examining reasons for this lack of male participation in the field, respondents widely cited stereotypes as a top challenge. Other barriers cited included a lack of career support, few male nurse educators and mentors, a feeling of being unwelcome in the clinical setting, and sex-related bias in obstetric rotations (Hart, 2004). Men have also reported experiencing microaggressions, particularly if they demonstrate an aptitude or interest in a STEMM field that is primary dominated by women, such as nursing. This can contribute to a lack of men pursuing careers in the field (Hart, 2004).
Similarly, there is male gender segregation in a number of medical specialties. This has been the case, in particular, in medical subspecialties where women are well represented and the field has thus faced a corresponding drop in prestige and pay. For example, despite being a female-dominated field, obstetrician-gynecologists who are women face barriers to advancement to leadership positions and earn $36,000 per year less than men in obstetrics and gynecology
(Hughes and Bernstein, 2018). Similarly, men in obstetrics and gynecology may be negatively affected by unconscious bias and socially prescribed roles for men and women (Hughes and Bernstein, 2018). In another study examining performance and gender representation in obstetrics and gynecology clerkships, male students reported that their gender negatively affected their experience during the clerkship (Craig et al., 2018). Additionally, there are fewer male students applying for obstetric and gynecology residency (Craig et al., 2018). As Hughes et al. (2018) notes, “For the obstetrician-gynecologist, sexism is not just a ‘women’s issue.’”
Women experience high rates of sexual harassment in science, engineering, and medical education and careers (NASEM, 2018b). Sexual harassment consists of three forms: gender harassment (verbal and nonverbal behaviors that convey hostility, objectification, exclusion, or second-class status about members of one gender); unwanted sexual attention (unwelcome verbal or physical sexual advances, which can include assault); and sexual coercion (when favorable professional or educational treatment is conditioned on sexual activity) (NASEM, 2018b).
Women commonly and disproportionately experience sexual harassment at multiple career levels. Surveys from a university system and a university with multiple campuses demonstrate that 20–50 percent of women students experience sexual harassment from faculty or staff, depending on their stage of education and field (Krebs et al., 2016; Swartout, 2018). The best meta-analysis of surveys to date indicates that more than 50 percent of women employees (faculty and staff) in academia experience sexual harassment (Ilies et al., 2003). Research shows that these numbers are far worse for women with intersecting marginalized identities (Buchanan et al., 2008; Clancy et al., 2017; Cortina, 2004; Cortina et al., 1998; Konik and Cortina, 2008; Rabelo and Cortina, 2014). Although men can and do experience sexual harassment (APA Handbook of the Psychology of Women, 2018; Berdahl, 2007; Rabelo and Cortina, 2014), they do so at considerably lower rates (Sexual Harassment in the Federal Workplace: Trends, Progress, Continuing Challenges, 1995; Ilies et al., 2003; Kabat-Farr and Cortina, 2014; Magley et al., 1999).
Of the three types of sexual harassment, gender harassment is the most common and can be as harmful as the other forms of sexual harassment (NASEM, 2018b). Examples of gender harassing behavior include comments that denigrate women as a group or as individuals in gendered terms, and comments about women that are crude or sexist. Gender harassing behavior can also include visual behavior such as leaving porn or lewd images in group spaces. Gender harassment is often ambient, meaning it is “not clearly targeted at any individual or group of individuals” (Parker, 2008) and can include behavior that extends beyond the direct target of the harassment (Glomb et al., 1997). Gender harass-
ment can include uncivil and disrespectful behavior sometimes described as “microaggressions” (Sue et al., 2007) (discussed more in-depth below). Additionally, sexual harassment often takes place within environments in which incivility occurs (Lim and Cortina, 2005).
A 2018 National Academies consensus study report concluded that the cumulative result of sexual harassment in academic sciences, engineering, and medicine is significant damage to research integrity and a costly loss of talent in these fields (NASEM, 2018b). Research across workplace sectors shows that sexual harassment “undermines women’s professional and educational attainment and mental and physical health,” leading to negative career outcomes (NASEM, 2018b). When women experience sexual harassment in the workplace, the professional outcomes include increases in job stress (Barling and Cooper, 2008; Fitzgerald et al., 1997); declines in job satisfaction, performance, or productivity (Bond et al., 2004; Cortina et al., 2002b; Fitzgerald, 1997; Glomb et al., 1999; Harned and Fitzgerald, 2002; Holland and Cortina, 2013; Lim and Cortina, 2005; Magley and Shupe, 2005; Morrow et al., 1994; Munson et al., 2000; Piotrkowski, 1998; Ragins and Scandura, 1995; Schneider, 1997), and withdrawal from the organization and disengagement from their work (Barling et al., 2001; Cortina et al., 2002a; Culbertson and Rosenfeld, 1994; Fitzgerald et al., 1997; Glomb et al., 1999; Holland and Cortina, 2013; Lonsway et al., 2013; Schneider et al., 1997; U.S. Merit Systems Protection Board, 1995; Wasti et al., 2000). When students experience sexual harassment, the educational outcomes include greater truancy, dropping courses, receiving lower grades, or dropping out (Duffy et al., 2004; Fitzgerald, 1990; Lee et al., 1996; Reilly et al., 1986). As the 2018 Sexual Harassment of Women report (NASEM, 2018b) concluded, sexual harassment is a significant factor influencing the recruitment, retention, and advancement of women in STEMM, and its persistence in the workplace and education environments is putting at risk the gains made in improving the representation of women in these fields.
“Microaggressions” refer to “the everyday verbal, nonverbal, and environmental slights, snubs, or insults, whether intentional or unintentional, which communicate hostile, derogatory, or negative messages to target persons based solely upon their marginalized group membership. In many cases, these hidden messages may invalidate the group identity or experiential reality of target persons, demean them on a personal or group level, communicate they are lesser human beings, suggest they do not belong with the majority group, threaten and intimidate, or relegate them to inferior status and treatment” (Sue, 2017). Though the gender harassing form of sexual harassment commonly overlaps with microaggressive behaviors (Sue et al., 2007), microaggression is a broader form of discrimination that can extend beyond gender into race, identity, religion, and
many other legally protected characteristics. Microaggressions can contribute to feelings of alienation, pressure to work twice as hard to receive recognition, and work environments in which one is under constant scrutiny and presumed incompetent (Johnson et al., 2011; McGee, 2016; McGee et al., 2019; Ong et al., 2011). Importantly, microaggressions have devastating short- and long-term effects on both targets and bystanders (Ruder et al., 2018; Torres et al., 2010; Wilkins-Yel et al., 2019; Yang and Wright, 2018).
The extent to which microaggressions specifically are present in STEMM is unclear (Harrison and Tanner, 2018), although a few research studies indicate that it is a barrier for women in these fields. In one study using interviews with 21 women in physics and astronomy programs, Barthelemy et al. (2016) found that the majority of subjects experienced microaggressions. In another study examining women STEMM faculty at a large Midwestern institution, 68.8 percent of those interviewed reported experiencing a workplace microaggression (Rockinson-Szapkiw and Wade-Jaimes, 2019). Furthermore, faculty rank did not predict faculty experiences with microaggressions, indicating that women experience microaggressions at all stages of their faculty career.
One reason there may not be many studies on microaggressions in STEMM conexts is that use of the term “microaggression” to characterize these behaviors is considered by some researchers to be misleading. Micro implies insignificant, minor, or imperceptible; many behaviors that are categorized as “microaggressions” are actually overtly offensive and extremely damaging. Further, aggression is a term most commonly reserved for behavior that carries intent to harm (Lilienfeld, 2017), which is not always the case with the behaviors included under the term “microaggressions”—as the definition above makes clear. An alternative term more commonly used in workplace aggression literature, and throughout the 2018 National Academies consensus study report on the Sexual Harassment of Women, is the term “incivility,” which refers to “low-intensity deviant behavior with ambiguous intent to harm the target, in violation of workplace norms for mutual respect” (Andersson and Pearson, 1999). Regardless of the term used—microaggressions or incivility—the behavior is harmful and a barrier to the progress of women and particularly women of color.
INTERSECTIONALITY AND THE DOUBLE BIND
Intersectionality can be defined as “the processes through which multiple social identities converge and ultimately shape individual and group experiences” (McCall, 2005; Museus and Griffin, 2011). Structural intersectionality refers to the ways in which multiple social systems intersect to shape the experiences of individuals (Crenshaw, 1991). Many employers, including those at educational institutions have adopted programs and the policies aimed at improving equity and diversity in STEMM without considering the complex, cumulative ways in which multiple intersecting identities influence outcomes of the interventions. For
women of color in particular, multiple forms of discrimination, such as racism and sexism, intersect to shape their experiences.
Research demonstrates that programs aimed at improving the representation of women in STEMM have largely benefited White women and that intersecting identities can influence the efficacy of interventions to achieve gender equity. Ong et al. (2011) suggested that the absence of sustained efforts to serve and support women of color in STEMM may be “possibly due to the misguided idea that burgeoning efforts by the NSF [National Science Foundation] and other institutions aiming to serve women or minorities would, consequently, serve minority women.” The authors further note that “history has borne out the reality that programs intended to serve women disproportionately benefit white women, and programs intended to serve minorities mainly benefit minority males” (Ong et al., 2011).
Women of Color4
Strategies and practices with potential for improving the retention, persistence, and achievement of women in STEMM, particularly women of color, have been developed and deployed. The strongest indicator of the effectiveness of such strategies and practices is the changing number of women of color entering and remaining in STEMM. That said, even though the share of science and engineering degrees earned by underrepresented minority women has more than doubled over the past two decades at all levels of higher education (bachelor’s, master’s, and doctorates) (see Figure 2-2), women of color (with the exception of Asian American women) remain underrepresented in these fields relative to their representation in the U.S. population (NSF 2013, 2017) (see Figure 1-2). Minority women have been awarded more STEMM degrees as measured in absolute numbers since the 1970s but remain underrepresented at advanced education and career stages in most fields relative to White women (Ong et al., 2011).
In a groundbreaking paper relating to underrepresentation of women of color, Malcolm (1979) presented the problem as a “double bind,” where women of color are excluded for biases related to both their gender and their race and ethnicity (Malcom, 1979). In spite of the attention called to the double bind over 40 years ago, it remains a major issue for women in STEMM. As recently as 2019, in a
4 We include in our definition of women of color African Americans, Hispanics, Latinas, American Indians, Asian Americans, Alaska Natives, Native Hawaiians, and other Pacific Islanders. Although Asian American women are overrepresented among STEM degree earners, they remain underrepresented in ranks of full professor and in university leadership (e.g., deans or university presidents) (U.S. Department of Education, 2019). Similarly, Asian American women are poorly represented on corporate boards of trustees and among managers in industry or government (Deloitte, 2018). For this reason, we include Asian American women in our analysis.
study of postdoctoral hiring bias, researchers examined how perceptions of race and gender influence evaluation of postdoctoral candidates (n = 251) from eight large research universities. Professors were asked to read one of eight identical curricula vitae (CV) of a hypothetical doctoral graduate applying for a postdoctoral position, and rate them for competence, hirability, and likeability. The candidate’s name on the CV was used to suggest race (e.g., Asian, Black, Latinx, and White) and gender (female or male). Physics faculty rated the CVs of Black women and Hispanic women lower than the CVs of women and men from any other racial/ethnic group (Eaton et al., 2019). That said, gender gaps in STEMM can vary in unpredictable ways across racial and ethnic lines. Women with multiple marginalized group identities, such as women of color, can experience both advantages and disadvantages compared with those with a single subordinate group identity (Purdie-Vaughns and Eibach, 2008). Black women may be overlooked or marginalized due to “intersectional invisibility”—a lack of visibility because they do not embody expectations of “women” or “Black people” (Bell, 1992; Davis, 1981; Purdie-Vaughns and Eibach, 2008). But this invisibility can also protect racial minority women by making them less conspicuous targets of common biases and stereotypes (Biernat and Sesko, 2013). Along the same lines, Black and Hispanic men often face negative stereotypes about tendencies for engaging in criminal or violent behavior; Black and Hispanic women encounter these stereotypes far less frequently (Ghavami and Peplau, 2012). As well, in a study on callbacks for jobs, Mullainathan and Bertrand (2004) found that Black women were less disadvantaged than Black men—although there was a bias against Black relative to White applicants.
In addition to experiences of heightened bias, women of color in STEMM frequently experience isolation (i.e., experience a sense of invisibility or hypervisibility), macro- and microaggressions, and a sense of “not belonging” in STEMM (Ong et al., 2011). Beyond feelings of isolation, there is evidence to indicate that both women of color and White women in STEMM have more limited social network supports than men, which can tangibly and negatively impact their career trajectory (Collins and Steffen-Fluhr, 2019; Etzkowitz et al., 1994; Feeney and Bernal, 2010). In other words, women may not only feel isolated but, may actually be isolated.
Experiences of bias, isolation, microaggressions, and not belonging in STEMM can lead to “racial battle fatigue,” a term coined by William Smith (Smith et al., 2007). Racial battle fatigue is the “cumulative result of a natural race-related stress response to distressing mental and emotional conditions” that adversely affects the health and achievements of students and faculty of color (Corbin et al., 2018; Smith et al., 2007). While this term was coined to describe the experiences of Black men in predominately White spaces, it has been since expanded to be inclusive of women of color in historically and predominantly White spaces, particularly for Black and Latina women (Corbin et al., 2018; Franklin et al., 2014). Women of color also experience more harassment than White women, which manifests as both
racial harassment and intensified forms of sexual harassment (Berdahl and Moore, 2006; Buchanan and Fitzgerald, 2008).
While most women report career-life balance as a challenge to working in STEMM fields, for women of color, this may be a critical factor contributing to why they remain underrepresented in these fields (Kachchaf et al., 2015). Academic STEMM work environments often require a commitment to the job through long hours, stagnant career trajectories, and constant availability and visibility—that is, the “ideal worker norm” (Acker, 1990; Williams, 2000). These expectations assume that there will be gendered separation of work and family duties (Traweek, 1988; Williams, 2000), which further reinforces the image of the scientist as a man. Adherence to these ideal worker norms disadvantages both men and women who have commitments and duties outside of work (Acker, 1990; Williams, 2000). The ideal worker norm has a disproportionate impact on women of color because their multiple identities and their small numbers in STEMM departments contribute to an even greater perception that they do not demonstrate the characteristics of the “ideal worker” (Turner, 2002).
Women of color may also contend with cumulative disadvantage, such as interest on debt, and disadvantages, such as lower salary and delayed promotion, which accrue over time. Those who differ from the norm experience a cycle of disadvantage—the further from the norm, the more cumulative the disadvantage (Kachchaf et al., 2015). Kachchaf et al. (2015) report that women of color must work harder, including working extended hours, to fit the ideal worker norm despite having had fewer role models who have successfully managed these expectations, fewer culturally competent mentors, and less access to informal professional networks.
Women with Disabilities
People with disabilities are underrepresented in STEMM from K-12 through higher education and continuing within the workforce. Although individuals with disabilities report nearly identical interest in pursuing STEMM as those without disabilities (~25 percent) (Thurston et al., 2017), far fewer persons with disabilities graduate with STEMM degrees (NSF, 2017).
Beyond the barriers faced by women in STEMM in general, women with disabilities encounter unique obstacles related to their disabilities that may be responsible for their disproportionate underrepresentation in STEMM careers. These barriers may include lack of physical access to laboratory and classroom spaces, lack of equipment that can be used by persons with sensory and motor disabilities, a shortage of disabled role models in STEMM, and a higher likelihood of negative mentoring interactions (Duerstock and Shingledecker, 2014). Teaching styles in undergraduate classrooms can also contribute to attrition; for example, large lecture-style courses, particularly when inclusive pedagogy is not prioritized, may serve as a “weeding” class for students with disabilities (Moriarty, 2007).
Despite these barriers, there are a number of direct interventions that can vastly improve outcomes for students with disabilities. For example, providing students with assistive and adaptive technology, such as software that makes printed pages more accessible and facilitates writing, equipment that supports auditory and visual comprehension, and laboratory environments that are designed to be accessible greatly improve the educational experiences for students with disabilities (Duerstock and Shingledecker, 2014; NASEM, 2019b). Other interventions include first-year college transition programs to provide supplemental support for students with disabilities, access to culturally sensitive mentorship, access to tutors (particularly for disciplines that require complex computational methods and concepts), and access to individualized advising (Duerstock and Shingledecker, 2014; NASEM, 2019b). However, few if any studies have identified factors that influence career retention for female scientists with disabilities.
Women with LGBTQIA Identities
Women who identify as LGBTQIA face significant barriers in STEMM, in part due to their intersectional identities—being both a woman and a sexual minority. What has been studied on this topic indicates that women who are LGBTQIA are particularly marginalized across STEMM fields, and that while some interventions and recruitment efforts have increased representation of this population, the reality is that the numbers are not improving, and, in some cases, are getting worse.
For example, Yoder and Mattheis (2016) conducted a survey of 1,427 individuals who identify as LGBTQIA working in STEM fields, known as the “Queer in STEM” survey. Participants completed a 58-item questionnaire to report their professional areas of expertise, levels of education, geographic location, and gender and sexual identities and rated their work and social communities as welcoming or hostile to queer identities. Almost one-half of participants identified as female (48 percent); 44 percent identified as male, 7 percent as transgender, 4 percent as androgynous, and 9 percent as genderqueer. LGBTQIA participants reported that they felt excluded from STEM workplaces and professional culture. Faculty level also appeared to be a factor in their comfort in being open to colleagues about their sexuality. The authors found that early-career academics (for example, postdoctoral researchers, medical residents, laboratory technicians, or managers) reported lower openness to colleagues than survey participants at later career stages (e.g., assistant, associate, and full professors, or emeritus/retired) (Yoder and Mattheis, 2016).
This level of openness also varied by STEM field. Participants working in earth sciences, engineering, mathematics, and psychology reported being less out to colleagues, and participants working in the life sciences, physical sciences, and social sciences reported being more out (see Figure 2-3) (Barres et al., 2017; Yoder and Mattheis, 2016). Similarly, in another study that examined LGBTQIA scientists in physics found that LGBTQIA scientists may feel the need to remain
closeted if they are unsure of their advisor’s perspective on their rights and personhood (Atherton et al., 2016).
Yoder and Mattheis’s (2016) research suggests that better representation of women in STEMM is associated with greater inclusion of those who are stereotyped as not conforming to gender roles, in that LGBTQIA scientists working in STEMM fields with better representation of women were more likely to disclose their identities to their colleagues.
In addition to openness, individuals with minority genders, sexual orientation, or both experience higher rates of sexual harassment and assault than cisgender straight women (Brewster et al., 2012, 2014; Eliason et al., 2011). In a recent survey of sexual and gender minorities (n = 474) in astronomy and planetary sciences, LGBTQIA women and gender minorities were more likely to experience homophobic and transphobic remarks from their peers, were more likely to feel unsafe at work due to their racial, gender, and/or sexual identities compared with cisgender straight women, and were more than twice as likely to experience assault at work. All of this leads to a loss of opportunity and contributes to the underrepresentation of LGBTQIA individuals in astronomy and planetary sciences (Richey, 2019).
Regarding retention in STEM, Hughes et al. (2018), using national longitudinal survey data, examined whether students who identified as a sexual minority were more or less likely to persist after 4 years in STEM fields, as opposed to switching to a non-STEM program compared with their heterosexual peers. The authors found that LGBTQIA students were 9.54 percent less likely to be re-
tained in STEM than their heterosexual peers. However, they also noted that this group was also far more likely to report participating in undergraduate research programs. In fact, LGBTQIA students were nearly 10 percentage points more likely to participate in undergraduate research than their heterosexual peers. This may indicate that LGBTQIA students are interested in participating in STEMM research at the undergraduate level, but are more likely than their non-LGBTQIA peers to leave these fields at later career stages (Hughes, 2018).
While the committee identified few interventions targeted to LGBTQIA women in STEMM, preliminary research indicates that raising intersectional bias awareness in college classes can encourage positive changes in attitudes and beliefs (Case and Lewis, 2012). It will be critical for future work to continue exploring how interventions impact this population, including individuals who identify as gender nonbinary.
Generally, the committee found few studies designed to examine and address the underrepresentation of LGBTQIA women in STEMM. More research is needed to understand the intersectional experiences of LGBTQIA women and practices that would be most effective to increase participation and retention of this group in STEMM.
International Women in STEMM in the United States
Research indicates that international women students in U.S. institutions, along with their male counterparts, face discrimination in STEMM fields as a result of their national origin and cultural differences. Overall, the number of nonnative men and women entering STEMM fields has generally increased over time (NSF, 2016). For example, the number of international students in U.S. doctoral programs in specific STEMM fields has been rising, particularly in computer science, engineering, and physics, where international students constitute 51 percent, 56 percent, and 45 percent of Ph.D. recipients in those fields, respectively (NSF, 2016). In fact, the percentage of international doctorate recipients has risen by over 30 percent since 2000 in almost all STEM fields (NSF, 2015). The majority of international students in Ph.D. programs in the United States come from China, India, and South Korea (NSF, 2010).
From 1996 to 2006, the number of doctorates awarded to temporary visa holders increased in every scientific discipline (Figure 2-4). During this time, the number of female doctoral recipients also grew from 45 percent to 51 percent among U.S. citizens and permanent residents, and from 23 percent to 34 percent among temporary visa holders (NSF, 2016). While the numbers vary by specific field, foreign-born women are anticipated to continue to join STEMM as students in the U.S. system at increasing rates.
Despite increasing numbers of women born and raised outside the United States joining STEMM, there is relatively little known about how they are faring and the barriers they are facing in these fields (Hayes and Bigler, 2015; King
>Miller, 2017). Additionally, little is known about the treatment of women by foreign born men. Due to the increasing number of international students in the U.S. STEMM enterprise, climate surveys that examine a cross-cultural perspective are important to better understand the experiences of these students. As described below, what is known about the issues facing these women in STEMM is complicated.
Cultural differences exist between U.S. and international students characterizing their experiences in STEMM. Hayes and Bigler (2013, 2015) found that women who are born and raised outside the United States, especially in regions marked by potentially less progressive gender roles, may have more traditionally feminine occupational values than their U.S. counterparts (Hayes and Bigler, 2013, 2015). The authors noted that the converse of this might also be true: “International women have presumably sacrificed a good deal to pursue STEM training in the U.S. (e.g., increased financial cost and separation from family) and thus they may be more similar to men in their occupational values than to their U.S.-born female colleagues” (Hayes and Bigler, 2015). Hayes and Bigler (2015) also found that, among international groups, women who are targets of gender discrimination in their department report lower satisfaction with their graduate training (Hayes and Bigler, 2013, 2015).
While the literature has highlighted the underrepresentation of African American and Hispanic women in STEMM, data are significantly lacking regarding the experiences of women of African descent in these fields (King Miller, 2017). As this group is racialized as Black, they experience similar struggles for inclusion as African American women (Burton et al., 2010; Fries-Britt et al., 2014; King Miller, 2017). For example, the data that are available do not differentiate African Americans from foreign-born Blacks. Therefore, data on the number of Black women in STEMM may not exclusively represent the percentage of African Americans present in STEMM in comparison with foreign-born members of African descent. Instead, these data may be representative of all Black women inclusive of immigrants employed in STEMM careers within the United States (i.e., Afro Caribbeans and Africans) (King Miller, 2017).
In terms of discrimination that Africans and Afro Caribbeans face in STEMM, the “racialization rooted in American society and the emphasis on race is often unfamiliar to black immigrants because their racial identity is shaped from a country other than the United States” (Fries-Britt et al., 2014; King Miller, 2017). In many of the countries that Black immigrants emigrated from, identity is shaped by ethnicity rather than race (King Miller, 2017).
Several studies suggest that the U.S. educational system may begin to marginalize students based on skin color once they become assimilated into American culture, resulting in a disparity between those who have recently immigrated and those who may look and sound like African Americans. For the second generation and those who arrived in the United States at a very young age, there may be “pressure from their African American peers to conform in speech and behavior” (King Miller, 2017; Woldemikael, 1989). In addition, because Afro Caribbeans, for example, share the same racial classification as African Americans, they are vulnerable to the same forms of racial discrimination (Rogers, 2006).
Similarly, Tseng (2006) found that first-, second-, and third-generation immigrants who were European, African, Afro Caribbean, Asian, and Latin American entered STEM fields at similar rates, but in the second and third generation, students from each ethnic group showed a significant decrease in selecting STEM. The authors posit that, as the immigrant population becomes more assimilated into the U.S. education system, the likelihood that they will abandon the pursuit of STEMM-related subjects and careers increases (Tseng, 2006). Generally, the lack of data on the experiences of Afro Caribbeans and Africans and other international students in U.S. STEMM fields highlights this as a critical research need.
DIFFERENCES ACROSS STEMM FIELDS
While many barriers to full and equitable participation are shared across all STEMM fields, their form varies with the history, culture, and context of disciplines. In disciplines where much of the work takes place outside traditional professional spaces—astronomical observatories where data collection takes place at night, remote field sites that require camping or extensive hiking, laboratory experiments that require daily attention, including weekends—incivilities, harassment, or assault can be more common (Clancy et al., 2014; Nelson et al., 2017). The culture of these disciplines also matters. In physics, astrophysics, and planetary science, for example, invited speakers are often interrupted during their talks, whereas in the biological sciences such interruptions are atypical (NASEM, 2018b). Biology and physics also have very different histories. As sociologist of science Joseph Hermanowicz (2009) writes, “Physicists possess a recognizable genealogy of immortals—the likes of Kepler, Newton and Einstein—who promote a sense of scientific heroism and define a ‘model’ career for those who follow.”
One of the downstream effects of these cultural histories is that success in physics is presumed to hinge on innate brilliance, whereas in biology success is
perceived to require effort and empathy (Leslie et al., 2015). In those disciplines where successful practitioners are expected to have “raw, innate talent” women are less well represented, a phenomenon that led to the development of “the field specific abilities hypothesis” by Leslie et al. 2015.
In a nationwide survey of academics across STEM and the humanities, Leslie et al. (2015) found “evidence that the field-specific ability beliefs hypothesis can account for the distribution of gender gaps across the entire academic spectrum.” The authors found that negative stereotypes about women’s innate aptitude in certain STEM (e.g., math, physics, computer science, engineering) and humanities fields (e.g., philosophy, economics) better explained their underrepresentation in these fields relative to other potential explanations, such as willingness to work long hours required in certain fields or selectivity of graduate programs (as evidenced by estimated percentage of graduate applicants admitted to the department each year and 2011-2012 Graduate Record Examination (GRE) scores). In fact, fields that were more selective (according to these criteria), appeared to have had higher representation of women, although the difference was not statistically significant. Additionally, the authors found that the field-specific abilities belief hypothesis could explain the patterns of underrepresentation of African Americans across STEM and humanities fields; they did not, however, publish intersectional data on African American women specifically. The authors did directly confront the question of whether “women and African Americans [are] less likely to have the natural brilliance that some fields believe is required for top-level success?” and conclude that “the case has not been made that either group is less likely to possess innate intellectual talent (as opposed to facing stereotype threat, discrimination, and other such obstacles).” The authors’ conclusion is supported by a detailed analysis in the 2007 National Academies report Beyond Bias and Barriers (NASEM, 2007).
The “ideal worker norm” culture of effort and “hustle” in biology is not necessarily healthier than one that relies on assumptions of raw talent. Cultures of hustle encourage a work-life blurring, as well as the transgressing of other boundaries (Clancy et al., 2014; NASEM, 2018b; Nelson et al., 2017). In the hustle culture, stereotypes about women who nurture, as well as actual responsibilities they may have for childcare or elder care, may violate the ideal worker norm expectations in certain STEMM fields. When professional boundaries are blurred, harassment and assault can be intentionally perpetrated in the name of collegiality or over-friendliness (NASEM, 2018b; Wurth, 2018).
These features lead to variability in women’s representation in STEMM disciplines and across educational and career stages. With the exception of biology and medicine, fields in which women are at parity at degree-granting stages, women are below parity in STEMM at all academic training and career stages. Although women are disproportionately underrepresented in computer science and engineering at all levels, women who pursue these fields have a high likelihood of persisting across the academic career trajectory, making up about 20 per-
cent of bachelor’s degree earners and professors. In contrast, in biology, physics, mathematics, and chemistry, attrition of women occurs at every additional step in the academic career pathway—from postdoctoral associate to assistant professor to associate professor to full professor (Mangurian et al., 2018). In medicine, like biology, women are overrepresented; as of 2018 women outnumbered men among medical school students (Mangurian et al., 2018). Notwithstanding their overrepresentation at early career stages, women remain underrepresented among senior leadership roles in medicine (see Figure 2-1). As of 2018, women accounted for only 18 percent of hospital chief executive officers and 16 percent of deans and department chairs (Mangurian et al., 2018).
Dichotomies in Diversity Issues
The fields defined in the statement of task—physics, engineering, computer science, mathematics, biology, chemistry, and medicine—can be divided into two broad categories: those in which disparities in participation arise by the time students enter college and those in which underrepresentation occurs primarily at more senior career stages. Computer science, engineering, and physics are fields where women and girls are underrepresented relatively early on (before graduate school), and the life sciences and chemistry are fields in which barriers and biases prevent equal representation at the faculty level and block advancement into leadership positions. Mathematics and medicine do not fit cleanly into either of these categories, so the committee has separated them out into their own sections.
Computer Science, Engineering, and Physics
Computer science, engineering, and physics are fields with extremely low representation of women (Cheryan et al., 2017). In 2016, fewer than 20 percent of bachelor’s degrees were awarded to women in both computer science and physics and 21 percent of bachelor’s degrees were awarded to women in engineering (NSF, 2019). However, within sub-branches of the physical sciences and engineering, there is significant variation in the representation of women. For example, astronomy has twice the percentage of women than physics does (Urry, 2015), and nearly half of all bachelor’s degrees in environmental and biomedical engineering were awarded to women (ASEE, 2016; Yoder and Mattheis, 2016).
There is a widespread perception that girls and women are uninterested in computing and programming (Fisher et al., 1997). This perception, however, is not supported by a large body of evidence. Research on cultural attitudes suggests that adolescent girls are bombarded with stereotypes that computer science is a masculine field (Urry, 2015). Both girls and boys receive the message that computer science is a field ideal for “geeks” who are by archetype male, brilliant, socially awkward, isolated, and fond of science fiction; there are
even physical elements of the archetype, including pale skin and myopia (Beyer, 2014; Cheryan et al., 2015; Master et al., 2016; Rasmussen and Håpnes, 1991). Such stereotypes influence the decisions students make about programs of study and classes to take, particularly if they have not had early exposure to these disciplines (Cheryan et al., 2017). For example, according to Nord et al. (2011), when computer science classes are offered in high schools, boys are more likely than girls to enroll. Such gender imbalances at the high school level lead to imbalances at the undergraduate level, an issue that first emerged 20 years ago (Fisher et al., 1997; Master et al., 2016). Although stereotypical attitudes contributed to the perceptions of girls that they do not belong in computer science, interventions to alter classroom environments, such as including more examples of female scientists in wall art, can counter impacts of stereotypes (Master et al., 2016). This finding suggests that interventions at the high school level may be effective in increasing the numbers of women who are interested in and prepared for computer science courses at the undergraduate level. However, the widespread male-dominated culture that prevails at the undergraduate level may still lead to departures of women from the field (Fisher et al., 1997) (see Chapter 3 for further discussion). In fact, historical surges in computer science enrollment are often followed by decreases in representation of women, indicating that the culture of academic computer science during periods of growth may crowd out of the field.
Other programs focused on increasing women in computer science include those that aim to change the masculine stereotypes of the discipline, changing disciplinary content, altering the educational environment to be more inclusive and less hostile, and increasing the numbers of women in computer science (Fisher et al., 1997; Lagesen, 2007; Roberts et al., 2002) (see Chapter 3). A number of nonprofit organizations, such as code.org, Girls Who Code, Black Girls Code, and TECHNOLOchicas, have as a mission to increase the number of women in computer science and change the stereotypical image of programmers by exposing adolescent girls to computer science and programming.
At the transition from undergraduate to graduate programs, White women are retained at high percentages; in 2017, for example, the percentages of White women awarded Ph.D.s (10.9 percent among all computer science Ph.D.s) was slightly higher than the percentage awarded bachelor’s degrees (8.35 percent of all bachelor’s degrees in computer science) (NSF, 2016a,b). However, the numbers for women of color are extremely low. For example, like White women, Asian/Pacific Islander women slightly increased representation at the Ph.D. level, comprising 2.5 percent of Ph.D.s awarded compared to 1.6 percent of bachelor’s degrees awarded (Ong et al., 2011). However, overall representation was much lower than the percentages of Asian/Pacific Islander men awarded bachelor’s degrees (6.8 percent) and Ph.D.s (10.5 percent) (Ong et al., 2011).
The absolute number of underrepresented minorities in computer science has decreased over time. A longitudinal analysis (1960-2009) of the U.S. computing labor force found that diversity decreased during this period. Whereas White
women were 69 percent less likely than White men to work in computing in 1980, by 2009 the odds against working in computing increased to 71 percent. Underrepresented minority women were also 71 percent less likely than White men to work in computing throughout the period.
Because the problem in computer science is often characterized as a “pipeline” problem (Alper, 1993; Berryman, 1983; Ivie and Ray, 2005), only a few studies offer insight into broader cultural problems in that field. Women of color in computer science experience isolation and are marginalized beyond what White women experience (Charleston et al., 2014; Ong et al., 2011). Women of color, particularly Black women, are challenged by their peers regarding their academic competence and credentials (Charleston et al., 2014). Moreover, Black women in computer science are marginalized by both White women and Black men who prioritize gaining acceptance from the White men who hold cultural capital (Charleston et al., 2014). For example, Charleston et al. (2014) noted that “despite sharing similar racial experiences, participants noted how Black men and women were not always valuable sources for social support or camaraderie. As one participant elaborated, ‘Just cause there’s another Black brother [in class] doesn’t mean they want to work with you either.’ In sum, participants felt that Black men placed a strong emphasis on developing relationships with White males, whereas Black women were less inclined to do so” (Charleston et al., 2014).
Like computer science, the engineering profession is characterized by stereotypes associated with masculinity and “geeky,” antisocial tendencies (Cheryan et al., 2015). The low representation of women, context of masculinity, and stereotypical expectations all perpetuate an atmosphere that can be hostile to women (Cheryan et al., 2015; Hunt, 2016). Compared with other science disciplines, engineering has been characterized as particularly resistant to diversity and inclusion efforts (Burack and Franks, 2004).
Early socialization provides the first departure point in gender disparities in engineering. In interviews, male engineers are more likely to report early experiences with building and taking apart toys, whereas women engineers are more likely to refer to role models who specifically encouraged them to pursue engineering, as well as targeted opportunities such as science camps and middle school competitions (Chanderbhan-Forde et al., 2012). At the high school level, stereotypical expectations play a role in deterring women from gaining the necessary prerequisite coursework, and girls generally receive less encouragement to apply to undergraduate programs in engineering than boys receive (Cheryan et al., 2015; Hunt, 2016).
At the undergraduate level, the masculine context and social exclusion create barriers. Although male engineering students also reported a dearth of mentoring, a majority were able to obtain mentoring from upperclassmen, whereas, as described above, female students relied more on family members who were engineers. This pattern suggests that the likelihood of pursuing a
career in engineering is higher for those with backgrounds that give them access to mentoring.
Women are also substantially underrepresented as engineering faculty and professionals (Bejerano and Bartosh, 2015). In the same positions as men, female engineers make less money, receive less support for their research and ideas, and have fewer opportunities for advancement (Bejerano and Bartosh, 2015; Hunt, 2016; Xu, 2008). Compared with other scientists in other STEMM fields, engineers are more likely to be employed in the field for which they were trained, but departures from the field are characterized by high rates of gender disparity (Fouad and Santana, 2017; Hewlett et al., 2010; Hunt, 2016). Women who leave engineering careers cite three major factors: (1) gender disparities in pay in conjunction with difficult working conditions, (2) dissatisfaction with the ways their experience and skills are underutilized, and (3) lack of recognition or advancement opportunities (Fouad and Santana, 2017). The gender disparities in pay and advancement opportunities point to patterns of underlying structural discrimination in hiring and promotion (Hunt, 2016).
Women of color in engineering experience intensified marginalization relative to men (Chanderbhan-Forde et al., 2012; Foor et al,, 2007; Ong et al., 2011; Tate and Linn, 2005). Black female students have few faculty role models with respect to both race and gender (Chanderbhan-Forde et al., 2012). Like Black women in computer science, Black women in engineering are likely to experience marginalization from both White female and Black male peers (Charleston et al., 2014). This intersectional status results in amplification of barriers to obtaining access to prerequisite education in high school as well as of messages of social exclusion at the undergraduate and graduate levels (Chanderbhan-Forde et al., 2012; Foor et al., 2007; Ong, 2011). Minority women, particularly Black and Latina women, are told throughout their careers, by their peers, colleagues, and students, and by the ambient environment, either explicitly or implicitly, that they do not belong (Foor et al., 2007). Because faculty and peers may be unwelcoming, minority women often seek social support from sources outside their discipline and create separate social and academic peer groups (Cross et al., 2017; Mendenhall et al., 2018; Ong et al., 2011; Tate and Linn, 2005).
Physics fields are strongly male dominated and are characterized by many of the preconceptions of aptitude and brilliance that occur in other male-dominated sciences (Leslie et al., 2015). As is the case in computer science and engineering, enrollment in undergraduate physics programs and preparation for these programs reflect high school experiences. Despite the fact that female students on average have higher high school grade point averages than their male counterparts, as well as equivalent mathematics preparation, male students enter introductory college physics courses with better preparation from high school physics classes (Hazari et al., 2006; Kost-Smith et al., 2010). This may be due to girls being discouraged from taking physics courses, or teachers
directing their pedagogy toward male students (Hazari et al., 2006; Kost-Smith et al., 2010).
Female physics undergraduates experience the same challenges as computer science and engineering undergraduates in finding mentors, role models, and peer support (Aycock et al., 2019). However, they experience other challenges that their male counterparts experience with much lower frequency; Aycock et al. (2019) reported survey results indicating that 74 percent of female physics majors experience sexual harassment, the majority of which is gender harassment perpetrated by their peers. Women in physics graduate programs experience frequent microaggressions, in which they are treated negatively compared with male graduate students, and receive demeaning comments from both peers and faculty; and their complaints about their experiences with differential treatment are often dismissed (Barthelemy et al., 2015, 2016). These experiences have both racial and gender components for women of color (Clancy et al., 2017; Johnson, 2017; Ko et al., 2014). In a survey of women in astronomy and planetary science (fields closely related to physics), 40 percent of women of color felt unsafe in their workplace environments, and both White women and women of color avoided professional events due to safety concerns (Clancy et al., 2017).
A 2020 report by the National Task Force to Elevate African American Representation in Physics and Astronomy (TEAM-UP), which examined the reasons for the persistent underrepresentation of African Americans in these fields (AIP, 2020), offers additional insights. By conducting student and department chair surveys, interviews with students, site visits to five high-performing physics departments, and a review of the relevant literature, the task force identified five key factors responsible for the success or failure of African Americans in physics and astronomy:
- Physics identity
- Academic support
- Personal support
- Leadership and structures.
The authors noted that the persistent underrepresentation of African Americans in these fields is due to “(1) the lack of supportive environments for these students in many departments, and (2) to the enormous financial challenges facing them individually, as well as the financial challenges faced by the programs that have consistently demonstrated the best practices in supporting their success” (AIP, 2020).
The task force discussed a number of barriers that African Americans, including women, face in these fields, particularly as a result of their intersecting identities. These can include stereotypes about who is interested or capable of entering physics or astronomy. To reduce some of these factors, student peers can
play an important role in improving a sense of belonging in physics by mitigating microaggressions, the imposter phenomenon, and stereotype threat. Retention in these fields is further improved by an increase in the number of faculty who get to know these students and support their success (AIP, 2020) (see Chapter 3 for additional discussion of the important role faculty can play in instilling physics identity in African American physics students).
In physics careers, gender differences in salary emerge mid-career. For recent physics graduates, there were no gender differences in salary 1 year after graduation; however, men had salaries that were 10 percent higher than salaries of women 10–15 years after graduating with a physics doctorate. Also, compared with men, women reported that their careers progressed more slowly and that they received fewer career resources and opportunities. In addition, women were more likely to make career compromises for family reasons (Porter and Ivie, 2019).
Biology and Chemistry.
Women are generally not underrepresented in biology and its sub-disciplines, including biophysics and computational biology, and chemistry at the undergraduate level. Yet, in both of these fields, the proportion of women declines at subsequent professional stages (Addessi et al., 2012; Crangle, 2009; Ledin et al., 2007; Nüsslein-Volhard, 2008).
The percentage of women among the students earning bachelor’s degrees in biology peaked at 62 percent in 2003–2006 (APS, 2018). Since then, it has remained steady, fluctuating between 59 and 61 percent from 2007 to 2017. The percentages of women obtaining master’s degrees and enrolling in doctoral programs in the biological and biomedical sciences in 2017 were 52.6 percent (NSF, 2018a), 45 percent (Martinez et al., 2007)5 and 38 percent (Plank-Bazinet et al., 2017), respectively.
Chemistry has gender parity at the undergraduate level (Grunert and Bodner, 2011); from 2000-2017, with the proportion of women among all students receiving undergraduate degrees in chemistry fluctuated between 48 percent and 52 percent (APS, 2018). The percentages of women obtaining master’s degrees and enrolling in doctoral programs in chemistry are lower than the percentages of women obtaining bachelor’s degrees: In 2008 women received 36.1 percent of chemistry doctorates and 23.6 percent of postdoctoral fellowships, and women comprised 18 percent of faculty applicants to research-intensive institutions (Grunert and Bodner, 2011; NSF, 2011).
The popular images of biologists and chemists are not as male-oriented as in fields such as computer science, engineering, and physics (Cheryan et al., 2017). Negative stereotypes about women’s innate abilities are also less prominent in biology and chemistry (Cheryan et al., 2017). In many institutions, however, biases and barriers undermine success and increase the desire to leave at later educational and career stages.
5 Data on postdocs and faculty refer to women in the biomedical sciences only.
Although women are not underrepresented in biology in general, they face biases and barriers that drive them out of the discipline over the course of their careers with greater frequency than their male counterparts. In 2014, the SAFE (Survey of Academic Field Experiences) team published its first findings on sexual harassment in the field sciences, most of which are biological sciences (Clancy et al., 2014). Researchers distributed an online survey through a variety of channels (e.g., Facebook, Twitter, LinkedIn, professional societies, and through science and service blogs in two waves: the first, aimed at biological anthropologists included 124 respondents, and the second (N = 542) that allowed respondents to provide their professional discipline). They found that most, or 72.4 percent, of women observed sexual harassment; a large number, 64 percent (N = 423/658), experienced sexual harassment, and a significant minority were sexually assaulted while conducting fieldwork. In this sample, when women were harassed the perpetrator was more likely to be senior to them in the workplace hierarchy; when men were harassed, the perpetrator was more likely to be a peer. Few (18 percent) respondents reported field experiences where there was a clear reporting mechanism for sexual harassment, and, of the hundreds who shared that they experienced sexual harassment, only a handful (N = 37) reported, and only seven were satisfied with the outcome of that report. One of the initial hypotheses proposed by Clancy et al. (2014) for why field researchers were so often sexually harassed was that the harassment was coming from “locals” in countries with sexist views about women; however, data demonstrated that women were much more often harassed by their co-workers, not the “locals” whom they encountered in the field.
In 2017, the SAFE team published a follow-up study based on 26 interviews with field scientists (Nelson, 2017). The team found evidence that the culture of the field sciences is characterized by unclear boundaries, few sanctions for bad behavior, and unequal access to resources for women. As respondents shared, many were implicitly or explicitly told that they could not communicate with colleagues about their experiences with inappropriate behavior at their field sites. Respondents who reported incidents, spurned advances, or otherwise fought back in their harassing environment faced significant personal and professional consequences, from sabotage of their research to posttraumatic stress symptoms that interfered with publication of the results of their field research.
Women in biology also face substantial hiring disparities. Sheltzer and Smith (2014) found that at the trainee level, on average, male principal investigators ran laboratories that had 36 percent female postdocs and 47 percent female graduate students, significantly lower than was observed in laboratories headed by women, who employed on average 46 percent female postdocs and 53 percent female graduate students. Women are also hired at the faculty level less frequently at prestigious institutions, and instead are directed toward less prestigious, more teaching-focused positions (Sheltzer and Smith, 2014). These findings point to lost opportunities for advancement for women in biology across the career stages.
Chemistry is a discipline that has gender parity at the undergraduate level (Grunert and Bodner, 2011), but has unique cultural challenges that act as barriers for women later in their education and careers. In a survey of British doctoral students in chemistry, female students reported more issues with lack of mentoring and social marginalization compared to male students (Newsome, 2008). Furthermore, women were more likely to perceive their research group’s culture as inhospitable (Newsome, 2008). They also raised concerns about the isolating nature of their doctoral study as well as about warnings they received suggesting that they would have to sacrifice relationships and family in order to remain competitive in the postdoctoral and academic job markets.
Many women who obtain academic positions in chemistry find them unwelcoming (Greene et al., 2010). Typically, men have higher salaries, are given better or larger research space, and are more likely to be promoted at all career stages (Greene et al., 2010). Men receive greater recognition from the university, are more respected by students, and have an easier time gaining administrative assistance. In contrast, women are more likely to have higher teaching and service loads (Greene et al., 2010), which they believed were among the departmental barriers to recruiting and hiring other women faculty (along with overt opposition to hiring female faculty).
Like computer science and physics, mathematics is a discipline where aptitude is assumed to be due to innate brilliance and this belief is compounded with culturally ingrained, sexist stereotypes (Cvencek, 2011; Leslie et al., 2015; Master et al., 2016). By second grade, children form implicit and explicit associations between boys and math, and girls are less likely to state explicitly that they like math (Cvencek, 2011). Even for girls and women who are motivated to study math and excel at math, negative stereotypes hinder their mathematical performance. Stereotype threat is the phenomenon where awareness of a negative stereotype about identity leads to anxieties about confirming that stereotype (Spencer et al., 1999; Steele and Aronson, 1995). Middle school girls who are told that they are taking a test that measures mathematical skills underperform on those tests when they are alone or in a mixed-gender group, but not when they are in a group of girls (Huguet and Régner, 2007). Asian American girls experience competing stereotypical pressures: on the one hand, some stereotypes associate female identity with poor mathematical aptitude, and on the other, some stereotypes associate Asian identity with excelling at math (Ambady et al., 2001). The impact of gender socialization and stereotypical association persist through high school and undergraduate studies.
Early studies indicate that gender differences in math performance emerged in high school, but that gap has now closed (Hyde et al., 1990a, 1990b, 2008). Because girls and boys are similarly prepared by high school math courses, undergraduate women are theoretically well prepared for mathematics curriculum at the college level. The number of undergraduate degrees awarded to women
in mathematics reached a high of 48 percent in 1999–2000, but since then the proportion of degrees awarded to women has steadily declined to 41 percent in 2017 (Herzig, 2004b; APS, 2018).
These patterns raise the question as to how a significantly greater proportion of women students graduate with undergraduate degrees in math than in other physical science STEMM majors, despite similar negative associations and stereotypes about women’s abilities. There are at least two possibilities that might work separately or in conjunction to explain why math is more gender balanced. First, negative stereotypes about girls’ abilities in math may be weaker than negative stereotypes in other more male-dominated fields (Cheryan et al., 2017). Second, pre-college math is mandatory in most curricula, and girls perform as well as boys in these courses. Mandatory courses—if taught well for girls—may reduce gender disparities in later participation because they provide girls with an opportunity to try a field (in a classroom that is gender balanced) instead of relying on stereotypes about girls or about the field to guide their decisions.
In academia, women are disproportionately lost from mathematics careers at two key points: the transition from undergraduate to graduate school and the transition from graduate school to faculty. In 1996, 46 percent of bachelor’s degrees in mathematics were awarded to women, yet women comprised only 33 percent of entering graduate students. By 2002, the proportion of women hired into tenure-track mathematics positions had declined to 22 percent. Representation of Black, Latina, and Native American women among doctoral recipients is extremely low. Like other STEMM fields, women leave mathematics at the doctoral and postdoctoral levels due to isolation and a lack of mentoring in their graduate experiences (Herzig, 2002, 2004b). Negative experiences from faculty in mathematics, such as exclusion from social networks and gender harassment, play a role in driving women to leave their profession (Herzig, 2004b). Those who stay in mathematics are more likely to access cultural capital and mentoring networks due to family members in mathematics or involvement in undergraduate research experiences (Herzig, 2002).
The absolute numbers of women who are medical school applicants, admitted students, and medical school graduates has steadily increased from around 10 percent in 1973 to gender parity today, although there are gender imbalances in some specializations (AAMC, 2016b). Generally, women are well represented in specialties that involve women and children or are associated with nurturing, whereas men are represented in higher proportions in specialties that require technical specialization (Carnes et al., 2008). Like biology, chemistry, and math, women are more likely to exit the profession at higher ranks. For example, in 2015, women comprised 51 percent of M.D. instructors, but their representation declined at the assistant professor level (43 percent), associate professors level (33 percent), and full professor level (20 percent) (Kenyon College, 2019). Moreover, a gender gap exists in several aspects of scholarly
publication. Women have low proportions of first authorships (29.3 percent) and senior authorships (19.3 percent) relative to male peers (Jagsi et al., 2006). As well, women in academic medicine receive less institutional funding and administrative support compared with male colleagues (Carr et al., 2003).
Women in the field of medicine, as in other STEMM fields, experience conflicts between biological and professional clocks, as well as challenges of traditional gendered division of domestic labor. Women physicians report challenges of optimally timing childbearing in relation to their careers and of obtaining childcare, particularly during residency years (Jagsi et al., 2007). Women physicians reported spending 8.5 more hours a week on domestic activities than male peers and were more likely than male peers to have spouses or domestic partners who were employed full-time (Jolly et al., 2014). Additionally, faculty with both childcare and clinical responsibilities were significantly more likely to report low satisfaction with work-life balance and career than were colleagues without children and/or clinical responsibilities (Beckett et al., 2015).
Women of color face a double bind in medicine and are more underrepresented at higher academic ranks. Although minority women make up 18 percent of the U.S. population, they represent 3.2 percent of full professors in medicine. Additionally, women of color face higher workplace discrimination rates and work-family conflict, contributing to a negative climate (Carapina et al., 2017).
Of the disciplines examined in the 2018 National Academy of Sciences report, Sexual Harassment of Women: Climate, Culture, and Consequences in Academic Sciences, Engineering, and Medicine, medical students had the highest rates of sexual harassment compared with students in engineering, the sciences, and the nonsciences (NASEM, 2018b). The hierarchical and hostile work environments in many academic medical centers lead to greater bullying, intimidation, and harassment from patients, peers, and superiors. In fact, one-third of women in academic medicine experience gender harassment, and many reported that discrimination hindered their careers (Foster et al., 2000). Harassment is a particular concern in medicine because it takes place in “environments with little structure or accountability for the faculty member, and a decreased ability for students to leave without professional repercussions” (NASEM, 2018b). In a qualitative study at 23 medical schools, members of the Group on Women in Medicine and Science and the Group on Diversity and Inclusion of the Association of American Medical Colleges were interviewed. Despite the increase in numbers of women in medicine since the 1980s, only modest improvements were seen in the academic climate for women over the past 20 years, and there was a reported lack of institutional oversight and substantial variations by department (Carr et al., 2015).
Bias and discrimination in medicine can directly harm those not even employed in the field. The hierarchical and hostile training landscape many physicians experience introduces considerable bias, which is harmful to both patients and the physicians themselves. Hostility and incivility have adverse effects on
medical teams’ efficacy in diagnosis and treatment, possibly even greater in magnitude than adverse effects caused by sleep deprivation (Riskin et al., 2015). Physician biases regarding weight, gender, race, and other factors lead to missed diagnoses, delayed treatments, and poorer outcomes for many patients. Women and non-White males frequently receive less aggressive care than White males (Dressler et al., 2005; Geronimus et al., 2006; Gravlee, 2009; Suite et al., 2007). The ways in which these hierarchical and discriminatory practices influence the treatment landscape likely also have consequences for women and non-White male physicians who work in these cultural contexts.
CONCLUSION: SYSTEMIC ISSUES IN STEMM CULTURE
Across the science enterprise, widespread instances of active and passive actions, as well as implicit and explicit biases, hinder women’s careers in STEMM. Women of multiple marginalized identities are confronted with amplified forms of these biases and barriers throughout all STEMM fields. The result is that many women leave STEMM or stall out in positions of lower educational attainment, rank, prestige, and pay compared to male colleagues. Based on many years of research, it is fair to conclude that many White women and women of color make rational decisions to leave environments in which they are subject to harassment, believe that their careers are stalled, and/or that they are discriminated against in pay and promotions. While research demonstrates that many White women and women of color report that they “feel” unwelcome, isolated, or unfairly held back, the objective reality is that they work in cultures and climates that often exclude them and push them out. The fact that these patterns exist broadly, including in fields that have large numbers of entering White women, points to systemic problems that cannot be solved simply by recruiting more women. Disrupting the forces at play in STEMM are necessary to create an inclusive environment that will enliven American science, engineering, and medicine. The chapters that follow describe the current state of research on educational and workplace interventions that show promise in supporting improved recruitment, retention, and advancement of women across STEMM, while pointing out where there are important gaps in knowledge.
FINDINGS: CHAPTER 2
FINDING 2-1: Evidence does not support the longstanding perception that women are underrepresented in STEMM because of a lack of innate ability in these fields.
FINDING 2-2: Implicit and explicit biases contribute to the underrepresentation of women in STEMM. These biases manifest in multiple ways at all stages of STEMM career life cycles. Across STEMM fields, biases often affect
women’s educational and career trajectories at critical junctures, such as recruitment into lab management positions, consideration for graduate admissions, consideration for postdoctoral positions, and in promotion decisions. Women of multiple marginalized identities (e.g., women of color, women with disabilities, LGBTQIA women) experience intensified forms of bias and discrimination in STEMM as a result of the complex, cumulative ways in which multiple forms of discrimination (e.g., racism, sexism) intersect.
FINDING 2-3: In addition to experiences of heightened bias, sexual harassment, and microaggressions, women of color in STEMM frequently experience:
- isolation (i.e., experience a sense of invisibility or hypervisibility) and exclusion from social network supports usually available to men
- a sense of “not belonging” in STEMM
- “racial battle fatigue,” which is the “cumulative result of a natural race-related stress response to distressing mental and emotional conditions” that adversely affects the health and achievements of students and faculty of color
- racial harassment
- cumulative disadvantage; such as interest on debt, and disadvantages, such as lower salary and delayed promotion, which accrue over time f. expectations that they must work harder, including working extended hours, to fit the ideal worker norm despite having had fewer role models who have successfully managed these expectations, fewer culturally competent mentors, and less access to informal professional networks
FINDING 2-4: There is less research on the factors that drive the underrepresentation of women with disabilities, LGBTQIA women, and international women, but the available research suggests that these groups face significant barriers in STEMM due to their intersectional identities.
FINDING 2-5: While bias, discrimination, and harassment exist across all STEMM disciplines, the form that these phenomena take varies with the history, culture, and context of the specific discipline, as does women’s representation. In fields such as physics, engineering, and computer science, disparities in participation are seen by the time students enter college. In contrast, in the fields of biology, medicine, and chemistry, women encounter barriers and biases that prevent equal representation at the faculty level and block advancement into leadership positions. Mathematics has a slightly different pattern of underrepresentation, in that women comprise around 40 percent of undergraduate degree earners, but are lost at the transition from undergraduate to graduate school and the transition from graduate school to faculty.
FINDING 2-6: The number of women medical school applicants, admitted students, and graduates has steadily increased from around 10 percent in 1973 to gender parity today, although gender imbalances still exist in some specializations. Generally, women are well represented in specialties that involve women and children or are associated with nurturing, whereas men are represented in greater numbers in specialties that require technical specialization.
FINDING 2-7: Notwithstanding their overrepresentation at early career stages, women in medicine remain underrepresented among senior leadership roles in medicine. As of 2018, women accounted for only 18 percent of hospital CEOs and 16 percent of deans and department chairs.
FINDING 2-8: Although women are disproportionately underrepresented in computer science, engineering, and physics at all levels, women who pursue these fields have a high likelihood of persisting across the academic career trajectory, making up about 20 percent of bachelor’s degree earners and professors. In biology, mathematics, and chemistry, attrition of women occurs at every additional step in the academic career pathway—from postdoctoral associate to assistant professor to associate professor to full professor.
FINDING 2-9: There is a long-standing cultural association between masculinity and objectivity in most segments of American society, which, in turn, underlies the associations of masculinity with STEMM. This expectation that STEMM professionals are White and male is implicitly conveyed in cultural portrayals of STEMM and STEMM education and these stereotypical associations shape the social and educational environments of children, as well as structural patterns that occur in STEMM professions. The popular images of biologists and chemists are not as male-oriented as in fields such as computer science, engineering, and physics.
FINDING 2-10: Cultural expectations and biases about which jobs and careers are held, or should be held, by women and men present biases and barriers that limit both women’s and men’s opportunities in STEMM (for instance, biases and stereotypes can limit men’s opportunities in certain medical fields).
FINDING 2-11: The culture of disciplines matters as one of the downstream effects of these cultural histories is that success in physics, engineering, and computer science is presumed to hinge on innate brilliance, whereas in biology success is perceived to require effort and empathy. In both STEMM and the humanities, disciplines where successful practitioners are expected to have “raw, innate talent,” women are less well represented—most likely because of sexist beliefs about women’s innate “brilliance” that is not supported by evidence (see discussion of the field-specific abilities hypothesis).
FINDING 2-12: Fields such as the biological sciences and medicine have a culture of “hustle” that encourages a work-life blurring. Such STEMM work environments often require a commitment to the job through long hours, stagnant career trajectories, and constant availability and visibility—i.e., the “ideal worker norm” that implicitly assumes that there will be gendered separation of work and family, further reinforcing the stereotypical image of the scientist or physician as a man. Adherence to these ideal worker norms disadvantages both men and women who have commitments and duties outside of work and has a disproportionate impact on women of color because their multiple identities and their small numbers in STEMM departments contribute to an even greater perception that they do not demonstrate the characteristics of the “ideal worker.”