In 2012, the U.S. Equal Employment Opportunity Commission released an enforcement guidance document cautioning employers on the use of criminal background checks in hiring. The EEOC advised employers that blanket use of convictions to deny employment could have a disparate impact against African American applicants, who statistically are convicted at a higher rate than members of other races. The agency recommended that employers look at each conviction to determine how it raises specific concerns over the applicant’s suitability for the job.

On Tuesday, the full Second Circuit Court of Appeals denied an en banc petition to review the dismissal of a class action that involved a disparate impact claim. In Mandala v. NTT Data, Inc., the employer allegedly disqualified applicants convicted of a felony automatically. An earlier panel of the Second Circuit concluded that the claim relied on nationwide criminal conviction statistics, not the conviction rates among the pool of applicants qualified for the particular jobs offered by the employer.

The full Second Circuit affirmed this decision. The majority opinion noted that statistical disparities in the general population cannot be used to demonstrate such discrimination in a particular industry or applicant pool. The plaintiffs must instead offer proof that the employer’s policy had a true disparate impact on the specific applicants who constitute the purported class. A number of Second Circuit judges dissented, noting the potential difficulties plaintiffs could face at the pleadings stage in trying to obtain such precise statistical information. Regardless, if this decision is followed by other federal appellate courts, it would make disparate impact claims more difficult to pursue based on the need to obtain more refined data to support discrimination claims.