Algorithms discriminate because people do, job agency says


Algorithms used in staff recruitment confirm existing biases, according to a Radboud University study commissioned by National jobs website Vacaturebank.

The employment agency matched 9,000 anonymous resumes to 12 million job descriptions using artificial intelligence. It revealed that women and men were matched with jobs in traditionally female and male sectors like childcare and construction, and that women were linked with jobs that paid far less money.

Part of the reason the algorithms are biased is because the system is based on data provided by people, said Gido Schoenmacker, head of data science at the agency. RTL News. “Algorithms amplify biases. What we find is that women are less likely to get certain jobs because it’s always the men who are selected for them.

The employment agency wants to use self-learning systems more often to match people to jobs, but sometimes they learn the wrong things, Schoenmacker said. ‘Our own research And the one of Radboud University will help prevent this.

An example of algorithms learning from people’s biases can be seen at Amazonwhere the algorithms were found to have a preference for white males.

Language usage is something the algorithms rely on, Schoenmacker said. Words such as “success-oriented” and “winning mentality” are found more often on men’s resumes and in job descriptions for higher-paying jobs.

If the algorithm is configured in such a way that it no longer finds clues to distinguish men from women, gender-based biases no longer play a role. One of the models applied in the research reduced the pay gap from €1,680 to €180.

The system can learn anything you want, Schoenmacker said, but whether it’s a good tool for matching people and jobs remains to be seen. The employment agency will start a trial using the system next year.

Thank you for your donation to

The team would like to thank all the generous readers who have donated over the past few weeks. Your financial support has helped us extend our coverage of the coronavirus crisis into evenings and weekends and ensure you are kept up to date with the latest developments. has been free for 14 years, but without the financial support of our readers, we would not be able to provide you with fair and accurate information on all things Dutch. Your contributions make this possible.

If you haven’t donated yet, but would like to, you can do it via Ideal, credit card or Paypal.

Sharon D. Cole