Researchers from ETH Zurich found that gender, origin and ethnicity sometimes unconsciously play a role in a manager’s decision to hire a candidate, even when the manager uses a hiring platform. While education, professional skills and qualifications for the job are the important things, these unconscious biases are often why otherwise competent candidates are passed. This violates equal opportunities and imposes long-term disadvantages on those affected.
The team says that it is crucial to understand who is discriminated against and why in order to fight such practices. To do this, they enlisted the help of an artificial intelligence (AI) algorithm. The new method allowed the researchers to study discrimination across a variety of professions and points in time and analyze the entire search process.
Source: Unsplash
During the study, the team accessed Job-Room, a large recruitment program in Switzerland, which has the profiles of over 150,000 job seekers. While using the platform, recruiters specify the criteria for the job and receive a list of applicable candidates and review their profiles which contain a person’s expertise, gender, nationality and language skills. Once the recruiter gathers their candidates, they can contact and invite them for an interview.
The team analyzed recruiters' actions on Job-Room over ten months, focusing on which candidates were contacted and how the recruiters made their choice. This novel approach enabled the researchers to determine how the origin or gender of a candidate influences the likelihood of conduct.
Researchers found that, on average, immigrant jobseekers were 6.5% less likely to be contacted than Swiss jobseekers with identical characteristics. Discrimination was more pronounced for migrants from the Balkans, Africa, Middle East and Asia. Foreign origin had a stronger negative impact when recruiters reviewed CVs at noon or in the evening. Recruiters also discriminate based on what gender they unconsciously believe should be in the role. In five occupations with high proportions of women, 13% of women were more likely to be contacted than men.
Online platforms are becoming an important tool for recruitment and it is important that recruiters remember that they can still be biased even when using an online tool. There is no evidence of more discrimination when using online platforms vs traditional recruitment. Discrimination is a structural and societal problem, not a technology problem.
This study was published in Nature.
