Big Data Blog Item

The saying that new technology is only as good as the people who are using it can also apply to companies that embrace “big data” analytics for talent acquisition. Big data is typically defined as data that comes from many sources, contains a wide variety of information and arrives at a fast pace.

“Big data can be very useful in employment analyses, but it needs to be emphasized that more data does not always result in a more accurate outcome if it’s not understood and interpreted correctly,” Paul F. White, Ph.D., a specialist in statistical analysis of employment practices, told Bloomberg BNA. “Big data can do more harm than good if the underlying model is not reflective of the decision-making process being analyzed,” White, a partner at Resolution Economics, LLC, added.

In a recent blog item, the labor economist weighed in on the Equal Employment Opportunity Commission’s proposal to require businesses to disclose compensation data. White, in an upcoming blog Q&A, will discuss pay reviews for the "outliners" within an organization.  

Bloomberg BNA: Precisely, what are the pros and cons of using big data analytics as it relates to complying with federal nondiscrimination and affirmative action laws?

White: The advantages of big data are primarily related to the old adage that more information is better than less.  For example, in the context of talent acquisition, it is increasingly common to obtain electronic information on an applicant’s prior work history and education.  It wasn’t long ago that most of that information was handwritten on a paper application, thus making it very labor intensive to build a database from those documents. 

Matching the applicant data to the employer’s HRIS [human resources information systems] data allows the analyst to track the hires throughout their employment and identify applicant characteristics that are correlated with successful employees. This can help reduce turnover costs among employees.

This leads to some of the disadvantages of big data. For instance, a posted position may attract hundreds or thousands of applicants. Rather than processing each and every application, an employer may use facially neutral criteria to exclude some of the applicants to narrow the applicant pool to a manageable size using criteria such as education attainment or prior work experience. 

This often leads to a disproportionate number of applicants from a protected group being excluded. Furthermore, if the applicant pool is large enough, only a small difference in the “pass” rates between protected and non-protected applicants may lead to a statistically significant and adverse finding. 

A naive implementation of big data tools could easily exacerbate disparate impact, and it’s not the tools that will determine the effect on compliance, but the use and misuse of the tools as they are implemented.

Take advantage of an authoritative resource to address the ever-evolving issues in labor and employment law with a free trial to the Labor & Employment Law Resource Center .