Less Discriminatory Alternative – Procedural
(June 20, 2012; rev. June 24, 2012)
This is sub-page addresses procedural issues concerning implications of the existence of a less discriminatory alternative to practice that has a discriminatory impact. Section A, which is a longer version (about 2340 words) of what would be published (at about 1880 words) as "Slip-Up in the Civil Rights Bill," Legal Times (Dec. 11, 1991),[i] deals with the fact that Congress, in responding to a Supreme Court decision, codified an aspect of that decision that made no sense whatever. Section B, which is only a sketch, discusses how a ruling/statutory provision might be dealt with by the courts and by entities wishing to comply with the law.
This page is related to the Less Discriminatory Alternative – Substantive sub-page of the Disparate Impact page, which addresses issues concerning determining whether a practice is less discriminatory particularly in the circumstances where the impact is measured in terms of relative differences in adverse outcomes and the alternative practices that would commonly be deemed less discriminatory tend to increase relative differences in adverse outcomes.
A. The Codifying of a Nonsensical Ruling
When the Supreme Court issued its decision in Wards Cove Packing Co. v. Atonio, 490 U.S. 642 (1989), the majority was much criticized for the cavalier manner in which it discarded eighteen years of precedent regarding the business necessity justification for employment practices having a disparate impact on minorities and women. The Court also deserved serious criticism, however, for including in its analysis one of the more curious passages to appear in an employment discrimination decision. The potential for confusion as a result of that passage would not be obviated by the Civil Rights Act of 1991. The act, indeed, would codify the confusion.
In Wards Cove the Court both relaxed the "business necessity" standard for justifying a practice with a disparate impact and placed on the plaintiff the burden of proof as to whether that relaxed standard had been met. Consistent with prior analysis, the Court went on to explain that even though the challenged practice has been justified, plaintiff would have the opportunity to show that there exists an alternative procedure that equally serves the employer's interest with a less discriminatory impact – an approach of increased importance in light of the Court's other holdings in the case. The Court went on, however, to indicate that liability would follow only if the defendant were then to refuse to adopt the alternative procedure, seeming to create the extraordinary state of affairs where liability turns on events occurring after trial.
To understand how the Court came to cause this situation, as well appreciate why Congress ought not to have followed suit, it is necessary to look back fourteen years prior to Wards Cove, to the. Supreme Court's decision in Albemarle Paper Co. v: Moody, 422 U.S. 405 (1975). In Albemarle, in addition to strongly reaffirming the basic disparate impact principle of Griggs v. Duke Power Co., 401 U.S. 424 (1971), the Court held that after the defendant had met its burden of proving the job-relatedness of the challenged practice, it remained open to the plaintiff to show that "other tests or selection devices without a similarly undesirable racial effect, would also serve the employer's interest in 'efficient and trustworthy workmanship.'" Although this was the first time the Supreme Court itself had held that a plaintiff could prevail through establishing the existence of a less discriminatory alternative, the holding was probably more significant in overruling circuit court precedent that had placed a burden of showing the absence of an alternative on defendant.
At any rate, of greater concern here is that Justice Potter Stewart's majority opinion would go on to observe that a showing that there existed a less discriminatory alternative "would be evidence that the employer was using its tests merely as a 'pretext' for discrimination." The suggestion in that observation that the ultimate issue in a disparate impact case might be one of intent, however, was utterly out of keeping with the Court's more explicit statements that the intent behind the use of an unjustified selection practice with a disparate impact simply did not matter. As the courts of appeals had reasoned, if there existed a satisfactory alternative, the challenged practice was not necessary, and was therefore unlawful regardless of the employer's motivation.
Nor did the remark make very much sense. Discriminatory motivation would, of course, render unlawful even a practice that had a strong business justification. But rarely would the existence of a less discriminatory alternative be very probative that the decision to use the procedures at issue was discriminatorily motivated. More likely, it would simply indicate that the employer was thoughtless or was mistaken in its view that the practice was necessary. In any case, the
remark was largely ignored in subsequent treatment of disparate impact claims, and, it would seem, properly so. For it was sufficiently at odds with established precedent to create one of those situations where the Court ought not to be understood to mean what it said.
Twelve years after Albemarle, however, Justice Sandra Day O'Connor would allude to this remark in her plurality opinion in Watson v. Fort Worth Bank and Trust, 427 U.S. 977 (1987), an opinion that contained several intimations of a perceived identity between disparate impact and intentional discrimination. But her allusion was still in the context of a suggestion that the availability of less discriminatory alternatives would go both to the issue of whether the challenged practice was justified and to the issue of whether it was employed purposefully to discriminate.
If Justice O'Connor's Watson plurality hinted that part of the Court was inclined to redefine disparate impact into a means merely of proving intentional discrimination, Justice Byron White's Wards Cove opinion went a good deal farther. In several subtle and not so subtle ways, the opinion appeared to be setting the stage for an eventual transformation of disparate impact analysis into a tortuous search for discriminatory intent. This was evident in several aspects of the analysis that reordered the burden of proof, as well as in the Court's relaxing the business necessity standard to a point where the absence of so minimal a justification might indeed suggest that the employer was engaged in purposeful discrimination.
The insinuative process of redefining disparate impact, however, was probably most evident in the discussion of the less discriminatory alternative. Relying directly on Albemarle, Justice White reasoned, not that a plaintiff's demonstration of a less discriminatory alternative would undermine a defendant's claims that a practice was necessary, but simply that such demonstration 'would prove that 'the employer was using its tests merely as a "pretext" for discrimination.'"
Possibly to justify treating as proof of discriminatory motivation what Justice Stewart's Albemarle opinion had regarded merely as evidence of such motivation, Justice White went on to deflect the objection that the existence of a less discriminatory alternative actually proves very little about the employer's intent. He reasoned: If [plaintiffs], having established a prima facie case, comes forward with alternatives to [defendants'] hiring practices that would reduce the racially disparate impact of practices currently being used, and [defendants] refuse to adopt these alternatives, such a refusal would belie a claim by [defendants] that their incumbent practices are being employed for nondiscriminatory purposes.”
That an employer refuses to adopt what has been proven to be a less discriminatory alternative is obviously probative of discriminatory intent in a way that the mere existence of such an alternative is not. So, seen as part of the evolutionary transformation of disparate impact analysis, the discussion of the employer's refusal to adopt the alternative makes some
sense. But in the context of a litigation where the plaintiff persuades the trier of fact that there exists a less discriminatory alternative that equally serves the employer's business needs, one must ask what on earth the Court's discussion of the employer's refusal to adopt the alternative could possibly mean.
One's first thought is that the Court may not actually be talking about the litigation context, but is envisioning circumstances where sometime in the past the employer refused to adopt a suitable alternative that was made known. But it is hard to imagine circumstances outside of litigation where the plaintiff would have demonstrated the suitability of the alternative to the employer. In any case, Justice White's words unequivocally pertain to a situation where the refusal follows a showing made in court.
Yet, given that unavoidable interpretation, one is left with conjuring up a variety of equally absurd scenarios. Did the Court, for example, envision that after the trial judge finds for the plaintiff with respect to the existence of a satisfactory less discriminatory alternative, the judge would ask the employer if it will implement the new procedure? And would a refusal be interpreted to reflect that the prior use of the policy was discriminatorily motivated, and hence that there is liability for intentional discrimination through the past use of the practice – which, incidentally, is the use that affected individuals ordinarily would be complaining about? Or is it only the future – post refusal – use of the challenged practice that the Court is talking about?
And let us not overlook that, though refusal to use what a court has found to be a less discriminatory alternative is more probative of discriminatory motivation than the mere existence of the alternative, it still is not all that probative. Few litigants are persuaded that a court's findings adverse to the litigant's position in fact reflect a truth to which the litigant previously had been blind. In actual practice, moreover, when asked by the judge what it plans to do about implementing the alternative, the employer would usually say that what it plans to do is to appeal. Leaving aside questions of whether the employer has yet won or lost or what it would appeal from, consider what we are to infer about the employer's discriminatory motivation from a refusal to adopt the alternative after a divided court of appeals upholds the district court's finding regarding the alternative, with even the majority able to say no more than that the finding was not clearly erroneous.
Possibly the refusal issue would be moot since the district court ought simply to enjoin use of the unlawful practice. But has there yet been an unlawful practice? Of course, faced with the injunction and any other liability because of the refusal, no employer actually will refuse; accordingly, it would seem, no employer actually would ever be found liable because of the existence of the less discriminatory alternative.
So, while it may be impossible to say what the Court could have meant, it seems clear enough that it could not have meant anything that makes any sense. This is one reason why we could expect the lower courts to ignore this cryptic passage just as they ignored the language in Albemarle. Keep in mind, moreover, that the refusal to adopt the alternative actually is relevant only to establishing discriminatory motivation in the use of the existing practice; it has no relevance to the way in which the existence of an alternative shows that the existing practice is not justified. At least until the Supreme Court had completed the process of limiting disparate impact to a means of proving intentional discrimination, this would be another reason for the lower courts to ignore the language about the employer's refusal to adopt an alternative.
Enter the Civil Rights Act of 1991. One thing that the bill clearly intends is to arrest the transformation of disparate impact into a means solely of proving intentional discrimination. And whatever the practical import of the actual amendments – which return to the defendant the burden of proof-on" the business necessity issue and state that the business necessity standard is whatever it was prior to Wards Cove – Congress will have precluded the Court from wholly redefining the disparate impact principle.
Yet, somehow one of the few things from Wards Cove that has been deemed worthy of codifying is that employer liability resulting from the existence of a less discriminatory alternative turns on the employer' refusal to adopt the less discriminatory alternative after the plaintiff proves that such an alternative exists. Aspects of Congress' treatment of the "alternative" issue, to be sure, are a bit on the vague side. The countless revisions to the disparate impact provision resulted in two cross-referenced subsections on the issue, each suggesting that the substance was to be found in the other. According to subparagraph (A)(ii) of a Title VII's new Section 703(k)(1), plaintiff can prevail on a disparate impact claim by "making the demonstration described in subparagraph (C) with respect to an alternative employment practice and the respondent refuses to adopt such alternative employment practice." Subparagraph (C) merely states that "the demonstration referred to by subparagraph (A)(ii) shall be in accordance with the law as it existed [prior to Wards Cove] with respect to the concept of 'alternative employment practice.'"
If the latter provision was meant to prevent the Court from requiring that the alternative practice serve the employer's interest just as well as the challenged practice, or to do anything else for that matter, the language chosen surely is not explicit enough. The former provision, however, is at least as explicit as the Wards Cove decision itself in indicating that the refusal that renders the employer liable comes after the plaintiff demonstrates the alternative procedure at trial. Indeed, the word "demonstrates" is defined by the amendments as "meet[ing] the burden of proof and persuasion."
A brief passage in an interpretative statement by Senator Kennedy attempts to deal with the implications of the refusal language by noting that the "employer cannot escape liability ... by adopting the [alternative] at a later time, such as during the trial of the disparate impact claim." Whatever "later time" is supposed to mean, such glossing cannot avoid the fact that the statutory language explicitly contemplates that the issue of the refusal does not even arise until after trial. In any case, a section of the bill specifically precludes the courts from using such statements in interpreting the disparate impact provisions.
The focus on less discriminatory alternatives is unlikely to be greatly diminished by the act. The act does place the burden of proving business necessity back on the employer. But instead of providing a definition of business necessity, the act merely codifies the law as it existed prior to Wards Cove – which the Wards Cove majority would say is precisely the interpretation it provided in Wards Cove. So proving the existence of less discriminatory alternative may continue to be a principal route by which plaintiffs may prevail in disparate impact suits. Figuring out what Congress could have meant in the language about the employer's refusal to adopt the alternative and how that meaning will be implemented in the context of a trial is unlikely to be any easier than figuring out what the Court could have meant in Wards Cove. The only difference is that courts have less leeway in ignoring things that make no sense in statutes than they have in ignoring things that make no sense in Supreme Court opinions.
B. Dealing With a Nonsensical Rule
As mentioned in the introduction, this page is only a sketch. I have not explored how courts have endeavored to deal with issues raised in Section A. One interesting thing about the situation is that while there exist various rules for interpreting cryptic congressional language, I am not sure how useful those rules are for interpreting a statutory provision that makes no sense. Another interesting thing involves the fact that there are increasing efforts to apply disparate impact doctrine in other contexts – and the statistical misperceptions underlying those efforts are discussed on the Discipline Disparities and Lending Disparities pages of this site. So let us assume that the incongruous ruling and statutory provision in Section A have to be dealt with in the employment context even if they make no sense, the issue may be different in context where the incongruous ruling and statutory provision do not explicitly apply.
In any case – again, while not yet exploring the issue – from what I can tell there seems yet to be little guidance on the matter. The April 25, 2012 EEOC Enforcement Guidance styled “Consideration of Arrest and Conviction Records in Employment Decisions Under Title VII of the Civil Rights Act of 1964” (Section V.C) discussed on the Less Discriminatory Alternative – Substantive sub-page merely refers one to the statutory provision and the Watson case. A disparate impact bulletin published on behalf of the American Bankers’ Association following the April 18, 2012 announcement of the Consumer Financial Protection Bureau that it would apply the disparate impact concept to the enforcement of the Equal Credit Opportunity Act (which announcement is the subject of my “’Disparate Impact’: Regulators Need a Lesson in Statistics” American Banker, June 5, 2012) emphasizes that in holding a lender liable on the basis of the existence of a less discriminatory alternative to a challenged practice, the alternative “must have previously been brought to the bank’s attention.” It also notes that banks cannot be required to test each combination of underwriting and pricing factors to test which has the least discriminatory effect. What any of this will mean in practice remains to be seen. Sorting it out, of course, will be especially difficult in any context where dichotomies are at issue and hence where whether a practice is less discriminatory will often turn on whether one examines the favorable or the adverse outcome.