Tuesday, December 20, 2016
Algorithmic Hiring: Cause of or Solution to Human Discrimination?
In June, I posted on Pauline Kim's forthcoming article arguing that employer use of algorithms in evaluating applicants and employees could hard-wire discrimination into the evaluation process. For a different take on the issue, I've teamed up with 3L David Savage in an article arguing that if algorithms are carefully used and periodically evaluated, they can avoid causing both disparate treatment and disparate impact discrimination and address the discriminatory concerns created by unconscious bias. The article is David D. Savage & Richard A. Bales, Video Games in Job Interviews: Using Algorithms to Minimize Discrimination and Unconscious Bias, 32 ABA J. Labor & Employment Law (forthcoming 2017); here's the abstract:
As the number of applicants for many job openings grows into the thousands, employers have searched for methods to efficiently sort through these applications and compile a shorter list of individuals to interview for open positions. One method growing in popularity is using algorithms to analyze statistical information and determine the candidates that will perform the best if hired based on factors such as cognitive ability, management skills, and workplace performance. Predictive analytics involving algorithms are being used by 8% of companies in the United States. Some of these employers have had applicants play video games created by developers that use these algorithms to analyze their performance and select the best candidates for the job. Scholars have argued that the use of algorithms in general and in video games may lead to discrimination in the workplace. Although any type of employment practice can cause discrimination, this article argues that the use of algorithms in video games to evaluate job candidates may be a cost-effective and beneficial business method that can help avoid discrimination. If created and administered carefully, video games using algorithms have the ability to minimize human bias, including unconscious bias, from the initial job hiring process.