Microsoft Word - CJRL Sharon F17 NEUTRAL IN NAME: ROTHE, THE ERROR OF ANTICLASSIFICATION, AND THE STATE OF RACE- NEUTRAL MEANS William Sharon1 This Note discusses the language state-actors use to create affirmative action programs, and the methods courts employ to determine their constitutionality. It describes the context and history of affirmative action jurisprudence, and explains the anticlassification method, the antisubordination method, and the former’s influence on the current tiered approach. This Note then discusses a 2016 D.C. Circuit decision, Rothe v. United States Department of Defense, where the court of appeals held that a statutory preference awarded to people “who ha[ve] been subjected to racial or ethnic prejudice” does not contain a racial classification. In so ruling, the court illustrated one reason why the anticlassification approach to affirmative action jurisprudence is untenable. While anticlassification is already an antiquated method of interpretation, this Note contends that, in light of Rothe, it is an unworkable one as well. I. INTRODUCTION .......................................................... 176 A. Rothe ............................................................... 177 B. Important Terms ............................................ 179 II. INTERPRETIVE METHODS IN ANTIDISCRIMINATION LAW: ANTICLASSIFICATION VS. ANTISUBORDINATION ................................................................................... 181 1 Wil Sharon is a J. D. Candidate 2018, Columbia Law School; B. A. 2015, Ohio State University. The author is grateful to Professor Kent Greenawalt for his advice throughout the drafting process, to Professor Jessica Bulman-Pozen for her feedback, and to the staff of the Columbia Journal of Race and Law for their edits. The author would also like to thank Lisa Sharon for reading drafts and discussing issues as they arose. 176 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 8:1 III. AFFIRMATIVE ACTION AND DISPARATE IMPACT IN THE CONSTITUTIONAL CONTEXT .................................................. 185 A. Affirmative Action .......................................... 186 B. Disparate Impact ............................................ 189 C. Race-Neutral Affirmative Action ................... 191 IV. THE CASE: ROTHE V. UNITED STATES DEPT. OF DEFENSE ............................................................................... 193 A. The Opinion .................................................... 193 B. The Dissent ..................................................... 197 V. THE IMPLICATIONS OF ROTHE AND THE PROBLEMS WITH ANTICLASSIFICATION ................................................... 199 A. The Practical Implications of the Rothe Reasoning ................................................................. 200 B. Rothe’s Implications for the Anticlassification Principle .................................................................... 203 VI. CONCLUSION ............................................................. 207 I. INTRODUCTION In the late 1970’s, equal protection law faced a new challenge.2 With affirmative action in its nascent stage, the Supreme Court considered whether to subject programs intended to benefit historically subjugated racial groups to the same suspicion it reserved for laws designed to oppress minorities.3 The Court opted for this heightened scrutiny.4 Since then, many affirmative action programs have failed to clear this hurdle.5 But state actors seeing the value of 2 See Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265 (1978). 3 Id. at 290. 4 Id. 5 See, e.g., Adarand Constructors, Inc. v. Pena, 515 U.S. 200 (1995); City of Richmond v. J.A. Croson Co., 488 U.S. 469 (1989); id. No. 4:175] NEUTRAL IN NAME 177 affirmative action have continued to strive for methods acceptable under this form of equal protection analysis.6 On June 23, 2016, the Supreme Court held that the University of Texas at Austin’s partially race-based admissions program did not violate the equal protection rights of a white person whose application the University denied.7 This was the second time the Court encountered the case, Fisher v. University of Texas at Austin, and the result was a victory for affirmative action.8 In the wake of Fisher, the U.S. Court of Appeals for the D.C. Circuit considered another case involving the issue of racial preferences.9 The case, Rothe v. United States Department of Defense, concerned a law allowing the Small Business Administration (“SBA”), when awarding government contracts, to favor businesses whose owners had “been subjected to racial or ethnic prejudice or cultural bias.”10 While, in Fisher, the Court explained the circumstances under which a university would be allowed to classify based on race in its admissions process, the court in Rothe bypassed that determination by deciding that the statute in question did not contain a racial classification. A. Rothe Rothe required the court to determine the constitutionality of 15 U.S.C. § 637(a),11 which authorizes the SBA to contract with federal agencies which can then subcontract to eligible small businesses.12 The statute provides that eligible businesses include those majority- owned by “socially and economically disadvantaged” individuals.13 § 637(a)(5) defines a socially disadvantaged 6 Fisher v. Univ. of Tex., 136 S. Ct. 2198, 2215 (2016). 7 Id. at 2215. 8 Id. 9 Rothe Dev., Inc. v. United States Dep’t of Def., 836 F.3d 57 (D.C. Cir. 2016). 10 Id. at 61. 11 15 U.S.C. § 637. 12 15 U.S.C. § 637(a)(5). 13 15 U.S.C. § 637(a)(1)(B). 178 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 8:1 individual as someone “who has been subjected to racial or ethnic prejudice or cultural bias because of their identity as a member of a group without regard to their individual qualities.”14 The plaintiff, Rothe, was a small computer-service business in Texas.15 It regularly bid on Defense Department contracts, including those specified by the statute.16 Rothe did not claim to be majority-owned by people who had experienced racial or ethnic prejudice or cultural bias.17 Thus, it argued, it could not compete for government contracts on equal footing with businesses awarded the preference.18 It challenged the law19 as violating its equal protection rights under the Due Process Clause of the Fifth Amendment,20 contending that the statute contained an unconstitutional racial classification.21 The district court granted the government’s motion for summary judgment, holding that the statute was constitutional.22 On appeal, the D.C. Circuit affirmed, reasoning that the law contained no facially race-based classification.23 Therefore, the court declined to apply “strict scrutiny,” instead reviewing the law under the “rational basis” standard.24 As that analysis is significantly more deferential to Congress, § 637(a) easily passed muster.25 14 15 U.S.C. § 637(a)(5). 15 Petition for Writ of Certiorari at 6, Rothe Dev. v. U.S. Dep’t. of Def., 836 F.3d 57 (D.C. Cir. 2016) (No. 16-1239). 16 836 F.3d at 61. 17 Id. at 62. 18 Petition for Writ of Certiorari, supra note 15. 19 Because Rothe did not challenge the SBA regulations, which apply § 637 in a racially classificatory manner, the court limited its analysis to the specific language of the statute. Id. at 62. 20 836 F.3d at 61. 21 Id. 22 Rothe Dev., Inc. v. Dep’t of Def., 107 F. Supp. 3d 183, 213 (D.D.C. 2015). 23 836 F.3d at 63. 24 Id. 25 Id. No. 4:175] NEUTRAL IN NAME 179 On October 16, 2017, the Supreme Court denied Rothe’s petition for certiorari.26 By declining to hear the case, the Court left unresolved questions concerning the possibilities stemming from the D.C. Circuit’s decision. This Note will address some of those questions and situate the Rothe reasoning within the context of affirmative action jurisprudence. B. Important Terms Because some of the important terms in this Note appear in different contexts, clarification is necessary. A decision, law, or plan containing a “racial classification” is one that explicitly identifies race as a factor for special treatment.27 Regardless of the purpose of the decision, law, or plan, if it requires different treatment of people due to their membership in a racial group, it has a facially race- based classification.28 For example, separating prison inmates by race constitutes a system of racial classifications, even when purportedly undertaken to prevent violence.29 A law or state action that is “race-neutral” is one that does not contain an explicit race-based classification.30 Nevertheless, it might have a race-based effect, and, as this Note will discuss, may even be intended to achieve that effect. An action, law, or plan that is “race-conscious” is one where race is a motivating factor in the enactment process.31 An action can be race-conscious whether or not it contains a 26 Rothe Dev., Inc. v. Dep’t of Def., No. 16-1239, 2017 WL 1375832, at *1 (U.S. Oct. 16, 2017). 27 See generally, Johnson v. California, 543 U.S. 499 (2005) (explaining that racial distinctions constitute suspect classifications regardless of whether they burden all races equally). 28 Id. at 499. 29 Id. 30 See generally Reva B. Siegel, Race-Conscious but Race-Neutral: The Constitutionality of Disparate Impact in the Roberts Court, 66 ALA. L. REV. 653 (2015) (explaining that facially neutral laws do not explicitly categorize based on race). 31 Id. 180 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 8:1 racial classification. Thus, a law can be both race-neutral and race-conscious.32 Of course, a facially race-based classification will almost invariably be race-conscious. Race-consciousness is subjective to the decision- maker. Drawing school districts with the intent of giving black children a better chance of attending good public schools would be race-conscious, even if the plan did not explicitly require that black people be given treatment different from people of other races.33 A plan intended to make Latino/a people less likely to be selected for juries would be race-conscious,34 even if the plan were facially race-neutral. But a plan resulting in fewer Latino/a people being selected for juries would not be race-conscious if the decision-maker did not consider race when enacting it.35 Historically, courts addressing equal protection questions have begun by attempting to discern “illicit” classifications.36 Those categories typically include race.37 When a court finds a proscribed classification, it will accord less deference to the decision-maker. Racial classifications typically trigger “strict scrutiny.” In order for a law or decision to survive that analysis, it must be narrowly tailored to achieve a compelling governmental interest.38 This Note argues that determining whether laws contain racial classifications should not be the chief concern in evaluating equal protection claims. Part I describes two 32 Id. 33 See, e.g., Parents Involved in Cmty. Sch. v. Seattle Sch. Dist. No. 1, 551 U.S. 789 (Kennedy, J., concurring in part and concurring in the judgment). 34 To the extent that equal protection jurisprudence conflates ethnic and racial characteristics. See, e.g., Gloria Sandrino-Glasser, Los Confundidos: De-Conflating Latinos/as’ Race and Ethnicity, 19 CHICANO- LATINO L. REV. 69, 138 (1998) (arguing that, in a case involving exclusion of Latino people from juries, the Court has conflated ethnic neutrality with racial neutrality). 35 See, e.g., Hernandez v. New York, 500 U.S. 352, 353 (1991). 36 Kent Greenawalt, Judicial Scrutiny of “Benign” Racial Preference in Law School Admissions, 75 COLUM. L. REV. 559, 563 (1975). 37 See generally id. 38 Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 227 (1995). No. 4:175] NEUTRAL IN NAME 181 interpretive methods of antidiscrimination law. Part II provides some of the constitutional background of affirmative action and disparate impact. Part III explains the holding in Rothe. Part IV describes the implications of the Rothe reasoning for affirmative action and the interpretive methods. As cases like Rothe demonstrate, the distinction between race-neutral-but-race-conscious and facially classificatory laws is usually one of form and not substance. Applying strict scrutiny to affirmative action plans because they classify based on race is a misguided method of legal interpretation and loses its purported value when race- neutral alternatives are neutral only in name. This Note will attempt to show that the classification-oriented approach creates a legal conundrum; it results either in courts striking down benevolent plans designed to aid minority groups, or, in light of cases like Rothe, it forces legislatures to contort statutory language in order to skirt an antiquated interpretive method. II. INTERPRETIVE METHODS IN ANTIDISCRIMINATION LAW: ANTICLASSIFICATION VS. ANTISUBORDINATION Since the enactment of the Fourteenth Amendment, several interpretive perspectives have battled for control of equal protection law. The two predominant theories are the Antisubordination principle and the Anticlassification principle. The former refers to the idea that a law may not “‘aggravate’ or ‘perpetuate’ ‘the subordinate status of a specially disadvantaged group.’”39 Instead, laws should attempt to rectify the historically secondary position of oppressed minorities.40 Anticlassification, on the other hand, is the idea that a law may not classify people based on 39 Jack M. Balkin & Reva B. Siegel, The American Civil Rights Tradition: Anticlassification or Antisubordination?, 58 U. MIAMI L. REV. 9, 10 (2003) (quoting Owen M. Fiss, Groups and the Equal Protection Clause, 5 PHIL. & PUB. AFF. 107, 157 (1976)). 40 Id. at 9. 182 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 8:1 certain characteristics (e.g., race).41 Essentially, the government must avoid grouping people on the basis of forbidden categories, regardless of whether the impetus is invidious or benign.42 Antisubordination and anticlassification frequently coincide. A law requiring black people to attend segregated schools would violate both.43 On the one hand, it would expressly classify based on race and therefore run afoul of the anticlassification principle. On the other, it would subjugate a historically disadvantaged minority group, thus violating the tenets of antisubordination. But as a practical matter, the concepts conflict in two scenarios. The first point of tension exists when laws or policies that do not classify groups based on proscribed characteristics nonetheless disproportionately affect people belonging to disadvantaged minorities. For example, a law prohibiting people living in a particular neighborhood from voting would be facially race-neutral. But if that neighborhood happened to be predominantly black, the law would disparately impact black people. Thus, the statute would not implicate the anticlassification principle, but it would run afoul of the doctrine of antisubordination.44 The second discord occurs when laws or policies classify based on proscribed characteristics, but benefit 41 Id. at 10. 42 Id. 43 Compare Parents Involved in Cmty. Sch. v. Seattle Sch. Dist. No. 1, 551 U.S. 746 (2007) (Roberts, C.J., arguing that the Court in Brown v. Bd. of Educ., 349 U.S. 294 (1955) struck down a law requiring black children to attend segregated schools on anticlassification principles. “[T]he Court held that segregation deprived black children of equal educational opportunities regardless of whether school facilities and other tangible factors were equal, because the classification and separation themselves denoted inferiority,” with Parents Involved, 551 U.S. at 803 (Breyer, J., dissenting) (arguing, based on antisubordination ideals, that Brown promised racially integrated education and encouraged local authorities to undertake efforts to bring about that aim). 44 But see Fiss, 5 PHIL. & PUB. AFF. at 171 (arguing that the [anticlassification] principle could conceivably be construed to apply to disparate impact cases). No. 4:175] NEUTRAL IN NAME 183 minority groups. The most common form of this phenomenon is affirmative action. Where a program awards a preference based on racial status, it violates the anticlassification principle. The anticlassification theorist would disapprove of explicit affirmative action based on the theory that certain classifications are inherently harmful.45 The issue, from that perspective, would concern means and not ends. Justice Thomas discussed affirmative action from an anticlassificationist angle in his concurrence in Adarand Constructors Inc. v. Pena, where the Court struck down a statute awarding contracting preference to racial minorities; “That these programs may have been motivated, in part, by good intentions cannot provide refuge from the principle that under our Constitution, the government may not make distinctions on the basis of race.”46 Proponents of antisubordination, conversely, would likely accept explicit affirmative action as presumptively valid.47 Laws and policies designed to benefit historically disadvantaged minorities would not raise concerns of aggravating subjugation or perpetuating perceived inferiority. Dissenting in Adarand, Justice Stevens argued that “[t]here is no moral or constitutional equivalence between a policy that is designed to perpetuate a caste system and one that seeks to eradicate racial subordination.”48 Essentially, the concern under the Equal Protection Clause should be the motivation or effect of a law, not whether it names racial groups. Since the 1970s, the Supreme Court has applied a relatively anticlassification-oriented approach to equal 45 See Balkin supra note 39 at 11 (noting that if the Court interpreted Brown to invalidate segregation in light of the anticlassification principle, affirmative action would be presumptively unconstitutional). 46 515 U.S. at 240 (Thomas, J., concurring in part and concurring in the judgment). 47 See id. (contending that a paternalistic impetus should not exempt racial classifications from strict scrutiny). 48 515 U.S. at 243 (Stevens, J., dissenting). 184 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 8:1 protection law.49 Affirmative action plans have survived application of this doctrine, but under limited circumstances.50 Many of the explicit affirmative action decisions have turned on the availability of “race-neutral alternatives.”51 This inquiry requires a reviewing court to assess whether the positive result sought to be achieved through racial classification could be achieved through measures that do not refer explicitly to race. For example, a race-neutral alternative to giving school admissions preferences to members of minority racial groups might be “drawing public school attendance zones with general recognition of the demographics of the neighborhood” in order to maintain school integration.52 While the Court has upheld laws and policies with facially race-based classifications as “last resorts” to achieving diversity,53 it has rejected them absent “serious, good faith consideration of workable race-neutral alternatives.”54 When the Court deems such options available, it is generally disinclined to accept even benign racial classifications. As the Court noted in Fisher, Though “[n]arrow tailoring does not require exhaustion of every conceivable race-neutral alternative” or “require a university to choose between maintaining a reputation for excellence [and] fulfilling a commitment to provide educational opportunities to members of all racial groups,” it does impose “on the university the ultimate burden of demonstrating” that “race-neutral 49 Balkin at 10. 50 See, e.g., Grutter v. Bollinger, 539 U.S. 306 (2003). 51 See, e.g., Parents Involved, 551 U.S. at 789 (Kennedy, J., concurring in part and concurring in the judgment). 52 Id. 53 Id. at 790. 54 Id. at 735 (Roberts, C.J., writing for the Court). No. 4:175] NEUTRAL IN NAME 185 alternatives” that are both “available” and “workable” “do not suffice.”55 But determining whether a statute contains a racial classification is not always easy. To begin with, “‘express racial classification’ is not a self-defining term.”56 Intuition might suggest that it refers to any law explicitly naming a racial group. But courts have applied the anticlassification principle inconsistently, often relying on subjective impressions.57 The decision in Rothe is the latest example of competing antidiscrimination law principles. On one hand, the D.C. Circuit applied an anticlassificationist reading to the statute in question. It determined that 15 U.S.C. § 637(a)(5) lacked a racial classification on its face, and that the law therefore warranted relatively lenient scrutiny. But in refusing to discern a classification, the majority implicitly relied on the antisubordination principle. III. AFFIRMATIVE ACTION AND DISPARATE IMPACT IN THE CONSTITUTIONAL CONTEXT Before moving to the implications of the court’s holding in Rothe, review of the legal history of affirmative action is important. Racial affirmative action is based on antisubordination reasoning.58 That is, it relies on racial parity not as a means but as an end. Treating people differently based on race is an acceptable method of pursuing ultimate equality. Frequently, this results in methods that are explicitly unequal.59 In that sense, antisubordination is a product of societal context. In a world without a history of 55 Fisher, 136 S. Ct. at 2208 (quoting Grutter, 539 U.S. at 339, and Fisher I, 570 U.S. at ––––, 133 S.Ct. at 2420). 56 Richard A. Primus, Equal Protection and Disparate Impact: Round Three, 117 HARV. L. REV. 493 (2003). 57 See Id. 58 Balkin, supra note 39, at 11. 59 See, e.g., Grutter, 539 U.S. 306; Parents Involved, 551 U.S. 701; Regents of Univ. of California v. Bakke, 438 U.S. 265 (1978). 186 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 8:1 racial inequality, there might be no need for the antisubordination principle. Without minority subjugation, there would be nothing from which to recover. But in a society where racial inequality is not only mired in the past but a fact of the present, the antisubordination principle is an important premise on which to base remedial efforts.60 A. Affirmative Action In the last thirty years, many industries have adopted affirmative action plans. Legislatures have enacted laws with facially race-based classifications in order to benefit minority groups.61 Schools have used race in their admissions criteria.62 One area where affirmative action has been particularly contentious is government contracting. In the 1989 case, City of Richmond v. J.A. Croson, the Supreme Court adopted a restrictive standard for affirmative action.63 In rejecting the City of Richmond’s plan setting aside 30 percent of public-works money for minority-owned businesses, the Court applied strict scrutiny to affirmative 60 The history of the Reconstruction Amendments supports an antisubordinationist view. To begin with, the drafters of the Equal Protection Clause were concerned primarily with protecting black people in their ability to exercise certain rights. See Michael J. Klarman, An Interpretive History of Modern Equal Protection, 90 MICH. L. REV. 213, 216 (1991). The purpose of the Fourteenth Amendment was not to proscribe racial classifications as such. Id. at 256. In fact, the Court did not create a presumption that racial classifications would be invalid under the Equal Protection Clause until it decided McLaughlin v. Florida in 1964 (379 U.S. 184, 192 (1964)), almost a century after the Amendment passed. In light of that history, many have argued that racial classifications should not be suspect when they are enacted to remedy the subjugation of minority racial groups (See, e.g., Gratz v. Bollinger, 539 U.S. 244, 298 (2003) (Souter, J., dissenting); Adarand, 515 U.S. at 243 (Stevens, J., dissenting); Joel K. Goldstein, Not Hearing History: A Critique of Chief Justice Roberts’s Reinterpretation of Brown, 69 OHIO ST. L.J. 791 (2008)). 61 See, e.g., Fullilove v. Klutznick, 448 U.S. 448, 448, (1980); Croson Co., 488 U.S. 469. 62 See, e.g., Bakke, 438 U.S. 265; Grutter, 539 U.S. 306. 63 488 U.S. at 511. No. 4:175] NEUTRAL IN NAME 187 action.64 It ruled that a remedial plan targeting the effects of “past societal discrimination” did not amount to a compelling interest that could pass muster under strict scrutiny.65 Instead, an affirmative action plan aimed at remedying discrimination would be valid only if the party defending the plan could demonstrate a “strong basis in evidence” that such remedial action was necessary.66 Such a showing required evidence of prima facie statutory or constitutional violations in the relevant industry and jurisdiction.67 This meant, essentially, that in order to create a facially race- based affirmative action plan, an employer or legislature needed to identify systemic disparate treatment within the industry.68 Seven years after J.A. Croson, the Court in Adarand affirmed the use of strict scrutiny review of facially race- based affirmative action plans.69 The Court rooted its holding in a desire to maintain consistency.70 Assuming all racial classifications to be inherently suspect, the Court chose to apply a uniform standard to distinguish justifiable 64 Id. at 493. 65 Id. at 505. 66 Id. at 500. 67 Id. 68 This requirement differs from the necessary showing under Title VII. Unlike the constitutional requirement announced in J.A. Croson, to enact affirmative action measures in accordance with Title VII, an employer or legislature need show only a “manifest imbalance in a traditionally segregated job category.” United Steelworkers of Am., AFL- CIO-CLC v. Weber, 443 U.S. 193, 197 (1979). The Court in Weber maintained that this disparity need not rise to the level of a prima facie statutory violation. In Johnson v. Transportation Agency, Santa Clara County., Cal., the Court clarified that standard: In some cases, of course, the manifest imbalance may be sufficiently egregious to establish a prima facie case. However, as long as there is a manifest imbalance, an employer may adopt a plan even where the disparity is not so striking, without being required to introduce the nonstatistical evidence of past discrimination demanded by the “prima facie” standard. . . . 480 U.S. 616, n. 11 (1987). 69 515 U.S. at 201. 70 Id. 188 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 8:1 classifications from invidious ones.71 The majority relied on the anticlassification principle, reasoning that expressly considering race is inherently dangerous and ought to require a compelling justification. Rather than judging classifications as invidious or benign and then conducting the requisite analysis, the Court purportedly applied strict scrutiny in order to differentiate invidious from benign classifications.72 Dissenting in Adarand, Justice Stevens argued that classificatory laws designed to harm minorities ought to receive a different analysis from those enacted to benefit protected groups.73 He explained that “a single standard that purports to equate remedial preferences with invidious discrimination cannot be defended in the name of ‘equal protection.’”74 In response to the majority’s claim that strict scrutiny should be used to make the initial distinction between invidious and benign, he compared the burden of discerning an illicit impetus to that of unearthing intentional discrimination in disparate impact cases.75 Distinguishing between invidious and benign classifications, he reasoned, would be no more difficult than differentiating between discriminatory and innocent motives, something the Court already did, without applying strict scrutiny, in its disparate impact line of cases.76 While the majority in Adarand adopted the anticlassification principle, the dissent championed antisubordination. Justice O’Connor, writing for the Court, endorsed the idea that racial classifications are inherently harmful.77 Justice Stevens, joined by Justice Ginsburg, 71 Id. at 228. 72 Id. 73 Id. at 243 (Stevens, J., dissenting). 74 Id. at 246 (Stevens, J., dissenting). 75 Adarand Constructors, Inc. v. Pena, 515 U.S. 245-46 (1995) (Stevens, J., dissenting). 76 Id. (Stevens, J., dissenting). 77 Id. at 234. No. 4:175] NEUTRAL IN NAME 189 worried about the effects of laws on racial minorities whether or not those statutes actually classified racial groups.78 But by adopting the strict scrutiny standard for all laws or plans containing racial classifications, the Court in Adarand heightened the requirements for constitutionality in the context of race-based affirmative action. Although the Court explained that application of strict scrutiny would not necessarily be fatal to a law or plan, the Court has since been reluctant to find racial classifications constitutional under that analysis.79 B. Disparate Impact In 1976, the Court decided the seminal constitutional disparate impact case, Washington v. Davis.80 The plaintiffs in that case were black people in the District of Columbia applying to be police officers. Their applications were rejected because they failed a written test.81 They sued to invalidate the test on the ground that it racially discriminated in violation of the Fifth Amendment.82 The test was facially race-neutral.83 Nothing about it expressly burdened black people more heavily than people of other races.84 In determining that the test was constitutional, the Court created a standard for facially neutral programs. The test would be invalid, it said, only if the plaintiffs could identify a racially discriminatory purpose.85 The Court qualified that standard by noting that a racially discriminatory purpose could be inferred from the disproportionate impact alone.86 But it ultimately concluded 78 Id. at 243 (Stevens, J., dissenting). 79 See, e.g., Peter J. Rubin, Reconnecting Doctrine and Purpose: A Comprehensive Approach to Strict Scrutiny After Adarand and Shaw, 149 U. PA. L. REV. 1, 27 (2000). 80 Washington v. Davis, 426 U.S. 229 (1976). 81 Id. at 233. 82 Id. at 234. 83 Id. at 235. 84 Id. 85 Id. at 240. 86 Washington v. Davis, 426 U.S. 241 (1976). 190 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 8:1 that its “cases [had] not embraced the proposition that a law or other official act, without regard to whether it reflect[ed] a racially discriminatory purpose, [would be] unconstitutional solely because it ha[d] a racially disproportionate impact.”87 This particular test, the Court said, sought to ascertain whether applicants had acquired a particular level of verbal skill, not to disproportionately keep black people off of the police force.88 Thus, the Court established that a state- imposed racial disparate impact, without more, was constitutional. The Court affirmed this standard three years later in Personnel Administrator of Massachusetts v. Feeney.89 It held that a law or other official act having a disparate impact would not necessarily be unconstitutional even if the legislature were aware of the disproportionate effect.90 The Court thus declined to apply the proposition that “a person intends the natural and foreseeable consequences of his voluntary actions.”91 In disparate impact cases under the Constitution, the Court ruled, a course of action would be invalid only if undertaken at least in part “because of, not merely in spite of” its adverse effects upon an identifiable group.92 The Court in Feeney noted that a facially neutral law would not violate equal protection absent an “invidious” purpose.93 In that regard, a facially neutral law designed to harm a minority racial group would be suspect. “[W]hen a neutral law has a disparate impact upon a group that has historically been the victim of discrimination, an unconstitutional purpose may still be at work.”94 This language implied that a facially neutral law intended to 87 Id. at 239 (emphasis added). 88 Id. at 245. 89 442 U.S. 256, 272 (1979). 90 Id. at 279. 91 Id. 92 Washington v. Davis, 426 U.S. 279 (1976). 93 Id. at 274. 94 Id. No. 4:175] NEUTRAL IN NAME 191 benefit a historically victimized group would be presumptively valid. The Court endorsed this theory in Texas Department of Housing & Community Affairs v. Inclusive Communities Project, Inc., where it recognized the existence of disparate impact claims under the Fair Housing Act.95 In that case, the Court noted that local housing authorities could “choose to foster diversity and combat racial isolation with race- neutral tools.”96 The Court accepted that racial diversity could be at least part of the motive for a facially neutral policy without that policy triggering strict scrutiny.97 Thus, it validated the contrapositive to Feeney. While the Court in Feeney held that a facially neutral law with an invidious purpose would be presumptively invalid, the Court in Texas Dept. of Housing explained that a similar law with a benign purpose would be presumptively valid, even if the disparate effects were intentional.98 C. Race-Neutral Affirmative Action The Davis/Feeney test applies in the context of affirmative action. In that setting, plans disproportionately benefitting minority racial groups are presumptively constitutional if they do not facially classify by race.99 The Court has referred to this sort of disparate-impact- affirmative-action as “race-neutral means.”100 Within that category, a benign policy is presumptively constitutional if it is facially race-neutral, even if it is motivated at least in part by race-consciousness.101 A school or employer can, for instance, create an admissions plan or application designed in part to accept more people belonging to minority racial 95 135 S. Ct. 2507 (2015). 96 Id. at 2525. 97 Id. 98 E.g., a statute drawing school attendance zones in light of neighborhood demographics. Id. at 2525. 99 See, e.g., Parents Involved, 551 U.S. at 789 (Kennedy, J., concurring in part and concurring in the judgment). 100 Id. at 790. 101 Id. at 789. 192 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 8:1 groups, as long as the plan does not expressly classify based on race.102 In fact, when a court applies strict scrutiny to an affirmative action plan that classifies based on race, one of the considerations at the “narrow tailoring” stage is whether the decision-maker considered race-neutral alternatives.103 A recent example, which the Court alluded to in Fisher,104 is the Texas Legislature’s adoption of a “Top Ten Percent Plan.”105 After the Fifth Circuit struck down the University of Texas at Austin’s facially race-based admissions program,106 the Texas Legislature enacted a policy whereby the University would admit any student from the top ten percent of his or her high school class.107 Because many Texas high schools had predominantly black student bodies, the plan resulted in the acceptance of a higher number of black students than would otherwise have been admitted.108 The Legislature did not deny that the plan was intended, at least in part, to have a racially-disparate effect.109 Nevertheless, the Court did not apply strict scrutiny analysis to the Top Ten Percent Plan.110 Rather, the Court treated the plan as a “race-neutral holistic review.”111 In that vein, the Court has suggested that “race- neutral” methods can still be race-conscious without triggering strict scrutiny. As one commenter notes, “The Court’s acceptance of the percent plan illustrates that government may act in race-conscious but facially neutral ways to promote equal opportunity, even where government 102 Id. 103 See Bakke, 438 U.S. at 357. 104 136 S. Ct. at 2213. 105 Tex. Educ. Code Ann. § 51.803. 106 Fisher v. Univ. of Tex. at Austin, 631 F.3d 213, 242 (5th Cir. 2011), vacated and remanded, 133 S. Ct. 2411, 186 L. Ed. 2d 474 (2013). 107 Fisher v. Univ. of Tex. at Austin, 133 S. Ct. 2411, 2416 (2013). 108 Id. 109 Fisher, 631 F.3d at 242. 110 Abigail Noel Fisher v. Univ. of Tex. at Austin, 136 S. Ct. 2198, 2213 (2016) [hereinafter Fisher II]. 111 Id. No. 4:175] NEUTRAL IN NAME 193 seeks to alter racial outcomes.”112 Justice Kennedy affirmed this idea in Parents Involved.113 Discussing hypothetical policies that would be race-neutral, he suggested that “[parties] are free to devise race-conscious measures to address [a lack of diversity] without . . . individual typing by race.”114 In sum, affirmative action is not constitutionally suspect if attempted through race-neutral means. But facially race-based classifications are inherently suspect. A public university policy automatically accepting fifteen black applicants every year would trigger strict scrutiny. Admitting anyone from the state in the top ten percent of his or her high school class would not, even if the purpose were partly to promote racial diversity on campus. But in some cases, distinguishing facially neutral policies from those containing racial classifications is not so easy. That determination was the issue in Rothe. IV. THE CASE: ROTHE V. UNITED STATES DEPT. OF DEFENSE A. The Opinion The plaintiff in Rothe challenged § 637(a) of the Small Business Act on the grounds that it contains a racial classification. It argued that the definition of “socially and economically disadvantaged” explicitly allocates a preference to certain racial groups. Rothe was not the first time a party had challenged § 637. A different section, § 637(d), was the focus of Adarand more than 20 years before.115 The Supreme Court in 112 Reva B. Siegel, Race-Conscious but Race-Neutral: The Constitutionality of Disparate Impact in the Roberts Court, 66 ALA. L. REV. 653, 674 (2015). 113 551 U.S. at 789 (Kennedy, J., concurring in part and concurring in the judgment). 114 Id. at 788-89 (Kennedy, J., concurring in part and concurring in the judgment). 115 515 U.S. at 207. 194 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 8:1 Adarand deemed that section unconstitutional.116 But the provision the D.C. Circuit addressed in Rothe differs slightly. The primary distinction between (d), which the Supreme Court considered in Adarand, and (a) which the D.C. Circuit dealt with in Rothe, was the fact that (d) expressly relied on the presumption that certain minority group members were socially and economically disadvantaged;117 “The contractor shall presume that socially and economically disadvantaged individuals include Black Americans, Hispanic Americans, Native Americans, Asian Pacific Americans, and other minorities, or any other individual found to be disadvantaged.”118 Section (a), on the other hand, contains no such language; “Whatever Congress’s reasons for directing private businesses to use race-based criteria under [§] (d)’s subcontracting clause, Congress authorized more nuanced implementation by the agency under [§] (a).”119 Nevertheless, the plaintiff in Rothe challenged § 637(a) as creating a similar racial presumption, and thus, a similarly suspect classification.120 In ruling that the statute is facially neutral, the court followed several lines of reasoning. First, the court determined that § 637(a)(5) envisions an “individual-based approach” rather than one reliant on group characteristics.121 The court noted that the statute could encompass people who were not part of minority racial groups, who had nonetheless experienced discrimination based on race, culture, or ethnicity.122 Moreover, the statute would not necessarily apply to people of minority groups who had not individually experienced racial discrimination.123 116 Id. at 204. 117 Id. at 207. 118 15 U.S.C. § 637(d)(3)(C)(ii) (2016). 119 Rothe, 836 F.3d at 69. 120 Id. at 61. 121 Id. at 64. 122 Id. 123 Id. No. 4:175] NEUTRAL IN NAME 195 The court distinguished the academic admissions policy in Regents of the University of California v. Bakke,124 where the University of California at Davis’ Medical School used a racial quota in its admissions process.125 In contrast, the Rothe court noted, § 637(a)(5) allows consideration of personal experience (e.g., subjection to racial prejudice) in gauging social disadvantage.126 Unlike Bakke, Rothe does not rely on race qua race, and therefore does not contain an express racial classification.127 Next, the Rothe court addressed § 637(a) from a disparate impact perspective.128 The court did not deny that, when it enacted the statute, Congress was aware of the racial effect § 637(a) would have.129 But Judge Pillard, writing for the majority, noted that congressional anticipation of disproportionate racial effects, absent a discriminatory purpose, does not condemn a law to strict scrutiny.130 “Mere foreseeability of racially disparate impact, without invidious purpose, does not trigger strict constitutional scrutiny.”131 In reaching that conclusion, the court applied the Davis/Feeney framework.132 Under Feeney, a facially neutral law triggers constitutional heightened scrutiny only if enacted “because of, not merely in spite of,” a discriminatory purpose.133 Addressing that point, the Rothe court adopted some of the reasoning from Texas Dept. of Housing and ruled that § 637(a) was not meant to discriminate invidiously, but to advance equal opportunity.134 Thus, even though the legislature may have intended § 637(a)’s disproportionate 124 Id. 125 438 U.S. at 276. 126 836 F.3d at 64. 127 Id. 128 Id. at 71-72. 129 Id. at 71. 130 Id. at 72. 131 Id. 132 Rothe Dev., Inc. v. United States Dep’t of Def., 836 F.3d 71-72 (D.C. Cir. 2016). 133 442 U.S. at 279 (internal quotations omitted). 134 836 F.3d at 72. 196 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 8:1 effect, that intent was not “discriminatory” in the sense the Court described in Feeney.135 The Rothe court also cited Justice Kennedy’s concurrence in Parents Involved.136 There, Justice Kennedy stressed that “[s]chool boards may pursue the goal of bringing together students of diverse backgrounds and races through [race-neutral] means.”137 Justice Kennedy rejected the application of strict scrutiny to those cases. He explained, “These mechanisms are race conscious but do not lead to different treatment based on a classification that tells each [person] he or she is to be defined by race, so it is unlikely any of them would demand strict scrutiny to be found permissible.”138 Finally, the Rothe court addressed prior affirmative action decisions.139 It quoted Justice Scalia’s concurring opinion in J.A. Croson, where he noted that a legislature can, for example, “adopt a preference for small businesses, or even for new businesses—which would make it easier for those previously excluded by discrimination to enter the field.”140 Differing effects along racial lines could be acceptable, the Rothe court determined, based on Scalia’s view that “[s]uch programs [could] well have racially disproportionate impact” and still be valid.141 The Rothe court also reiterated the controlling concurrence in Parents Involved, where Justice Kennedy pointed out several presumably valid race-neutral, yet disproportionately effective ways of pursuing diversity in education.142 Relying on the reasoning in those cases, the Rothe court determined 135 Id. 136 Id. 137 Parents Involved, 551 U.S. at 789 (Kennedy, J., concurring in part and concurring in the judgment). 138 Id. (Kennedy, J., concurring in part and concurring in the judgment). 139 Rothe, 836 F.3d at 72. 140 Id. (quoting J.A. Croson, 488 U.S. at 526). 141 Id. (quoting J.A. Croson, at id.). 142 Id. (citing Parents Involved, 551 U.S. at 789 (Kennedy, J., concurring in part and concurring in the judgment)). No. 4:175] NEUTRAL IN NAME 197 that § 637(a) constitutes a race-neutral alternative as envisioned by the Court in the affirmative action context.143 B. The Dissent The dissenting judge in Rothe argued that § 637(a) contains an express racial classification.144 In reaching that conclusion, she relied in part on paragraph (a)(8); “All determinations made pursuant to paragraph (5) with respect to whether a group has been subjected to prejudice or bias shall be made by the Administrator after consultation with the Associate Administrator for Minority Small Business and Capital Ownership Development.”145 To the dissent, the use of the word “group” in this provision suggested a presumption based on race, thereby undermining the view that the challenged provision applies to individuals.146 If a group had been subjected to racial or ethnic prejudice or cultural bias, then its members were presumptively socially disadvantaged.147 It did not, as the majority claimed, provide for an individual inquiry into whether each applicant met the definition separately. “The message is clear—groups suffer discrimination and therefore persons who are members of those groups are socially disadvantaged.”148 To support that view, the dissent pointed to the congressional findings section of the statute, which note (A) that the opportunity for full participation in our free enterprise system by socially and economically disadvantaged persons is essential if we are to obtain social and 143 Id. 144 Id. at 74 (Henderson, J., concurring in part and dissenting in part). 145 Rothe Dev., Inc. v. United States Dep’t of Def., 836 F.3d 76 (D.C. Cir. 2016) (Henderson, J., concurring in part and dissenting in part (quoting 15 U.S.C. § 637(a)(8))). 146 Id. (Henderson, J., concurring in part and dissenting in part). 147 Id. at 76-77 (Henderson, J., concurring in part and dissenting in part). 148 Id. at 76 (Henderson, J., concurring in part and dissenting in part). 198 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 8:1 economic equality for such persons and improve the functioning of our national economy; (B) that many such persons are socially disadvantaged because of their identification as members of certain groups that have suffered the effects of discriminatory practices or similar invidious circumstances over which they have no control; [and] (C) that such groups include, but are not limited to, Black Americans, Hispanic Americans, Native Americans, Indian tribes, Asian Pacific Americans, Native Hawaiian Organizations, and other minorities . . . .149 Thus, she argued, the statute creates a floor for participation in the § 637(a) program.150 It classifies by race, and then allows the Small Business Administration (“SBA”) to select from within that class.151 “[I]f not rebutted, the SBA must presume members of [racial groups and other minorities] are socially disadvantaged.”152 The majority dismissed these concerns, arguing that the congressional findings were not dispositive.153 There are many reasons Congress might have identified certain racial groups when announcing the policy behind the [§ 637(a)] program. Congress might have wanted to offer paradigmatic examples of the problem or to 149 Id. (Henderson, J., concurring in part and dissenting in part (quoting 15 U.S.C. § 631(f)(1)(A) – (C) (footnote omitted))). 150 Id. at 77 (Henderson, J., concurring in part and dissenting in part). 151 Rothe Dev., Inc. v. United States Dep’t of Def., 836 F.3d 77 (D.C. Cir. 2016) (Henderson, J., concurring in part and dissenting in part). 152 Id. at 78 (Henderson, J., concurring in part and dissenting in part). 153 Id. at 66. No. 4:175] NEUTRAL IN NAME 199 send a signal of responsiveness to Americans of minority backgrounds, many of whom felt they lacked a fair shot at the American dream.154 The court also rejected the idea that the findings section creates a presumption that racial minorities are socially disadvantaged.155 “Congress’s findings that individual business owners may have been unfairly subjected to race-based disadvantage do not . . . supplant the race- neutral definition of social disadvantage found in section 637(a)(5).”156 The case ultimately turned on whether § 637(a) does in fact presume that racial minorities are socially disadvantaged. While the Court in Adarand held that § 637(d) created such a presumption,157 the court in Rothe held that § 637(a) does not.158 Providing a benefit to a person who has experienced racial prejudice is acceptable; assuming that a member of a racial minority has experienced such prejudice by virtue of minority status is not. V. THE IMPLICATIONS OF ROTHE AND THE PROBLEMS WITH ANTICLASSIFICATION Regardless of whether the D.C. Circuit was correct in determining that § 637(a) does not contain a race-based presumption, Rothe expanded the possibilities of what counts as “race-neutral.” Whether Justice Kennedy envisioned a law like the one in Rothe at the time of Parents Involved, and whether he would have called § 637(a) “race-neutral” is unclear. Regardless, following the Rothe holding, a classification can come very close to being a race-based preference without triggering strict scrutiny. 154 Id. 155 Id. at 67. 156 Id. at 66. 157 515 U.S. at 203. 158 836 F. 3d at 65. 200 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 8:1 So what does the Rothe holding say about affirmative action? On the one hand, if followed, it expands the possibilities for race-conscious decision-making in many areas. From government contracting to college admissions, the sweep of the Rothe reasoning may be far-reaching. On the other hand, the Rothe holding pushes the anticlassification principle to the limits of its usefulness. A. The Practical Implications of the Rothe Reasoning From a practical perspective, Rothe seems to create a space for affirmative action removed from strict scrutiny. By allowing a policy that ventures so close to racial classification to slip through relatively unchallenged, the court in Rothe expanded the zone of what constitutes race-neutral means. Doctrinally, the Feeney/Davis standard is central to analyzing the D.C. Circuit’s holding. Because laws and policies are exempt from strict scrutiny if their motive is benign and they do not expressly classify, facially neutral laws intended to benefit minorities are likely to receive rational basis review. Since statutes facing that test will generally pass constitutional muster, Rothe’s restrictive view of what constitutes a classification is particularly important. Consider the possibilities; an agency using language similar to that in § 637(a)(5) could enact a law or plan favoring “people who have been discriminated against because of their identity as member[s] of a group without regard to their individual qualities.”159 In fact, several U.S. statutes have adopted the language from § 637(a) and applied the provision to other contracting agencies.160 159 15 U.S.C § 637(a)(5). 160 See, e.g., 51 U.S.C. § 30304 (applying § 637(a)’s definition of “socially and economically disadvantaged individuals to National Aeronautics and Space Administration contracting); 42 U.S.C. § 4370d (applying the definition to Environmental Protection Agency subcontracts). No. 4:175] NEUTRAL IN NAME 201 To be sure, without a presumption of eligibility, § 637(a) is not necessarily the ideal method of furthering contracting opportunities for minority-owned businesses. Without a presumption that the statute applies to all members of specific racial groups, proving eligibility will likely require business owners to show that they have personally faced discrimination. Depending on what proof is needed to make such a showing, that demonstration may be difficult, even for those who fit within the category. For example, people who have suffered the effects of implicit bias161 (maybe without realizing it) might not be able to provide proof of having faced racial discrimination in those cases. Furthermore, people who have suffered the consequences of past segregation and discrimination, but cannot point to specific instances in their own lives, might not fit the definition in § 637(a)(5). Although agencies implementing preferences like § 637(a) might attempt to define “discrimination” to include such nonspecific instances or the effects of implicit biases, proving that sort of general prejudice presents its own difficulties. For instance, people who have experienced the effects of disparate housing opportunities might not be able to prove that they “have been subjected to racial or ethnic prejudice or cultural bias.” Even if subjection to inferior housing options fits an agency’s definition of “discrimination,” individuals might struggle to show that they have personally faced such circumstances. The Rothe court did not address the requisite proof demonstration under statutes like § 637(a). In that regard, the holding leaves several important questions unanswered. For example, if a presumptive homeowner who is black is 161 Occurring when “[U]nconscious attitudes (including culturally learned associations or generalizations that we tend to think of as stereotypes) introduce unjustified assumptions about other people and related evidence that can distort a person’s judgment and behavior.” Jennifer K. Elek & Paula Hannaford-Agor, First, Do No Harm: On Addressing The Problem of Implicit Bias in Juror Decision Making, 49 CT. REV. 190, 190 (2013). 202 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 8:1 shown houses in only minority neighborhoods, has she faced sufficient discrimination to fit the definition? Has someone whose parents were denied fair housing opportunities on account of race suffered racial prejudice by virtue of the difficulties involved in growing up in inferior housing? Has someone whose ancestor was a slave faced racial discrimination by virtue of that fact? To be sure, if the entity enacting the facially race- neutral affirmative action plan is also the one implementing it,162 then a wider range of instances might be more likely to fit the definition. But that is not always the case. For example, a legislature might create a law like § 637(a)(5) and entitle third parties to determine whether applicants meet the “having faced discrimination” definition. In that instance, the implementing agency might adopt a very specific and difficult proof standard, especially if the agency opposes the plan to begin with. Ultimately, the Court in Adarand suggested that a presumption of eligibility for people belonging to certain racial groups would be tantamount to a classification based on race.163 Thus, although a legislature can unequivocally say that a statute is intended to favor people of minority races, it cannot, without a sufficient justification under strict scrutiny, prefer those people by entitling them to a presumption of eligibility. In Rothe, that meant that § 637(a) could create a disparate impact favoring people belonging to minority races and that the legislature could unambiguously espouse that effect as its purpose. But it also meant that the more definitively the statute achieved the intended end (e.g., by presuming that the people it wanted to help were those entitled to participate in its aid program), the more suspect it would become. Nevertheless, benevolent racial classifications and statutes like § 637(a)(5) should yield different results only in 162 For example, a university adopting language like § 637(a)(5) in its own admissions policy and then deciding which applicants have faced racial prejudice. 163 515 U.S. at 213. No. 4:175] NEUTRAL IN NAME 203 marginal cases. Moreover, as Justice Souter noted in his dissent in Gratz v. Bollinger, the inconsistent treatment of the two methods encourages legislatures and state decision- makers to “hide the ball.”164 In Gratz, the Court applied strict scrutiny to the University of Michigan’s use of racial preferences in undergraduate admissions.165 Justice Souter argued in dissent that mere acknowledgment of race- consciousness should not trigger increased suspicion.166 He explained that “it seems especially unfair to treat the candor of the admissions plan as an Achilles’ heel.”167 As Justice Souter recognized, if the methods are the problem, the incentive is to minimize transparency through coyly worded legislation, not avoid a particular result. When courts are willing to tolerate facially neutral race-consciousness, this sentiment highlights the hypocrisy of anticlassification all the more. In sum, courts will allow legislatures to further benevolent race-conscious goals as long as they use facially race-neutral means. But as soon as Congress, intending to favor people from historically subjugated racial groups, classifies those it intends to help, the laws become suspect. One implication seems inescapable; the anticlassification principle is the wrong way to further the goals of equal protection. B. Rothe’s Implications for the Anticlassification Principle Ultimately, the Rothe holding indicates the infeasibility of anticlassification as an interpretive method. The anticlassificationist argument against facially race- based affirmative action relies on the belief that classifying people based on race is itself an evil prohibited by equal protection. This was Justice Roberts’ concern in Parents Involved, where he argued that “[t]he way to stop 164 539 U.S. at 298 (Souter, J., dissenting). 165 Id. at 270. 166 Id. at 297. 167 Id. 204 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 8:1 discrimination on the basis of race is to stop discriminating on the basis of race.”168 From that perspective, the Reconstruction Amendments were intended only to prohibit racial distinctions (and those based on membership in other protected classes). It is a “race-conservative” point of view.169 Classifying people based on race is undoubtedly problematic in many instances. When the Court decided Loving v. Virginia in 1967, it said that classification was harmful even if both white people and black people were labelled.170 But racial distinctions were not the most concerning issue for the Court. What the Court truly cared about in Loving was the reason why the particular classification was harmful. Classification, in the context of miscegenation, was wrong because of the way it subjugated black people. It was wrong because it carried with it sentiments like, Almighty God created the races white, black, yellow, malay and red, and he placed them on separate continents. And but for the interference with his arrangement there would be no cause for such marriages. The fact that he separated the races shows that he did not intend for the races to mix.171 Of course, the anticlassificationist would argue that this is precisely the sort of problematic language at the heart of equal protection. And in part, the anticlassificationist would be correct. But also, she would overlook the deeper issue. Pointing out alleged differences between the races was not the crux of the problem in Loving. The real concern was the motive and rationale. The belief that the races ought to be separate arose from ideas concerning inferiority and superiority. The racial classification in Loving was 168 551 U.S. at 748. 169 See Reva B. Siegel, From Colorblindness to Antibalkanization: An Emerging Ground of Decision in Race Equality Cases, 120 YALE L.J. 1278, 1282 (2011). 170 388 U.S. 1, 12 (1967). 171 Id. at 3. No. 4:175] NEUTRAL IN NAME 205 suspect because it was designed to perpetuate a history of racism. Moreover, if § 637(a) is a “race-neutral” alternative, there doesn’t seem to be much use for the anticlassification principle. By enacting § 637(a), Congress intentionally favored people because of race. It may not have explicitly labelled those people, but it did not attempt to hide its motive. If facially race-based classifications are inherently discriminatory (as seems to be Justice Roberts’ opinion in Parents Involved), the statute the D.C. Circuit addressed in Rothe does not seem much less discriminatory. What, then, does the Rothe holding ultimately mean for anticlassification? On one hand, Rothe is a test case for anticlassificationists. It forces those who support the theory to consider the bounds of its usefulness. When anticlassification extends only to the label and has little practical effect on the law, is it still a valuable interpretive method? If the legislature can act for race-based reasons without expressly classifying by race, then the only thing left to consider is whether the absence of a label is worth forcing Congress to hide the ball. But confusingly, the Court has suggested that when an agency successfully avoids classifying, it need not even bother to conceal its race-based motive. A school is free to select students for race-conscious reasons as long as it does not use a racial label in its criteria.172 When a decision- maker can act for the explicit purpose of achieving a race- oriented outcome, it is difficult to see how only a racial label implicates equal protection, or, conversely, why a racial label automatically triggers strict scrutiny. As Justice Ginsburg noted in her dissent in Gratz, “the Constitution, properly interpreted, permits government officials to respond openly to the continuing importance of race. Among constitutionally permissible options, those that candidly disclose their consideration of race seem to me preferable to those that 172 See, e.g., Fisher, 136 S. Ct. at 2209 (2016) (discussing the Top Ten Percent plan as a valid race-neutral alternative). 206 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 8:1 conceal it.”173 In light of Justice Ginsburg’s point, if openly considering race is preferable to surreptitiously awarding racial preferences, why is the degree of openness so troubling? Where does candor cross the line from desirable to suspicious? But modern statutes rarely classify based on race in order to subjugate minority groups. The days of state- endorsed Jim Crow racism have given way to subtler methods.174 In the pre-Brown era, a law classifying based on race was likely to be the product of anti-minority sentiment. Now, after Brown, a law containing a racial classification will almost invariably be a method of aiding historically subjugated groups. Racism is undoubtedly alive in the modern age, but in the political world it has, for the most part, been relegated to areas like implicit bias, disparate impact, and surreptitious action. This is the case because, to begin with, state actors are likely to be aware that actions overtly intended to harm racial minorities will be unconstitutional.175 If a legislature harbors such malign intent, it is likely to act without classifying at all. Thus, even if distinguishing invidious from benign classifications was once a difficult task (because the former were prevalent), it is a much more straightforward inquiry now (as overt racism in law-making has, for the most part, given way to implicit forms). In some instances the determination might be more difficult than in others,176 but, nevertheless, discerning invidious race-based intent is well within the province of the judiciary.177 173 Gratz, 539 U.S. at 305 n. 11 (Ginsburg, J., dissenting) (Internal citations omitted). 174 See Ian F. Haney López, Post-Racial Racism: Racial Stratification and Mass Incarceration in the Age of Obama, 98 CAL. L. REV. 1023, 1072 (2010). 175 See, e.g., Kent Greenawalt, The Unresolved Problems of Reverse Discrimination, 67 CAL. L. REV. 87, 93 (1979). 176 See, e.g., United States v. Starrett City Assocs., 840 F.2d 1096, 1102 (2d Cir. 1988) (discussing a plan designed to integrate a community by limiting the number of black people able to rent homes). 177 See Feeney, 442 U.S. at 279. No. 4:175] NEUTRAL IN NAME 207 VI. CONCLUSION Some authors have proposed amending the equal protection inquiry to consider factors other than classifications.178 Some contend, for example, that courts should focus instead on whether laws or plans perpetuate a history of past discrimination or affect access to the political process.179 Under that approach, a law would not be constitutionally suspect merely because it contains a facially race-based classification. Instead, a court would first determine whether the law perpetuates one of the social ills most often associated with minority subjugation.180 This Note does not necessarily promote a particular alternative. It suggests only that the absence of racial labels should not be as significant a factor in determining the constitutionality of laws, plans, or other state actions. Distinguishing classificatory laws from facially-neutral-but- race-conscious ones creates a false dichotomy. The two are essentially the same, and if the reasoning in Rothe takes hold, becoming more so. Reliance on the classification-oriented method of analysis is an antiquated form of legal interpretation. It is relatively rare that plans and laws enacted for discriminatory purposes will explicitly subjugate racial groups. These days, when a state actor expressly classifies by race, it is unlikely that the classification will be invidious. Ultimately, Rothe demonstrates that racial classifications are not the crux of the issue. If a legislature wants to target a group, it can do so through facially neutral methods. The relevant concern is effect. The Court has recognized this in the disparate impact context as purposeful discrimination and has granted protection under the Fifth 178 Lauren Sudeall Lucas, Functionally Suspect: Reconceptualizing “Race” As A Suspect Classification, 20 MICH. J. RACE & L. 255, 281 (2015). 179 Id. 180 Id. 208 COLUMBIA JOURNAL OF RACE AND LAW [Vol. 8:1 and Fourteenth Amendments.181 According to disparate impact doctrine, there is little practical difference between an invidious discriminatory law that classifies based on race, and one that does not. Courts are likely to find both unconstitutional. But there is currently a constitutional difference between a benign law that classifies based on race and one that does not. This is where cases like Rothe are illustrative. Because race-consciousness can be as overt as that in Rothe, or the Top Ten Percent plan in Fisher, then why does it matter whether those racial concerns are manifest through classifications? Moreover, if classifications are so difficult to distinguish from facially neutral means, why are courts so focused on labels in the first place? If a facially neutral law intended to benefit minorities will not trigger strict scrutiny, so too should a benign racial classification with the same effect be free from such suspicion. 181 See, e.g., Feeney, 442 U.S. at 274.