When is it lawful under anti-discrimination laws for an employer to intentionally discriminate against members of one protected class, to avoid a disparate impact claim by individuals in another protected group? The U.S. Supreme Court addressed that question in Ricci v. DeStefano, 2009 DJDAR 9567 (June 29, 2009). The court’s decision explains the interplay between two branches of anti-discrimination laws: disparate treatment and disparate impact.
Here is an oversimplified summary of the two types of employment discrimination claims. In lawsuits alleging “disparate treatment” discrimination under civil rights laws such as Title VII of the Civil Rights Act of 1964, or California’s Fair Employment and Housing Act, the plaintiff claims an employer based a decision on an illegal criterion (such as race, sex, religion, disability, etc.). The employer avoids liability if its decision was based on legitimate, non-discriminatory reasons.
The law also prohibits decisions based on facially neutral or objective criteria, but which have a “disparate impact” on members of protected groups. The plaintiff must show the challenged practice has a statistically significant, negative impact on a protected group, typically as compared with a similar group of white applicants or employees. The employer prevails if it demonstrates the practice is “job related and consistent with business necessity,” and the plaintiff cannot establish the existence of a less discriminatory alternative.
‘Ricci v. DeStefano’
New Haven, Conn. developed a test for evaluating candidates for promotion in its fire department. The test results revealed minority examinees were unsuccessful at a significantly disproportionate rate compared with white applicants.
Minority firefighters threatened to sue the city, on the ground that reliance on the test was a discriminatory practice because of the statistically skewed results. After holding several hearings, the city’s civil service board refused to certify the test results. That action deprived the successful examinees a place on the promotion list.
The city believed it could develop a new program to evaluate candidates’ suitability for promotion, but without the profound disparity in scores between white and minority test-takers. The successful examinees believed the city intentionally discriminated against them because of their race. So, 18 firefighters (17 white and one Hispanic) sued the city, claiming the decision constituted a violation of the U.S. Constitution’s Equal Protection Clause and several civil rights statutes, including Title VII.
The firefighters and city each filed cross-motions for summary judgment. The city argued that its refusal to certify the test results was based on its good faith belief that doing so would have resulted in unlawful “disparate impact” discrimination. The district court granted New Haven’s motion. The court held the city’s decision to throw out all the test scores was not discriminatory against the plaintiffs based on their race. The court also held it is not unlawful discrimination per se to take race into account to avoid a disparate impact claim. A 2nd Circuit panel decision affirmed the lower court’s judgment without issuing its own opinion. The full court refused re-hearing en banc by a narrow 7-6 margin, over strong dissents by two circuit judges.
The Supreme Court’s Decision
The Supreme Court reversed the lower courts’ judgment. Justice Anthony Kennedy, writing for a 5-4 majority, held unequivocally that the city’s decision to cancel the test because of the racial composition of the successful test-takers was a prima facie violation of Title VII’s anti-discrimination provisions.
The court then turned to whether the city’s justification was a legitimate defense to the disparate treatment claim. The city argued it needed only a “good faith belief” that the test created an unlawful disparate impact. The plaintiffs argued an employer can never take race into account to cure a “disparate impact” or, alternatively, that the employer must actually know the practice is going to result in disparate impact liability.
But the majority rejected both parties’ arguments as too extreme. Borrowing from cases decided under the Equal Protection Clause, the court held that employers are entitled to remedy a practice’s disparate impact by making a race-based decision, but only when the employer has a “strong basis in evidence” that it will be liable for disparate impact.
Writing for the dissent, Justice Ruth Bader Ginsburg would have adopted the “good faith” standard the city advocated. The majority’s objection to the dissent’s view was that employers would too frequently discard legitimate test results or, worse, would affirmatively attempt to engineer a desired outcome under the guise of avoiding disparate impact liability.
What Is a Strong Basis in Evidence?
The lower courts will have to develop the contours of the majority’s “strong basis in evidence” standard. But the majority’s rejection of the city’s justifications provides clues regarding what is an inadequate basis for an employer’s discriminatory decision to avoid disparate impact liability.
A statistical disparity alone is not “a strong basis in evidence.” It was undisputed in Ricci that minority firefighters would have established a prima facie case of disparate impact discrimination. But without proof the test was not job-related, or there were less discriminatory alternatives available, there was no “strong basis in evidence” for throwing out the test. The majority reviewed the record and found no triable issue of fact regarding either of these issues.
The court also rejected the notion that the fear of disparate impact litigation alone justifies a race-based decision. Intentionally treating employees in protected groups more favorably than others is discrimination, whether or not motivated by litigation avoidance. The court also opined that the minority firefighters’ threatened disparate impact claim necessarily was doomed because it was based on statistics alone.
‘Ricci’s’ Ripple Effects
Ricci is a reminder that even well-intentioned employment decisions are unlawful if they are principally based on protected status such as race. But beyond that general proposition, it would be a mistake to assume that Ricci has no bearing on private-sector employers. For example, private-sector employers routinely develop neutral hiring criteria, such as skills-based or physical agility tests. Group layoff decisions, too, create the potential for disparate impact claims.
The Ricci majority discussed in detail New Haven’s efforts to ensure its exam was fair to minority applicants. The majority pointed out that Title VII does not prohibit employers from designing neutral evaluation tools that give all applicants and employees a fair chance to succeed. As the facts in Ricci demonstrate, sometimes even the most carefully crafted practice has a statistically significant adverse effect.
After Ricci, though, employers should be cautious when deciding what to do if unanticipated statistical disparities arise. The court held that once New Haven implemented its assessment protocol, the city was not privileged to alter it retroactively to improve the racial mix of the successful candidates. Thus, the court rejected New Haven’s contention that lowering the weight of the written test and increasing the weight of the oral exam would have constituted a less discriminatory alternative. The majority held that would have been an illegal, intentionally discriminatory decision.
This discussion may give employment lawyers pause. Consider, for example, an employer that decides to select employees for layoff based on performance appraisal scores and other criteria. Before the layoff, the employer evaluates whether its selection criteria could create disparate impact liability. The employer concludes its reliance on performance reviews may result in the layoff of “too many” females, as compared with the overall employee population. So, may the employer lawfully change its layoff criteria to ameliorate the statistical disparity without having a “strong basis in evidence” that the performance appraisal scores would result in disparate impact liability? Only time, and Ricci’s progeny, will tell.