Abstract
Many corporations in the United States are under pressure to increase gender diversity in professions such as software engineering that have long been dominated by men. Current sociological theory suggests that in response to such pressure, hiring decision makers value women job candidates as potential contributions to gender diversity and incorporate this value into their initial hiring screening decisions. However, it remains unclear whether and how decision makers take diversity into consideration in later hiring stages, namely, during offer decisions, and particularly when there is a strong culture of meritocracy. In this article, the author examines a critical case study of software engineering hiring at a midsized high-technology firm. Drawing on qualitative data, the author finds that diversity considerations do play a role in job offers, but as an implicit motivational force during job offer deliberations rather than an explicitly stated factor. A quantitative analysis complements the qualitative analysis: the author finds no female penalty on receiving an offer where one might be expected, which the author argues is due to implicit diversity considerations during hiring deliberations. The findings of this study have implications for theories of hiring discrimination, cultural fit as a legitimating tool, and motivating reasoning under strong pressures to diversify.
U.S. corporations often face pressure to increase the representation of women in professions such as software engineering that are traditionally filled by men (Dobbin and Kalev 2021; Han and Tomaskovic-Devey 2021). Recent sociological research suggests that under strong pressure to diversify, corporate decision makers may value women job applicants for their potential contribution to firm diversity and incorporate this value, often explicitly, into their initial screening decisions (Weisshaar, Chavez, and Hutt 2024). However, it remains unclear whether an applicant’s value to firm diversity, so relevant during the initial applicant screen, is also relevant during the job offer decision. Indeed, there is reason to assume that diversity considerations have a limited role, given that decision makers are averse to the explicit consideration of gender during interview evaluations and job offer decisions (Rivera 2012a, 2015b). In this article, I ask, under pressure to diversify, do corporate decision makers incorporate an applicants’ value toward organizational diversity into their job offer decisions when hiring for male-dominated occupations? If so, how?
I address these questions by examining a case study of software engineering hiring at a midsized high-technology firm. Software engineering is a profession in which women are vastly underrepresented and which has experienced intense calls to increase gender diversity; it is also a profession that is dominated by White and Asian men (both Asian immigrants and Asian Americans) (Neely, Sheehan, and Williams 2023). Drawing on comprehensive data from in-depth interviews with hiring decision makers, participant observation of hiring deliberations, and archival data of interviewer and recruiter notes, I describe a process in which applicant diversity value—that is, an applicant’s worth as a contributor to organizational diversity—contributes to the job offer decision. However, and importantly, unlike previous research focusing on hiring screening decisions (Weisshaar et al. 2024), diversity value acts as implicit motivation during offer deliberations rather than an explicitly stated consideration.
In this study, I follow job candidates as they enter the in-person interview stage of hiring. During the interviews, women are generally evaluated as less technically able than men, but better cultural fits because of their perceived communality. These gender differences in evaluation are unsurprising, given that both evaluations are subjective and likely prone to bias. At the same time, decision makers express clear, if somewhat ambivalent, motivation to hire women for their diversity value. I argue that this motivation influences the hiring deliberations: despite decision makers’ insistence that strong cultural fit—particularly fit on the basis of communality—should not compensate for weak technical evaluations, decision makers put in effort to justify job offers for some candidates who fit that very profile, a disproportionate number of whom are women. I argue that women’s diversity value is an implicit motivation behind that extra effort. Moreover, despite the substantial representation of Asian women among job candidates, I do not find that decision makers perceive or treat Asian women differently from White women in this process. I support the qualitative findings with a path analysis drawing from interview evaluations and offer decisions of 1,082 job candidates.
The InGen case study provides insights into the role of diversity value during offer decisions for women candidates but offers limited perspective on its role for Black and Latino candidates. InGen decision makers consider Black and Latino workers to contribute to firm diversity, but so few Black and Latino candidates enter the interview stage that analytical focus on these candidates is not possible. I detail additional study limitations in the discussion section.
Despite these limitations, the findings of this study make three key contributions. First, I extend the explanatory reach of theories regarding hiring and diversity value (Weisshaar et al. 2024) to the offer decision stage of hiring where the explicit use of diversity value is perceived as illegitimate. Instead of explicitly incorporating diversity value into the offer decision—which would violate meritocratic values—decision makers work to justify giving job offers to a small, but noteworthy, number of women in a more legitimate way: by arguing that the candidates’ communal strengths mean that the candidates are motivated to improve their technical skills. Second, I contribute to literature on cultural fit in hiring and, in particular, how the ambiguity of cultural fit allows decision makers to justify their offer decisions (Nichols, Pedulla, and Sheng 2023; Rivera 2012b). Finally, I contribute to literature on gender stereotypes and interactional evaluations (Correll et al. 2020; Ridgeway 2011), as well as literature on motivated thinking (Norton et al. 2006, 2008), by demonstrating how decision makers, when motivated to increase gender diversity, positively frame their stereotypical perceptions of women to justify their offer decisions (e.g., Norton, Vandello, and Darley 2004).
Pressures to Diversify, Diversity Value, and Offer Decisions
For U.S. corporations, pressure to increase workforce diversity comes from many sources: antidiscrimination laws to which corporations must demonstrate compliance, corporate employees and clients, and the corporations’ own legitimacy concerns (see reviews in Dobbin and Kalev 2021). In professions that are dominated by men, pressures to increase gender diversity are particularly strong, as it is precisely in these occupations where women are vastly underrepresented (Han and Tomaskovic-Devey 2021). In the past, firms might have withstood pressures to diversify by displaying a public commitment to diversity while refraining from fundamentally altering hiring procedures or other personnel processes (Meyer and Rowan 1977). Yet as organizations have become increasingly accountable and transparent, they have become more responsive to diversity pressures, as evidenced by the growing adoption of formal practices meant to increase diversity (Bromley and Powell 2012).
Recent research gives some insights into how pressures to diversify may alter personnel decisions such as those at the hiring interface. In particular, diversity commodification theory argues that under pressures to diversify, employers explicitly assess workers’ “diversity value”—meaning their worth for contributing to organizational diversity—and incorporate their diversity value into screening decisions along with other biases and preferences employers may possess. In this way, corporations may reduce discrimination against traditionally excluded groups during screening processes and, potentially, increase the representation of such groups in the workforce, not by reducing their biased assessments of ability but by incorporating diversity value into their selection calculus (Weisshaar et al. 2024).
By highlighting the role of diversity value in decision making, diversity commodification theory helps explain patterns of gender and racial discrimination at the initial hiring screen. However, there is less empirical insight into whether diversity value influences the eventual job offer decision. Previous research on hiring norms and procedures tends to highlight two general criteria that seem to matter during the offer decision–and are often assessed during in-person interviews: an applicant’s ability to perform on the job, and an applicant’s “cultural fit,” which can encompass shared values and work style (e.g., Chatman 1991), or even a shared “styles of play” between an applicant and the current employees, the interviewer, or the organization (Rivera 2012b). Moreover, these assessments of an applicant’s performance ability and cultural fit may be biased by stereotypes of men and women, particularly when the criteria are ambiguous and subjective (Correll et al. 2020; Ridgeway 2011). The role, if any, of diversity value in offer decisions remains a blind spot in the current literature.
There is reason to think that diversity value may not be a legitimate consideration during offer decisions in contrast to during the initial hiring screen. For instance, in a study of hiring in a financial firm, Rivera (2015b) found that recruiters consider the diversity of the applicant pool during recruitment, but they refrain from explicitly incorporating diversity considerations into offer decisions; diversity is only relevant when decision makers are selecting between two seemingly equivalent candidates, and only half-heartedly after prodding from human resources professionals. The illegitimacy of diversity value during offer decisions may explain why explicitly gender-conscious “best practices,” such as targeted recruiting and diversity hiring panels, seem to exist in the early stages of hiring but are absent in latter stages (Dobbin and Kalev 2022). The explicit incorporation of diversity value into offer decisions is even more likely to be illegitimate in professions associated with a culture of meritocracy, given the common, and often unfounded, belief that efforts to increase diversity come at the expense of meritocracy and hiring the best people for the job (Heilman and Welle 2006; Konrad, Richard, and Yang 2021).
If diversity value is not explicitly incorporated into offer decisions, it may still influence the decision-making process in a less explicit way. Literature on motivated reasoning suggests that when employers are motivated to favor an applicant, they may, consciously or unconsciously, shift how and the extent to which their evaluations influence their selection decisions so as to benefit their favored applicants (Norton et al. 2006, 2008). To be sure, employers often engage in motivated reasoning to penalize groups that are historically disadvantaged in a given context (e.g., Uhlmann and Cohen 2005). However, under pressures to diversify, employers may be motivated to advantage historically disadvantaged groups. In a series of experiments, Norton et al. (2006, 2008) found that evaluators adjust their admissions standards for acceptance into college to favor racial minority students when racial diversity is valued in the setting. Employers may incorporate diversity value into their offer decisions, not by explicitly using diversity value as an evaluation criterion, but by subtly adjusting their offer justifications in response to diversity value thus altering their offer decisions.
Diversity Value in Software Engineering Hiring
Software engineering is a useful profession to examine whether and how diversity commodification of women candidates occurs during job offer decisions. First, software engineering is a highly technical profession dominated by White and Asian men. Currently, only 19 percent of software engineers in the United States are women, a dismal statistic that has remained consistent in the past few decades (Neely et al. 2023; Zippia 2022). Women software engineers, who are typically numerical tokens among their fellow engineers, often face a chilly climate at work in part because of the perception that an ideal engineer embodies masculine-typed characteristics such as technical and analytical skills, competitiveness, and an orientation toward things (Robinson and McIlwee 1991; Wynn and Correll 2018). Although women are vastly underrepresented in software engineering, it is also a profession with an overrepresentation of Asian workers, both Asian Americans and Asian immigrants (Neely et al. 2023). At Google (2023), for instance, 51 percent of workers in technical positions identify as Asian, and 42 percent identify as White. 1 Indeed, Asian workers are often not considered to be underrepresented minorities in the software engineering context (Chow 2024).
Second, software engineering has faced intense criticism for its lack of gender diversity. Most relevant for this project, public pressure to increase gender diversity surged in 2013 following grassroots efforts to publicize the lack of female representation among software engineers in Silicon Valley’s leading firms (Chou 2013). In response to increasing pressures to address diversity, top Silicon Valley firms officially publicized the gender and racial representation of their workforces and vowed to improve (Gutman 2018). Third, the rising pressure to address gender diversity in software engineering coincides with the long-established association between the software engineering profession and a culture of meritocracy (O’Mara 2019). For software engineers, technical ability is central to their professional identity, and they often perceive their success as purely the result of their abilities and talent (Seron et al. 2018). Although cultural fit still plays a role when employers decide whom to hire (Chavez 2021), software engineering is a context in which aversion to the incorporation of diversity value into offer decisions is theoretically strong.
And yet there is some indication that diversity value may play a role in job offer decisions for software engineers. In a study across 441 high-technology firms, Fernandez and Campero (2017) found no evidence of discrimination against women engineers in receiving a job offer, conditional on making it to the interview stage. In a context in which gender stereotypes regarding technical competence are exceptionally strong and influential, the lack of gender discrimination during selection decisions may hint at a countervailing force provided by diversity value. And yet there is not a clear understanding of how diversity value would, if at all, influence selection decisions in this context. I next turn to the case study.
The InGen Case Study
The Software Engineering Hiring Process at InGen
I first provide contextual detail about the case study, InGen, on the basis of data gathered from interviews with InGen employees and management. 2 I fully describe the data collection and analytical process in the following section. InGen is a midsized high-technology firm in Silicon Valley. Women constituted 12 percent of InGen’s roughly 300 software engineers and 28 percent of InGen employees overall. By comparison, women represented 23 percent of software engineers in Silicon Valley’s technical workforce the year of data collection (2014 American Community Survey five-year sample, author’s calculations).
I focus on InGen’s software engineering hiring process. Recruiters first screen workers who apply to InGen software engineering positions. After this initial screen, candidates undergo a short technical exam, typically administered over a video-sharing application, during which InGen engineers test the applicants’ coding and algorithmic fundamentals. Candidates who pass the technical screen move on to in-person interviews, in which they have four to five one-hour, one-on-one interviews with InGen engineers at the InGen facility, with a short coffee break. The hiring manager, who is also a software engineer, conducts one of the interviews. During each interview, interviewers ask a technical question, and candidates have roughly 45 minutes to work on a whiteboard as the interviewer stands alongside questioning the candidates’ decisions.
Interviewers evaluate the candidate’s technical ability and cultural fit. Technical ability is the primary hiring criterion and core to an engineer’s capacity to perform. Interviewers evaluate technical ability on the basis of the candidate’s answers to the interview questions—their written code, algorithms, design, and technical knowledge. However, assessment of the candidate’s technical output requires interpretation, as there are usually many correct answers to the technical questions, and regardless, candidates often do not solve the problem completely. To inform their technical assessment, interviewers draw on the candidates’ presentation of self, looking for signals of deep knowledge and ability, and their projected confidence.
Interviewers also evaluate an applicant’s “cultural fit.” There is no formal definition of cultural fit at InGen. Instead, interviewers base their evaluation of cultural fit on the chemistry they feel during the interview (see Rivera 2015a). One hiring manager offered this simple heuristic: “Would you want to drive from here to L.A. with this person in the car? Would that just be the worst thing in your life, or would it be fine?” On the basis of my observations, two implicit archetypes of cultural fit receive high marks. One archetype is an engineer with a passion for, dedication to, and intimacy with, technology and engineering who is seemingly obsessed with coding projects in their free time. The second archetype of high cultural fit is an engineer who is interpersonally warm and enthusiastic: someone with whom interviewers imagine socializing and having fun. Engineers are generally excited to be around both archetypes.
Interviewers evaluate technical ability and cultural fit on two separate four-category scales. 3 An interviewer also assesses the candidate’s cultural fit during a coffee break.
The evaluation scale for each criterion is as follows: 1 = “Would not recommend for hire, and would fight for a rejection”; 2 = “Would not recommend for hire, but could be persuaded otherwise”; 3 = “Would recommend for hire, but could be persuaded otherwise”; and 4 = “Would recommend for hire, and would fight for candidate.”
After the candidate’s visit, the interviewers meet for 15 minutes to deliberate whether to offer the candidate a job. Average scores of 3 for both technical and cultural evaluations represent a rule-of-thumb threshold for an acceptable offer. During the deliberation, interviewers explain their technical and cultural evaluations and answer other interviewers’ clarifying questions. The hiring panel then deliberates whether to give the candidate an offer. The hiring manager records the offer decision along with the official offer or rejection justification.
During offer deliberations, the rule of thumb is to err in rejecting technically strong candidates rather than risk hiring technically subpar candidates. As repeated to me multiple times, decision makers fear that InGen’s rapid growth—and rapid pace of hiring—may decrease the quality of the engineering core, a fate supposedly met by other Silicon Valley companies. Because of this fear of lowering the technical quality, poor cultural evaluations may be a legitimate reason to reject a candidate, but strong cultural evaluations on the basis of interpersonal warmth and enthusiasm alone are not a legitimate basis for an offer. Extending an offer to a candidate with strong cultural evaluations but substandard technical evaluations is perceived as improper, and is something that, as one engineer said in a hiring deliberation, interviewers “usually, in the past, say ‘no’ to.” As another engineer told me, when deciding to extend an offer, technical evaluations have priority, and cultural evaluations are “weighted lower just implicitly by everyone.”
Data
With InGen management’s permission, I had access to comprehensive data, allowing an in-depth look into the InGen hiring process. 4 From August 2013 to August 2014, I used snowball sampling to recruit and interview 50 InGen employees involved in the hiring process (see Table 1). All interviews took place at InGen and were recorded and transcribed with the respondents’ consent. Interviews were semistructured, meaning that I developed a standard interview protocol that evolved during the data collection process and allowed flexibility to follow up on spontaneous interview topics. I designed interviews to be in depth; each interview lasted about an hour, and I interviewed key respondents multiple times.
Descriptive Statistics of Interview Respondents.
Note: Managers and directors are software engineers who are in manager or director positions. Gender and ethnicity are self-reported.
I also observed 69 hiring deliberations, 12 for women candidates, over 11 months. For the first month of observation, I introduced myself as a graduate student researcher working with the recruiting team to learn about the hiring process and asked for permission to observe and take written notes on the conversation. After about a month, I became a regular fixture in the meetings. To supplement my data from participant observation of hiring deliberations, I gathered hiring manager and recruiters’ notes from 331 hiring deliberations stored in InGen’s records, including the justifications behind the offer or rejection decision. Thus, in total, I had qualitative insights into 400 hiring deliberations.
Finally, I constructed a quantitative dataset drawing on the average technical and cultural in-person interview evaluations and the offer decisions for each of the 1,094 candidates who reached the interview stage from June 2011 to February 2014. I supplemented these data with information on candidates’ work and educational history collected from resumes and publicly available online databases (e.g., LinkedIn), and other information collected by InGen, such as how they were recruited (e.g., employee referral), and the level and type of software engineering position to which they applied. Candidate gender and ethnicity were inferred by their names (see Appendix for more detail). The final analytical dataset consists of 1,082 candidates, as 12 candidates were missing education or work-related information.
Analytic Strategy
The main analysis is based on the qualitative data from the in-depth interviews, and qualitative data from the offer deliberations based on my notes from participant observation and notes taken by the hiring manager or recruiter present during the meeting. To analyze these data, I drew on an abductive approach, which allowed creative findings to emerge through an iterative process of qualitative data collection and analysis (Timmermans and Tavory 2012). During data collection, I highlighted emergent themes in the incoming data, elaborated on those themes in written analytic memos, and aggregated themes into broader concepts for analysis using the qualitative analysis software Dedoose (Charmaz 1983). I used the analytical technique of data matrices to compare themes by candidate gender (Miles and Huberman 1994), and I relied on triangulation across the three qualitative data sources to strengthen the reliability of the findings. The abductive technique of “alternative casing” (Timmermans and Tavory 2012) was particularly useful for discovering that the empirical findings fit well as a case of diversity commodification.
I support the main qualitative analysis by drawing on a quantitative data. 5 Specifically, I use structural equation modeling to determine whether the relationships uncovered in the qualitative analysis between applicant gender, the technical ability and cultural fit evaluations, and the job offer are reflected in the quantitative data. Generalized structural equation modeling is ideal for a supporting analysis, as it allows for a binary outcome (the offer decision) but also continuous mediators (the technical and cultural evaluations). I control for additional applicant characteristics including ethnicity, educational background and work experience, and hiring process factors such as the recruitment method, position level, position type, and whether at least one woman was on the interview panel. I estimate the model using the gsem package in Stata. Technical and cultural evaluations are centered and standardized.
Findings and Analysis
In this section, I first establish how InGen decision makers tend to evaluate candidates’ technical ability and cultural fit in line with gender stereotypes. I then demonstrate how decision makers appear to value women candidates for their contribution to diversity, and how this value subtly influences offer rationalizations during deliberation. I then turn to the quantitative data to conduct a complimentary analysis to the qualitative findings. Finally, I describe the curious lack of substantial variation in the findings by candidate ethnicity, even though Asian women are well represented among candidates in this setting.
Stereotype-Biased Evaluation during In-Person Interviews
The interview respondents describe how, from their perspective, men and women candidates differ in their technical ability and cultural fit during the interviews in ways that align with gender stereotypes. To be sure, interviews are social events between the interviewer and applicant, and gender stereotypes frame both participants’ behavior and their interpretation of the others’ behavior (Ridgeway 2011). The key observation is that decision makers emerge from the interaction perceiving men and women candidates to differ in key ways, which, in a context strongly emphasizing technical ability, favor men.
Perceived Technical Ability
For respondents, the most salient difference between men and women is not technical ability per se but the difference in the technical confidence displayed during the interview. Presenting confidence in one’s ability during the interview—and that includes pushing back on interviewers’ assertions and asking good clarifying questions—signals an aspect of technical ability that respondents prize (Cech et al. 2011). Engineer Henry Chu describes the importance of “communicating complicated things to other teammates, arguing, scheduling, coordinating, all that stuff. . . . In industry, convincing other people that [a given solution] is the right way to go is probably more important [than in school].” Technical confidence is an important element of technical ability.
Respondents describe women as displaying lower technical confidence than men during interviews. Senior engineer James Hoffman tells me that women seem more interested in what I think of their solution than coming up with the right solution. . . . It’s a lot of looking [backward]. . . . I have a dog [that] loves to be with people . . . but is always looking back to make sure you’re still following.
James’ comments illustrate interviewers’ tendency to judge women as exhibiting “approval-seeking behavior,” or being too deferential. Men candidates, on the other hand, tend to display confidence in their technical skills, at times to the point of being overconfident. Senior engineer Sofia Lomonso agrees that women are too deferential during the interview compared with men—but men, she tells me, are perhaps too confident, as she has never seen a man apologize for an error during the interview, unlike women, who tend to apologize for small errors in their code.
Respondents describe how women receive lower technical evaluations than men during interviews because of their lower technical confidence. As hiring manager Ethan Russo explains, at all levels, men tend to come off more confident, and women . . . come off as a little less sure of their answer. . . . It’s just faint and slight, but that may weigh into some people’s technical score where they’d be like, “She got the right answer, but I’m not sure if she would push for her right answer.” There’s a certain level of confidence you hope that they [have] when they’re explaining their problem.
For Ethan and other interviewers, the perception of women’s lower technical confidence plays into their lower technical evaluations relative to men.
Perceived Cultural Fit
Respondents also perceive that women generally have higher interpersonal warmth and enthusiasm than men, which leads to women having generally higher cultural evaluations. Interviewers attribute these stereotypical communal strengths to socialization and upbringing, with women being “just naturally more social creatures. . . socially programmed to be more open to taking other people’s advice” (hiring manager Andrew Morgan). In a typical portrayal of a woman with high cultural fit, engineer Justin Shi describes a recent woman candidate as a valuable potential hire because of her “bubbly” enthusiasm: She was super excited and also pretty bubbly, so just her attitude going into the interview was really positive. She seemed like she really wanted the job and liked InGen. Unfortunately, we ended up not giving her an offer. I say unfortunately because she seemed like a great person to work with.
Interviewers perceive some men candidates as interpersonally warm and enthusiastic, but in general, men are seen as less interpersonally warm than women. Women, senior engineer James Hoffman says, are “more used to engaging with people . . . [but the] cold fish . . . [who] won’t look you in the eye, give you the very limp handshake . . . tend to be male.” Interviewers argue candidates who are too cold or introverted will disrupt collaboration once hired. Those socially disruptive workers, who no one wants to interact with, are almost always men.
However, men have another avenue to high cultural fit that women, in general, do not: the perception that they think deeply about technology, have deep knowledge about it, or have a deep passion for it. Engineer Anthony Fang expresses the importance of displaying technical passion during the interview. “[I] think of my interviews as a nerdy conversation. . . . I really like arcane C++ things that [candidates] can dive into, and I like to see how far they go with that by themselves.” Passion for technology and tinkering with computers has long been documented as a key characteristic of an ideal masculine engineer (DePalma 2021; Robinson and McIlwee 1991; also see Rao and Neely 2019). During the interview, signs of a candidate’s passion for technology create emotional chemistry and heighten perceptions of underlying ability; by comparison, interpersonal warmth and enthusiasm do not connote an intimate relationship to technical ability. Importantly, when interview respondents describe examples of candidates with deep technical passion, they only recall candidates who were men.
Women Are Considered Valuable for Organizational Diversity
Interview respondents are well aware of the widespread pressure to increase the representation of women in software engineering. As a result of this pressure, decision makers consider women software engineering candidates as valuable for their contribution to organizational diversity. “We definitely want more [gender] diversity [but] we really don’t know how to attract it,” Sofia Lomonosov tells me. “It’s an ongoing problem, but we definitely strive for it. There is a conscious effort to look for diverse candidates.” 6 To be sure, respondents consider Black and Latino software engineers as contributing to diversity but increasing the representation of women is a far more salient issue at InGen. Respondents stress that InGen is one of many companies vying for female talent in response to the pressure to increase female representation in software engineering. As senior engineer John Brown says, “everybody’s compensating. Everybody’s hiring a bunch of female engineers, so the good ones get jobs and don’t get to InGen, because we’re not as big as [other companies]. . . . The good ones are getting snatched up.” For John, the reason for this is clear: “people are trying to offset the gender thing.” Women’s diversity value often manifests as excitement at the prospect of an applicant potentially increasing the gender diversity of the engineering core (Weisshaar et al. 2024). For instance, senior engineer Lisa Roberts expresses her enthusiasm for women candidates: “when I’m interviewing women, I’m always more excited. Part of me is cheering for them a little more.”
Although respondents value women for their potential contribution to organizational diversity, they simultaneously express commitment to meritocracy, and how the explicit use of diversity value would be illegitimate during evaluation and selection decisions. Amal Moshrif, senior engineer tells me, “I just give the score based on the performance. I don’t really look . . . I do not correlate gender with performance.” Another senior engineer, Samir Reddy, shares the same sentiment: “it doesn’t affect me one way or the other. It’s just whether or not you can do the interview [that] matters. Gender is not a concern.” And yet, adherence to meritocracy rests uneasily with interviewers’ expressed excitement over the prospect of hiring a woman software engineer. Senior engineer John Brown reveals this conflict when he tells me, we want to hire more females because it is imbalanced. When there’s a female candidate, you want them to do better. Even though internally I’m like, okay, cool, a female engineer, I don’t want it to make a difference, and it doesn’t make a difference in terms of how I rate them or how I might accept them or not.
John exhibits his excitement about hiring a woman engineer but catches himself and says that he cannot incorporate that value into his evaluations and offer decisions.
Strong meritocratic norms present a clear impediment to the explicit use of diversity value during offer decisions. And yet some respondents give me insights into what many may have left unsaid: that interviewers still incorporate diversity value into their decisions, but in a more subtle and unspoken way. As senior engineer Andy Farkas tells me, I wouldn’t be surprised if there’s a little bit of a bias to try and . . . it’s almost like affirmative action. I don’t want to say that we accept lower quality in terms of technical skill for women, but it just might be the thing where it’s like, “Oh, she wasn’t amazing technically, but [she’s] a female.”
Andy’s paraphrase of the underlying offer decision logic—“she wasn’t amazing technically, but [she’s] a female”—would be an unacceptable reference to diversity value if said explicitly during formal deliberation. As I show in the next section, decision makers instead use tortuous justifications during deliberations in lieu of explicitly drawing on diversity value.
Offer Justification and Effort during Hiring Deliberations
As previously described, respondents insist that hiring candidates with subpar technical evaluations, but strong cultural fit, is unusual, as it goes against meritocratic norms. And yet I find that such decisions do occur. In this section, I draw on qualitative data from the offer deliberations for which I have notes from participant observation or notes written by hiring managers or recruiters present in the meetings. On the basis of these data, Table 2 categorizes candidates by gender, the offer decision, and the strengths and weaknesses of the candidate as described by decision makers during deliberation. Of the 400 offer deliberations for which I have data—337 for men and 63 for women—I find that 20 of the 102 men who receive an offer (20 percent) are described as demonstrating substandard technical ability but strong cultural fit during the deliberation. By comparison, 11 of the 20 women who receive an offer (55 percent) fall under the same description. In Table 3, I display the bases for the candidates’ strong cultural fit and substandard technical evaluation, as discussed during the deliberation, and the justification decision makers use to eventually make the offer. Decision makers tend to draw on different offer justifications for men and women because the bases of their cultural evaluations differ: decision makers focus on women’s high cultural marks for their enthusiasm, excitement, or good energy but focus on men’s high cultural marks for passion for technology and deep technical knowledge. As a result, justifications for women often require far more effort to make. I argue that women’s diversity value motivates decision makers to make this effort.
Candidate Strengths and Weaknesses as Described during the Offer Deliberation, by Offer Decision and Candidate Gender.
Note: The distribution of offer and rejection profiles significantly differs for men and women (p < .001, Fisher’s exact test). Twenty-two percent of the candidates are ethnically Indian, 46 percent are ethnically Chinese, 7 percent are another Asian ethnicity, and 26 percent are other (non-Asian) ethnicity. Forty-four percent of the sample received an undergraduate degree from a foreign country. There are no significant differences across candidate gender in the distribution of ethnicity (p = .213, χ2 test), or place of education (p = .604, χ2 test).
Detailed Candidate Strengths, Weaknesses, and Offer Justification for Candidates Described as Strong in Cultural Fit but with a Substandard Technical Performance, by Candidate Gender.
Note: N = 20 men candidates and 11 women candidates. I also document other justifications decision makers used. For instance, decision makers might simply excuse the candidate’s technical weakness as less damaging than it might have appeared in the interview or attribute the weakness to something else, such as the candidate being tired.
Justification Effort for Women
Decision makers rely on the following argument to justify the job offer for 7 of the 11 women who receive an offer after being described as demonstrating strong cultural fit but substandard technical ability: the candidate’s enthusiasm and interpersonal warmth means she is motivated to learn and improve her technical skill, often with help from mentors.
7
Senior engineer Andy Voung explains this justification explicitly: A girl might not be up to par technically, but she’s very excited and seems really engaged. If she was bad technically, we’d never consider it. [I’ve seen it] two or three times where you know this person seems really engaged, and she seems like she could be good if we just trained her a little bit. Everyone was like, “I really wanted her to do well.” When that’s the consensus across the board, we’re like, “okay, let’s just try it out.” She’s not up to par with what we normally expect, but . . . she seems clearly motivated to learn.
Using the logic of a motivation to learn because of enthusiasm and warmth requires a good deal of deliberation and hand-wringing among decision makers, in part, because it is generally perceived as risky. As the basis for the justification is not rooted in the impression that the candidate has strong underlying technical ability, it is often unclear whether the candidate will, in fact, learn the necessary skills. Hiring manager Richard Preston’s summary of the offer justification shows the committee’s weighing of risk when deciding whether to hire a woman candidate with mediocre technical evaluations (2.5 average), but strong cultural evaluations (3.5 average). Richard writes, “Everyone likes Emma’s energy and enthusiasm—3s and 4s on culture-fit. Technical[ly] she was good, though she does lack depth of experience in some areas. Generally seemed smart and wrote good code . . . I had to think long and hard about it [italics added]. The candidate performed well on two of the technical questions and received high cultural scores for her communication, passion for learning, and InGen in general, so I thought we should make her an offer. I talked to other hiring managers about this type of situation before, and it turned out well in other circumstances . . . .” A representative of upper management writes: “I’m in favor of hiring smart people who are enthusiastic learners. I’m comfortable with this.” (Hiring deliberation, recruiters’ notes, November 12, 2012)
As this passage reveals, interviewers emphasize the candidate’s enthusiasm and energy to compensate for the interviewers’ technical concerns, but only with a serious weighing of risk. Richard has little experience as a hiring manager and proceeds only after confirming precedent from the more experienced hiring managers. Importantly, Richard and the other decision makers put a lot of effort into justifying the offer to themselves since it is a more logically tortuous rationale: “thinking long and hard about it” and talking to other hiring managers.
Decision makers spend much effort convincing themselves that enthusiastic women will be motivated to learn the needed technical skills, which sometimes entails referring to previous times the same rationale was used to hire men and women candidates who later became successful employees. In one hiring deliberation, interviewers are wary of a woman candidate’s technical skills and hesitant to give her an offer based on the argument that her enthusiasm will result in learning the skills she needs (hiring deliberation, participant observation notes, June 6, 2014; average technical evaluation = 2.1, average cultural evaluation = 3.1). One interviewer has the group recall a current employee with infectious enthusiasm who they hired despite his middling technical evaluations. As he became a successful engineer, so too might the woman candidate. This argument—pointing to a rare, but successful case—is the final piece of convincing they need.
Finally, decision makers put effort into defending their offer justifications to upper management who disproportionately scrutinize offer decisions for women candidates. Despite only 20 women in the deliberation data receiving an offer, InGen management questions the offer decisions for five women candidates, including the above example from Richard Preston. In three of those instances, upper management questions the technical ability of women candidates who are given offers with an enthusiasm-based justification. In one instance, upper management asked, “wait . . . all the coding comments were weak on this one . . . why so enthusiastic?” In response, the hiring manager reiterates the offer justification, emphasizing that she performed well, and that “she also had a great attitude and some of her needing help was related to being nervous in an interview setting.” Management upholds all offer decisions for women, but only after pushback from the hiring manager and teams. There are only four instances of upper management questioning the offer justification for men, and none involved an enthusiasm-to-learn offer rationale. These instances of resistance from management highlight the atypicality of the enthusiasm-to-learn argument, as well as the general scrutiny of women candidates.
Justification for Men
During deliberations, decision makers rarely use enthusiasm-to-learn justifications to offer jobs to men who are described as demonstrating strong cultural fit but substandard technical ability, with only one instance in the deliberation data. Instead, decision makers are more likely to argue that candidates’ passion for technology means they will quickly update their technical skill, or that their underlying technical ability means they either already have the skills that were thought missing or will quickly learn them. Decision makers spend less effort on making these types of justifications, compared with enthusiasm-based justifications, as there is a clearer logical connection between a passion for technology or (assumed) deep technical knowledge, and actual or future technical ability. Take the following hiring deliberation (May 1, 2014) in which decision makers expressed concern over a candidate’s coding ability (average technical = 2.6, average cultural = 3), yet emphasized the candidate’s perceived underlying ability as the evidence that he will quickly learn the missing technical skills.
We had a really good high-level conversation. He was . . . pretty good at talking about technology. It seems like . . . he naturally thinks about [it].
My feeling is that he would be a solid contributor.
I wish I could see his code.
He is conceptually strong, [but] I didn’t see enough of his coding to say that he was a strong coder.
There is potential there to be a strong person, if he improves, and can actually back up his conceptual understanding with coding.
Kseny, what is in your gut?
I think it is worth the risk.
Interviewers override troubling concerns about the candidate’s coding ability—many did not see his actual code—by citing the candidate’s “natural” ability to learn the technical skill. The underlying assumption among interviewers is that, despite his mediocre technical evaluations, he is a deeply technical person. The deliberation is quick and ends with the hiring manager trusting her gut that the candidate has or will soon have the technical ability for the position.
In other deliberations, interviewers draw on men’s underlying ability or passion for technology to argue that they will quickly learn the necessary skills with a little mentoring. In one deliberation, the hiring manager frames a candidate’s strength as “he loves reading [about] software principles . . . [and] new software productivity practices. He is basically perfect when it comes to attitude, excitement, and drive.” The manager justifies the offer by saying that “technically [the candidate] was on the cusp, but with his very high cultural scores . . . he seemed like a good fit for the role. Naturally, he will require a bit of mentorship to really excel in the role.” There is no need for much effort to justify the offer: the connection between passion for technology and future technical skill is clear.
Triangulation of the Qualitative Findings
I triangulate the above qualitative findings with a quantitative analysis of 1,082 software engineering candidates who entered InGen’s interview stage. Table 4 displays the descriptive statistics of the sample. Only 13 percent of candidates are women, which is surprisingly given decision makers’ emphasis on increasing gender diversity. Asian workers are well represented, constituting 71 percent of candidates. Moreover, men and women candidates differ significantly in ethnic composition (p < .001, χ2 test): while 69 percent of men are categorized as Asian (Indian, Chinese, or “other” Asian), Asian workers constitute 87 percent of women candidates.
Descriptive Statistics of Quantitative Dataset of In-Person Candidates by Candidate Gender.
Note: In “passive” recruitment, InGen recruiters contact engineers who are not actively looking for work.
Average evaluation received by the candidate. The reported values represent the average of the mean technical or cultural evaluation received by candidates. For instance, on average, men candidates receive an average in-personal technical evaluation of 2.45.
Among candidates with noninternship work experience.
p < .05 and **p < .01 (two-tailed t-test); †††p < .001 (χ2 test).
I conduct a generalized structural equation analysis to capture the relationships between candidate gender, the technical and cultural evaluations, and the job offer as described in the qualitative findings, accounting for the candidate characteristics, recruitment method, position characteristics, and interview panel characteristics listed in Table 4. Figure 1 displays the results, which are consistent with three key qualitative findings. 8 First, the qualitative analysis suggests that being a woman, compared with a man, is negatively associated with the technical evaluation, but positively associated with the cultural evaluation, all else equal. As seen in Figure 1, the quantitative analysis supports this finding. Second, the qualitative analysis suggests that interviewers are motivated by women’s diversity value, but they do not explicitly apply women’s diversity value to the offer decision. The quantitative analysis supports this finding: I find no significant direct effect of candidate gender on the job offer after accounting for the technical and cultural evaluations.
Significant pathways from candidate gender to the job offer.
Third, the qualitative analysis suggests that decision makers frown upon giving job offers to candidates who have substandard technical evaluations even if they are enthusiastic and warm. As women tend to receive high cultural marks for warmth and enthusiasm, while men receive high cultural marks for their passion for technology and deep technical knowledge, one might expect the cultural evaluation to have a smaller effect on the offer decision for women candidates compared with men. However, the qualitative analysis also reveals that decision makers put in effort to justify and defend offers to women who receive high cultural fit marks for enthusiasm and excitement but substandard technical evaluations, thus counteracting this would-be penalty. The quantitative analysis supports the idea that decision makers efforts counteract a potential female penalty: I find no significant interaction effect of candidate gender and the cultural evaluation on receiving a job offer. It is in the absence of a gender effect on the job offer decision where circumstantial evidence for the role of diversity value lies.
The Primacy of Gender in Diversity Commodification at InGen
The preceding analyses suggest a subtle process of diversity commodification of gender during job offer decisions at InGen, but one in which race and ethnicity are curiously absent. First, there is little emphasis on Black and Latino diversity at InGen. The interview respondents acknowledge that Black and Latino workers would contribute to firm diversity, but they are far more comfortable, even enthusiastic, when talking about the importance of gender diversity. The lack of attention to Black and Latino diversity parallels an almost complete absence of Black and Latino candidates who enter the in-person interviews: interview respondents had little to say about their perceptions of Black and Latino candidates, primarily because most had no personal experience to draw on. 9 The complete avoidance of Black and Latino diversity as an issue at InGen may be symptomatic of an implicitly White organization (Ray 2019) and a product of Silicon Valley organizational culture before the Black Lives Matter movement made race more prominent in discussions of diversity (see Weisshaar et al. 2024).
Second, decision makers appear to perceive and treat Asian women and White women similarly, according to the data available. Interview respondents tend to talk about women candidates as women, rather than differentiate between White women and Asian women. Data from hiring deliberations do not provide decisive evidence of heterogeneity in the treatment of women candidates across ethnicity. And, in the quantitative analysis, I found no substantial variations in the gendered pattern across candidate ethnicity. 10 This nonfinding, although perhaps curious given that general stereotypes of Asian women differ from White women (Benard et al. 2023), resonates with current research suggesting that for Asian women in software engineering, gender is far more salient than Asian ethnicity because of the hypermasculine context in which there is pressure to increase the representation of women and where Asian workers are generally well represented (Chow 2024). Ethnic divisions among women may simply not salient in this context.
Discussion
Previous research on diversity commodification suggests that under strong pressures to diversify, corporate decision makers value women for their contribution to diversity and incorporate that value into their hiring screening decisions alongside their assessments of candidate ability biased by gender stereotypes (Weisshaar et al. 2024). And yet it is an empirical question whether such a process occurs during the in-person evaluations and offer deliberation, particularly as explicit incorporation of diversity value is often perceived as unmeritocratic and thus illegitimate at those stages. Given this puzzle, I ask two questions: Under pressure to diversify, do employers incorporate diversity value into their offer decisions when hiring for male-dominated occupations? If so, how?
Drawing from a case study of software engineering hiring at a midsized high-technology firm, I find evidence consistent with the influence of gender stereotypes on the two evaluation criteria during in-person interviews: technical ability and cultural fit. At the same time, InGen decision makers value women candidates for their contribution to diversity while adhering to the norms of meritocracy during evaluation and offer decisions that make explicit incorporation of diversity value illegitimate. It is in this context that I document how diversity value motivates decision makers to put in extra effort to justify the offer for women who receive strong cultural fit evaluations on the basis of warmth and enthusiasm, but substandard technical evaluations. These findings suggest that diversity commodification occurs during the offer stage of hiring, with diversity value acting as an implicit motivating force rather than an explicit consideration.
These findings most clearly contribute to current theory of diversity commodification in the hiring setting by offering empirical evidence that applicant diversity value may influence decision makers during offer deliberations. Current empirical work on diversity commodification focuses primarily on the initial screening decision. Although other research has documented instances of formal, infrequent, and rare diversity considerations at the offer stage (e.g., Rivera 2015b), it was empirically unclear how diversity value may influence the offer decision more generally, if at all, particularly when norms of meritocracy are high. This study fills this empirical gap and offers theoretical insight into how diversity value influences the offer decision: as a motivational factor that works in tandem with gender stereotypes of men and women’s performance capacities to influence job offer decisions. In highlighting the role of gender stereotypes in biasing assessment and diversity value, this study also contributes to literature on gender inequality in performance evaluations which focuses primarily on stereotypes and biased assessments as a driving factor of gender-based discrimination while neglecting diversity value (Correll et al. 2020; Ridgeway 2011).
Second, this research contributes to literature on cultural fit during offer decisions. Previous literature argues that cultural fit is an important and often legitimate consideration during offer decisions, and that the use of cultural fit results in increasing inequality, such as by class (Rivera 2012b) and ethnicity and immigrant status (Chavez 2021). Decision makers may also use cultural fit to discriminate against unwanted groups in a seemingly legitimate way (Nichols et al. 2023). I contribute to this literature by demonstrating how, under pressures to diversify, decision makers may use cultural fit to legitimately discriminate in favor of underrepresented groups, in this case, women. Cultural fit may indeed be a “hiring tool,” but the extent to which decision makers use that tool to help or hinder traditionally marginalized groups depends on context.
And finally, this research contributes to literature on motivated reasoning (Norton et al. 2004, 2006). Previous literature suggests that when decision makers desire a given outcome, they may consciously or unconsciously shift evaluation criteria to achieve their desired result. The process that I describe can be categorized as a type of motivating reasoning: decision makers want to hire more women because of their value toward diversity, and in response make and defend tortuous justifications. In many cases, motivated reasoning results in discrimination against groups typically disadvantaged (Biernat and Kobrynowicz 1999; Phelan, Moss-Racusin, and Rudman 2008). But I find, in line with previous experimental studies (e.g., Norton et al. 2004), that motivating reasoning under pressures to diversify may reduce inequalities in outcomes rather than accentuating them.
I argue that the diversity value of women is a driving force behind the justification process found during offer deliberations at InGen. And yet there are two alternative explanations worth considering. First, decision makers, who are by and large men, may push for women to be hired because of a romantic attraction. Such an explanation is not far-fetched; previous research suggests that emotional chemistry, even romantic chemistry, is an important motivation during hiring decisions (Rivera 2015a). A second alternative explanation is that women decision makers, not their men colleagues, feel an affective connection with women candidates, resulting in women interviewers evaluating women candidates as better cultural fits and leading the effort to justify the offer for women candidates. Both alternative explanations arguably suggest either heterophily or homophily between interviewer and candidate explain the empirical patterns I find. In an auxiliary quantitative analysis of interview-level data, I do find some evidence of homophily in that women interviewers evaluate women candidates significantly higher in cultural fit than men interviewers (see the Appendix for details). Yet importantly, I find that both men and women interviewers evaluate women candidates higher in cultural fit than men candidates. Moreover, the qualitative data suggest that both men and women put in effort to justify the offer for women during hiring deliberations. Although homophily effects may add nuance to the main argument, they do not negate it.
There are several limitations to this study to be addressed in future research. First, the case study design provides insights into a particular organization but is inherently limited in addressing the extent to which the findings differ across contexts. For instance, although implicit incorporation of diversity value may occur in other settings, the exact process may require the use of cultural fit as a formal hiring criterion. Moreover, it is not clear whether the lack of differentiation among women candidates found at InGen holds in contexts in which Asian workers are less prevalent. Second, the data, although comprehensive in some ways, are limited in others. For instance, although I benefit from having a quantitative measure of cultural fit, separate evaluations for technical passion and interpersonal warmth would have made the quantitative analysis more precise. The age of the data may also limit the relevance of the findings, as I collected the data in 2013, before the Black Lives Matter movement placed race more squarely as a diversity concern. The conspicuous lack of attention to Black and Latino identity among InGen decision makers may have differed had the data been from a more contemporary period. Finally, future research should determine whether factors such as immigrant status moderate the gendered process described here. 11
Conclusion
U.S. corporations often face pressures to diversify professions that are dominated by men. Understanding how organizational decision makers respond to such pressure is important. Decision makers may not respond at all, which would suggest a failure of such pressure to make a difference. If they do respond, understanding how may give us insights into how successful such efforts will be. In this article, I focus on software engineering hiring and demonstrate how decision makers value women’s contribution to diversity and, in response to that value, justify giving offers to women with strong cultural fit evaluations but substandard technical evaluations. In one sense, this practice may benefit women. At least in this case study, women candidates who enter the in-person interview stage are equally likely to receive an offer compared with men. However, the longer term consequences are unclear. Gender stereotypes are still strong in this context, and women are perceived stereotypically as having lower technical ability than men during the interviews. It is possible that such perceptions of substandard technical ability fueled by gender stereotypes, or the stigma of being hired for reasons other than exceptional ability, may follow women workers into the workplace once hired.
At the time of this writing, corporations are facing growing political backlash against their diversity, equity, and inclusion efforts, and some leading firms are eliminating their formal goals to increase diversity (Isaac, Frenkel, and Conger 2025). Thus, a real question is whether the process I find here will still occur in the near future. If indeed the signs of growing backlash portend a change in pressures to diversify, decision makers may be less influenced by diversity value and, without a counterweight to gender stereotypes and biased evaluation, may be more likely to discriminate against women. Alternatively, if the pressure to diversify remains strong among company employees, job applicants, and other constituencies on which organizations depend (see Weisshaar et al. 2024), decision makers may continue to believe that increasing diversity gives their organization a competitive advantage and, in turn, continue to allow diversity value to influence their offer decisions. Because the process I find does not involve the explicit incorporation of diversity value into decision making, the influence of diversity value may continue even if organizations face backlash against their diversity efforts.
Finally, this study provides insights for policymakers interested in increasing gender diversity in organizations. The process of diversity commodification represents a hands-off approach to diversity management in that decision makers do not receive any formal, top-down instruction to incorporate diversity value during the offer decision; they do so out of their own volition. Organizations may benefit from this arrangement in that they can retain an image of meritocracy and objectiveness in their hiring processes while depending on individual workers to incorporate diversity value in decision making. On the other hand, this approach has clear disadvantages. First, gender bias may still influence candidate evaluations in subtle ways. Second, even if more women receive an offer, the overall consequence of this process on gender diversity is quite small: women remain very underrepresented among job candidates at InGen and among InGen engineers. More formal and explicit practices may indeed be more effective at increasing gender diversity.
Supplemental Material
sj-docx-1-srd-10.1177_23780231251323372 – Supplemental material for Gendered Pathways to the Job Offer: Stereotypes and Diversity Value during Job Offer Decisions in Software Engineering Hiring
Supplemental material, sj-docx-1-srd-10.1177_23780231251323372 for Gendered Pathways to the Job Offer: Stereotypes and Diversity Value during Job Offer Decisions in Software Engineering Hiring by Koji Chavez in Socius
Footnotes
Supplemental Material
Supplemental material for this article is available online.
Notes
Author Biography
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
