The Museum of Questionable Medical Devices displays dozens of “amazingly useless gadgets,” including such oddities as the Battle Creek Vibratory Chair, which shook patients violently to increase oxygen flow, and the Cosmos Bag, a radioactive cloth sack designed to cure arthritic joints with radium-laced water. Modern medicine would easily identify these devices as absurd and unworkable. A clever pitch man claiming professional credentials can make something up and sell it to a gullible audience, willing to try anything to get the desired result.   

Trying anything to get the desired result, however, is not a good strategy for engineering social change through litigation, as the Equal Employment Opportunity Commission was recently reminded (again) by a federal Court of Appeals.   

ADVERTISEMENT

In EEOC v Freeman, the commission once again relied on the statistical analysis of an industrial psychologist to try to prove that an event planning company discriminated against black male job applicants when it ran credit and criminal background checks.  On Feb. 20, the Fourth Circuit Court of Appeals held that the statistical case presented by the EEOC was “rife with analytical errors,” “completely unreliable,” and contained a “mind-boggling number of errors and unexplained discrepancies.”  The Fourth Circuit judges tossed the EEOC’s case, just as the Sixth Circuit did in 2014, when the EEOC relied on the same expert whose report was rejected for containing the same types of fatal flaws. 

For years, the EEOC has doggedly pursued litigation against companies whose background check procedures it deemed too expansive and discriminatory. The EEOC’s strategy relies on the observation that, under certain circumstances, the use of criminal background and credit checks as a pre-hiring screen tends to eliminate minority applicants in disproportionate numbers.  The EEOC claims that this violates Title VII. 

But wait a minute.  Just because your arthritis got better doesn’t mean it was because of the Cosmos Bag.  Correlation and causation are not the same thing.   

Statistical data show that a higher proportion of minorities have criminal records than non-minorities.  The NAACP reports that African Americans and Hispanics makeup 58 percent of prisoners, despite making up only a quarter of the general population.  There are undoubtedly a variety of factors that influence those numbers, but employers who perform routine background checks on prospective hires have no control over whom police arrest, governments prosecute, or juries convict.  If there is injustice in the justice system, employers performing background checks are not the ones to blame. 

The reason that criminal background checks reveal more convictions for minorities is because more minorities have been convicted, not because employers who run pre-hire background checks are engaging in unlawful discrimination.  

The EEOC’s insurmountable hurdle, therefore, is proving race discrimination, which the agency must do through statistics.  If a criminal background check were racist, it would have to screen out minorities in a higher proportion than the rate at which minorities have been convicted of crimes.   

To date, the EEOC’s efforts to prove disparate impact through statistics have had roughly the same level of success as the Battle Creek Vibratory Chair.  

Last year, the Sixth Circuit called the EEOC’s statistical case against an employer “a homemade methodology, crafted by a witness with no particular expertise to craft it, administered by persons with no particular expertise to administer it, tested by no one, and accepted only by the witness himself.”  The commission then used the same expert again – with the same result. 

If the EEOC’s theory were statistically supportable, it would not be so hard for the EEOC to find statistical support.  Yet, as Judge Roger Titus wrote when dismissing the EEOC’s case at the trial court level, the commission’s race discrimination argument is a “theory in search of facts to support it.”  Court after court has panned the EEOC’s approach. 

The goal of promoting employment for qualified, rehabilitated individuals with criminal records is a worthwhile one.  The EEOC, however, has no statutory mandate allowing it to pursue that goal.  Strong-arming companies by filing unprovable disparate impact lawsuits is not the best model for engineering social change. 

Fortunately, there is a better way.  States like Georgia have enacted laws that allow reformed ex-offenders to obtain a certificate of rehabilitation.  Employers who hire these ex-offenders receive immunity against negligent hire claims. 

The Georgia program addresses head-on the worthwhile goal of promoting employment for rehabilitated ex-offenders and does so in a way that provides critical legal safeguards for employers. This carrot-based approach provides ex-offenders a path for re-entry into mainstream society, while giving employers the legal protection they need so they can reasonably take a chance on giving someone a second chance.

That is a much more common sense approach than the EEOC’s discredited statistical gadgetry. 

In 2002, the Museum of Questionable Medical Devices closed, and its collection was donated to the Science Museum of Minnesota, where curiosity seekers can still hook themselves up to a phrenology machine that measures intelligence and morality based on the size of bumps on the head. 

It remains to be seen whether the Fourth Circuit’s recent decision will prompt the EEOC to change its much-maligned approach to background checks.  If the EEOC is finally ready to retire its repeatedly discredited expert statistical papers, the museum is accepting donations

 

Lebowitz is a Cleveland-based employment law partner with BakerHostetler.  He can reached at tlebowitz@bakerlaw.com.