“Gendered wording” refers to using words that carry unconscious stereotypical traits. Such words or formulations appeal to stereotypical male or female attributes. Example: The word “compassionate” is female coded as it unconsciously calls on to a stereotypical female attribute. “Leader” on the other hand is male coded. (1)
Gendered wording is subtle, yet frequently used in job advertisements. And it has a wide reaching effect: “Women were deterred from masculinely worded jobs, finding them less appealing” (2). Job ads worded in a way to call on to “female” attributes attract women stronger and do not deter men (3). But Gaucher et al. found a 38% higher usage of masculine wording in male-dominated occupations (4) than in female-dominated occupations. Thereby sustaining gender inequality in these jobs.
Removing masculine gendered wording in job advertisements increases responses from women considerably: 42% more responses were measured when gender-neutral wording was used (5). Job advertisers use gendered words undeliberately and commonly in classic job recruitment processes. Especially senior level positions tend to be male biased (6).
The hidden bias in gendered wording acts as a barrier to the inclusion of women and more generally as a barrier to diversity - especially in Tech, an industry that is today male-dominated. The careful usage of gender-neutral language can help against underrepresentation and inequalities.
Unconscious bias in recruitment and promotion
Different scientific works shed light on the influence of unconscious biases and gender role expectations as barriers to women’s careers. For example, there is a smaller probability that women without management experience will be employed in a leadership position than men without the same management experiences (7).
Underrepresentation of women in traditionally male-dominated fields or management positions can be explained by unconscious biases against women’s capacity for the area (8). The biases can also lead women to consider themselves not fitting for the job. Another study finds that in a mixed team, credits for success are given to the male participants rather than the female participants (9). These findings lead to the question of how to mitigate and counteract the negative consequences of unconscious bias and unconscious gender role expectations.
Countering unconscious bias at the organisational level: Starting with the job ad
The research of Welpe et al. 2014 suggests prevention of unconscious bias mostly on the organisational level. Because on the individual level, training on one’s own unconscious biases does not sustainably change recruitment decisions. Rather processes should be defined in a way to become more independent from individual decisions. Thereby neutralizing individual biases.
In recruitment, the starting point is often a job advertisement. Specific and concrete criteria for the open position need to be defined in advance by the company according to the needs of the job. Then, enhanced attention must be given to the formulation and words used since those have a significant impact on who is reached and attracted by the job ad. In order to equally appeal to both genders a balanced choice of words has to be considered.
Already in 2011, a study identified gendered wording in the recruitment process as a reinforcing factor to underrepresentation of women in traditionally male-dominated occupations (10). Analysing 4,000 job advertisements in both traditionally male-dominated and traditionally female-dominated occupations, they were looking for gendered wording. In job advertisements for male-dominated occupations they found 38% more stereotypically masculine words compared to advertisements for traditionally female-dominated areas (11). For example, they found a more frequent use of adjectives that are typically associated with male stereotypes (such as dominant, competitive) in job advertisements that promote positions traditionally dominated by men. But this effect was only found in one direction: stereotypically feminine words were equally present across male- and female-dominated occupations. The British job platform Totaljobs.com analysed over 75,000 job ads posted on their platform. They found for example 70,539 mentions of “Lead” (a male-coded word) and 83,095 mentions of “Support” (a female-coded word). Especially senior level positions tend to be male biased (12).
Wording impacts behavior and numbers of applications
Do gendered words used in a job advertisement make a job less appealing to individuals? And therefore, does gendered wording contribute to the perpetuation of discrimination and inequality?
Welpe et al. 2014 point out that women react differently on a job advertisement depending on the wording: wording calling on stereotypically female attributes exerced a bigger attraction to women (for example team-oriented working method). The feeling of belongingness can be diminished or magnified by the use of words with a gender connotation. The feeling that one fits in with others within a domain affects motivation and engagement and even trust in and comfort with a company (13). They found that “increases in masculine wording were sufficient to decrease women’s job appeal ratings and their anticipated belongingness in specific occupations.”(14) For men, no effect of gendered wording was detected, meaning that it did not affect men’s anticipated feelings of belongingness (15).
The American employment platform ZipRecruiter found in an analysis that gender-neutral wording leads to 42% more responses to a job listing (16). Their data scientists had over 9 mio postings (17) at their disposal to extract gendered wording indicating both female and male stereotypes. Their results prove that removing gendered wording in job listings increases applications and therefore helps to find the right talent for the position.
These results provide evidence for a perpetuation of gender inequality in male-dominated fields by gendered wording in the recruitment process. Gendered wording had the largest effect on women, affecting them as individuals. Men’s feeling of belongingness to a job was only slightly changed by different words used in the advertisement.
Because the wording difference is subtle, it is particularly difficult to detect and to reduce. Thus it is a potent contributor to preserve inequality. Gendered wording is “common in male-dominated fields, and contributes to the division of traditional gender roles by dissuading women’s interest in jobs that are masculine worded.” (18)
Gender-neutral wording is therefore key to a culture of diversity and inclusion. Especially in Tech - a now male dominated industry - if the industry in general or companies in particular want to attract more female talents, the choice of words in job advertisements needs to be meticulously monitored in order to increase gender diversity.
(1) see Appendix A in Gaucher et al. 2011 "Evidence that gendered wording exists", p. 17
(2) Gaucher et al. 2011, p. 10; Welpe et al. 2014 "Wenn Gleiches unterschiedlich beurteilt wird - Die Wirkung unbewusster Vorurteile."
(3) Gaucher et al. 2011
(4) Results Study 1: Gaucher et al. 2011, p. 5
(5) https://www.ziprecruiter.com/blog/removing-gendered-keywords-gets-you-more-applicants/ , accessed 04.08.19
(6) https://www.totaljobs.com/insidejob/gender-bias-decoder/ , accessed 04.08.19
(7) see Bosak & Sczesny 2011 "Gender bias in leader selection? Evidence from a hiring simulation study."
(8) Welpe et al. 2014
(9) see Heilman & Haynes 2005 "No credit where credit is due: attributional rationalization of women's success in male-female teams."
(10) Gaucher et al. 2011
(11) Results Study 1 + 2: Gaucher et al. 2011, pp. 5-6
(12) https://www.totaljobs.com/insidejob/gender-bias-decoder/ , accessed 04.08.19
(13) see Vaughns et al. 2008 "Intersectional invisibility: The distinctive advantages and disadvantages of multiple subordinate-group identities"
(14) Gaucher et al. 2011, p. 9
(15) Gaucher et al. 2011
(16) https://www.ziprecruiter.com/blog/removing-gendered-keywords-gets-you-more-applicants/ , accessed 04.08.19
(17) ZipRecruiter on Youtube, accessed 04.08.19
(18) Gaucher et al. 2011, p. 12