Algorithmic tools are being utilized in hiring processes to enhance fairness and diversity by implementing constraints like gender-balanced candidate shortlists.
Enforcing equal representation in shortlists may not always result in increased diversity in final hires, especially if there is a high correlation between the algorithm's screening criteria and the human hiring manager's evaluation criteria.
A study involving nearly 800,000 job applications in technology firms revealed that promoting equal shortlists has limited effects on hire diversity when the algorithm closely reflects the hiring manager's preferences.
Proposing a new algorithmic approach that aims to diversify shortlists by selecting candidates who might be overlooked by managers but are still competitive based on evaluation criteria, leading to enhanced gender diversity in final hires without compromising hire quality significantly.