Gender bias is pervasive at work and in organisations, creating inequalities at every stage of the employment cycle. Gender-based stereotypes affect which candidates get recruited for certain roles and which do not, which candidates get selected for those roles and why, how salaries are negotiated, how managers provide feedback to their employees, and which employees receive career development opportunities and career encouragement and which do not. Each of these factors compounds across women’s careers, producing and sustaining gender inequality from recruitment to selection to promotion.
Decades of research has made one thing clear: gender biases are nearly always present in employment decisions, subtly influencing our assessments about who is the ‘right’ or ‘best’ person for the job. This insight paper highlights some of the research examining how gender bias operates at work and provides evidence-based suggestions for creating more equitable recruitment and promotion systems.
Gender bias in recruitment and selection
Gendered wording in job advertisements. It has long been unlawful to advertise positions specifically for women or men or to discriminate on the basis of sex*. However, research has shown that the language used in job advertisements can subtly filter men and women into separate occupations and industries.
Advertisements in male-dominated occupations and industries are more likely to contain masculine-stereotyped words like ‘competitive’ and ‘dominant’ than advertisements for jobs in female-dominated occupations and industries (Gaucher et al., 2011). When presented with such advertisements, women are more likely than men to rate these jobs as unappealing; they also tend to believe that women are not well represented in these roles. Use of such language is unlikely a deliberate attempt to exclude. However, gender-coded language in job advertisements may not only deter individual women from applying for certain jobs, it can also reinforce wider social beliefs about who ‘belongs’ in those jobs and who does not (Gaucher et al., 2011).
Gender bias in interview invitations. Multiple studies have demonstrated that when women apply for jobs, they receive fewer interview invitations than equally qualified men – an effect that is compounded for older women, women with children and women from certain ethnic or racial groups (Abrams et al., 2016; Bertrand and Mullainathan, 2004; Correll et al., 2007; Moss-Racusin et al., 2012; Neumark, 2010; Riach and Rich, 2006; Weichselbaumer, 2016). Research from Australia and the United Kingdom has shown that men are similarly disadvantaged when applying for entry-level roles (Riach and Rich, 2006) or jobs in heavily female-dominated occupations (Booth and Leigh, 2010).
In each of these studies, researchers sent out matched pairs of applications, which differed only in candidates’ identifying information, to thousands of actual job advertisements and monitored interview invitation rates. Because this approach effectively controls for other factors, such as work history or education, it is widely seen to be the most reliable indicator of gender-based employment discrimination in the initial recruitment stage (Neumark, 2010). If all else is equal, and men receive more interview invitations than identically-qualified women, logic dictates that managers must be using gender to assess candidates’ employability (Foley and Williamson, 2018).
* Sex Discrimination Act 1984 (Cth
Gender bias in evaluation of credentials. Research has shown how women face tougher evaluation of their credentials during the initial recruitment stage. A Yale University study asked 127 science faculty at research-intensive universities to rate the job application materials of a prospective laboratory manager who was randomly assigned either a female or male name (Moss-Racusin et al., 2012). Both male and female evaluators rated the male candidate as more competent and ‘hireable’ than the female candidate, even though both applications were otherwise identical. Evaluators also recommended higher starting salaries for the male applicant and offered more mentoring opportunities.
In another study, 238 academic psychologists in the United States were randomly assigned to evaluate one of two identical resumes (either ‘Karen Miller’ or ‘Brian Miller’) for a candidate applying for an academic role (Steinpreis et al., 1999). Although participants said they liked the female candidate better than the male candidate, they were more likely to recommend the male candidate for the job. Interestingly, participants were four times more likely to write cautionary remarks in the margins of the female candidate’s application, questioning the independence and legitimacy of her accomplishments, such as: ‘I would need to see evidence that she had gotten these grants and publications on her own’.
These findings suggest that women face more rigorous scrutiny and are held to higher standards than equally qualified men and that, consciously or unconsciously, managers associate candidates’ genders with their likelihood of job-related success (for a comprehensive and accessible account of gender differences in job evaluations, see Bohnet, 2016; Fine, 2011).
The ‘likeability’ penalty in job interviews and salary negotiation. Women are sometimes encouraged to ‘lean in’ (Sandberg, 2013) by displaying confidence and assertiveness in their careers. However, studies have repeatedly demonstrated that women who appear to be overly confident or assertive are frequently judged to be less ‘likeable’ than women who conform to a more traditional feminine stereotype (Eagly and Karau, 2002; Heilman, 2001; Rudman and Glick, 2001).
Job interviews. Researchers at Rutgers University videotaped male and female actors interviewing for a role as a computer lab manager (Phelan et al., 2008). In one experimental condition, the actor-candidates presented as self-confident, ambitious and competitive; in the other condition, the actor-candidates presented as competent, but modest and cooperative.
A sample of 428 participants, who believed they were evaluating real job candidates, were then asked to rate the candidates by competence, social skills and employability. When assessing employability, the participants rated competence above social skills for all candidates except for the female candidates who presented as self-confident, ambitious and competitive. For these ‘agentic’ women, social skills were deemed to be the most important employability factor. Thus, relative to other applicants, confident women were penalised for violating traditional feminine stereotypes (Phelan et al., 2008).
Men also pay a penalty when acting against stereotypes. A subsequent study found that men applying for a managerial role that required both technical and social skills were penalised relative to women if evaluators perceived the men to be overly modest (Moss-Racusin et al., 2010).
Salary negotiations. Women also face a ‘likeability’ penalty if they attempt to negotiate for higher pay. In a series of controlled experiments, researchers at the universities of Harvard and Carnegie Mellon found that female candidates who advocated on their own behalf for better salaries were judged less favourably than men who did the same, particularly if the evaluator was male (Bowles et al., 2007). When researchers examined candidate behaviour, they found that women were significantly less likely than men to consider asking a male manager for a better salary package, suggesting that women are well attuned to the social risk of attempting to negotiate for a better deal.
Given that men hold most senior management positions in Australia and therefore control the allocation of financial rewards in many organisations, these findings are problematic for women seeking to address the gender pay gap on their own. If ‘women are justifiably less inclined than men to initiate negotiations with men, then they may have fewer opportunities to increase their compensation and promotion potential’ (Bowles et al., 2007, p. 99).
Gender bias in the application of selection criteria. Evaluators frequently use selection criteria to determine which candidate to hire or promote. However, research has shown that when selection criteria are not given preassigned weightings, evaluators may subtly shift the importance of certain criteria to favour candidates who fit a gender stereotypical profile.
In one U.S.-based experiment, researchers asked 93 participants to consider both male and female applicants for a construction manager position (Norton et al., 2004). When the applicants’ genders were concealed, participants rated education as a more important criterion for success on the job than prior work experience and were more likely to prefer the more educated candidate. However, when the applicants’ genders were revealed, and the highly educated applicant with low work experience was a woman, participants rated prior work experience as more important than education, to the advantage of the male applicant (Norton et al., 2004).
A similar experiment by Yale University researchers asked participants to evaluate one of two hypothetical candidates for the role of police chief (Uhlmann and Cohen, 2005). One of the candidates was a streetwise but well-liked police officer with little formal education, while the other candidate was a highly educated, media savvy official who lacked street-based experience. Participants were asked to rate the candidates by the various education or street experience criteria. Next, participants were asked to rate the importance of each criterion to the role of police chief. Participants who evaluated the male candidate (‘Michael’) amplified the importance of formal education or street-experience when he possessed these attributes and downplayed their importance when he lacked these attributes. The female candidate (‘Michelle’) received no such accommodation, regardless of her qualifications or experience.
The participants did not overtly downgrade Michelle because of her gender. Instead, evaluators simply reframed ‘merit’ – the skills, experience and attributes deemed essential to the job – to fit Michael’s credentials. The evaluators may have ‘felt that they had chosen the right man for the job, when in fact they had chosen the right job criteria for the man’ (Uhlmann and Cohen, 2005, p. 479). In a subsequent study, the researchers found that evaluators shifted selection criteria to favour female applicants over male applicants when the job in question was stereotypically feminine – i.e. a professorship in Women’s Studies – regardless of the male applicant’s skills or experience (Uhlmann and Cohen, 2007).
Evaluators in all three studies justified their prejudicial decisions by emphasising different selection criteria depending on the gender of the candidate. Interestingly, evaluators who engaged in the process of shifting criteria were more likely to rate their decisions as more ‘objective’ than evaluators who did not shift criteria to fit the gender-typical candidate (Uhlmann and Cohen, 2007). These findings suggest that selection processes grounded in flexible criteria may consciously or unconsciously enable evaluators to justify or conceal prejudicial decisions under the guise of objectivity.
Should organisations anonymise job applications?
Many organisations have turned to anonymous recruitment procedures, removing names and other identifiers from job applications and resumes, in order to minimise the effect of biases in hiring, but this approach has produced mixed results. Although some studies have shown that anonymous recruitment can increase the proportion of women receiving interview invitations (Åslund and Nordström Skans, 2007; Krause et al., 2012b), other studies have shown that anonymous recruitment can result in fewer women being selected for interviews (Hiscox et al., 2017; Krause et al., 2012a).
As a bias-reduction strategy, anonymous recruitment is grounded in the assumption that managers cannot rely on gender-based stereotypes in their assessment of candidates’ employability if the primary trigger for such stereotypes (i.e. candidates’ names) are removed from applications. However, one recent Australian study found that hiring managers may be able to deduce information about candidates’ genders from implicit cues embedded in resumes (Foley and Williamson, 2018). For example, when managers noted extended periods of mid-career leave in anonymised applications, they perceived those resumes as likely belonging to female candidates, thereby making gender salient to their assessments and reintroducing the potential for bias.
Although anonymous recruitment aims to shift managers’ focus away from applicants’ genders to their job-related skills and qualifications, the process does not account for structural factors (such as women’s disproportionate concentration in part-time work, or time out of labour market) that may affect their ability to accumulate job-related skills and qualifications at an equal rate. Thus, anonymous recruitment practices may subtly benefit candidates who are already advantaged in the labour market (Foley and Williamson, 2018; Rinne, 2014). Anonymisation may also prevent hiring managers from considering male and female candidates’ achievements relative to time in the labour market and caring responsibilities, or from promoting diversity by engaging in affirmative action (Hiscox et al., 2017).
Further research is needed to assess the conditions under which anonymous recruitment is most effective at reducing discrimination. However, merely removing names and other identifiers from traditional job applications and resumes does not appear to be sufficient to remove bias from recruitment processes.
Gender bias in promotion
Gender biases also affect women’s ability to advance within organisations. Gender differences can also be observed in the performance evaluation feedback women and men receive, the factors contributing to women’s relatively lower representation in senior leadership roles, and the opportunities women receive for career development.
Language in performance evaluations. A study at Stanford University’s Clayman Institute for Gender Research found that women were systematically less likely than men to receive ‘specific feedback tied to outcomes’ (Correll and Simard, 2016), regardless of whether the feedback was positive or negative. In an analysis of more than 200 performance reviews conducted in a large U.S. technology firm, the researchers found that 57 per cent of performance reviews of female employees contained vague feedback, such as ‘you had a great year’, compared to 43 per cent of reviews of male employees. Men were significantly more likely than women to be given a clear picture of what they were doing well, and specific guidance on how to improve (Correll and Simard, 2016).
Similarly, the language in performance evaluations can be strongly gendered. Researchers in the United States evaluated a large military dataset comprised of 81,000 performance evaluations of 4,000 employees (Smith et al., 2018). In positive evaluations, men were most often described as ‘analytical’, while women were described as ‘compassionate’. In negative evaluations, men were most likely to be criticised for being ‘arrogant’, while women most likely to be accused of being ‘inept’.
The language used in performance evaluations is a clear indicator of what is valued and respected in organisations (Smith et al., 2018). Although ‘analytical’ and ‘compassionate’ are both positive descriptors, the former is more likely than the latter to be regarded as integral to the mission and strategy of a military institution. Similarly, arrogance is more likely than ineptitude to be dismissed as an unfortunate but forgivable by-product of strong leadership. Thus, on both positive and negative feedback measures, women are penalised (Smith et al., 2018).
Australian research has shown that female employees perceive these differences. A survey of 4,481 employees in large and small Australian businesses, government and non-profit organisations found that women were twice as likely as men to report being told that they needed to display ‘more confidence’, and 30 per cent more likely to report being told that they needed ‘more experience’ to be ready for promotion (Sanders et al., 2017). However, only half of the women said they were given the opportunity to gain the experience required or received clear and specific feedback about what experience they needed to be deemed ready for promotion. Men were 50 per cent more likely than women to say that they had received clear and specific feedback about how to improve (Sanders et al., 2017).
Women receive fewer challenging assignments. Managers often perceive that women have lower levels of career motivation than men and may seek to ‘protect’ their female employees from overwork. Such assumptions contribute to women receiving fewer career development opportunities, such as challenging work assignments, training and development and career encouragement.
Researchers surveyed 112 employees and 52 managers at a Fortune 500 company in the United States to assess how managers perceived the career motivations of their employees, and how those perceptions influenced the distribution of career development opportunities (Hoobler et al., 2014). Managers perceived that female employees had lower motivation levels than male employees, irrespective of their seniority, tenure or promotion histories, and found a strong correlation between managers’ perceptions of employees’ career motivation and employees’ reports of receiving career development opportunities. In other words, managers gave more career development opportunities to employees they perceived as being more motivated – mostly men. Moreover, employees who reported being given more career opportunities reported higher career aspirations, suggesting that career development opportunities feed career motivation and aspiration. When women are overlooked for career development opportunities because they are deemed to be less committed or motivated, their career aspirations suffer.
A series of studies in the United States has also found that women and men are equally likely to express interest in receiving challenging work, but male managers are less likely to give women challenging assignments out of a benevolent desire to protect women from overwork (King et al., 2012). Taken together, these findings suggest that stereotype-based beliefs about women’s lower work commitment or need to be ‘protected’ – which run counter to findings from research (AWWF, 2017) – can contribute to women receiving fewer career development opportunities necessary to progress their careers.
Leadership aspirations and self-promotion. Australian research has shown that women and men have similar aspirations and ambition when it comes to their careers, but women are more likely than men to believe that ‘feminine’ leadership styles are undervalued in workplaces. Women also avoid self-promotion because they perceive – often correctly – that such behaviour will reflect poorly on them.
A survey of 842 members of the Australian business community found that 74 per cent of women and 76 per cent of men aspired to leadership roles (Sanders et al., 2011). However, significant differences emerged in in respondents’ beliefs about why there were fewer women in senior management roles across Australia. Among respondents, 61 per cent of men said that women were less represented in senior leadership roles due to competing work-life priorities, compared to 22 per cent of female respondents (Sanders et al., 2011). In contrast, 78 per cent of female respondents said the shortage of women in senior leadership roles could be attributed to a lack of appreciation for stereotypically ‘feminine’ leadership styles. Respondents in this group said that men were more likely to ‘appoint or promote someone with a similar style to their own’ and that women were less likely than men to advocate on behalf of their own experience and capabilities, or self-promote (Sanders et al., 2011).
The perception that women are less likely than men to promote their accomplishments has been shown in multiple studies (for a review, see Budworth and Mann, 2010). However, women’s reticence to engage in self-promotion may be a form of self-preservation. Evidence suggests that women may be well aware of the penalty faced by agentic women and may downplay their own achievements to avoid backlash (Eagly and Karau, 2002; Heilman, 2012, 2001; Rudman and Glick, 2001). Indeed, research has shown that women avoid self-advocacy at work if they perceive that doing so will reflect unfavourably on them (Battle, 2008 cited in Chrobot-Mason et al., 2019).
Sustained experiences of gender inequality in recruitment, selection and evaluation may contribute to women’s under-representation in senior management (Brands and Fernandez-Mateo, 2017). Using a combination of field, survey and experimental data from the United Kingdom and the United States, researchers at the London Business School recently found that senior professional women were less likely than men to consider a job opportunity if they had been rejected by the firm in the past, and that experiences of rejection were more likely to trigger questions of belongingness in senior professional women, compared to men of similar professional rank (Brands and Fernandez-Mateo, 2017). The authors argued that women’s experience of gender inequality at various stages in their careers may influence their tendency to scale back their ambitions – to ‘lean out’ – when considering senior leadership roles.
Gender bias and the myth of ‘meritocracy’
Many organisations pride themselves on employing and promoting candidates based on their individual skills, attributes and qualifications, irrespective of gender or race. However, the evidence presented in this paper demonstrates the way in which cultural assumptions and stereotypes can sway managers’ recruitment, selection and promotion decisions in a non-conscious fashion. Because we are all subtly influenced by these pervasive biases, it can be difficult for managers to determine which candidate truly has ‘merit’ and which does not.
In a series of three experiments involving 445 participants with managerial expertise, researchers in the United States have demonstrated that managers in organisations that explicitly promote themselves as meritocratic – recruiting, rewarding and promoting the ‘best’ people based on their individual merits – are more likely to favour men over equally qualified women (Castilla, 2008; Castilla and Benard, 2010). This so-called ‘paradox of meritocracy’ (Castilla and Benard, 2010) occurs because managers in these organisations are not forced to examine the subjective nature of their decisions, and how they might be affected by biases.
Unchallenged faith in the objectivity of merit can also act as a barrier to managerial support for policies and programs aimed at supporting women’s entry into and progression within certain occupations and industries. A substantial body of research has demonstrated that belief in the fundamental ‘rightness’ and ‘fairness’ of merit is the primary source of opposition to affirmative action, such as gender targets and quotas (for an excellent discussion, see Son Hing et al., 2011, 2002). Even when managers accept that recruitment and promotion processes are plagued with biases and subjectivities, they nevertheless hold the ideal of merit paramount when presented with affirmative action measures, such as gender targets, as a potential solution (Foley and Williamson, 2019).
However, Australian research has shown that women are significantly less likely than men to perceive that their organisations are governed by merit. In one large survey of Australian employees, only 45 per cent of women said that they believed their organisations were truly meritocratic, compared to 61 per cent of men (Sanders et al. 2017). Surveys have repeatedly shown that only 30 per cent of women believe that qualified men and women have equal opportunity to be recruited or promoted into management and executive positions, compared with 60 per cent of men (Sanders et al., 2017, p. 7). There is clearly a gender gap in how employees perceive meritocracy in Australian organisations.
This paper has summarised key facets of gender bias that serve to undermine equality for women in the recruitment and selection process. This bias can be overtly and covertly reinforced through the gendered ‘coding’ of job advertisements; the gendered review of performance; and in the skewed ways in which applications are tracked and credentials compared in ways which benefit male applicants. Women can also experience significant barriers to career development because gender bias has limited their opportunities to undertake challenging job assignments; and because workplace culture either directly or indirectly discouraged them from assuming leadership styles at variance to men.
- Gender-coded wording in job advertisements may dissuade women from applying for some jobs and may reinforce stereotypes about which genders are better suited to certain roles.
- In many occupations and industries, women must send out substantially more applications to receive the same number of interview invitations as equally qualified men, an effect which is compounded for older women, women with children and women from certain racial or ethnic groups. Male applicants may also be disadvantaged when applying for roles in female-dominated occupations and industries.
- Women and men are held to different standards in the recruitment processes, with women more likely to face tougher evaluation standards, or to have their achievements and qualifications more closely scrutinised.
- Gender biases affect the way applicants are perceived in job interviews, and whether they are deemed to be likeable and competent.
- Women are more likely than men to be penalised for attempting to negotiate for higher pay.
- When selection criteria are not given pre-assigned weights, managers subtly shift the importance of certain selection criteria to favour candidates who fit a stereotypical profile, allowing managers to justify prejudicial decisions under the guise of objectivity.
- Anonymous recruitment appears to be an attractive policy option to address hiring discrimination, but various experiments and trials have produced mixed results. Further research is necessary to establish the conditions under which anonymous recruitment is most effective.
- Reviewers use different language to evaluate the performance of men and women, an effect which is clearly perceived by women workers. Vague, inconsistent or negative feedback can have significant impact on women’s ability to be hired or to progress within their careers.
- Stereotypical beliefs that women are less committed to their work or need to be protected from overwork limit women’s exposure to the career development opportunities necessary to advance to senior leadership roles.
- Women and men in Australia do not have materially different aspirations to senior leadership positions, but women are more likely than men to believe that ‘feminine’ leadership styles are undervalued in workplaces.
- Many organisations that purport to be meritocratic are not, because the notion of merit provides a veneer of objectivity which allows biased decisions to go unchallenged. Furthermore, merit acts as a barrier to acceptance of gender targets and other affirmative action programs designed to circumvent biases in recruitment and promotion.