Dr Kiki Leutner
Diversity starts with recruitment. Organisations are notoriously bad at it, employing more males than females. And it gets progressively worse with seniority. Women make up 47% of entry level roles, but only 20% of c-suite leaders. At every promotion step, women, and especially women of color, lose out to men. This is because women’s potential is underestimated. In a hiring scenario where there is just one woman in the hiring pool, her chance of being hired is zero. Compared to this, a man’s chance of being hired when he is the only man is 33%. When humans are making hiring decisions, they are affected by bias.
Can recruitment technologies help? Recruitment technology has time and cost saving benefits, but a potentially more impactful benefit is its ability to minimize the impact of bias in recruitment and progression decision. Unstructured interviews are commonly used in selection, yet interviewers are bad at picking up on job relevant skills and attributes during interviews. Equally, structured interviews and psychometric tests are considered the gold standard for recruitment practices, but they are far less common than they should be in application due to their high financial and time costs. Recruitment technologies make these best practices more cost effective and accessible. What can be done to ensure recruitment technology is used ethically and improves representation?
Is it benefitting job seekers, and benefitting all job seekers equally?
Any adopted technology should have clear benefits for job seekers. This can include a chance to get evaluated fairly, or the opportunity to learn about themselves through meaningful feedback. It can also be tangible benefits like decreasing the workload of job applications.
Can job seekers provide informed consent?
Any transaction with candidates need to be transparent. Candidates need to be informed about how their data is being used, what data is being recorded and who will have access to the data. Informed consent is only meaningful if candidates have a true choice. If women can choose only between recruitment practices that are all biased against them, there is no meaningful choice. Recruitment technologies increase the presence of good choices in the recruitment market: Standardised recruitment processes such as structured interviews used in video interviewing analytics compare favourably in terms of fairness to the common unstructured interview.
Is data being protected and treated confidentially?
Recruitment technology must protect and preserve the confidentiality, anonymity, and data protection of candidates. The access to recruitment data should be limited to required decision makers. However, with recruitment, there is also a case for retaining candidate data, in particular to evaluate diversity performance of recruitment tools. For example, to monitor whether a selection process is bias free, companies need aggregate level recruitment data. Making this data available is in the interest of job seekers.
Are candidates receiving feedback and an opportunity to increase self-awareness?
It is ethical to return as much information as possible to candidates, whether they are offered a job or not. Organizations should engage with rejected candidates, explaining how they evaluated them, and why their profile was deemed a poor fit with the role. Science-based assessments (with or without AI) allow companies to address candidate concerns in a meaningful way, because they can explain what candidates were evaluated on and how.
Is the technology used explainable?
Explainability is an important features of many AI ethics frameworks. It needs to be applied to any recruitment technology used. Providers should be able to not only show that their technology predicts job performance or achieves desired outcomes, but also explain why it does so. For example, organisations should favour tools that boost their understanding of what someone is like by measuring job relevant skills and competencies, rather than simply predicting that they are the right candidate.
In summary, technology can help organisations implement recruitment best practices in a cost and time effective way. In this sense, they are an important tool for increasing diversity in orgranisations. To ensure that the technology in question will indeed achieve this goal, determine that it is beneficial to all job candidates, allows meaningful consent, protects data and privacy where needed, gives back to candidates through feedback and learning, and is transparent and explainable.
