People from ethnic minority groups are less likely to be contacted by job recruiters than people from the majority group, according to an analysis of users on a Swiss public employment website.
Dominik Hangartner of ETH Zurich in Switzerland and colleagues studied the actions of more than 43,000 recruiters who conducted 450,000 searches of 17.4 million jobseekers’ profiles between March and December 2017. They tracked every click to see how recruiters interacted with the profiles, which include information on ethnicity, age and nationality inserted by case workers at the Swiss national employment agency, similar to Job Centre Plus in the UK.
How often Swiss nationals born in the country and from the majority ethnic group were contacted by recruiters was used as the baseline for the analysis, with the probability of recruiters clicking a button to contact job applicants based on ethnicity calculated relative to that. The team found that people from immigrant and ethnic minority groups were up to 19 per cent less likely to be contacted.
Recruiters spent only 0.3 seconds less, on average, on profiles of ethnic minority jobseekers, which the researchers say means the result cannot be entirely explained by recruiters consciously discriminating against people based on ethnicity.
But the time recruiters spent on a person’s profile varied depending on the time of day: between 9am and 10am, they spent 10.5 seconds on average per profile, and 12 per cent less time on those from jobseekers from minority ethnic backgrounds. Between 5pm and 6pm, they spent 9.5 seconds on the average profile, and 14.7 per cent less on ethnic minority accounts. Similar variations are found just before lunch breaks. The team found no significant difference based on the gender of applicants for the average job.
“Around 20 per cent of the anti-[ethnic minority] discrimination we see is driven by the time of day, when arguably recruiters are more exhausted and tired,” says Hangartner. The discrimination is calculated by monitoring the amount of time spent on individual profiles, and the likelihood of being contacted by recruiters. He thinks the rest of the discrimination may be unconscious bias that particularly rears its head when users are tired.
“This is the kind of data analysis that shows us racial discrimination is still a deeply entrenched practice,” says Safiya Umoja Noble of the University of California, Los Angeles. “What we need is rigorous monitoring of systems to ensure such systems do not make discrimination even more opaque.”
Hangartner hopes the data can be used to redesign such websites to mitigate the impact of implicit bias against people from ethnic minority groups.
Journal reference: Nature, DOI: 10.1038/s41586-020-03136-0
More on these topics:
Source: Humans - newscientist.com
