Reflections on Ricci
||Eric M. Dunleavy
We previewed the Supreme Court ruling on Ricci v. DeStefano (2009) in the April 2009 issue of TIP and discussed the actual ruling in the October 2009 issue. Briefly, the New Haven Civil Service Board (CSB) refused to certify firefighter promotion exams for lieutenant and captain because of its “good faith belief” that it would lose an adverse impact challenge to minority applicants. The Supreme Court, in a narrow 5–4 decision, ruled that the CSB needed more, namely a “strong basis in evidence” to act on that fear. We will not belabor the ruling here; we said plenty in the October 2009 column. Rather, our purpose below is to update you on some Ricci-related cases and to reflect on what the Ricci ruling has meant—so far.
Let’s start by looking back at Hayden v. Nassau County (1999), a controversial pre-Ricci case. In our April 2009 preview, we saw Hayden as a precursor to Ricci. After all, there were several commonalities. Both cases were decided by the 2nd Circuit Court, both featured public safety employees, and both featured (for lack of a better term) postadministration test manipulations based on racial motives. We felt strongly that no matter how Ricci was decided, it would feature and reflect on Hayden. Surprisingly (at least to us), Hayden was never mentioned in the Ricci ruling. Upon reflection, we think there is an obvious reason why.
First the facts. In Hayden, a technical design advisory committee (TDAC) created a 25-component hiring exam for police officers and administered it to 25,000 candidates. This exam, in its entirety, produced “severe” adverse impact. To reduce the adverse impact, the TDAC eliminated 16 of the 25 components, prompting the claim by 68 unsuccessful candidates they would have passed if all 25 components were used. The 2nd Circuit acknowledged there was a racial motive for “redesigning” the test but, nevertheless, supported the final product on grounds it was scored in a race-neutral fashion. Despite numerous calls for the Supreme Court to review this case, that never happened. We wondered why. We think reasoning behind the Ricci ruling provides the answer, albeit, after the fact.
The Hayden case began in 1977 after the DOJ sued Nassau County for— what else—adverse impact against minorities. The two parties entered into a consent decree in 1982 to construct an exam that either produced no adverse impact or was valid “in accordance with Title VII and the Uniform Guidelines.” However, exams developed in 1983 and 1987 also resulted in adverse impact (and two new consent decrees). So it was in that context that in 1990 the DOJ and Nassau jointly appointed the TDAC to do its thing. In hindsight, we can’t imagine a “stronger basis in evidence” for the actions of Nassau County than an obvious and substantiated fear of losing an adverse impact challenge to minority applicants over nearly a 20-year period. Stated differently, if our reasoning is correct, Ricci actually supports the Nassau County ruling. However, this is hardly a license for municipalities to follow the TDAC plan absent the type of pressure faced by Nassau County (from the DOJ and by a court-sanctioned consent decree).
Next, let’s fast forward to Briscoe v. New Haven (2010 & 2011). Michael Briscoe was an unsuccessful Black applicant for lieutenant in the Ricci case. He was the top scorer among 77 applicants on the oral exam, which accounted for 40% of the total score. However, in the end, he was 24th overall because of his poor performance on the written test, which accounted for 60% of the total score. Briscoe wanted to test for director of training, a position open only to lieutenants and captains. He sued to be included in the testing on grounds he would win an adverse impact claim against the city and, therefore, would be a lieutenant. He claimed that if the weighting was reversed to 70% oral and 30% written, it “would be equally good or better at identifying the best-qualified candidates for promotion and would have less disparate impact on racial minorities.”
At the district court level (Briscoe v. New Haven, 2010), Judge Charles S. Haight, Jr. noted that after the Supreme Court’s ruling the original exam scores from Ricci were certified by the district court judge in the Ricci case (Janet Bond Arteton). The city argued it now had a strong basis in evidence for believing it would lose a disparate treatment claim if it did not follow Judge Arteton’s order. Judge Haight agreed. However, as important, he did not discard the weighting issue. Rather, he emphasized that the proper times for Briscoe to challenge the weighting were in 2003, when the exams were administered, and in 2004, when the Ricci plaintiffs sued. Indeed, he stressed the narrowness of his ruling as follows:
It is important to emphasize the narrow boundaries of this opinion. I am concerned only with the effect of Ricci upon Briscoe’s disparate-impact challenge to the 2003 examinations. That limited reach is dictated by the fact that the complaints of the Ricci plaintiffs and Briscoe relate solely to the 2003 examinations. Nothing in this opinion would foreclose or diminish the rights of Briscoe or any other firefighter to challenge a subsequent NHFD promotional examination on the same grounds that Briscoe seeks to assert in this case with respect to the 2003 examinations.
Interestingly, in the appeal to the 2nd Circuit (Briscoe v. New Haven, 2011), New Haven abandoned its “preclusion” theory and argued instead that the “strong basis in evidence” test for disparate treatment applies equally to adverse impact claims. The city’s basis for this argument is from that part of the Supreme Court ruling in Ricci that states:
If, after it certifies the test results, the City faces a disparate-impact suit, then in light of our holding today it should be clear that the City would avoid disparate-impact liability based on the strong basis in evidence that, had it not certified the results, it would have been subject to disparate-treatment liability.
Although seemingly a clear statement, a three-judge panel of the 2nd Circuit questioned the generality of this “one sentence dicta.” In an opinion written by Circuit Court Judge Dennis J. Jacobs, the court ruled:
The city characterizes this one sentence of dicta as establishing a symmetrical companion to Ricci’s earlier holding that an employer may avoid disparate-treatment claims based on a “strong basis in evidence” of disparate-impact liability. That is, the city argues that an employer may defeat a disparate-impact claim if it had a strong basis in evidence that it would have been subject to disparate-treatment liability. The city argues that Briscoe’s suit was properly dismissed not because it was precluded but because the Supreme Court’s Ricci mandate itself supplied the strong basis in evidence of disparate-treatment liability (for not certifying the results). [emphasis by authors]
Judge Jacobs conceded that the Supreme Court anticipated Briscoe’s lawsuit. Nevertheless, he ruled:
[W]e would have to conclude that the Supreme Court intended to effect a substantial change in Title VII disparate-impact litigation in a single sentence of dicta targeted only at the parties in this action….Ricci did not substantially change Title VII disparate-impact litigation or preclusion principles in the single sentence of dicta targeted at the parties in this action. [emphasis by authors]
In other words, Judge Jacobs saw nothing in the Ricci ruling that alters the preexisting case law on adverse impact—a factor we alluded to in our October 2009 column. More importantly (at least for Michael Briscoe), the case now returns to the district court for trial on it merits relating to alternatives with less adverse impact.
As if that wasn’t enough, on October 7, 2011, seven Black plaintiffs sued the City of New Haven and the International Association of Firefighters Local 825 for racial discrimination in relation to the promotion exams targeted in the Ricci case (Tinney v. New Haven; see http://firelawblog.com/files/2011/11/
Tinney-Complaint.pdf). Although seemingly a technicality, Briscoe’s suit focused on alternative selection procedures that produce less or no adverse impact, an important consideration after job relatedness is proven. Tinney asserts there is no basis in evidence for supporting the 60–40 split favoring written over oral exams, meaning the tests were not job related to begin with. A key concern here is that while trying to undo their original exams the New Haven CSB never received a final validity report from the test maker. Therefore, if this lawsuit goes forward, the validity of this 60–40 split will be a central issue in determining if the testing process, in its entirety, is job related and consistent with business necessity, let alone the best procedure for the promotion process.
But wait—there is more. The Tinney plaintiffs are also suing for disparate treatment on grounds that the 60–40 split is the product of 20 years of collective bargaining by a union whose members are primarily White. In other words, the claim is that the union has a racial motive for maintaining an arbitrary split based on the knowledge that it would limit the number of minorities promoted. This is a novel complaint because proof of an illegal motive is unnecessary in an adverse impact claim.
One final point before we close. In a FAQ (frequently asked questions) news release (see http://www.dol.gov/ofccp/regs/compliance/faqs/Ricci_FAQ.htm), the OFCCP answered several anticipated questions relating to the Ricci ruling. The agency made it clear that Ricci did nothing to alter how it will conduct compliance reviews of contractors. More specifically, in response to the FAQ, “Does the Supreme Court’s decision in the Ricci case change how OFCCP will conduct compliance evaluations of contractors’ employment practices?” the OFCCP responded as follows:
No. The Ricci decision does not affect how OFCCP examines the use and impact of selection procedures, such as tests. OFCCP will therefore continue to assess whether a contractor’s use of its particular selection procedures complies with the Uniform Guidelines on Employee Selection Procedures (UGESP)
And to the FAQ “Does the Ricci decision change contractors’ affirmative action obligations or their obligations regarding the use and validation of tests?” the OFCCP responded as follows:
No. Ricci does not change a contractor’s affirmative action obligations under the mandates enforced by OFCCP. Likewise, a contractor’s obligation to comply with UGESP when using a test as part of its selection process remains the same. If a test has a disparate impact on a particular race, ethnic group, or gender, the test must be validated as to the particular job for which it is being used. The contractor must also investigate alternative selection procedures and must use an alternative procedure if it would result in less adverse impact and would be valid for the job in question.
Similar answers were given to related FAQs. The bottom line is that the OFCCP sees no changes in adverse impact precedents that predated the Ricci ruling. From what we have seen, it is reasonable to assume that EEOC and DOJ are on the same page. There are several other cases in which Ricci is cited. However, none of these cases suggest that Ricci has affected disparate impact standards in any way.
As the above case law review demonstrates, selection procedure users and developers haven’t learned much from Ricci or related case law on legal defensibility best practices in selection. In fact, the issue of reasonable alternatives is as ambiguous now as it has ever been. The issue of component weighting is something that selection procedure users and developers should always consider along with decisions between compensatory and multiple hurdle approaches. We are hopeful that the Briscoe ruling will provide some ground rules for the standards used to compare different weighting schemes.
Consider also a brief written by five SIOP Fellows (Aguinis, Cascio, Goldstein, Outtz & Zedeck)1 on behalf of the defendant, which questioned whether other methods like assessment centers are superior to more traditional written tests and/or interviews. From a case law perspective, it is unclear whether the additional time, effort, and expense to develop such measures will be ruled as “reasonable” relative to more traditional methods. However, as the brief demonstrated, research supports that assessment centers are often strongly job-related and will produce less adverse impact than written tests.
It is not a new idea that selection procedure users and developers should have evidence to support the choice of particular assessments and how these assessments are used. However, it may not be as intuitive that these decisions can be framed in the reasonable alternatives context. We suggest that vendor and assessment searches be documented and included as part of an introduction to a validation study to support why particular constructs were measured and tools were used. It may often be persuasive to include a separate section in a validation report that covers implementation considerations, and supports decisions related to weighting schemes, compensatory versus multiple hurdle strategies, and cut scores. Such documentation would satisfy both UGESP and CRA-1991 requirements, and may head off preliminary challenges on this issue. Until case law provides clearer legal standards related to reasonable alternatives, the strategies suggested above may be the best proactive measures available to the reasonable alternatives challenge.
Briscoe v. City of New Haven (2010) 2010 U.S. Dist. LEXIS 69018.
Briscoe v. City of New Haven (2011) 654 F.3d 200.
Hayden v. Nassau County (CA2 1999) 180 F.3d 42.
Ricci v. DeStefano (2009) 129 S. Ct. 2658.