October 17, 2016

The Trojan Horse: Poor Data Management or Disparate Impact?

A September 26th OFCCP News Release indicated that OFCCP is suing Palantir Technologies for systemic discrimination against Asian job applicants in three job titles: Quality Assurance Engineer, Software Engineer, and Quality Assurance Engineer Intern. A couple of interesting issues related to the administrative complaint are:

  • Surprisingly, the referent group was identified as “non-Asian applicants”. As you may recall, in the 2013 OFCCP v. VF Jeanswear ruling, the ALJ noted that “non-Asian” is neither a race nor an ethnic group defined by EO 11246 and the Uniform Guidelines on Employee Selection Procedures. Why or how is OFCCP litigating a matter that has already been decided as a matter of law by an ALJ?
  • Selection rates were very low, based on OFCCP’s numbers reported in the Press Release. They ranged from less than 1% for the Quality Assurance Engineer to just over 2% for the Quality Assurance Engineer Intern. When applicant numbers are large and selection rates low, small absolute differences in total selections across protected class subgroups can produce statistical disparities.
  • This case underscores the importance of good record keeping and use of appropriate disposition codes. A total of 7 QA Engineers were selected from among 730 qualified applicants. Of the 562 Asian applicants, one was hired. It is surprising to see such a high number of qualified applicants. This may be a reflection of inaccurate/incomplete disposition codes and/or a failure to appropriately apply the Internet Applicant Rule (e.g., Asian applicants not meeting minimum qualifications may have been included as part of the applicant pool).
  • Employee referrals were given preference. However, according to the OFCCP, the referral system disproportionately excluded Asians. Although a referral program can serve as a great recruitment and selection source, as with any other selection tool, periodic monitoring and analysis is crucial to identify if there is disparate impact. Such monitoring is especially pertinent if demographic diversity is limited or certain groups are underutilized in particular roles.

Our observations related to this case are not likely new revelations to federal contractors. Rather, they serve to remind clients of recruitment, selection, and data management best practices and to highlight the very real consequences of not employing them. Stay tuned as this case develops!

By Yevonessa Hall, Senior Consultant and Kayo Sady, Senior Consultant at DCI Consulting Group

Categories:

DCI Alerts

Submit your email address to receive free industry news updates


Type the above number:

You have Successfully Subscribed!

Bitnami