This week the U.S. Equal Employment Opportunity Commission (EEOC) convened a number of experts to discuss how employers have begun and will increasingly continue to use “Big Data” to make employment decisions. These uses of Big Data include using algorithms, “data scraping” of the internet, and other means to evaluate tens of thousands of data points. Employers use these techniques to determine who to hire, who to promote, how to determine whether an employee is performing well, and make other employment decisions.
“Big Data has the potential to drive innovations that reduce bias in employment decisions and help employers make better decisions in hiring, performance evaluations, and promotions,” said EEOC Chair Jenny R. Yang. “At the same time, it is critical that these tools are designed to promote fairness and opportunity, so that reliance on these expanding sources of data does not create new barriers to opportunity.”
The experts saw the potential for the use of Big Data to make fairer employment decisions, reduce the role of implicit bias, and promote equality. At the same time, if not used properly, the use of Big Data could continue, and possibly worsen, inequities in employment decisions. As one expert said, “algorithms may be trained to predict outcomes which are themselves the result of previous discrimination. The high-performing group may be non-diverse and hence the characteristics of that group may more reflect their demographics than the skills or abilities needed to perform the job. The algorithm is matching people characteristics, rather than job requirements.”
Employers can avoid the misuse of Big Data if they retain experts who can help them collect and use the data properly. If employers do not use Big Data properly and their use of Big Data results in disparate impact against protected groups like disables people and women, they run the risk of liability. If an employer has a non-diverse workforce, the risk of misusing Big Data in ways that have disparate impact on protected groups is high. For example, if a construction company employs mostly men, it may find through studying data on its workers that high percentages of its best performers can lift 100 lbs. over their heads. But that doesn’t mean a worker needs to be able to lift 100 lbs. over his head to be a superior performer. It could be that workers can do just as good a job if they can lift 50 lbs. over their heads. Requiring a 100 lb. lifting requirement would likely have a disparate impact against women and might be illegal if the job does not actually require workers to lift 100 lbs. over their heads. Experts in using data to make employment decisions, such as industrial occupational psychologists, can help ensure that employers use data properly and avoid these types of mistakes.