What to Read Next
It turns out that gender bias in hiring and advancement is more pervasive than we thought.
While progress has certainly been made toward workplace gender parity — some companies, for example, are writing more gender-balanced performance reviews — the reality is that women are still underrepresented in private-sector leadership positions. There are likely multiple drivers of this. Outright discrimination — denying women jobs on the basis of their gender rather than their skill sets — is certainly one. But another, harder-to-detect factor can contribute to the leadership gap: the tendency of some organizational decision makers to subtly dissuade women from pursuing leadership roles.
Research Updates From MIT SMR
Get weekly updates on how global companies are managing in a changing world.
Please enter a valid email address
Thank you for signing up
Specifically, our research found that decision makers who are more ideologically conservative tend to present leadership positions less favorably to female candidates than to male candidates compared with more liberally minded decision makers. As a result, women may never even pursue these roles, having been “talked out of it” based on how the opportunities were presented. This can be explained by psychology research that has found that conservatives are generally more protective of the status quo than their liberal counterparts and feel anxious about disrupting it. The problem is that the status quo they’re protecting already includes more men in leadership roles, partly as a result of past bias, which creates a systemic cycle of bias-fueled advancement that’s difficult for women to break.
Proving Subtle Bias
We conducted a series of studies to examine this form of subtle gender bias.
First, we asked participants to imagine themselves as decision makers tasked with staffing a leadership position and to write a job description for the role — aimed at either a male (“David”) or female (“Sarah”) candidate. We gave them eight pieces of information (such as office location and schedule flexibility) to choose from in writing the description: four pieces of information that were positive and four that were negative. The participants also indicated their political ideology on a 7-point scale ranging from 1 (extremely liberal) to 4 (moderate) to 7 (extremely conservative). Our results showed that conservatives included fewer pieces of positive information when describing the position to Sarah than to David; there was no such difference for liberal participants. In a follow-up study, we found that the more-positive job descriptions (written for David) were more attractive to potential candidates and generated higher interest in applying for the position compared with the less-positive descriptions (written for Sarah).
We then brought all of the elements together in a subsequent study: We asked participants to indicate their political ideology and then presented them with either Sarah or David as a prospective leadership candidate. We then asked them to rate how anxious they would feel interacting with that individual, on a scale from 1 (“not at all”) to 5 (“very much”). Afterward, we asked them to write a job description aimed at the candidate. The description’s linguistic attributes (such as the use of impersonal pronouns and positive emotion words) were analyzed using the Linguistic Inquiry and Word Count text analysis program. The data showed that 29.4% of conservatives experienced high levels of anxiety in anticipation of interacting with Sarah, whereas only 13.75% liberals experienced high levels of anxiety. This, along with the text analysis, indicated that their increased anxiety led them to write a less positive job position description for Sarah than for David.
Breaking the Cycle
How can we break the bias-fueled cycle that prevents women from advancing into leadership?
Our final study sought to examine this question. We used the same setup as in our first study (job descriptions for Sarah or David and self-rating of political ideology) but this time introduced a third condition: Participants anticipating interacting with Sarah also were informed that the presence of women in leadership correlates with better organizational performance. The idea was to alleviate conservative participants’ anxiety about disrupting the status quo. Indeed, incorporating this additional information resulted in conservative participants presenting the position to Sarah as favorably as they did to David.
Our results suggest that organizations can take several mutually reinforcing measures to mitigate bias and its effects, both by sharing information and fine-tuning the hiring process. To communicate the value of women in leadership and heighten awareness of gender bias, organizational leaders can provide decision makers with information about the value of having women in leadership roles and the need to rethink the status quo. This could take a form as simple as sharing research articles (like this one you’re reading) or as involved as companywide training. Organizations can also use the increasing number of tools aimed at removing bias from the hiring process. GapJumpers, for example, enables employers to pick candidates based only on their performance on a work-skills test, with no gender information shared. And tools like Textio can help employers avoid gender-biased language in job postings.
Overall, our research suggests that organizational bias remains pervasive in the 21st century and reveals itself in direct and subtle ways throughout recruiting and talent-management processes. Luckily, practical steps can mitigate or even remove bias in hiring. Organizations of all stripes would be well served to understand the nature of the bias we uncovered and work to eradicate it, creating a more objective hiring and advancement system for the benefit of all.