Virtually all employers in our country use algorithms to recruit and select job applicants. However, the fact that the use of algorithms can lead to discrimination and exclusion is less well known. In addition, employers hardly check their systems for fairness.
This is the conclusion of the Human Rights Board based on research by research firm SEO Economic Research (1).
Recruiting through digital job platforms like LinkedIn is the norm in the Netherlands. As many as 96 percent of all employers say they recruit candidates for vacancies this way. Nearly three-quarters of employers (71 percent) say they actively search for candidates themselves. Nearly one in eight employers (12 percent) say they use algorithms to select and assess candidates.
By posting job openings on platforms such as LinkedIn, algorithms are searching for suitable candidates for a position. This is something that employers and employees mostly do not realize, according to Jacobine Geel, president of the College. "After all, LinkedIn's algorithms and resume databases determine who sees your job posting and who shows up on your search query."
Algorithms play a major role in the development of recruitment technologies, according to Geel. "These are developed to assess applicants faster and on a larger scale. Think of personality or competency tests that calculate automatic test scores."
In itself, these are all good technological developments to speed up the application process. But it also carries risks: think discrimination and exclusion. Because little research has been done on this topic, the Human Rights Board approached SEO Economic Research. The research firm was commissioned to investigate whether and how employers use algorithms to recruit and select people. And whether they are aware of the dangers thereof. For the study, questionnaires were submitted to 896 employers in the Netherlands.
In their search for new personnel, almost all employers, consciously or unconsciously, make use of algorithms. Platforms like LinkedIn we just mentioned as an example. In turn, employers who decide to outsource the application process to external recruiters often deploy algorithms as well. One in six employers approaches an external party for the selection process.
Companies are increasingly using digital tools to select and assess job applicants. These include applying via social media (46 percent), a video application (14 percent) or algorithms (12 percent). Online assessments, personality tests and competency tests are also increasingly popular. Most employers still rely on the classic way of applying for a job: the curriculum vitae and a cover letter.
All stages of the recruitment and selection process contain risks of exclusion and discrimination. Another risk is that people with disabilities, such as the visually and hearing impaired, may not be able to take the digital tests offered even though they may be suitable for a position.
Online advertising is not the solution either. It can prevent suitable candidates from getting the job posting to be. That risk is greater when job boards offer the ability to target ads to demographic factors such as gender and age.
Employers who use algorithms to find candidates are also often unaware of the risks associated with this method of recruitment. In half of the cases, employers do not know whether candidates drop out because they are not passed through automated systems or algorithms, or whether a recruiter has rejected the person in question. "They hardly test recruitment technologies for fairness," says the Human Rights Board.
Geel says, "We advise employers to keep a close eye on whether any groups are left out and, if so, whether that is justified. It is important to distribute vacancies through multiple channels. Employers should also tell applicants in advance what role algorithms play in their recruitment and selection."
The latter is still barely happening. Furthermore, employers barely test whether the algorithms they use are fair and do not create inequality of opportunity. "Does everyone have a fair chance of getting a job, and are certain groups not excluded in advance? Specifically, do women also qualify for the position of forklift truck driver, or are they excluded from the outset because the algorithm has taught itself that forklift truck drivers are usually male and therefore men are more suitable for the job?" reads the research report.
The Human Rights Board believes it is the government's duty and responsibility to ensure that everyone has a fair chance at a job, and that employment discrimination must be countered. In addition, there should be means to test algorithms for fairness. Finally, employers should be able to explain to job seekers how algorithms select and evaluate candidates.
What many job applicants do not know is that they may demand that they not be subject to automated decision-making. Article 22 of the General Data Protection Regulation (AVG) prohibits that. If so, you are allowed to force a potential employer to look at your application again.
https://www.mensenrechten.nl/actueel/nieuws/2022/08/31/vrijwel-alle-werkgevers-gebruiken-algoritmes-voor-sollicitaties-bewustzijn-risico-op-discriminatie-en-uitsluiting-laag
