The future of disparate impact

Michigan Law Review

Volumn 108, Issue 8, 2010, Pages 1341-1388

  Primus, Richard ^a  

^a UNIVERSITY OF MICHIGAN LAW SCHOOL   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 77954437133     PISSN: 00262234     EISSN: None     Source Type: Journal    
DOI: None     Document Type: Article

References (190)

1. 129 S. Ct. 2658, 2683 (2009).
2. Ricci, 129 S. Ct. 2658; Adam Liptak, Sotomayor Case Draws Scrutiny, N.Y. Times, June 6, 2009, at A1.
3. Ricci, 129 S. Ct. at 2666-71 (describing the series of meetings held by the New Haven Civil Service Board to discuss the "significant disparate impact" of the written exams).
4. One of the Ricci plaintiffs, Benjamin Vargas, was Latino. Many accounts of the case have therefore spoken of the plaintiffs as a group of nineteen white firefighters and one Latino firefighter. See, e.g., Sotomayor Embracing Affirmative Action, Then and Now, Los Angeles Times, June 15, 2009; Firefighters' Case Called Civil Rights "Threat," New Haven Register, March 26, 2009. That said, "Latino" and "white" are not mutually exclusive categories, and according to published reports Lt. Vargas falls into both categories. E.g., Adam Liptak, Supreme Court Finds Bias Against White Firefighters, N.Y. Times, June 30, 2009, at A1 (describing all the plaintiffs as white firefight-ers and one plaintiff as also being Hispanic).
5. Ricci, 129 S. Ct. at 2664.
6. Id. at 2664, 2673; see 42 U.S.C. § 2000e-2(k) (2006).
7. Ricci, 129 S. Ct. at 2681. My use of the phrase "formal or intentional" is intentionally ambiguous: disparate treatment doctrine often conflates these two conceptions of discrimination, but there is value for present purposes in noticing that they are not the same. See infra Part I.
8. Ricci, 129 S. Ct. at 2676, 2681.
9. 515 U.S. 200 (1995).
10. 551 U.S. 701 (2007).
11. Ricci, 129 S. Ct. at 2682 (Scalia, J., concurring).
12. See, e.g., Paul Brest, Foreword: In Defense of the Antidiscrimination Principle, 90 Harv. L. Rev. 1, 4-5, 22-26 (1976); Owen M. Fiss, Groups and the Equal Protection Clause, 5 Phil. & Pub. Aff. 107, 141-46 (1976).
13. See Washington v. Davis, 426 U.S. 229, 248 (1976).
14. Richard A. Primus, Equal Protection and Disparate Impact: Round Three, 117 Harv. L. Rev. 493 (2003).
15. See generally id.
16. Ricci, 129 S. Ct. at 2674 ("We consider, therefore, whether the purpose to avoid disparate-impact liability excuses what otherwise would be prohibited disparate-treatment discrimination.").
17. Id. at 2676.
18. See infra Part I.
19. Ricci, 129 S. Ct. at 2682 (Scalia, J., concurring).
20. See Ronald Dworkin, Justice Sotomayor: The Unjust Hearings, N.Y. Rev. Books, Sept. 24, 2009, at 37, 39.
21. See infra Section II.B.
22. See infra Section II.C; see also Lawrence Lessig, Erie-Effects of Volume 110: An Essay on Context in Interpretive Theory, 110 Harv. L. Rev. 1785 (1997) (describing the ways in which an issue's moving from the background to the foreground of public consciousness can change constitutional doctrine's approach to that issue).
23. See, e.g., Brian T. Fitzpatrick, Strict Scrutiny of Facially Race-Neutral State Action and the Texas Ten Percent Plan, 53 Baylor L. Rev. 289 (2001).
24. See, e.g., Ian Ayres & Sydney Foster, Don't Tell, Don't Ask: Narrow Tailoring After Grutter and Gratz, 85 Tex. L. Rev. 517, 518 (2007).
25. See, e.g., Parents Involved in Cmty. Schs. v. Seattle Sch. Dist. No. 1, 551 U.S. 701, 787-89 (2007) (Kennedy, J., concurring in part and concurring in the judgment); Shaw v. Reno, 509 U.S. 630, 647 (1993) (stating that in some equal protection cases, "appearances do matter"); City of Richmond v. J.A. Croson Co., 488 U.S. 469, 507 (1989).
26. See In re Employment Discrimination Litig., 198 F.3d 1305, 1315 (11th Cir. 1999) (ex-plaining that the principal disparate impact remedy is enjoining the employer against future use of the challenged practice). As the above analysis suggests, a court could adopt the institutional and visible-victim readings simultaneously. See infra note 27.
27. The institutional and visible-victims readings are easily combinable in practice because the remedies that courts standardly provide for disparate impact violations all avoid creating visible innocent victims. Those remedies include injunctive relief against using the challenged practice in the future and equitable relief like backpay. These remedies run against the employer only and do not visibly burden determinate third parties, even if they necessarily have downstream distributional consequences. Assuming that future judicially ordered disparate impact remedies conform to this pattern, the law could therefore adopt the institutional and visible-victims readings of Ricci simultaneously. Alternatively, it could officially adopt only the institutional reading but also satisfy the concerns of the visible-victims reading as a consequence.
28. See, e.g., Local No. 93, Int'l Ass'n of Firefighters v. City of Cleveland, 478 U.S. 501, 515 (1986).
29. See Brief for Respondents at 17-18, Ricci v. DeStefano, 129 S. Ct. 2658 (2009) (Nos. 07-1428 & 08-328).
30. Ricci v. DeStefano, 129 S. Ct. 2658, 2674 (2009).
31. See Primus, supra note 14, at 566-67.
32. See Richard Primus, Double-Consciousness in Constitutional Adjudication, 13 Rev. Const. Stud. 1 (2007).
33. See Plessy v. Ferguson, 163 U.S. 537, 551 (1896) (upholding Louisiana's segregated-car law) ("We consider the underlying fallacy of the plaintiff's argument to consist in the assumption that the enforced separation of the two races stamps the colored race with a badge of inferiority. If this be so, it is not by reason of anything found in the act, but solely because the colored race chooses to put that construction upon it.").
34. This description of New Haven's promotion process is adapted from the district court's opinion in Ricci. See Ricci v. DeStefano, 554 F. Supp. 2d 142, 142-47 (D. Conn. 2006).
35. All the applicants for promotion were male.
36. Forty-one applicants took the captain exam. Twenty-two passed, thus becoming eligible for promotion to one of the seven vacant captain positions. Three of those who passed were African American. Given the Rule of Three, however, the seven vacant positions all had to be filled from the nine highest scoring applicants on the combined index, and none of the three African Americans with passing scores was in the top nine. Accordingly, the system as designed would have promoted no African Americans to the rank of captain. The situation with the lieutenant exam was similar: seventy-seven applicants took the tests, and thirty-four did well enough to be deemed qualified for promotion, six of them black. But under the Rule of Three, the eight vacant positions all had to be filled from the top ten scorers, and none of the qualified black applicants was within the top ten. Ricci, 554 F. Supp. 2d at 145.
37. See Ricci v. DeStefano, 129 S. Ct. 2658, 2671 (2009).
38. See id.
39. See Ricci, 554 F. Supp. 2d at 150.
40. Id. at 163.
41. Ricci v. DeStefano, 530 F.3d 87 (2d Cir. 2008), reh'g en banc denied, 530 F.3d 88 (2d Cir. 2008).
42. Ricci, 129 S. Ct. at 2681.
43. See id. at 2673.
44. See id. at 2674.
45. See id. at 2675.
46. See id. at 2677.
47. See id. at 2678 (recognizing the "four-fifths rule" of 29 C.F.R. § 1607.4(D) (2008), under which federal enforcement agencies generally find evidence of disparate impact for the purposes of Title VII if a selection mechanism results in a pass rate for one racial group that is less than 80 percent of the pass rate for another racial group).
48. 42 U.S.C. § 2000e-2(k)(1)(A)(i) (2006) (stating that no unlawful employment practice based on disparate impact is established in cases where the respondent can "demonstrate that the challenged practice is job related for the position . . . and consistent with business necessity").
49. Ricci, 129 S. Ct. at 2678-79.
50. Id. at 2673 ("Our analysis begins with this premise: The City's actions would violate the disparate-treatment prohibition of Title VII absent some valid defense."); see id. at 2674 (recognizing that the need to comply with Title VII's disparate impact doctrine would constitute such a defense).
51. Title VII prohibits discrimination on grounds of "race, color, religion, sex, or national origin." 42 U.S.C. § 2000e-2(a)(1). In this Article, however, I am concerned with an issue of race, and I will generally use language that is limited to issues of race.
52. See, e.g., Int'l Bhd. of Teamsters v. United States, 431 U.S. 324 (1977) (finding a dispa-rate treatment violation when an employer assigned black and Hispanic truck drivers to less-desirable positions than white truck drivers).
53. See, e.g., McMullen v. Warner, 416 F. Supp. 1163 (D.D.C. 1976) (finding disparate treatment when, in order to prevent a black applicant from filling a position, the position was eliminated, thus denying it to all applicants). Conversely, a showing of illicit motive is not required to make out a disparate treatment claim: a showing of formally disparate treatment in the ordinary-language sense will suffice. See, e.g., UAW v. Johnson Controls, Inc., 499 U.S. 187, 199 (1991).
54. See, e.g., Palmer v. Thompson, 403 U.S. 217 (1971) (addressing the decision of city officials in Jackson, Mississippi, to close municipal swimming pools entirely rather than permit African Americans to swim there).
55. Given its aggregate effects, it is also likely to be actionable under the doctrine of disparate impact.
56. See, e.g., Ky. Ret. Sys. v. EEOC, 128 S. Ct. 2361, 2366 (2008) (describing what must be proved by a plaintiff who claims " 'disparate treatment' (i.e., intentional discrimination[)]"). This terminological oddity is a product of the way that the Supreme Court organized antidiscrimination law in the 1970s. For a long time, official doctrine in American law had long wobbled among three accounts of the locus of actionable discrimination: motive, form, and impact. On the motive-based account, an action is discriminatory because of the actor's state of mind. On the form-based account, an action is discriminatory on the basis of the overt or visible aspect of the action. On the impact-based account, an action is discriminatory on the basis of the consequences that the action produces. See Kathleen M. Sullivan, After Affirmative Action, 59 Ohio St. L.J. 1039, 1046 (1998) (distin-guishing the three accounts). To be sure, these three concerns can flow into one another, such that many phenomena in antidiscrimination law cannot be fully understood as falling into one category but not the other two. But the law often tries to distinguish among them. When the disparate impact doctrine became the repository of the impact-based account of discrimination within Title VII, the other two accounts were grouped together as "not-disparate-impact": courts began classifying all cases of intentional discrimination and all cases of overt or formal differentiation as falling into a single category, and they extended the term "disparate treatment" to cover both kinds of cases de-spite its semantic awkwardness for the purpose. The term has stuck well enough that today we rarely notice the awkwardness at all. We simply understand that "disparate treatment" in Title VII is a term that covers both formal differences in the treatment of people of different groups and unlawful em-ployer motives.
57. Cases fitting the pattern here described and on which courts have declined to find dispa-rate treatment under Title VII include, for example, Oakley v. City of Memphis, 315 F. App'x 500 (6th Cir. 2008); Hayden v. County of Nassau, 180 F.3d 42 (2d Cir 1999); Byers v. City of Albuquerque, 150 F.3d 1271 (10th Cir. 1998).
58. See U.S. Comm'n on Civil Rights, Affirmative Action in the 1980s: Dismantling the Process of Discrimination 17 n.20 (1981).
59. See, e.g., United Steelworkers of Am. v. Weber, 443 U.S. 193, 194 (1979) (stating that Title VII was intended to be compatible with race-conscious affirmative action to help improve the position of African Americans).
60. See, e.g., Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265 (1978) (plurality opinion) (agreeing that a medical school could aim at admitting a racially diverse student body but disapprov-ing of the method used to achieve that aim).
61. See Deborah C. Malamud, Values, Symbols, and Facts in the Affirmative Action Debate, 95 Mich. L. Rev. 1668, 1676 (1997) (noting that affirmative action was controversial but disparate impact doctrine was not). The contrast between disparate impact and affirmative action parallels the distinction between overt disparities in treatment, strictly construed, and disfavored motives. Like disparate impact doctrine, affirmative action proceeds from motives that were broadly considered acceptable thirty years ago. But unlike disparate impact doctrine, most affirmative action programs engage in the disparate treatment (strictly construed) of particular persons. Accordingly, the Supreme Court long ago classified those forms of affirmative action that were acceptable under Title VII as exceptions to the general prohibition on disparate treatment. See, e.g., United Steelworkers of Am., 443 U.S. at 201-08 (acknowledging that a facially classificatory affirmative action plan was within the language of Title VII's prohibition on disparate treatment but that a legitimate affirmative action plan constituted a valid defense to liability). The Ricci premise extends this way of thinking to disparate impact doctrine for the first time.
62. See Malamud, supra note 61, at 1693 (noting the near-universal acceptance of disparate impact theory as a valid part of antidiscrimination law).
63. See, e.g., Parents Involved in Cmty. Schs. v. Seattle Sch. Dist. No. 1, 551 U.S. 701, 748 (2007) ("The way to stop discrimination on the basis of race is to stop discriminating on the basis of race.").
64. Compare id. (disallowing consideration of race as a tiebreaker in a small number of school assignments as part of a school district's attempt to maintain racial diversity in its schools), with Swann v. Charlotte-Mecklenburg Bd. of Educ., 402 U.S. 1 (1971) (upholding a wholesale busing remedy designed to integrate a school system).
65. See, e.g., Grutter v. Bollinger, 539 U.S. 306 (2003) (upholding a limited affirmative action plan in university admissions when that plan gave sufficiently individualized consideration to all applicants).
66. See Ricci v. DeStefano, 129 S. Ct. 2658, 2674 (2009).
67. See id. at 2674.
68. Id. at 2675 ("This suit does not call on us to consider whether the statutory constraints under Title VII must be parallel in all respects to those under the Constitution."); id. at 2681 (declining to address the plaintiffs' constitutional claim).
69. Equal protection doctrine covers all government actors, whether or not they are employ-ers, but it reaches no private parties. See Civil Rights Cases, 109 U.S. 3, 11 (1883). Title VII reaches only employers, but it covers all employers, private or public, over a certain size. 42 U.S.C. § 2000e(a) (2006) (specifying that "person[s]" include "governments, governmental agencies, [and] political subdivisions"); 42 U.S.C. § 2000e(b) (2006) (defining as covered employers all persons "engaged in an industry affecting commerce who have fifteen or more employees for each working day in each of twenty or more calendar weeks in the current or preceding calendar year"); 42 U.S.C. § 2000e-16 (2006) (extending coverage to federal government employees).
70. Compare 42 U.S.C. § 2000e-5 (2006) (setting forth administrative filing requirements and enforcement procedures under Title VII), with 42 U.S.C. § 1983 (2006) (listing the necessary components of a cause of action alleging the deprivation of constitutional rights), and Bivens v. Six Unknown Named Agents of Fed. Bureau Narcotics, 403 U.S. 388 (1971) (governing requirements for lawsuits raising causes of action under the Equal Protection Clause).
71. Note that equal protection doctrine, like disparate treatment doctrine, houses both the form-based and the motive-based accounts of discrimination-that is, everything but the concern with impact, which is carved off and placed elsewhere. See, e.g., Washington v. Davis, 426 U.S. 229 (1976) (regarding equal protection); supra Part I (regarding disparate treatment). The ambiguity between form and motive has animated a parallel set of conflicts at the statutory and constitutional levels. Just as the Court has divided deeply over whether formal racial classifications are offensive to equal protection even when not motivated by racial animus, see, e.g., Johnson v. California, 543 U.S. 499 (2005); Adarand Constructors, Inc. v. Pena, 515 U.S. 200 (1995), the Court has divided deeply over whether a statutory disparate treatment claim should lie when an employer deploys a disfavored classification in the course of advancing an administrative scheme not motivated by animus against any category of workers. Compare Ky. Ret. Sys. v. EEOC, 128 S. Ct. 2361 (2008) (majority opinion) (privileging motive), with id. at 2371 (Kennedy, J., dissenting, joined by Scalia, Ginsburg, & Alito, JJ.) (privileging form). As the examples of Johnson v. California and Kentucky Retirement Systems indicate, it is not always the same Justices who rest on form and the same Justices who rest on motive. A focus on constitutional affirmative action cases might encourage the generalization that the more conservative Justices are more focused on form than their liberal counterparts, but on fuller consideration the reality is more complex. Particular decisionmakers can be on either side under different circumstances, and the Court's familiar liberal and conservative blocs do not always cohere on the question. Whichever way the Court leans in a particular case, however, the ambiguity it confronts is the same in disparate treatment as in equal protection.
72. Suppose that a police department wants to assign an undercover officer to infiltrate the Russian mafia and considers only white officers for the position on the grounds that the target organization is composed exclusively of white people (i.e., ethnic Russians), such that a nonwhite officer could never pass as a member. If a black officer brought an equal protection claim alleging racial discrimination, a court could (and surely would) find against him on the grounds that choosing a white officer for this job was narrowly tailored to a compelling governmental interest. But if a black officer brought a Title VII disparate treatment claim, the police department would have no defense. (Title VII recognizes a bona fide occupational qualification (?BFOQ?) defense in cases where a person?s religion, sex, or national origin is actually necessary to performance of a job, but the statute recognizes no BFOQ defense to claims of disparate treatment of the basis of race. See 42 U.S.C. ? 2000e-2(e)(1) (2006).) In two respects, however, this example of a divergence in the coverage between Title VII and equal protection only serves to emphasize how thoroughly the two rubrics reproduce each other as a general matter. First, finding this difference requires resort to the fanciful: in real life, police officers do not sue to be permitted to undertake quixotic suicide missions like the one imagined here. Second, even this divergence between statutory and constitutional coverage arises from a difference in the defenses that apply in each sphere, not a difference between what Title VII and equal protection reach as an initial matter.
73. As noted earlier, I have explored the potential tensions between equal protection and disparate impact doctrine at length elsewhere. See Primus, supra note 14. Readers interested in the full analysis should see that discussion. But that article demonstrated that the relationship between equal protection and disparate impact was substantially indeterminate, such that the question of their compatibility would depend on which of several possible views of equal protection, and of disparate impact, an adjudicating official would ultimately adopt. The project of this Article is to show how Ricci narrows the range of views that the Supreme Court is likely to adopt.
74. See, e.g., Burlington N. & Santa Fe Ry. Co. v. White, 548 U.S. 53, 57 (2006).
75. E.g., Vincent v. Brewer Co., 514 F.3d 489, 493 (6th Cir. 2007).
76. E.g., Dorsey v. Morgan Stanley, 507 F.3d 624, 627-28 (7th Cir. 2007).
77. E.g., Nilsson v. City of Mesa, 503 F.3d 947, 952 (9th Cir. 2007).
78. th Cir. 2007).
79. E.g., Nichols v. S. Ill. Univ.-Edwardsville, 510 F.3d 772, 780 (7th Cir. 2007).
80. Cf. Williams v. Bristol-Myers Squibb Co., 85 F.3d 270, 274 (7th Cir. 1996) (explaining that Title VII does not protect against every action "that an irritable, chip-on-the-shoulder employee did not like").
81. Compare, e.g., Reed v. Unified Sch. Dist. No. 223, 299 F. Supp. 2d 1215, 1226 (D. Kan. 2004) (stating that refusal to give a letter of recommendation is adverse employment action because it risks harming the employee's future ability to get a job), with Enowmbitang v. Seagate Tech., Inc., 148 F.3d 970, 973-74 (8th Cir. 1998) (holding that giving a poor evaluation is not an adverse employment action if no further consequence immediately flows from it).
82. I assume that the canceled process and the hypothetical future process would have sorted roughly the same applicant pool. Obviously, the two processes would have sorted that pool some-what differently. But many highly qualified applicants probably would have succeeded under both processes, assuming that both were valid measurements of qualification.
83. If New Haven had replaced the original test with a system that wound up yielding ex-actly the same set of promotable candidates, then the plaintiffs' average probability of promotion would be unaffected. But it seems highly unlikely that New Haven's chosen replacement system would have yielded that result. After all, we can assume that New Haven would have replaced the original test with a system designed, among other things, to change the pool of promotable candi-dates.
84. See Lujan v. Defenders of Wildlife, 504 U.S. 555, 560-61 (1992) (setting forth the gen-eral criteria for constitutional injury cognizable in Article III courts).
85. See, e.g., Shaw v. Reno, 509 U.S. 630 (1993).
86. 411 U.S. 792 (1973).
87. Id. at 802-03.
88. See Deborah C. Malamud, The Last Minuet: Disparate Treatment After Hicks, 93 Mich. L. Rev. 2229 (1995).
89. Ricci v. DeStefano, 554 F. Supp. 2d 142, 151 (D. Conn. 2006) ("Because plaintiffs allege intentional discrimination, the familiar McDonnell Douglas three-prong burden-shifting test ap-plies."); id. at 151-60 (conducting the McDonnell Douglas analysis).
90. See Price Waterhouse v. Hopkins, 490 U.S. 228, 271 (1989) (O'Connor, J., concurring) (stating that the rationale for the McDonnell Douglas proof framework is that "direct evidence of intentional discrimination is hard to come by").
91. See generally Fisher v. Vassar Coll., 114 F.3d 1332, 1354-61 (2d Cir. 1997) (en banc) (Calabresi, J., concurring in part and dissenting in part).
92. See, e.g., Rogers v. City of Chicago, 320 F.3d 748, 754 (7th Cir. 2003).
93. See Fernandes v. Costa Bros. Masonry, 199 F.3d 572, 582 (1st Cir. 1999) (canvassing varying understandings of direct evidence).
94. Ricci v. DeStefano, 554 F. Supp. 2d 142, 145 (D. Conn. 2006) (noting that the parties strenuously disputed the legal issues in the cases but largely agreed on the facts).
95. Price Waterhouse, 490 U.S. at 228; see, e.g., Desert Palace, Inc. v. Costa, 539 U.S. 90 (2003) (applying Price Waterhouse). The facts of Ricci made the case a natural candidate for mixed-motive analysis: the parties disputed whether New Haven had acted for the mere purpose of complying with Title VII or for the purpose of gratifying an important racially defined political constituency, and one possible answer was "a little of both." Justice Alito's concurrence suggests that at least three Justices found the racial politics explanation plausible, at least in part. See Ricci v. DeStefano, 129 S. Ct. 2658, 2683-88 (2009) (Alito, J., concurring). Moreover, New Haven's deci-sion was the joint product of more than one decisionmaker, as many municipal decisions are, and decisions with multiple decisionmakers are regularly proper subjects for mixed-motive analysis because different decisionmakers may have acted for different reasons, or different combinations of reasons. The majority opinion suggested that New Haven acted for a combination of motives, rather than for any single purpose to the exclusion of all others. See id. at 2681 (majority opinion) (stating that "the raw racial results became the predominant rationale" for the city's decision to set aside the tests) (emphasis added)). But the parties did not raise a mixed-motive argument in the district court, Ricci v. DeStefano, 530 F.3d 88, 89 (2d Cir. 2008) (Calabresi, J., concurring in the denial of rehear-ing en banc), so there is a straightforward explanation for the absence of mixed-motive analysis in the Supreme Court's opinion.
96. The closest the Court has come has been to hold that direct-evidence cases are different from McDonnell Douglas cases in the context of the Age Discrimination in Employment Act. See Trans World Airlines, Inc. v. Thurston, 469 U.S. 111, 121 (1985); see also Wards Cove Packing Co. v. Atonio, 490 U.S. 642, 667 (1989) (stating, in the context of a disparate impact case, that McDon-nell Douglas provides "[t]he means for determining intent absent direct evidence"); Price Waterhouse, 490 U.S. at 270 (O'Connor, J., concurring) (distinguishing McDonnell Douglas cases from cases involving "direct evidence" of employer decisionmaking on the basis of a forbidden factor). These statements are a sufficient foundation for applying the idea in a disparate treatment case, but giving them force for the first time merits some discussion in light of prior practice to the contrary.
97. See Johnson v. Transp. Agency, 480 U.S. 616, 626 (1987).
98. See Daniel R. Ortiz, The Myth of Intent in Equal Protection, 41 Stan. L. Rev. 1105 (1989) (distinguishing several modes of intent analysis in equal protection).
99. See Kim Forde-Mazrui, The Constitutional Implications of Race-Neutral Affirmative Action, 88 Geo. L.J. 2331, 2333 (2000).
100. See Primus, supra note 14, at 539-44.
101. See, e.g., League of United Latin Am. Citizens v. Perry, 548 U.S. 399, 474 (2006) (Stevens, J., concurring in part and dissenting in part); Vieth v. Jubelirer, 541 U.S. 267, 284-86 (2004); Miller v. Johnson, 515 U.S. 900, 911-13 (1995); Primus, supra note 14, at 545. Predominant motive is an imprecisely defined term, if it is defined at all. Its use in these cases seems intended to signal that racial consideration should not assume undue importance relative to other, less problematic factors in the decisionmaking process.
102. Ricci v. DeStefano, 129 S. Ct. 2658, 2674 (2009) ("Petitioners next suggest that an employer in fact must be in violation of the disparate-impact provision before it can use compliance as a defense in a disparate-treatment suit.").
103. Id. at 2674-75.
104. Id.
105. Cf. Nelson Tebbe & Robert L. Tsai, Constitutional Borrowing, 108 Mich. L. Rev. 459 (2010) (describing the practice and hazards of using doctrinal tools and tropes from one context in another context).
106. Ricci, 129 S. Ct at 2675 (quoting Wygant v. Jackson Bd. of Educ., 476 U.S. 267, 277 (1986) (plurality opinion)).
107. Id. (considering City of Richmond v. J.A. Croson Co., 488 U.S. 469, 500 (1989); Wygant v. Jackson Bd. of Educ., 476 U.S. 267, 277 (1986) (plurality opinion)).
108. See Ricci, 129 S. Ct at 2675.
109. See id. at 2682 (Scalia, J., concurring).
110. See Dworkin, supra note 20, at 38-39.
111. See supra Section I.B.
112. See generally Andrew Kull, The Color-Blind Constitution (1992).
113. Technically , it would be possible for an employer to avoid disparate impact liability without any race-conscious action. An employer who ensured that all of his employment practices met the business necessity test would be able to defend against any claim that might arise if some of those practices had racially disparate impacts. But one should not invest too heavily in this possible resolution. Litigation is expensive, so most employers most of the time would rather avoid being in the position where a plaintiff could make a prima facie case of disparate impact which would then need to be answered by a business necessity defense. The way to avoid going down that road is, of course, not to cause statistically disparate impacts in the first place.
114. See, e.g., Grutter v. Bollinger, 539 U.S. 306, 326 (2003) (("[T]he Fourteenth Amendment 'protect[s] persons, not groups[.]' ") (second alteration in original) (emphasis added) (quoting Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 227 (1995)); Miller v. Johnson, 515 U.S. 900, 911 (1995) ("At the heart of the Constitution's guarantee of equal protection lies the sim-ple command that the Government must treat citizens as individuals.") (internal quotation marks omitted).
115. I specify public employers rather than employers generally because, given the state ac-tion requirement, only public employers can violate the Equal Protection Clause. A private employer trying to cure a disparate impact problem could violate Title VII, but its actions could not be uncon-stitutional.
116. See generally Mitchell N. Berman, Constitutional Decision Rules, 90 Va. L. Rev. 1, 48-50 (2004).
117. See, e.g., Katzenbach v. Morgan, 384 U.S. 641 (1966).
118. See Lawrence Gene Sager, Fair Measure: The Legal Status of Underenforced Constitutional Norms, 91 Harv. L. Rev. 1212 (1978).
119. See David A. Strauss, The Ubiquity of Prophylactic Rules, 55 U. Chi. L. Rev. 190 (1988).
120. See Richard H. Fallon, Jr., Judicially Manageable Standards and Constitutional Meaning, 119 Harv. L. Rev. 1274 (2006).
121. See Evan H. Caminker, Miranda and some Puzzles of "Prophylactic" Rules, 70 U. Cin. L. Rev. 1, 25 (2001); Daryl J. Levinson, Rights Essentialism and Remedial Equilibration, 99 Colum. L. Rev. 857 (1999).
122. 334 U.S. 1 (1948) (holding that judicial enforcement of a racially restrictive covenant in a title deed constituted a denial of equal protection under the Fourteenth Amendment).
123. Equal protection has a similar tolerance for nonjudicial governmental actors executing something like the judicial function of remediation. For example, an administrative office evaluating an internal grievance alleging racial discrimination in a government agency would be permitted to consider the race of the complainant in much the same way that a court could take note of the race of a Title VII plaintiff. Interestingly, neither the caselaw nor the literature contains a full account of why colorblindness is subject to this limit. One possibility is that the individualist ideals that moti-vate colorblindness require at least this much color-consciousness for their enforcement. But this is not a complete explanation, because one could easily ask why what is required is this much, rather than a little more or a little less. Whether this conundrum is a problem for prevailing practices or for the theory of colorblindness is a question for another day.
124. This is not to deny that judges, like everybody else, can suffer from biases, nor is it to deny that a judiciary whose members are recruited disproportionately from certain segments of the population might show biases in predictable directions. But the design of the office is based on an aspiration to neutrality.
125. See Michael Selmi, Was the Disparate Impact Theory a Mistake?, 53 UCLA L. Rev. 701, 756-57 (2006).
126. See, e.g., Griggs v. Duke Power Co., 401 U.S. 424 (1971); Parham v. Sw. Bell Tel. Co., 433 F.2d 421 (8th Cir. 1970); Young v. Edgcomb Steel Co., 363 F. Supp. 961 (M.D.N.C. 1973).
127. See, e.g., Boston Chapter, NAACP v. Beecher, 504 F.2d 1017 (1st Cir. 1974) (Boston); Vulcan Soc'y of N.Y. City Fire Dep't v. Civil Serv. Comm'n, 490 F.2d 387 (2d Cir. 1973) (New York); United States v. City of Chicago, 385 F. Supp. 543 (N.D. Ill. 1974) (Chicago); Officers for Justice v. Civil Serv. Comm'n, 371 F. Supp. 1328 (N.D. Cal. 1973) (San Francisco); Harper v. Mayor of Baltimore, 359 F. Supp. 1187 (D. Md. 1973) (Baltimore). There are recent examples as well. See, e.g., United States v. City of New York, 637 F. Supp. 2d 77 (E.D.N.Y. 2009) (granting summary judgment against the New York City Fire Department in a Title VII suit alleging that a written examination for selecting entry-level firefighters had an unlawfully disparate impact on black and Hispanic applicants).
128. Selmi, supra note 125, at 756-57.
129. See, e.g., Diane Cardwell, Racial Bias in Fire Exams Can Lurk in the Details, N.Y. Times, July 24, 2009, at A22 (describing a recent suit where fire department entrance exams tested knowledge of technical jargon, thus favoring those from traditional firefighter families or English-speaking families).
130. See Selmi, supra note 125, at 756.
131. See, e.g., African American Mayors: Race, Politics, and the American City 4-6 (David R. Colburn & Jeffrey S. Adler eds., 2001); Jon Teaford, The Twentieth-Century American City: Problem, Promise, and Reality 147 (1993); Clarence Stone, Regime Politics: Governing Atlanta, 1946-1988 247 (1989).
132. As should be obvious, the shift in political demographics did not mean that urban office-holders no longer had incentives to protect the interests of white ethnic groups in the allocation of public employment. Often those incentives remained. But similar incentives also obtained with respect to the employment of nonwhites. The precise balance of incentives in any particular case, or for any particular official, is a function of specific circumstances within the relevant polity.
133. See Selmi, supra note 125, at 764.
134. See Ricci v. DeStefano, 129 S. Ct. 2658, 2683-88 (2009) (Alito, J., concurring).
135. Ricci's willingness to let employers escape disparate treatment liability with a strong basis in evidence for believing that the alternative is a disparate impact violation-rather than requiring a completely clear showing that the alternative is such a violation-indicates some measure of willingness to give public employers margin for error. Clearly, the Court does not see every pub-lic employer as bent on subverting the law, and the institutional reading of Ricci does not require such a dim view. It requires only that courts see a greater need for checking public employers than there is for checking courts.
136. For one excellent modern distillation of this idea in the Supreme Court's jurisprudence, relying partly on James Madison, see Hamdi v. Rumsfeld, 542 U.S. 507, 545 (2004) (Souter, J., concurring in part, dissenting in part, and concurring in the judgment) ("For reasons of inescapable human nature, the branch of the Government asked to counter a serious threat is not the branch on which to rest the Nation's entire reliance in striking the balance between the will to win and the cost in liberty on the way to victory; the responsibility for security will naturally amplify the claim that security legitimately raises. A reasonable balance is more likely to be reached on the judgment of a different branch, just as Madison said in remarking that 'the constant aim is to divide and arrange the several offices in such a manner as that each may be a check on the other-that the private inter-est of every individual may be a sentinel over the public rights.' ") (quoting The Federalist No. 51 (James Madison) (J. Cooke ed. 1961)).
137. I thank Trevor Morrison for suggesting this example.
138. 42 U.S.C. § 2000e-2(k)(1)(A)(i) (2006).
139. Id.
140. 129 S. Ct. at 2678 (noting and rejecting New Haven's assertion that the promotion test was not job related and consistent with business necessity).
141. Or, more broadly, perhaps the Court reallocated the burden not because of any particular sense it had about this case but because courts have been informally reallocating that burden as a matter of course for years, partly in response to the shift in incentives here described. See Selmi, supra note 125, at 749.
142. 42 U.S.C. § 2000e-2(k)(1)(A)(i) (stating that a violation of Title VII is established when a statistically disparate impact is shown and "the respondent fails to demonstrate that the challenged practice is job related for the position in question and consistent with business necessity") (emphasis added).
143. See, e.g., Larry D. Kramer, Foreword: We the Court, 115 Harv. L. Rev. 4, 146-47 (2001).
144. See Primus, supra note 14, at 539-44.
145. See Parents Involved in Cmty. Schs. v. Seattle Sch. Dist. No. 1, 551 U.S. 701, 787-89 (2007) (Kennedy, J., concurring in part and concurring in the judgment).
146. See Barry Friedman, The Will of the People 361 (2009).
147. See Hopwood v. Texas, 78 F.3d 932 (5th Cir. 1996).
148. Brief for the United States as Amicus Curiae Supporting Petitioner at 14-18, Grutter v. Bollinger, 539 U.S. 306 (2003) (No. 02-241).
149. See Friedman, supra note 146, at 340-41 (quoting activists who hold this view).
150. See, e.g., Ayres & Foster, supra note 24, at 518.
151. See, e.g., Parents Involved in Cmty. Schs. v. Seattle Sch. Dist. No. 1, 551 U.S. 701, 787-89 (2007) (Kennedy, J., concurring in part and concurring in the judgment); City of Richmond v. J.A. Croson Co., 488 U.S. 469, 507 (1989) (plurality opinion).
152. 100 U.S. 303, 308 (1880) (describing the practice of excluding blacks from juries as "practically a brand upon them . . . an assertion of their inferiority").
153. 347 U.S. 483, 494 (1954) (explaining that legal segregation was "usually interpreted as denoting the inferiority of the negro group").
154. Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 228-29 (1995) (arguing that racial classifications, even when made with "good intentions," raise equal protection problems because they will be perceived to rest on stigmatizing assumptions about the benefited groups); id. at 241 (Thomas, J., concurring in part and concurring in the judgment) ("So-called 'benign' discrimina-tion teaches many that because of chronic and apparently immutable handicaps, minorities cannot compete with them without their patronizing indulgence. Inevitably, such programs engender atti-tudes of superiority or, alternatively, provoke resentment among those who believe that they have been wronged by the government's use of race. These programs stamp minorities with a badge of inferiority . . . ."); Shaw v. Reno, 509 U.S. 630, 643 (1993) (explaining that legal classifications by race "threaten to stigmatize individuals by reason of their membership in a racial group"); Croson, 488 U.S. at 493-94 (plurality opinion) (focusing on the danger of stigmatic harm resulting from racial classifications).
155. See, e.g., Donald R. Kinder & Lynn M. Sanders, Divided by Color: Racial Politics and Democratic Ideals 295-303 (1996) (using public opinion data to demonstrate the divide between the ways that whites and blacks perceive the meanings of legal policies).
156. See, e.g., Grutter v. Bollinger, 539 U.S. 306, 343 (2003) (expressing the hope that race-conscious policies necessary in 2003 would not be necessary in the future); Croson, 488 U.S. at 495 (plurality opinion) (stating that equal protection should be construed so as to diminish the relevance of race in American life over time).
157. See Christopher L. Eisgruber, Democracy, Majoritarianism, and Racial Equality: A Response to Professor Karlan, 50 Vand. L. Rev. 347, 355-56 (1997).
158. The race-conscious electoral districting at issue in cases like Shaw v. Reno, 509 U.S. 630 (1993), may be an example: it is notoriously difficult to identify the determinate individual victims of such practices. See generally Samuel Issacharoff & Pamela S. Karlan, Standing and Misunder-standing in Voting Rights Law, 111 Harv. L. Rev. 2276 (1998). I thank Nathaniel Persily for pressing this point.
159. See, e.g., Ricci, 129 S. Ct. at 2678 (explaining that the municipal consultant entrusted with designing the test made sure that "minorities were overrepresented" among the people designing the test).
160. See Ricci, 129 S. Ct. at 2677 (characterizing each test-taker's interest in having the test results applied as originally planned as a "legitimate expectation not to be judged on the basis of race"); Petitioners' Brief on the Merits at 2, Ricci, 129 S. Ct. 2658 (No. 07-1428) ("Our Constitution envisions a society in which race does not matter and individuals are judged on the strength of their character."); id. at 3 ("Petitioners qualified for promotion under a race-blind, merit-selection proc-ess.").
161. Ricci v. DeStefano, 129 S. Ct. 2658, 2676 (2009) (emphasis added). From a different perspective, it is misleading to say that the city acted in "sole" reliance on race-based statistics. If one credits the city's account, it acted on race-based statistics in combination with its understanding of its legal obligations under federal statute. But this point may not affect what the firefighters "saw" from their own perspective.
162. See In re Employment Discrimination Litig., 198 F.3d 1305, 1315-16 (11th Cir. 1999) (explaining remedies).
163. See id.
164. To sum up: On the general reading of Ricci, the compliance interest might underwrite a constitutional defense for public employers but not private ones. On the institutional reading of Ricci, the compliance interest might again offer a defense for public employers, but it is hard to imagine a court that is attracted to the institutional reading of Ricci also being willing to credit the idea that the compliance interest is compelling, because the two stances imply sharply different attitudes toward public employers. On the visible-victims reading, the compliance interest might fail narrow-tailoring analysis. See infra notes 179-181 and accompanying text.
165. Primus, supra note 14, at 520-21.
166. The same is true if the challenged practice has a disparate racial impact and is justified as a matter of business necessity but the plaintiffs demonstrate the existence of an alternative employment practice that is equally good at meeting the business's economic needs and the employer refuses to adopt that alternative.
167. See Primus, supra note 14, at 520-21.
168. City of Richmond v. J.A. Croson Co., 488 U.S. 469, 498-99 (1989) (rejecting the idea that an attempt to redress societal discrimination can rise to the level of compelling interest).
169. See Ricci v. DeStefano, 129 S. Ct. 2658, 2682 (2009) (Scalia, J., concurring) (implying this view).
170. Selmi, supra note 125, at 716, 768-69.
171. 129 S. Ct. at 2678-79.
172. That said, the lack of a compelling interest defense would be much less damaging to disparate impact doctrine if the Court adopts the institutional or visible-victims reading of Ricci than if the general reading prevails.
173. Bush v. Vera, 517 U.S. 952, 990 (1996) (O'Connor, J., concurring); id. at 1033 (Stevens, J., dissenting, joined by Ginsburg & Breyer, JJ.); id. at 1046 (Souter, J., dissenting, joined by Ginsburg & Breyer, JJ.).
174. League of United Latin Am. Citizens v. Perry, 548 U.S. 399, 475 n.12 (2006) (Stevens, J., concurring in part and dissenting in part, joined by Breyer, J.); League of United Latin Am. Citizens, 548 U.S. at 485 n.2 (Souter, J., concurring in part and dissenting in part, joined by Ginsburg, J.); id. at 518-19 (Scalia, J., concurring in the judgment in part and dissenting in part, joined by Roberts, C.J., Thomas & Alito, JJ.).
175. See, e.g., City of Boerne v. Flores, 521 U.S. 507 (1997) (striking down the Religious Freedom Restoration Act).
176. See League of United Latin Am. Citizens, 548 U.S. at 518-19 (Scalia, J., concurring in the judgment in part and dissenting in part) (explaining this rationale). For further discussion and some trenchant criticism of Justice Scalia's reasoning, see Nathaniel Persily, Strict in Theory, Loopy in Fact, 105 Mich. L. Rev. First Impressions 43, 44-46 (2006). To date, no Supreme Court case squarely holds that compliance with a federal antidiscrimination statute constitutes a compelling interest for a state or local official. But the dicta of eight Justices-seven of them still sitting- seems a pretty good indication of how the Court would approach the question.
177. First, the tendency toward judicial abstention sometimes runs particularly strong in vot-ing and election cases, and recognizing a compelling interest in compliance with the Voting Rights Act is a way of leaving more of that sphere to the ordering of other institutions. See, e.g., Ellen Katz, Withdrawal: The Roberts Court and the Retreat from Election Law, 93 Minn. L. Rev. 1615 (2009). There is no parallel rubric of extraordinary deference in employment law. Second, there is a long history of judicial skepticism toward Title VII's disparate impact doctrine that is not matched by anything in the history of the Voting Rights Act. Title VII as a whole may have a sacred status similar to that of the Voting Rights Act, but the disparate impact prong of Title VII has never much shared in that status. See Selmi, supra note 125. Recognizing a compelling interest in compliance with a statute that is widely regarded as sacred may be much easier than recognizing a compelling interest in compliance with a doctrine that many judges have at best tolerated for many years.
178. The Court in Ricci showed some sensitivity to the importance of giving employers some room to maneuver when facing a partly analogous compliance dilemma at the statutory level: Ricci holds that an employer must have a strong basis in evidence for believing that one of its practices violates Title VII's disparate impact doctrine before it can avail itself of an exception to disparate treatment doctrine, but it does not hold that the employer must actually have committed a disparate impact violation. See Ricci v. DeStefano, 129 S. Ct. 2658, 2675 (2009). That margin of difference reflects reluctance to subject public employers to situations in which a good faith desire to comply with antidiscrimination law will predictably lead to other antidiscrimination violations. Of course, the dilemmas are not fully analogous. One involves two pieces of a single statute, and the other involves a statute and the Constitution; one involves the fine line between two doctrines, and the other involves two doctrines that might actually demand conflicting behaviors.
179. The limitation to public employers here is simple at first glance, but on second look it invites a trip down a particularly dark doctrinal rabbit hole. Given that the entire possibility of a successful compelling interest defense based on the compliance dilemma is limited, it may not be worth readers' time and effort to work this puzzle through to the end. But for those who are so in-clined, here we go. At one level, it would seem that public employers are the only employers who could avail themselves of a compliance-based compelling interest defense-or, indeed, of any com-pelling interest defense. Private employers, not being state actors, are not subject to constitutional claims, and a party that cannot be sued on a constitutional claim is not a party that can raise a consti-tutional defense, or even a party that would want to. But matters are not so simple. Imagine a case in which a private employer takes some action necessary to comply with Title VII's disparate impact doctrine and in consequence is sued for disparate treatment, also under Title VII. (That is, imagine a case just like Ricci, but with two variations: the employer is private rather than municipal, and the employer's action was uncontroversially required by Title VII's provisions on disparate impact.) The employer, citing Ricci, defends on the ground that an action required by Title VII's disparate impact provisions cannot be a violation of Title VII's prohibition on disparate treatment. In response, the plaintiff argues that the disparate impact provisions are unconstitutional. That move, if successful, would deny the private employer its proffered defense. So the private employer might choose to defend the constitutionality of the disparate impact provisions-or at least its own compliance with those provisions. At this point, one might be tempted to say that the private employer would be in the same position as a public employer making the compliance-interest argument and should be entitled to its protection on the same terms: the private employer, like a public one, faces a nasty dilemma, one in which a lack of clarity in the law forces him to choose between violating two dif-ferent demands of antidiscrimination law. But the situation is not fully analogous. The idea that a public official's compliance interest rises to the level of the compelling is, after all, partly founded on special solicitude for public officials. In the end, the private employer's situation is no different from any situation in which uncertainty in the law makes it hard for some party to know how to escape liability. There are appropriate canons of construction to apply in such cases: concerns about lenity and notice and vagueness all come to mind. But to make all such cases involving constitu-tional law into sources of compelling interest arguments is to work an unnecessary universalization of an exceptional rule.
180. Parents Involved in Cmty. Schs. v. Seattle Sch. District No. 1, 551 U.S. 701, 788-89 (2007) (Kennedy, J., concurring).
181. The analysis above proceeds as if narrow tailoring is essentially a balancing test: the question is whether fulfilling the compelling interest is worth the required costs. To be sure, that is not the only way that narrow tailoring analysis operates: it also sometimes asks whether the measure taken is strictly required for achieving the end. The present narrow tailoring question would be harder, and more complex, on that model. But a court that read the Ricci premise in terms of visible victims would probably opt for the balancing model simply because of the strength of its concern that the cost of creating such victims was too great to justify a compelling interest argument.
182. See Nw. Austin Mun. Util. Dist. No. 1 v. Holder, 129 S. Ct. 2504 (2009) (upholding section 5 of the Voting Rights Act after showcasing powerful reasons for considering it unconstitutional).
183. I have tried to sort out such complexities in another place. Primus, supra note 32.
184. See Civil Rights Cases, 109 U.S. 3 (1883).
185. Ricci v. DeStefano, 129 S. Ct. 2658, 2674 (2009).
186. See, e.g., Johnson v. Transp. Agency, Santa Clara County, 480 U.S. 616 (1987); Local No. 93, Int'l Ass'n of Firefighters v. City of Cleveland, 478 U.S. 501, 515 (1986).
187. One might wonder whether the conclusion that voluntary compliance is the greatest threat to the doctrine's continued constitutionality is in tension with the possibility, discussed above, that compliance could be regarded as a compelling interest. It is not. The underlying dynamic here is the conflict between the desire to give public officials room to maneuver and the aversion to visible victims, and my assumption throughout is that the latter force is more powerful. Thus, I earlier concluded that the compliance interest probably could not succeed in underwriting a constitutional defense of a public employer whose compliance created visible victims, even if it might shield a public employer whose compliance avoided that result. See supra Section III.B. Here, I am similarly contending that compliance that produces visible victims might provoke a generally negative view of disparate impact doctrine.
188. Ricci, 129 S. Ct. at 2683 (Scalia, J., concurring).
189. One case still in litigation that might fit this bill is United States v. City of New York, 637 F. Supp. 2d 77, at 82-83 (E.D.N.Y. 2009) (granting summary judgment against the New York City Fire Department in a Title VII suit alleging that a written examination for selecting entry-level fire-fighters had an unlawfully disparate impact on black and Hispanic applicants). To be sure, the New York City Fire Department is in many ways a sympathetic litigant. But the facts of this case show the Department in a poor light. That the lawsuit was commenced by the Department of Justice under the Bush Administration suggests that the case lends itself to mainstream intuitions about improper discrimination.
190. The Justices themselves have substantial agency to choose the ground through the certio-rari process. It follows that if the Court were a unified decisionmaker with a clear prior view of the constitutional question, other people's strategic behavior might be relatively unimportant. The Court would simply deny review of cases raising the issue in a posture unfavorable for reaching its desired result and wait for a better vehicle. But the process is not necessarily this simple, because the Court is composed of nine different decisionmakers who may see the issue differently. Some probably have already formed the view that Title VII's disparate impact doctrine should be held unconstitu-tional across the board, and some have probably already formed the view that the doctrine should be upheld, and others may be still working through the question. Given the Rule of Four for granting Supreme Court review, one could accordingly imagine four Justices with a clear view forcing their colleagues to confront the constitutional question in the setting most favorable for their own pre-ferred perspective. Other permutations are also possible.