Printer Friendly

Affirmative action.

Affirmative action is so burning it's boring. Why? Partly because so much of what one sees and hears on the subject smacks of hypocrisy or speaking in code. Exhibit A: the name of California's affirmative-action-killing referendum, the "Civil Rights Initiative." Exhibit B: the term "affirmative action." Exhibit C: "I oppose affirmative action because it's harmful to minorities." Exhibit D: "Diversity is not counter to merit; it's an aspect of merit." Exhibit E: "[T]he fact that he is black and a minority has nothing to do with this in the sense that he is the best qualified at this time."(1)

This Essay rethinks the constitutionality of race-based, governmental affirmative action measures. There are, I know, a thousand essays on the same topic already. I make one promise: Readers who persevere will learn something new. First, although it is a matter of public record, most lawyers and judges are unaware that Congress in the 1860s repeatedly enacted statutes allocating special benefits to blacks on the express basis of race (and I am not referring to the well-known Freedmen's Bureau Acts,(2) which did not rely on express racial classifications). Accordingly, to be true to their principles, two of the five Justices in the prevailing anti-affirmative action majority--Justices Scalia and Thomas, whose commitment to original understandings and practices is also a matter of record--should drop their categorical opposition to race-based affirmative action measures.

Second, strict scrutiny doctrine, as it has been applied to affirmative action, can no longer survive strict scrutiny. I don't mean this statement figuratively. I mean that strict scrutiny doctrine, understood as it has been in the recent affirmative action cases, can no longer satisfy its own doctrinal requirements. Current affirmative action law may be the first instance in our jurisprudence of a constitutional doctrine unconstitutional under itself.

Finally, and most important, the Court's recent affirmative action decisions have consummated a remarkable but unremarked-upon transformation in the entire analytic structure of heightened scrutiny doctrine. One powerful function of strict scrutiny has always been that of "smoking out" invidious purposes masquerading behind putatively legitimate public policy. But under today's affirmative action doctrine, strict scrutiny has become altogether different. It has become a cost-benefit test measuring whether a law that falls (according to the Court itself) squarely within the prohibition of the equal protection guarantee is justified by the specially important social gains that it will achieve.(3)

This shift in the use and understanding of strict scrutiny is momentous, not only for equal protection law, but for every field of constitutional law in which the compelling state interest test figures. Or rather it would be momentous, if it really were the law. But it cannot be. Strict scrutiny cannot serve as a general escape hatch through which reasons of state may trump acknowledged constitutional injuries. As a smoking-out device, heightened scrutiny is sensible. As a cost-benefit justificatory test, it is indefensible. Or so I will argue.

This argument, however, will not decide affirmative action's constitutionality. Part of the problem with current doctrine is its effort to pack far too much of the difficult work of equal protection analysis into a determination of the appropriate "standard of review." Straightening out strict scrutiny in the affirmative action cases can only strip away a certain false doctrinal mesh, leaving exposed the contending claims of color-conscious and colorblind justice.

But the debate over colorblindness in constitutional law, whose thrusts and parries are so well known, will raise very different questions when the cost-benefit approach of current doctrine is systematically stripped away. For example, the Justices who have found against affirmative action programs repeatedly have done so on the ground that affirmative action threatens inadvertently to entrench racial thinking and to stigmatize minorities. Defenders of affirmative action tend to respond to this assertion by denying the reality of these harms or by arguing that they are outweighed by affirmative action's benefits. But this entire set of arguments, both for and against, is in fact constitutionally irrelevant.

Throughout Fourteenth Amendment jurisprudence, inadvertent harm to minorities, without more, is rejected as a basis of constitutional invalidity or even of heightened scrutiny.(4) Equal protection jurisprudence, outside the arena of affirmative action, generally does not engage in cost-benefit analysis. It does not purport to measure up and balance the social gains and losses a law will produce. The constitutional question is instead whether a law embodies an invidious or otherwise constitutionally impermissible purpose. And this must be the constitutional question with respect to affirmative action as well. The degree to which affirmative action inadvertently entrenches racism or harms minorities is of fundamental importance to affirmative action's merits as a matter of policy. But it is irrelevant to affirmative action's constitutionality. The ultimate constitutional question presented by race-based affirmative action--by no means an easy question--is whether whites' equal protection rights are violated when the government purposefully acts to assist blacks and other minorities by granting them special opportunities.

Part I of this Essay describes the failure of those who champion strict construction to live up to their principles when it comes to affirmative action. Part II analyzes current doctrine's treatment of strict scrutiny as a cost-benefit test rather than as a device for smoking out constitutionally illegitimate purposes. Part III confronts the debate over colorblindness as it would be presented if strict scrutiny were restored to its proper footing and concludes that standard affirmative action programs are in fact constitutional, however unwise they may be as policy in many contexts. Part IV adds a postscript on Romer v. Evans(5) and California's Civil Rights Initiative;(6) the former confirms the equal protection analysis offered in this Essay, and the latter, I suggest, is constitutional for the same reasons that affirmative action is. Although the Initiative may be a racial classification under the extant precedent, neither the Initiative nor standard affirmative action plans met the criterion of invidious purpose necessary to render a racial classification unconstitutional.

I. STRICT CONSTRUCTION

In a recent televised address, Senator Orrin Hatch--overseer of the nation's judicial confirmation process--vilified "activist" judges while lionizing those who have held the line against affirmative action.(7) He was applauded roundly for both points. Evidently his audience felt that these two positions could be held by one person at the same time. And so they can, to judge from the opinions of some of our best-known judicial figures, such as Justices Scalia and Thomas and former Judge Bork. All three are pupils in that interpretive school that famously instructs judges to adhere to the letter of the law, to the "original understanding,"(8) or, if these are ambiguous, then to "the most specific level at which a relevant tradition protecting, or denying protection to, [an] asserted right can be identified."(9) On this view, if at the time of enactment there was a specific understanding about the permissibility of a certain kind of law, judges today have no business altering that result.(10) That would be government by "authoritarian judicial oligarchy."(11)

When it comes to affirmative action, these strict constructionists(12) are strict indeed, holding that no governmental program "that operates on the basis of race" is "in accord with the letter" of the Constitution.(13) Pro-affirmative action holdings are for them just another example of "the politics of ultraliberalism . . . driving the law."(14) It follows, surely, that a strict rule condemning race-based affirmative action comports with strict construction, that such a rule either must reflect the original understanding of the 1860s, or at the very least must be the rule evinced by the most specific, relevant historical practices. For how else could these honorable men take the position they take?

In July 1866, the Thirty-Ninth Congress--the selfsame Congress that had just framed the Fourteenth Amendment--passed a statute appropriating money for certain poor women and children.(15) Which ones? The act appropriated money for "the relief of destitute colored women and children."(16) In 1867, the Fortieth Congress--the same body that was driving the Fourteenth Amendment down the throat of the bloody South--passed a statute providing money for the destitute in the District of Columbia.(17) (And remember that Congress is the constitutional analogue of a state legislature for the District of Columbia.) What classification did Congress adopt in this poor-relief statute? Relief was to be given to the destitute "colored" persons in the nation's capital.(18) Year after year in the Civil War period--before, during, and after ratification of the Fourteenth Amendment--Congress made special appropriations and adopted special procedures for awarding bounty and prize money to the "colored" soldiers and sailors of the Union Army.(19)

These statutes are not like the well-known Freedmen's Bureau Acts (20) of the same period,(21) directing benefits to blacks but using classifications that were formally race-neutral.(22) On the contrary, these statutes expressly refer to color in the allotment of federal benefits. Nor are these statutes buried in archives deep within the Library of Congress. They are, if not well-known, at least knowable by anyone who takes three minutes with the United States Statutes at Large (look up "colored" in the indexes for more).(23) What do they prove? Only that those who profess fealty to the "original understanding," who abhor judicial "activism," or who hold that the legal practices at the time of enactment "say what they say" and dictate future interpretation,(24) cannot categorically condemn color-based distribution of governmental benefits as they do.

I am no originalist, so I cannot regard the practices of Congress in the 1860s as dispositive of affirmative action's constitutionality. In fact, nearly no one today is a true equal protection originalist, because true equal protection originalism would repudiate Brown v. Board of Education.(25) Hence the point is not to foreclose argument by citing old statutes. It is to begin the argument with a little more candor. The colorblind contingent must begin by recognizing that they are calling on courts to render the kind of judgment about justice (beyond the letter of the law, beyond original intent) that elsewhere they deplore.

II. SCRUTINIZING STRICT' SCRUTINY: RECOVERING THE PURPOSE OF HEIGHTENED REVIEW IN EQUAL PROTECTION LAW

Apart from the debate over judicial activism, the prevailing doctrinal analysis of affirmative action occurs within the well-known three-tiered framework of equal protection review. A great deal of the argument concerning affirmative action's constitutionality has taken the form of a debate about selecting the proper place of affirmative action in this three-tiered framework.(26) As a result, a more systemic point has been missed. The framework itself, in current form, is unsound. And it is precisely the Court's treatment of affirmative action that has made it so.

Recent affirmative action cases have turned heightened equal protection review into a cost-benefit test. This cost-benefit understanding of equal protection review is untenable. To restore strict scrutiny to its proper footing, the Court must reorient it to its former purpose: not balancing state interests against conceded constitutional violations, but smoking out concealed constitutional violations.

A. Background: Classificationism

Before 1995, affirmative action programs implemented by the federal government needed to pass only "intermediate scrutiny" to be constitutional.(27) In Adarand Constructors, Inc. v. Pena,(28) the Court changed course, holding that strict scrutiny applies to all governmental affirmative action measures.(29) Supporters of affirmative action raised a cry, but, as Justice O'Connor pointed out in her opinion for the Court, Adarand's holding followed inexorably from principles of equal protection jurisprudence well established in other contexts. In particular, Justice O'Connor observed that, apart from the affirmative action cases, the Court had held for decades that (1) strict scrutiny applied categorically to all governmental measures classifying persons on the basis of their race;(30) and (2) there is no difference between the equal protection duties incumbent upon the states and those incumbent on the federal government.(31) The Court in Adarand merely honored the logic of this classification-driven framework, eliminating the exception that had been carved out for affirmative action.

As is often observed, strict equal protection scrutiny is almost always fatal. Indeed, when the Court has applied strict scrutiny to a race-conscious measure designed to assist minorities, it has never upheld the measure. The Court has struck down race preferences in governmental contracting(32) and has invalidated majority-minority voting districts.(33) Lower courts, following the Justices' new lead, have dismantled race-based preference plans in public university admissions and elsewhere.(34)

The critical holding of Adarand was that all laws employing a racial classification must undergo strict scrutiny, with no exception made on the basis of allegedly benign intentions.(35) The classification itself is the constitutionally suspect feature of the law, the feature that triggers heightened scrutiny, regardless of which race happens to be burdened, and regardless of the particular burdens imposed. "`[T]he standard of review under the Equal Protection Clause is not dependent on the race of those burdened or benefited by a particular classification ....'"(36)

As you might expect, opinions dissenting from this classificationist approach have argued that it makes no sense to treat the "classification as a talisman which--without regard to the rights involved or the persons affected--calls into effect a heavier burden of judicial review." As you might not expect, the words just quoted were written by Justice (now Chief Justice) William Rehnquist.(37)

Rehnquist was dissenting from Craig v. Boren, a case involving an Oklahoma statute prohibiting the sale of certain alcoholic drinks to men under twenty-one, while permitting women to buy these drinks at eighteen.(38) Justice William Brennan, writing for the Court, took advantage of the case to announce for the first time that all laws using a "gender-based" classification would be subject to intermediate scrutiny.(39) The Court struck down the statute under this test.(40) Justice Rehnquist (together with Chief Justice Burger) dissented, protesting the Court's "conclusion that men challenging a gender-based statute which treats them less favorably than women may invoke a more stringent standard of judicial review than pertains to most other types of classification."(41) Yet, a few years later, Rehnquist began joining opinions holding that whites challenging a race-based statute which treats them less favorably than blacks may invoke the most stringent standard of judicial review.(42) Without a word of explanation, he has written and joined opinions treating the "classification as a talisman which--without regard to the rights involved or the persons affected--calls into effect a heavier burden of judicial review."(43)

The point, once again, is not to stop debate with a charge of inconsistency (might not a similar charge, in reverse, be rung up against Justice Brennan's account?). Rather, the point is that Rehnquist was correct in his earlier decision. The Court's classification-driven framework is illogical and untenable.

One very familiar way to argue this point is to say that the Court has in effect created "a `moral [and] constitutional equivalence' ... between laws designed to subjugate a race and those that distribute benefits on the basis of race in order to foster ... equality.(44) But inasmuch as opponents of affirmative action affirmatively embrace this equivalence--the phrase just quoted belongs to Justice Thomas and is preceded by "I believe that there is"(45)--let's put this argument aside for now. We can demolish the classification-as-talisman framework without it.

A single example demonstrates how. Imagine a state imposing legal disabilities on the poor--defined in terms of some dollar amount of immediate-family wealth or income. For example, suppose that California adopts by referendum a "War on Poverty Initiative" barring the poor from voting, from attending educational institutions beyond high school, from accepting employment other than manual labor, and from checking in at nice hotels. Consider only whether this law should be held unconstitutional as a violation of equal protection in the absence of any alleged infringement of other constitutionally fundamental rights (such as the right to vote). Certainly it should. A court should strike it down as a blatantly unconstitutional form of legally mandated, status-based inequality relegating the poor to a second-class citizenship wholly unacceptable under the Fourteenth Amendment.

But how would this law be analyzed under current doctrine? The Court has expressly held that "poverty, standing alone, is not a suspect classification."(46) With good reason. For under the current framework, if the Court were to deem wealth classifications suspect, or even "semi-suspect," then every law incorporating such a classification would be subject to heightened equal protection scrutiny. It would make no difference whether the classification burdened the rich or the poor. Accordingly, the entire vast array of meanstested assistance programs, federal and state, would become presumptive violations of equal protection. Progressive taxation would almost certainly fall. Libertarians might envy such doctrine, but no sane judge today would suggest that the Equal Protection Clause guarantees the rich and the poor an equal right to receive food stamps.

If, however, wealth classifications are subjected to mere rational basis review, California's War on Poverty Initiative might well stand. Poverty doubtless correlates rationally with illiteracy, and literacy is a constitutionally permissible qualification for suffrage.(47) Barring the poor from matriculating at universities or checking in at nice hotels would be rationally related to the goal of screening out those who cannot pay their way. In addition, forcing the poor into manual labor might create (coupled, perhaps, with an elimination of the minimum wage) a superabundance of low-wage employees profitable to large sectors of the national economy. Under the rational basis test, these conceivable advancements of legitimate state interests should be sufficient to validate some or all of the Initiative.

I do not mean that today's Court would uphold the War on Poverty Initiative. I have no doubt that the Court would strike it down. The point is that this result would not square with current classificationist equal protection analysis, which cannot easily embrace both the invalidity of California's hypothetical poor law and the validity of our actual poor laws.

B. Smoking

What has gone wrong here is straightforward. An equal protection doctrine requiring an exact fit between the measures adopted by state actors and the interests these measures are said to advance makes good sense when a law singles out a particular class of persons for adverse treatment and there is reason to fear that the law seeks to achieve an impermissible purpose relating to this group, a purpose that the legislators will try to conceal behind noninvidious, group-neutral public policy explanations. For example, in the not-too-distant past, a law banning blacks from voting might have been defended in terms of assuring literacy among voters. The manifest overbreadth and underbreadth of such a law (so defended) would demonstrate that the law's true purpose was not in fact race-neutral. Used this way, strict scrutiny serves as a test of ulterior state interests. Its function, to paraphrase John Ely, is to smoke out illegitimate purposes that cannot be a valid basis for state action under the Equal Protection Clause.(48) Heightened scrutiny of a racial classification permits a court to conclude, in effect, "If the state were really interested in race-neutral purpose x, it would not have done what it did."

But the same analysis makes little sense where the race-based purposes behind a law are undisputed. In those circumstances there is nothing to smoke out. No strict scrutiny would be necessary to invalidate a law banning blacks from taking designated jobs if the law's professed purpose was to maintain white supremacy. (But of course that sort of purpose will rarely be professed.) Similarly, no strict scrutiny is called for when the undisputed purpose of a law is to assist members of one race at a cost to members of another. The sole question is whether the conceded race-based purpose is constitutionally legitimate. This is why a court cannot resolve the constitutionality of group-assistance measures such as welfare or affirmative action through classification-driven heightened scrutiny. Determining whether the relevant state interests are legitimate or illegitimate must precede the application of heightened scrutiny. Because it elides this critical inquiry, classificationism cannot properly vindicate equal protection principles.

In other words, for the application of heightened scrutiny to a group classification to make sense, two things must be true. First, some legislative interest related to this group must be illegitimate. Second, there must be good reason to believe that the purpose of a law using this group classification is to further the illegitimate group-related interest (despite the state's claims that other, legitimate interests underlie the law). Hence laws that assist the poor properly attract no constitutional attention. Helping the poor has never been held an unconstitutional state purpose, and there is no adequate reason to suspect any other, invidious objective behind such laws.(49) By contrast, a law like the War on Poverty Initiative ought to be regarded as highly suspect, for a state could hardly claim that such a law is designed to help the poor,(50) and there would be every reason to suspect that an illegitimate state purpose is being carried out. A group classification as such is insufficient to trigger strict scrutiny. Rather, as Justice Rehnquist observed in Craig, the propriety of heightened scrutiny depends on "the rights involved and the persons affected."(51)

C. Costs and Benefits

An objection. "Now just a minute, Lester," someone may say. "You've got it all wrong. You're treating race classifications as a signal for something else, which, if present, would make the law constitutionally out of bounds. But the race classification is itself the constitutional evil, regardless of the intentions behind it. That's the difference between race and wealth classifications: Race classifications are always constitutionally odious, whereas wealth classifications are only sometimes so. Strict scrutiny of race classifications is not a hidden-purpose test. It is a test to determine whether this constitutional evil--legally classifying persons on the basis of race--is justifiable in the case at hand. Strict scrutiny tests whether there exist sufficiently important state interests that can be achieved only by what would otherwise be an unconstitutional classification."

This objection, if correct, would be decisive. For on this objection, strict scrutiny would not be a smoking-out device at all; rather, it would be a cost-benefit justificatory test. Strict scrutiny would serve to determine whether a law that causes acknowledged constitutional harms is justified by sufficiently important benefits that a less constitutionally costly ("better tailored" or "less restrictive") law could not have achieved. Recast this way, strict scrutiny for race classifications would appear to make sense after all. All race classifications would properly demand strict scrutiny, because all such classifications, regardless of their purpose, would be deemed a constitutional evil that could not stand in the absence of compelling justification.

Recasting strict scrutiny as a cost-benefit justificatory test, rather than a smoking-out device, is exactly what must be done, and it is what the Court has in fact been obliged to do, in order to render its classificationist logic coherent. Thus, in Adarand, Justice O'Connor's opinion for the Court held:

[W]henever the government treats any person unequally because of his or

her race, that person has suffered an injury that falls squarely within the

language and spirit of the Constitution's guarantee of equal protection....

The application of strict scrutiny, in turn, determines whether a compelling

governmental interest justifies the infliction of that injury.(52)

Contrast this formulation with the corresponding sentence from Justice O'Connor's 1989 opinion in Croson: "Absent searching judicial inquiry into the justification for such race-based measures, there is simply no way of determining what classifications are `benign' or `remedial' and what classifications are in fact motivated by illegitimate notions of racial inferiority ...."(53) This earlier formulation reflects the smoking-out view. The opinion says so explicitly: "[T]he purpose of strict scrutiny is to `smoke out' illegitimate uses of race."' In other words, strict scrutiny in this Croson formulation serves to determine whether a law was "in fact motivated" by assumptions "of racial inferiority" or "illegitimate racial prejudice."(55) By contrast, the Adarand formulation makes no pretense of discovering whether an affirmative action measure "in fact" rests on "illegitimate racial prejudice." Instead, all racial classifications are said to inflict a constitutional "injury"--an injury that "falls squarely within the language and spirit" of the Equal Protection Clause--and strict scrutiny is necessary to test whether this injury is justified in a given case.

Not that Justice O'Connor's two opinions are irreconcilable. But the Adarand formulation represents what may be the Court's first unequivocal embrace of the justificatory view--an embrace logically necessary to make sense of the current treatment of affirmative action. It is logically necessary because the idea that the proponents of affirmative action programs are "in fact motivated" by "illegitimate notions of racial inferiority" is so implausible that the Court itself has never taken it seriously. In no case has the Court held or even suggested that an affirmative action plan's failure to satisfy strict scrutiny shows that its proponents were actually secret adherents of The Bell Curve.(56) Instead, the Justices have stressed the danger that affirmative action will "foster harmful and divisive stereotypes" or "reinforce" racialist thinking, even though these costs may have been entirely "unintended."(57) With respect to such unintended costs, strict scrutiny can be justified only as a justificatory test--a test measuring whether these constitutional costs are justified in a given case by offsetting social benefits.

At issue, then, are two very different understandings of strict scrutiny and its "compelling state interest" test. On the smoking-out view, to conclude that a law passes strict scrutiny is to hold that the law does no constitutional harm--does not violate the relevant constitutional norm--at all. This is so because the constitutional injury would have consisted in the government's pursuit of a constitutionally impermissible objective, and strict scrutiny serves to test the government's claim that a permissible state interest fully explains the law at issue. But on the cost-benefit view, to conclude that a law passes the compelling state interest test is to find that the law is permissible even though a constitutional injury is inflicted--even though a general constitutional norm or principle (such as the norm or principle of colorblindness) has been violated.

This shift in the meaning of strict scrutiny represents a little-noticed but profound transformation in equal protection law. To put it bluntly, current doctrine uses strict scrutiny as an escape hatch through which government can, with impunity, violate equal protection principles in the name of more important state interests. Strict scrutiny is no longer a means of smoking out concealed violations of constitutional principles. It is a means of "justif[ying]" a conceded constitutional "injury."(58) I do not mean that the Court has fallen into a foolish semantic trap, telling us that a law violating the Equal Protection Clause sometimes does not violate the Equal Protection Clause. Obviously, the Court's official formulation is that a law, when it passes strict scrutiny, does not violate the Equal Protection Clause at all. But strict scrutiny now serves, expressly and self-consciously, as a justificatory test in which important state interests are permitted to outweigh acknowledged constitutional injuries.

"And what is so wrong with this understanding of strict scrutiny?" someone may ask. "It seems not only sensible, but clearly correct. The fact that strict scrutiny, when survived, permits the perpetration of a constitutional injury is the very reason that strict scrutiny is, and should be, fatal in nearly every case. Racial classifications do indeed violate the letter and spirit of the Equal Protection Clause. Moreover, they do indeed cause harms that are of special constitutional significance. Therefore they should never be permitted unless the law is narrowly tailored to further a compelling state interest."

On the contrary, if race classifications do in fact fall squarely within the language and spirit of the Equal Protection Clause, they should be held unconstitutional even if they are narrowly tailored to further a compelling state interest. There is something deeply wrong with the cost-benefit picture of strict scrutiny. Economizing equal protection is unacceptable.

Suppose a state embarked on a program of wholesale racial expulsion and apartheid. Backers of the law might offer to demonstrate, with impeccable social scientific data, that their program will sharply reduce crime and eradicate racial violence, saving billions of dollars and x lives per year. Moreover, they offer to prove that only these extreme measures can achieve the desired results. How are the courts to respond?

Perhaps someone will say that judges must scrutinize the validity of the state's empirical showing. Perhaps the courts can find data suggesting that racial expulsion and segregation will actually increase violence, or perhaps judges can identify less race-restrictive means to reduce violence by the same amount. But what if no one can effectively rebut the state's empirical claims? Is the court to hold that saving x lives is not a compelling state interest? Surely the constitutionality of the state's program under the Equal Protection Clause cannot really turn on someone's ability to puncture the state's empirical showing. A law whose express purpose is racial apartheid or expulsion is unconstitutional per se, because racial purification of society is an objective that no legislature can pursue under the Fourteenth Amendment--period. The Constitution does not permit a state to treat members of a particular race as viral agents within the body politic--period. This prohibition does not rest on an empirical proposition. It does not rest on the proposition that apartheid or expulsion would fail to be cost-justified or would produce no compelling benefits. It rests on a principle of equality under law that cannot be economized.

Offsetting state benefits cannot "justify" a law violating an individual's equal protection rights. That is what it means to have an equal protection right; the right is not subject to any ordinary cost-benefit calculus. Treating an ethnic group as a menial class may serve any number of compelling state interests. Most peoples since the dawn of time have thought as much. Racial subjugation might even, conceivably, produce the greatest happiness for the greatest number. But the Fourteenth Amendment blocks every state action directed to this end, whatever interests it might serve.

To this conclusion, someone might object that the harms of racial expulsion and apartheid are so great that virtually no state interests are sufficient to outweigh them. If so, then the fact that racial expulsion and apartheid are categorically unconstitutional does not show the untenability of the cost-benefit justificatory approach. Rather (it might be said), the example is rigged, because it invokes a law whose constitutional costs are so high that almost nothing could justify inflicting them.

Observe that this objection implies that racial classifications vary in the constitutionally cognizable harms that they cause and that the weightiness of the state interests they must serve should vary accordingly. This idea is precisely what current doctrine rejects. A generalized sliding-scale balancing approach weighing differently the harms threatened by different kinds of racial classifications would obliterate the classificationist framework and instead vindicate the claim that everything depends upon the rights at stake and the parties affected.

Moreover, the objection is incorrect. The logic of the preceding argument turned not on the "high costs" of racial expulsion and apartheid, but on the manifest constitutional impermissibility of ethnic cleansing as a state purpose. The same analysis would have applied to any law whose purpose is to recognize or perpetuate the superiority of a particular race. Thus, a law forbidding blacks to enter universities is also categorically unconstitutional, no matter what compelling interests it might be said to further; so is a law requiring blacks to sit in the back of the bus. Unconstitutional race discrimination is unconstitutional not because its costs generally outweigh its benefits, with strict scrutiny serving to pick out the exceptions. Rather, unconstitutional race discrimination is unconstitutional because laws with racist purposes are constitutionally forbidden, with strict scrutiny helping to determine whether such racist purposes are in play.

In fact, the entire three-tiered framework of equal protection review becomes almost unintelligible on the cost-benefit view of heightened scrutiny. For example, on the cost-benefit view, what are we to say of the cases that evaluate claims of sex discrimination through intermediate Scrutiny?(59) The cost-benefit view would have to hold that gender classifications as such inflict a constitutional injury, but that this injury could be overridden by merely "significant," rather than "compelling," state interests. How can this lower standard of review be defended? Would the Court say that a law denying women rights enjoyed by men does not quite fall "squarely within the language and spirit of the Constitution's guarantee of equal protection"?(60) Or is it that women's equality rights are not worth quite as much? For that matter, what are we to say of rational basis review? Shall we say that every law inflicts a little constitutional injury, but that the costs of this injury are so small that a merely rational relation to a legitimate state interest is sufficient to offset them?

Rational basis review is not justificatory. Rather, it tests whether a law can be rationally explained only by reference to an illegitimate purpose. Its lenience is explained by the absence of indicia providing reason to suspect that the law's true purpose is something other than the advancement of legitimate state interests. Similarly, the lower standard of review for sex discrimination is premised on the theory that there is less reason to fear invidious discriminatory purposes in the case of laws differentiating on the basis of sex than in the case of laws differentiating on the basis of race. This proposition is at least arguable; think of sex-segregated bathrooms as compared to race-segregated bathrooms. By contrast, the idea that women's equal protection rights are worth less than others', which is the implication of intermediate scrutiny on the justificatory view, is risible. A law that violates "the language and spirit" of the Equal Protection Clause cannot be justified by significant governmental interests.

This argument does not rule out the possibility that an ethnic or racial group might temporarily he subjected to adverse laws in the face of imminent catastrophe. If a virus poised to kill tens of thousands were containable by (and only by) immediately quarantining every Jewish person in the nation, such a quarantine would presumably be constitutional. This result would follow on the analysis proposed here because there would be no illegitimate state purpose being pursued.(61) But strict scrutiny cannot be the violation-justifying test the Court's affirmative action cases suggest it is. Rather, strict scrutiny ought to be what it always was: a test for smoking out ulterior, unconstitutional state purposes.

D. Implications

Returning strict scrutiny to its proper function would have profound implications not only for equal protection law, but for every domain of constitutional law where the compelling state interest test figures.(62) Exploring these more general implications will have to await another day. But what conclusions follow for affirmative action? Granting everything I have argued thus far, a court would not yet have reason to find affirmative action constitutional or unconstitutional. Straightening out strict scrutiny cannot answer the ultimate question of affirmative action's constitutionality. Returning strict scrutiny to its smoking-out function cannot even determine which level of scrutiny should apply to affirmative action programs. The appropriate standard of equal protection review in any given context must follow from a prior determination of the legitimacy of the relevant state interests.

The admitted purpose of most affirmative action programs is to assist blacks and other minorities by granting them opportunities denied to whites. If this is not a permissible governmental purpose, then the machinery of strict scrutiny would be entirely appropriate in affirmative action cases. Why? Because once the Supreme Court established the impermissibility of this race-conscious objective, defenders of all subsequently litigated affirmative action programs would defend their measures in terms of race-neutral interests, such as achieving a diversity of backgrounds or perspectives in a particular institution. This race-neutral defense of race-classifying measures would quite properly call for heightened scrutiny, under which most affirmative action measures would probably be struck down. But if affirmative action's avowed race-conscious purpose is permissible, then in the absence of reason to believe that a given program actually served other, unconstitutional purposes, no heightened scrutiny would apply. There would be nothing to smoke out.

The ultimate question of affirmative action's constitutionality, therefore, depends first and foremost on whether the government may permissibly seek to aid blacks or other minorities by granting them opportunities denied to whites. This question calls for the exercise of judgment--normative, interpretive, constitutional judgment. It cannot be answered through the proxy of any standard of review, heightened or lowered, because it precedes the determination of which standard of review to apply. Affirmative action's constitutionality demands a confrontation with the debate over colorblindness in constitutional law. But as we shall see, within this debate, the proper function of strict scrutiny, together with the impropriety of an equal protection jurisprudence that tries to measure racial costs and benefits, will reemerge as an extremely important issue.

III. THE COLOR OF LAW

Almost every move and countermove in the colorblindness debate is by now well known. One camp says that affirmative action is racial discrimination all over again.(63) The other replies that offering a long-oppressed group special opportunities cannot be regarded as the moral equivalent of the racism it seeks to redress.(64) On the one hand, racial classifications are said to be divisive and inconsistent with the ideal of equality under law.(65) On the other, colorblindness is said to be a deceptive and unjust neutrality given the facts, past and present, of race discrimination in America.(66) One side says that affirmative action stigmatizes, entrenches invidious stereotypes, and threatens to undermine minority success in the long run.(67) The other replies that minorities are already stigmatized, stereotyped and undermined, and that affirmative action at least attempts to do something about it.(68)

Shall we try to adjudicate these contending propositions? Shall we try, as so many have, to make the best case we can for one of these two sets of claims, either ignoring or trying to demolish the other? To do so would assume that the rival claims are mutually exclusive. The truth, however, is that they are all undeniable.

The task for constitutional law is not to amass votes for one or another partial truth about affirmative action. It is to figure out what the Constitution requires in light of the whole truth. Imagine: Not one of the Justices has been able to resolve the great questions surrounding affirmative action's justifiability. Of course they haven't: It isn't their task to do so. Rather, the Court must begin by determining which of the undeniable truths about affirmative action states a constitutional claim for or against it. By now, the framing of a distinctly constitutional inquiry into affirmative action should have become quite sophisticated. Surprisingly, it hasn't. Indeed, many of the most important arguments on which current doctrine rests are not constitutional arguments at all.

After all the nonconstitutional considerations are cleared away, a serious constitutional question remains, but one that I think admits a pretty clear answer. Standard affirmative action programs--the term "standard" to be specified below--are constitutional. At least this conclusion follows, I shall try to show, if the analysis of equal protection scrutiny offered in Part II is correct. The argument proceeds in three stages. Section III.A argues that affirmative action's unintended consequences--including, for example, the stigmatization of minorities--harmful though they may be, are constitutionally irrelevant. If affirmative action doctrine is to be consistent with the rest of equal protection law, only an illegitimate racial purpose can render affirmative action unconstitutional. Section III.B discusses affirmative action's purposeful prejudicial treatment of whites (or others excluded from its benefits) and concludes that this treatment can indeed amount to a constitutional violation, but that it does not do so in standard affirmative action plans. Section III.C returns to the problem of suspect classifications in equal protection law and suggests general rules establishing what kind of affirmative action programs should be held constitutional and what kind should not.

A caution before proceeding. Every substantive consideration for and against affirmative action mentioned hereafter will no doubt be familiar to readers already. From this point on, I try only to present a new way to think through and embrace what readers already know.

A. Unintended Evils

There are three basic grounds on which to argue affirmative action's unconstitutionality: (1) its harm to society as a whole; (2) its injury to blacks or other racial minorities to whom preferences are extended; and (3) its injury to whites or others excluded from its benefits. This section addresses the first and second grounds.

The most prevalent harm-to-society argument against affirmative action is that it polarizes; the most prominent harm-to-minorities argument is that it stigmatizes. Justices ruling against affirmative action have relied expressly and repeatedly on both these arguments to justify strict scrutiny: (1) "Unless [classifications based on race] are strictly reserved for remedial settings, they may in fact promote notions of racial inferiority and lead to a politics of racial hostility."(69) (2) "[W]e subject racial classifications to strict scrutiny precisely because that scrutiny is necessary to determine whether they are benign . . . or whether they misuse race and foster harmful and divisive stereotypes without a compelling justification."(70) (3) "[R]acial paternalism' and its unintended consequences can be as poisonous and pernicious as any other form of discrimination. So-called `benign' discrimination teaches many that because of chronic and apparently immutable handicaps, minorities cannot compete with them without their patronizing indulgence."(71)

Observe that the evils imputed here to affirmative action are unintended evils. No one claims that legislatures or other state actors adopt affirmative action programs in order to foster hate or to stigmatize. On the contrary, the claim is that affirmative action "may in fact promote"(72) invidious racial thinking, even though such thinking "is not the actual predicate"(73) of the law. Strict scrutiny is deemed necessary because of affirmative action's "unintended consequences."(74)

Does affirmative action "in fact promote notions of racial inferiority and lead to a politics of racial hostility"?(75) Without doubt. Of course, affirmative action's critics tend to forget that the relevant question on this point is whether affirmative action fosters more racial hostility and stereotyping than would exist without it. (To have extremely few black students at some of our most prestigious academic institutions would also promote notions of racial inferiority. There is something extremely odd going on when this fact is omitted from the analysis. It is as if one were to oppose seat belt laws on the ground that seat belts can lead to physical injury in the event of an accident--without even trying to assess whether the alternatives one supports would result in more injuries.) Let's grant, however, that affirmative action inadvertently but significantly increases racial polarization and stigmatization. Does this fact state a constitutional claim against it? Contrary to conventional wisdom and judicial opinion, it does not.

Consider how strange the Justices' arguments would sound if situated somewhere other than in the superheated zone of affirmative action law. Suppose, for example, that a district judge began striking down the nation's welfare laws under strict equal protection scrutiny on the ground that welfare leads to a politics of class hostility. Such reasoning would make no sense. Grant that welfare laws are politically divisive; whose equal protection rights do they therefore violate? Those of the poor? Of the rich? Or is the violation unattached to any individual--a sort of equal protection violation in the air? The Equal Protection Clause is not a judicial license to hold unconstitutional, or to apply heightened means-ends scrutiny to, laws deemed to embody overly divisive forms of politics. If it were, Lochner(76) would have been rightly decided,(77) and the Fourteenth Amendment would not even have overruled Dred Scott.(78) On the contrary, the Fourteenth Amendment would have supplied the missing constitutional rationale on which Dred Scott could have rested with impunity.

Our district judge's decision would be no better if he added that welfare laws foster negative stereotypes about the poor or create unhealthy dependencies among them. Those who believe that welfare is bad for its intended beneficiaries may be right or wrong. But at least they have had the good grace not to argue that putting bread on a poor man's table violates his equal protection rights. When laws grant benefits to certain individuals, those excluded from the law's benefits have the logical form of an equal protection claim. But we are dealing here with alleged harms to affirmative action's beneficiaries. Some people think that handicap-accommodation laws unhealthily lead the disabled away from self-reliance; perhaps so, but who would argue that building a wheelchair access ramp violates the equal protection rights of those in wheelchairs?

Against these points, there are three principal potential objections. The first claims that affirmative action is "paternalist" and therefore requires a different analysis. The second claims that affirmative action involves race and therefore requires a different analysis. The third is an argument from precedent; it claims support for the stigmatization rationale from Brown v. Board of Education.(79) Let's consider these objections in turn.

1. Paternalism

The paternalism argument, as illustrated by Justice Thomas's opinion in Adarand,(80) claims that the preferential treatment affirmative action bestows upon minorities is in fact "poisonous" and that affirmative action is constitutionally noxious to members of the very groups it ostensibly benefits. Justice Thomas is surely right in implying that paternalist laws can give rise to a constitutional claim on the part of their ostensible beneficiaries. Thus, women have an equal protection claim against a law that excludes them from the practice of law, even if this exclusion is said to be in their best interests. The reason is that the legal "benefit" granted in such a case is in reality a denial of a good or freedom enjoyed by others, leaving the "beneficiaries" with a colorable equal protection claim that this "benefit" is in fact a constitutional injury.

Affirmative action programs, however, are not typically paternalist. Let's define a standard affirmative action plan as one that offers preferential treatment to minorities in the allocation of desirable opportunities, but leaves them free to opt out of this preferential treatment if they so choose. Such a plan is not paternalist. On the contrary, if a government official prevented minority group members from taking advantage of such a plan, claiming that affirmative action was against their own interests, that would be an instance of racial paternalism.

To be sure, whenever a law grants to some persons (the poor, veterans, billion-dollar corporations) valuable benefits denied to others, the law can be called "paternalist" in the sense that it reflects a legislative decision that the beneficiary group needs or ought to receive special state assistance. But this sort of "paternalism" is not constitutionally actionable. At least it gives no equal protection claim to the law's beneficiaries. The charge of paternalism states an equal protection claim when officials have denied to certain individuals goods or liberties enjoyed by others on the putative ground that these individuals cannot be trusted to act in their own best interests. Hence, in the constitutional sense, standard affirmative action measures cannot be charged with paternalism.

2. Racial Classifications and the Potential Self-Invalidation of Current Strict Scrutiny Doctrine

Turn, then, to the much stronger defense of the harm-to-society and harm-to-minorities arguments against affirmative action: the defense that race is different. Racial divisiveness and racial stigmatization, it may be said, differ categorically from the divisiveness and stigma potentially caused by welfare laws or indeed by any nonracial laws. This categorical difference (the argument would go) is attributable both to the special dangers of racism in American society and to the special role of race in the history of the Equal Protection Clause. Surely that Clause authorizes courts to be peculiarly sensitive to the evils of race discrimination. Surely there is nothing illogical in a court's decision to impose strict scrutiny on all racial classifications because of their propensity to foster racial polarization and stigmatization.

Yes, there is. Consider the primary rival to affirmative action in most settings: reliance on standardized tests. Such tests are also racially polarizing. They have almost certainly done far more than affirmative action to foster notions of racial inferiority. Would any opponent of affirmative action conclude that state use of standardized tests should therefore be held unconstitutional or subject to strict scrutiny? Of course not. Or to return to welfare: A number of the most prominent welfare programs have proven to be divisive and stigmatizing not only along class lines, but also along racial lines. Is strict scrutiny therefore applicable to welfare as well?

To be sure, standardized tests and welfare programs are facially race-neutral, and strict scrutiny does not apply to race-neutral laws in the absence of some showing of an invidious racial purpose. But this point of law does not support current affirmative action doctrine. Instead, it undermines the stigmatization and polarization rationales that have played so large a part in the opinions holding affirmative action unconstitutional.

The well-established law throughout equal protection jurisprudence is that unintended racial harms are not actionable. In other words, state action with palpable, adverse disparate impact upon racial minorities is perfectly constitutional--and not subject to any heightened scrutiny--so long as this impact is unintended.(81) How then can strict scrutiny apply to affirmative action because of affirmative action's unintended racial consequences? Isn't it just a little strange that the only laws in our entire system invalidated because of their inadvertent harm to minorities should be laws that disparately help minorities and only speculatively hurt them?

A supporter of current doctrine has to explain why affirmative action's unintended racial harms trigger strict scrutiny when hornbook equal protection law holds that unintended racial harms do not trigger strict scrutiny. He cannot argue that the hornbook rule is wrong. For only the hornbook rule saves him from the conclusion that standardized tests, which are also racially divisive and which also engender notions of racial inferiority, are as unconstitutional as affirmative action. As a result, only one line of defense remains available to him.

He must say that race-classifying laws as a general matter are far more likely than race-neutral laws to engender intolerable, inadvertent racial consequences such as polarization and stereotyping. Some race-neutral state action, he will concede, such as the use of standardized tests, might inadvertently cause invidious racial consequences. But judges cannot be expected to evaluate the racial consequences of every law on a case-by-case basis. In the great run of cases, race-classifying laws threaten pernicious racial consequences to a far greater extent than do any other laws. Hence they are properly categorically subject to strict scrutiny. This line of reasoning is the only way a supporter of current doctrine can preserve the claim that affirmative action's inadvertent racial consequences call for strict scrutiny, while defending the line of cases absolving race-neutral laws from strict scrutiny even when their racial consequences are just as severe.

But this line of defense fails. Understood as a bulwark against inadvertent polarization and stigmatization, strict scrutiny of race classifications cannot survive strict scrutiny. Measured against this interest, strict scrutiny of race classifications becomes a blunderbuss tool, leaving untouched a good deal of state action that undoubtedly promotes notions of racial inferiority, and striking down some state action that is not particularly polarizing or stigmatizing. Defended in terms of unintended racial consequences, strict scrutiny of racial classifications uses race both over-inclusively and under-inclusively. It thereby becomes a kind of logical monstrosity, a snake consumed by its own jaws. Strict scrutiny of race classifications may be the first instance in our jurisprudence of a constitutional doctrine unconstitutional under itself.

To be strictly accurate, we cannot know whether current affirmative action doctrine is self-invalidating without knowing exactly how current doctrine defines racial classifications. Under the pertinent precedent, though, the Court has indeed found a racial classification, requiring strict scrutiny, in state action that imposes special obstacles on race-conscious laws, especially when these laws have been designed to help minorities.(82) If this line of cases remains good law, then judicial insistence on strict scrutiny of all racial classifications would itself be a racial classification within its own meaning. If so, then insofar as the doctrine is defended as protecting against inadvertent racial consequences, current strict scrutiny doctrine is indeed self-invalidating.

From a historical point of view, this paradoxical result is unsurprising. A law that provides, "Whatever privileges or immunities are enjoyed by whites, must not be denied to blacks or members of any other race," would seem to be race-classifying on its face. But this proposition captures exactly the kind of law that the Fourteenth Amendment was enacted to bring about; indeed the Fourteenth Amendment might itself be redescribed as containing at its core exactly this legal proposition. Yet current doctrine, insofar as it is based on unintended racial polarization or stereotyping, would seem obliged to regard such a law as deeply constitutionally suspect. Nothing was more divisive of American politics than the Civil War. Nothing was more certain to lead to a racial politics than the rights given to blacks by the Thirteenth, Fourteenth, and Fifteenth Amendments. Any principle derived from the Civil War, from emancipation, and from those amendments must come perilously close to self-invalidation if it deems constitutionally suspect all governmental action that inadvertently threatens harmful racial consequences.

Regardless of whether current strict scrutiny doctrine officially requires the strictures of strict scrutiny to apply to itself (and hence officially invalidates itself), its inability to satisfy these strictures remains telling. For whenever state action cannot satisfy strict scrutiny in matters directly addressing race relations, the inference becomes available that the true purpose of the state action is not what the state actors claim it is. This inference applies here.

Opponents of affirmative action do not really demand strict scrutiny because of the potential fostering of notions of racial inferiority. If they did, they would demand strict scrutiny of standardized tests as well. At the very least, they would have to regard the racial disparities in standardized test scores as a strong (even if not compelling) reason against the use of such tests, which they nearly never do.(83)

It is precisely in its unintended consequences reasoning (whether of the harm-to-minorities or harm-to-society variety) that current doctrine commits itself to the cost-benefit view of heightened scrutiny. For after all, if inadvertent consequences trigger strict scrutiny, then strict scrutiny cannot be understood as smoking out concealed purposes. It must instead be understood as testing whether the unintended harms at issue are justified by offsetting benefits. The reason that the unintended consequences defense of strict scrutiny cannot survive strict scrutiny is that it mistakes the true purpose of heightened scrutiny, which is not to protect against inadvertent effects but to smoke out unacknowledged purposes.

The disparate impact cases confirm this conclusion. These cases do not rest on the idea that the unintended consequences of race-neutral laws will tend, over the great run of cases, to be less harmful to minorities or to society than those of race-classifying laws. They rest rather on the "basic equal protection principle that the invidious quality of a law claimed to be racially discriminatory must ultimately be traced to a racially discriminatory purpose."(84) The Court has "repeatedly" affirmed the "principle that an invidious purpose must be adduced to support a claim of unconstitutionality."(85) This "basic equal protection principle" cannot be reconciled with striking down affirmative action (or applying strict scrutiny to it) on the basis of its unintended invidious racial consequences. It can be reconciled only with an analysis of affirmative action that examines the permissibility of affirmative action's purposes.(86)

Central though it has been in the Justices' reasoning, the argument that affirmative action inadvertently polarizes or stigmatizes has no place in equal protection analysis. Laws whose purpose is to help minorities by granting them opportunities denied to whites cannot be invalidated by reason of unintended racial costs or inadvertent harms to the intended beneficiaries.

3. Brown

But how can this conclusion be squared with Brown v. Board of Education?(87) Brown rested on the express ground that separate-but-equal public schooling, no matter how equivalent in tangible respects, generated in black children "a feeling of inferiority as to their status in the community" that the Constitution did not tolerate.(88) Doesn't Brown refute the conclusion arrived at a moment ago? If affirmative action is in fact stigmatizing, isn't the Court right under Brown to invalidate it, or at the very least to treat it as constitutionally suspect?

No, because Brown did not involve stigmatization as a law's unintended consequence. Separate-but-equal did not inadvertently foster negative racial stereotypes. To be sure, the Court denied the invidious purpose behind separate-but-equal in Plessy v. Ferguson,(89) but in Brown and the decisions that followed Brown, the Court decisively repudiated Plessy, vindicating Justice Harlan's lone dissent in that case--a dissent that chastised the majority precisely for refusing to acknowledge the true purpose of Louisiana's separate-but-equal statute.(90)

No one will doubt, I take it, that American separate-but-equal laws were in fact and by design a legal expression of black degradation. "The thin disguise of `equal' accommodations," as Justice Harlan wrote in Plessy, "will not mislead any one."(91) Hence the existence of an invidious racial purpose in separate-but-equal laws cannot really be in dispute. The only question, then, is whether to understand Brown's repudiation of Plessy as committing the Court to striking down laws with invidious racial effects (purposeful or not) or to striking down laws with invidious racial purposes.(92)

By stressing the elementary school social science referred to in Brown,(93) one can certainly generate an effects-only reading. But the sweeping abolition of the entire American apartheid regime in the cases following Brown(94) took the Court far beyond the special susceptibilities of elementary schoolchildren. In these decisions, the Court expressly recognized the invidious purpose underlying racial separation laws. The true "purpose" of these laws, as the Court wrote in Loving v. Virginia,(95) was "invidious racial discrimination."(96) They were "measures designed to maintain White Supremacy."(97)

Brown has to be read in light of the great truth for which it came to stand: that separate-but-equal was nothing other than a set of laws designed to keep blacks in their place and that it was, as such, categorically unconstitutional. This conclusion is dictated not only by the entire line of post-Brown separate-but-equal cases, but also by the entire line of post-Brown disparate impact cases as well. The effects-test reading of Brown cannot be squared with the "basic equal protection principle"(98) repeatedly affirmed in the disparate impact cases. As discussed above,(99) this basic principle makes invidious purpose a necessary element of unconstitutionality and holds that a law is not constitutionally suspect just because it has unintended consequences adverse to members of racial minorities. If under Brown state action fostering notions of racial inferiority ought to undergo strict scrutiny, then not only should affirmative action be held unconstitutional, but so too should its principal alternative: reliance on the SAT, the LSAT, and all the other standardized tests that, far more than affirmative action, "teach[] many that because of chronic and apparently immutable handicaps, minorities cannot compete with them without their patronizing indulgence."(100)

B. Injury to Whites

Let's turn finally to the injury that affirmative action inflicts on whites (or others excluded by race from affirmative action's benefits). How can it be constitutional for a state actor ever to discriminate among persons on the basis of race? If the Equal Protection Clause means anything, surely it means that government may not treat one person worse than another because of the color of his skin.

This is the rock bottom claim of those who oppose affirmative action. It is by far their most solid argument, and it is their only constitutional argument. Here at last they make a claim sounding not in unintended consequences, but in purposeful invidious discrimination. How would this claim be decided on the analysis of equal protection scrutiny offered here?

It would have to be decided by determining what impermissible purposes make state action an instance of unconstitutional race discrimination under the Equal Protection Clause. Does that Clause categorically forbid state actors from purposefully differentiating persons on the basis of color, or is the proscribed purpose to be defined more narrowly? The contours of this debate are well known. Everyone understands that the case for affirmative action's constitutionality depends ultimately on a claim that the Equal Protection Clause stands not for a principle of categorical colorblindness, but for a principle banning only those color-based measures that meet a further criterion of invidiousness. In what follows, I will not undertake anything like a full elaboration of this debate. I address only the question of how the preceding arguments of this Essay would bear on it.

1. Interpretive Method: Paradigm Cases

Part I of this Essay showed that the contemporaneous acts of the Framers of the Fourteenth Amendment do not support a categorical anti-affirmative action position. Nor can this position be said to follow either from the "plain meaning" of the Equal Protection Clause, or from the process-based theory so influentially expounded by John Hart Ely.(101) But it would be feckless to go through various interpretive methodologies, to line them up against colorblindness, and to conclude that the case is thereby closed. To do so would be to miss altogether the appeal and force of colorblindness in constitutional law. To rebut constitutional colorblindness, it is essential first to understand the arguments in its favor.

One argument that might be said to favor the Court's current position is the argument from simple morality. Can't everyone see the manifest unfairness and injustice of denying an individual a job, a place at a school, or any other valuable opportunity just because the person has the wrong skin color? This approach to affirmative action might in turn point to a line of academic thought holding that constitutional interpretation should largely consist of seeking the best results that moral thinking can deliver.(102) On this view, affirmative action's constitutionality becomes a matter of moral philosophy.

But if this were so, and if we were to credit the work of the best moral philosophers who have turned their attention to affirmative action, it would seem that current doctrine is again unsupported. For the philosophers inform us that from a moral point of view, the standard traits used to allocate valuable positions--traits like intelligence and athleticism--are as irrelevant and undeserved as skin color.(103) Affirmative action, it is argued, inflicts no more moral injury on a white who loses out to a black than athletic programs inflict on a short person who loses out to a seven-footer.(104) Tough luck (short white man), this argument goes, but don't start crying the beloved Constitution.

I have never seen a believer in colorblindness remotely impressed by this argument. Perhaps this is because true believers are never impressed by logical argument. Or perhaps it is because the argument profoundly understates the degree or nature of the unfairness suffered by those whose futures are diminished due to the adoption of race-discriminating standards. Either way, however, the inquiry into affirmative action's morality misses the constitutional point.

Affirmative action surely is unfair to whites, sharply and deeply so. But there is a peculiar notion at work in the conventional moral case for and against affirmative action's constitutionality. Those who argue that affirmative action is unfair and therefore unconstitutional--or moral and therefore constitutional--seem to think that the rest of the legal system would not have to be radically overhauled if constitutional law condemned all governmental unfairness. Our society is massively unfair. If justice writ large were the question, the Supreme Court should begin by striking down the entire set of laws under which thousands of our children are born into an inferno of poverty and despair, while others are born into comfort and security. There is plenty of good reason, but no particular justice or fairness, in distributing goods on the basis of the marginal market value of talents. I am far from holding that considerations of justice ought to play no role in constitutional law. But constitutional law is not moral philosophy, and unfairness is not unconstitutionality.

The constitutional case for colorblindness cannot be grounded with any confidence in pure morality (and neither can the case against it). But if not from moral philosophy, nor from originalism, nor from textualism, nor from processualism, where then does colorblindness get sufficient strength to command a majority position in the Supreme Court? Colorblindness in equal protection doctrine is testimony to the signal importance of paradigm cases in constitutional interpretation.(105) All the paradigmatic applications of the Equal Protection Clause--from the black codes of the 1860s, to the separate-but-equal train cars that should have been struck down under Justice Harlan's dissent in Pleasure,(106) to race-segregated schools, to the miscegenation laws finally invalidated in Loving(107)--are race cases. In every one of them, the core evil was racial discrimination. From what source does current affirmative action doctrine draw its strength? From Justice Harlan in Plessy: "Our constitution is color-blind."(108) From Loving: "At the very least, the Equal Protection Clause demands that racial classifications ... be subjected to the `most rigid scrutiny.'"(109) And from declarations, even if not always lived up to, like this one: "Distinctions between citizens solely because of their ancestry are by their very nature odious to a free people whose institutions are founded upon the doctrine of equality."(110)

Friends of affirmative action cannot hide from these paradigm cases and paradigmatic utterances, which provide the core meaning of the equal protection of the laws. If the Equal Protection Clause means anything, it means that the black codes, separate-but-equal laws, and racial miscegenation statutes were unconstitutional. Equal protection jurisprudence is centrally a task of saying what it means to honor the nation's commitment to abolish all such laws. Any reading of the Equal Protection Clause that does not accord these paradigm cases pride of place--any interpretation that cannot, without bending or breaking, embrace these paradigm cases at its core--is not a satisfactory account.

That is why, despite their lack of grounding in any of the academically privileged interpretive methodologies, the colorblinders still have the strongest possible constitutional starting point. Colorblindness offers itself as a conclusion drawn at great price from a certain constitutional struggle. Alexander Bickel's formulation is the most arresting:

The lesson of the great decisions of the Supreme Court and the lesson of

contemporary history have been the same for at least a generation:

discrimination on the basis of race is illegal, immoral, unconstitutional,

inherently wrong, and destructive of democratic society. Now this is to be

unlearned and we are told that this is not a matter of fundamental principle

but only a matter of whose ox is gored.(111)

Moral philosophy cannot unteach this "lesson." Nor can any interpretive methodology that attempts to bypass the "great decisions" to which Bickel refers. If this lesson can be rebutted at all, it can be rebutted only by recourse to the very paradigm cases on which it purports to stake its claim.

2. Interpreting the Paradigm Cases

Thus the question becomes whether the partisans of colorblindness can state a constitutionally impermissible purpose that captures the Equal Protection Clause's paradigm cases. And here we find a strange thing. There is a profound dissonance between constitutional colorblindness as it is currently understood and the paradigm cases from which it is drawn.

In order to capture Brown and all the other separate-but-equal cases, partisans of colorblindness cannot say merely that it is unconstitutional for any state actor purposefully to treat a person worse because of his color. Brown stands for the proposition that racial separation laws are unconstitutional even if they treat every member of every race to the identical norms and identical facilities. Hence partisans of colorblindness must say that the very act of purposeful racial classification--the very act of deliberate differentiation on the basis of color--is the constitutional evil. And this is of course what they do say. Professor Van Alstyne, for example, has given this principle a characteristically admirable statement:

The point of emphasis here is fundamental.... It is not that when race is

used, all persons identified to each race must be as well regarded as all

persons identified to some other race.... The thing condemned, rather, is

the assignment of entitlements by race. It is the impropriety of the basis of

the assignment ... that constitutes the government's offense.

...

... [O]ne gets beyond racism by getting beyond it now: by a complete,

resolute, and credible commitment never to tolerate in one's own life--or in

the life of practices of one's government--the differential treatment of

other human beings by race.(112)

Now consider: An implication of this principle is that the segregation laws upheld in Plessy and struck down in Brown violated the equal protection rights of whites just as they did those of blacks. There is nothing illogical in this proposition; after all, segregation laws prevented whites from sitting with blacks just as much as they prevented blacks from sitting with whites. Nor is there anything strictly illogical in the further implication of colorblindness, that blacks' equal protection rights are violated by affirmative action just as those of whites are. But if colorblindness betrays no illogic in these propositions, it begins to betray a strange affiliation with Plessy itself.

The Plessy opinion, it will be remembered, rested precisely on the claim that the treatment of whites and blacks was equivalent under the law and hence that there was no unique, constitutionally cognizable injury that separate-but-equal inflicted on blacks:

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.... We imagine that the white

race, at least, would not acquiesce in this assumption.(113)

Justice Harlan, as noted earlier,(114) gave the lie to this hypocrisy:

It was said in argument that the statute of Louisiana does not discriminate

against either race, but prescribes a rule applicable alike to white and

colored citizens. But ... [e]very one knows that the statute in question

had its origin in the purpose, not so much to exclude white persons from

railroad cars occupied by blacks, as to exclude colored people from coaches

occupied by or assigned to white persons.(115)

The true issue in Plessy, as Justice Harlan wrote, was the constitutionality of "state enactments which, in fact, proceed on the ground that colored citizens are so inferior and degraded that they cannot be allowed to sit in public coaches occupied by white citizens."(116) Colorblindness cannot recognize this issue. To set itself against the Plessy outcome, it must perversely align itself with the Plessy opinion. It must, with the Plessy Court, refuse to consider the true meaning--the true purpose--of race-classifying laws. In this sense, today's colorblindness recapitulates the wrongs of Plessy, only in reverse. Under Plessy, no race classification was presumed invidious; the invidious purpose of such a classification was irrelevant; instead, a race classification would be deemed legally injurious to blacks only if it treated blacks tangibly worse than others. Under current doctrine, every race classification is presumed invidious; the noninvidious purpose of such a classification is irrelevant; instead, a race classification is deemed legally injurious to blacks even if it treats them tangibly better than whites. Under Plessy, no race classification as such was constitutionally injurious to blacks, even if degrading blacks was its very purpose. Under current doctrine, every race classification as such is constitutionally injurious to blacks, even if assisting blacks is its very purpose.

But how can this be? How can colorblindness be put on the same side as the Plessy majority, when it was Justice Harlan who, in dissent, wrote that the "constitution is color-blind"?(117) Doesn't Harlan's opinion, together with Brown and Loving,(118) stand for the principle that a law may be found unconstitutional because of its racial classification as such, without more?

Certainly they stand for such a principle, but it must be remembered that all these cases, from Plessy to Loving, were racial separation cases. In all these cases, the mere fact of separate treatment--hence the mere classification itself, without more, without any more tangible inequality being imposed--had an intangible quality that did indeed render it "inherently unequal"(119) and even "odious to a free people."(120) This intangibility has been described in terms of stigma or badges of inferiority, but another term more accurately captures it. The relevant intangibility was nothing other than intangibility--untouchability--itself.

Untouchability is a universal marker of caste. A regime of untouchability prohibits commingling between specified classes of persons, one of which is deemed unclean, degraded, unfit to be mixed with. Emblematically, untouchability forbids intercourse between persons. Not for nothing did the Plessy Court cite "[l]aws forbidding the intermarriage of the two races" as a primary example of legislation that might go by the boards if blacks could constitutionally force their way into Louisiana's all-white compartments.(121) Not for nothing did the Plessy Court hold that the Fourteenth Amendment "could not have been intended ... to enforce ... a commingling of the two races upon terms unsatisfactory to either."(122)

Opponents of affirmative action have seized upon the principle that correctly emerged from the separate-but-equal decisions, but they have distorted it. Racial classifications as such, without more, are abominable where they legalize untouchability and hence legalize caste. Without the concept of caste or second-class citizenship, all the force is lost from the idea that distinctions based on color are "odious to a free people." (For caste distinctions, like titles of nobility, are indeed repugnant to the ideal of a people of equal citizens engaging in a collective project of self-government.) Without its imputation of untouchability, separate-but-equal is not--in the only words of Brown really worth remembering--"inherently unequal." And without caste, Justice Harlan's assertion of the Constitution's colorblindness must be seized from its context, and his recognition of the true constitutional evil of the legislation in Plessy must be suppressed. For Harlan's statement of colorblindness was already a statement of the anticaste principle:

The white race deems itself to be the dominant race in this country. And so

it is, in prestige, in achievements, in education, in wealth, and in

power... But in view of the Constitution, in the eye of the law, there is

in this country no superior, dominant, ruling class of citizens. There is

no caste here. Our constitution is color-blind ....(123)

Thus, the earliest--and possibly the most cited--statement of constitutional colorblindness itself explained the imperative of eliminating race classifications in terms of a principle against caste legislation.

In what follows, I will assume the availability of this anticaste alternative to--or more accurately, this anticaste interpretation of--constitutional colorblindness, but I will offer no further elaboration of it. Pathbreaking work was done by Owen Fiss twenty years ago,(124) and others have mined the field well since then.(125) add only one qualification. Proponents of an anticaste principle could in principle take an effects-test view, arguing against all laws whose consequence, purposeful or not, is to perpetuate the diminished social status of a particular group.(126) One difficulty with such a view is the fact that the basic legal framework of a market economy has this consequence. In any event, for all the reasons discussed earlier, the analysis here would reject an unintended consequences view of the anticaste principle. As I use it, the anticaste principle would apply only to state action whose purpose was to subject individuals to second-class citizenship or lower-caste status.

Under such a principle, it is pretty clear that standard affirmative action programs would be valid. If there is a colorable argument that whites are purposefully reduced to lower-caste status when the proportion of white students at prestigious public universities declines from ninety-five percent to eighty-five percent, I have yet to hear it. If there is a colorable argument that whites become second-class citizens when an urban police force takes measures to have some substantial number of black officers on the streets, I have yet to hear it. If there is a colorable argument that any white person is stamped with a badge of status inferiority when a white President and a white Senate appoint a black man to the Supreme Court in part because of the color of his skin, I have yet to hear it. The purpose of such actions has nothing to do with expressing caste animosity or imposing caste status. They would not, therefore, count as invidious racial discrimination under the Fourteenth Amendment.

3. Consistency

This conclusion, however, is not yet complete. The preceding section was designed to show that an anticaste principle captures the Equal Protection Clause's paradigm cases better than does a principle of colorblindness. But nothing was said against adopting both principles.

If a partisan of categorical colorblindness adds to his basic principle the recognition that racial classifications reflecting caste distinctions inflict a worse or supplemental constitutional injury, then has he not answered all the objections raised against him in the previous section? For now he can reaffirm that all racial classifications are unconstitutional, while recognizing the special injury that separate-but-equal laws inflicted on blacks. Now he can say that his two-principle approach achieves a fundamental consistency in the law that the anticaste principle does not. For under his approach, government must treat all persons alike, regardless of color, while on the anticaste approach, government may inflict some harms on whites that it could not inflict on blacks. Surely this consistency, he might argue, is more respectful of the Constitution's command to give every person the equal protection of the laws.

Let's assume that this two-fisted approach to racial classifications would capture the paradigm cases as well (even if not as parsimoniously) as would the single-principle anticaste approach. It remains the case--if the purposive view of equal protection scrutiny discussed above is correct--that categorical colorblindness must be rejected. For the consistency that colorblindness claims is inconsistent with vast quantities of equal protection law, including the law it most wants to sustain.

Suppose a state agency decides that certain workers require protection from ultraviolet light. The agency adopts regulations requiring the relevant employers to provide such protection, and these regulations rationally make the degree of requisite protection depend on the worker's skin color. What does colorblindness have to say to this banal hypothetical?

The answer cannot be that the regulations would stand because they are rational. Under current doctrine, mere rationality review would not apply; and in any event, no one disputes that affirmative action is at least rationally related to legitimate state purposes. The answer also cannot be that the regulations would satisfy strict scrutiny. For one thing, the categories adopted might be rough and imperfect, improving the health and safety of all affected workers, but not with the niceness of narrow tailoring. But whether or not this was so, what function would strict scrutiny serve here?

Once all the rhetoric of unintended consequences and all the trappings of cost-benefit analysis are stripped away, strict scrutiny has no role to play after a court has determined that the state has acted with a constitutionally proscribed purpose. And the state would have so acted, if that purpose is defined as the deliberate differentiation of persons on the basis of color. On the view that the Constitution forbids government from purposefully differentiating among persons on the basis of color, there is no call for strict scrutiny of race-classifying laws. There is nothing to smoke out. Every such law would be unconstitutional per se.

By contrast, the anticaste principle might well be consistent with imposing heightened scrutiny on the state's regulations. It is conceivable, after a, that the state's worker-safety rationale for the regulations is a smoke screen masking an ulterior purpose. The only way to test for that possibility, short of smoking-gun evidence unlikely to exist, would be to smoke it out by measuring the fit between the regulations and the claimed worker-safety rationale.

To be sure, an opponent of affirmative action could offer further qualifications to rescue his position. Where a color classification is directed at demonstrable biological differences, he might say, the classification need not be unconstitutional per se, and it can be permitted to stand if it passes a means-ends test calculated to determine whether the law is genuinely directed at demonstrable biological differences. We might then ask why the race-conscious integration measures enacted after Brown were constitutional; they deliberately differentiated on the basis of race, and they were not directed at any biological differences. Here the affirmative action opponent would probably say that the strictly remedial nature of such measures--righting identifiable race-based violations of law--distinguishes them from affirmative action.

All these qualifications are coherent, but they surrender the principle of colorblindness. The anti-affirmative action position now would not be that the Constitution equally proscribes all purposeful state differentiation on the basis of color. Instead, it would be that the Constitution proscribes such state action only when done with impermissible (nonbiological, nonremedial) purposes. To be sure, the opponent of affirmative action might try to flip this conclusion around, saying that the Constitution allows racial classification only when there is compelling justification for it. This response, however, (1) would fail to make good sense of the ultraviolet light regulations mentioned above, whose constitutionality does not plausibly depend on how crucial the protection is, but rather depends on the absence of any invidious purposes; and (2) would merely resurrect the cost-benefit understanding of heightened equal protection scrutiny, which at this point in the argument we have already rejected.

In any event, through these qualifications, the opponent of affirmative action will have conceded that deliberate racial classification is not a constitutionally impermissible purpose as such. Instead, the ultimate question will be one of determining what kinds of purpose render a racial classification permissible and what kinds render such a classification an act of invidious racial discrimination under the Equal Protection Clause. At that point, colorblindness is no longer the operative principle, and a given act of racial classification can be ruled unconstitutional only by a further inquiry into purpose. Though this proposition will chagrin the champions of colorblindness, it is merely a matter of consistency--of not making an exception for affirmative action. Consistency does not reside in treating all race classifications alike. On the contrary, that kind of consistency is inconsistent with the equal protection treatment of virtually every legal classification outside the arena of affirmative action.

Wealth classifications have been struck down only when used to deny fights to the poor, never when used to assist diem. A law denying opportunities to the handicapped might be suspect, while laws granting them special accommodation are perfectly constitutional. State laws granting special assistance to veterans are constitutional, but it is unimaginable that courts would let pass any state law singling out U.S. veterans for adverse treatment.

When government assists the poor, the handicapped, or veterans, why is there no cognizable equal protection claim for the better off, the able-bodied, or those who never served? It is no answer to say "because the mere rationality test applies to these classifications." The very question at issue is why state laws denying benefits to the unoffending rich, the innocently able-bodied, or the law-abiding civilian population, are subject to little or no equal protection scrutiny. (And in any event, if a state were to pass a law denying rights or privileges to the poor, to the handicapped, or to veterans, I assure you that the true mere rationality test would not be in play.) In other words, we are asking the question that precedes the determination of the appropriate level of scrutiny. The answer is that the charitable purpose behind these assistance laws absolves them from heightened equal protection scrutiny. They are efforts to help the hard-pressed, in no way imposing on them or anyone else a lower-caste status.

The harm to those excluded from affirmative action is real. In standard affirmative action plans, the principal harm takes the form of opportunity costs inflicted on those denied positions because of their race. These costs are high, they are demoralizing, and they deserve to be taken very seriously by all who must decide whether to support or oppose affirmative action as a matter of policy. But they do not differ in a constitutional sense from the harms inflicted on the better-off by programs that offer special opportunities to the poor, on the able-bodied by laws that require special accommodations for the handicapped, or on the civilian population by state action that grants preferences to veterans. All these programs are instances of affirmative action. It is impossible that the only kind of affirmative action made unconstitutional under the Civil War Amendments is the kind that would offer assistance to blacks.

C. Consequences for Equal Protection Law: From Suspect Classifications to Suspect Classes

If everything written to this point were accepted, what form would equal protection doctrine take with respect to strict scrutiny and affirmative action? The shift from a smoking-out to a justificatory use of strict scrutiny parallels a more obvious shift in the Court's equal protection jurisprudence: from classes to classifications. The tiered framework of equal protection review originally developed not with the notion of "suspect classifications," but with the recognition of "suspect classes." When laws explicitly imposed burdens on certain "suspect classes" of persons, the Court held, the suspicion that something constitutionally forbidden was afoot justified more stringent scrutiny. Which classes of persons were "suspect" in this way? One characteristic repeatedly held necessary to make a class "suspect" was a "history of purposeful" discrimination.(127)

At this point in the doctrine's development, strict scrutiny made sense. For when a state singles out a class of persons that has been subject to widespread, invidious prejudice and denies to members of this class rights or liberties that others enjoy, there is excellent reason to fear that the government has acted deliberately to reduce these persons to a second-class legal status. As a result, whites never could have been deemed a suspect class under equal protection doctrine as that concept was consistently developed and articulated prior to the affirmative action cases. Today, in effect, whites are a suspect class--even though the Court has never explained this result, which contradicts everything the Court ever said about the criteria necessary to establish a class as suspect for equal protection purposes.

The original purpose behind suspect class strict scrutiny explains why the integration measures adopted in the wake of Brown were not and should not have been evaluated under strict scrutiny. Everyone understood in advance that such measures had a racial purpose. They were supposed to integrate the races, and they were supposed to do so for the particular benefit of blacks. Accordingly, there was nothing to smoke out. Everyone understood that strict scrutiny did not apply to a measure of which no one supposed that there was an ulterior, racially invidious objective.

Shifting the focus from suspect classes to suspect classifications, current affirmative action shifts from the smoking-out to the cost-benefit view of strict scrutiny, telling us that equal protection rights are not rights at all, but interests possessed by individuals that must give way when countervailing (and not necessarily very extraordinary) state interests outweigh them. Current doctrine tells us that violations of the letter and spirit of the Constitution are perfectly constitutional. Current doctrine either cannot explain why integration was constitutional or else cannot explain why racial expulsion, suitably tailored, would be unconstitutional. (It might explain one or the other by broadening or contracting the compelling state interest test, but it cannot explain both.) And current doctrine cannot explain why affirmative action's unintentional racial costs trigger strict scrutiny, while everywhere else unintended racial harms are not actionable. This is what follows when strict scrutiny loses its moorings.

Justice Rehnquist was right in Craig v. Boren:(128) Classifications cannot be talismanic.(129) A court is obliged to consider the "rights involved and the persons affected"(130) before deciding what level of scrutiny to apply. One doctrinal formulation consistent with this admonition would be as follows. Where a law denies members of a suspect class any rights or privileges enjoyed by others, heightened scrutiny applies. And where an untouchability law treats the races formally equally, no special scrutiny applies; the law is unconstitutional--period. But standard affirmative action measures do not deny members of suspect classes any rights or privileges enjoyed by members of nonsuspect classes, and they are not untouchability laws.

Nevertheless, some circumstances still would exist in which affirmative action would be subject to strict scrutiny. For example, if in a particular jurisdiction members of a minority race had legislative control, and if they adopted laws burdening whites, strict scrutiny might then be appropriate because in such a case there might be ground for considering whites a suspect class.(131) Or there might be reason to believe that a particular affirmative action program secretly was intended to harm members of a suspect class (for example, a program granting special priority for minorities in an involuntary medical study). Or if, in the name of affirmative action, someone proposed to deny whites eligibility altogether at a university, or to subject whites to random police searches and seizures, or to do anything else to make whites "see what it feels like," the underlying objective could have crossed into invidiousness. Indeed, I think it would be improper for an affirmative action program to use race to take from whites (or anyone else) something they already have, as opposed to denying them an opportunity for which they are merely applying. The reason would be--to venture a hazardous distinction--that, on the level of the intangible, such a program would pass from merely withholding a benefit on the basis of race to penalizing individuals on the basis of race.(132) On this view, although deliberately offering individuals special opportunities on the basis of race would sometimes be constitutionally permissible, no state actor could ever deliberately penalize any individual--where penalizing is understood through a concept of taking away something (other than an unvested opportunity) that a person already has--on the basis of race. Thus, standard affirmative action plans would be constitutional, but a special tax levied only on whites or members of any other race would not.

Without doubt, racial discrimination is the paradigmatic Fourteenth Amendment violation. But that is because racial discrimination is itself paradigmatic of treating individuals as members of a lower order. Race-discriminatory laws, as Justice Harlan wrote, "proceed on the ground that colored citizens are so inferior and degraded" that they deserve to be treated as untouchable, as inferior, or as second-class members of the nation.(133) Not all racial classifications treat persons that way. If a racial classification treats no one that way, it is not purposefully invidious within the meaning of the Fourteenth Amendment.

IV. POSTSCRIPT: ROMER AND THE CIVIL RIGHTS INITIATIVE

This postscript briefly notes the implications of the preceding arguments for two issues beyond the constitutionality of affirmative action.

A. Romer v. Evans

In Romer v. Evans,(134) the Supreme Court invalidated a Colorado constitutional amendment that essentially made it a state constitutional right for anyone (with the possible exception of common carriers) to discriminate against persons on the basis of homosexuality.(135) I am not going to provide another academic defense of Romer on grounds other than the ones asserted by the Court itself. On the contrary, I want briefly to show how the Court's actual reasoning illustrates the analysis offered in this Essay.

The Romer Court observed that the Colorado law seemed to defy efforts to review the rationality of the fit between its provisions and the asserted governmental interests it was supposed to advance. "Amendment 2 fails, indeed defies, even this conventional inquiry."(136) The law, the Court declared, "confounds this normal process of judicial review."(137) Within the reigning framework of equal protection review, these statements are strange. They do not quite say that the law bears no rational relationship to any conceivable, legitimate state purpose (the legal test under mere rationality review). Rather, without adopting a higher standard of review, the Court suggested that Colorado's law was somehow almost invalid on its face, regardless of any possible fit between means and ends. "[T]he amendment has the peculiar property of imposing a broad and undifferentiated disability on a single named group, an exceptional and ... invalid form of legislation.... It is at once too narrow and too broad."(138) The strangeness of these statements was not dispelled by the fact that the Court went on expressly to evaluate (and to strike down) the Colorado law under mere rationality review, holding that the law could be explained only by reference to a constitutionally illegitimate purpose.(139)

Now, Romer's critics are doubtless correct to say that the Colorado law was (as almost any law is) rationally related to conceivable, legitimate state purposes.(140) But the Court was correct in suggesting that mere rationality review was not in fact applicable--even if neither of the other standards of review was applicable either. How can a law be reviewed at all, if not under one of the three standards of equal protection scrutiny? As I have said throughout: When a law's purpose is constitutionally illegitimate, none of the well-known standards of review applies. It makes no difference whether such a law satisfies the rational basis test, intermediate review, or even strict scrutiny. And the impermissible purpose of a law can sometimes be read from its face.

This was so in Brown v. Board of Education itself. The Court in Brown did not apply any of the three standards of review. The opinion contains no analysis of the fit between racial segregation and any of the putative legitimate state interests (avoiding racial conflict in the classroom, satisfying "associational" preferences, and so on) that might have been advanced thereby. In 1950s America, the untouchability imputed to blacks by separate-but-equal was visible on the face of the law. Once the Court became willing to recognize the true purpose of separate-but-equal, no means-ends scrutiny applied, for the very purpose of equal protection means-end review--in all three varieties--is nothing other than to assist in determining a law's true purpose. Hence, the true question in Brown, as in Plessy, was (as noted earlier) solely whether a state could constitutionally enforce a law the purpose of which was to legalize black untouchability. The purposive nature of equal protection review, indicated implicitly by Brown, is explicitly confirmed in Romer.

Moreover, the purpose that the Romer Court found constitutionally out-of-bounds confirms the analysis offered in this Essay. The Court did not hold that a state is constitutionally forbidden to criminalize homosexual conduct. Whether the Court was right or wrong in this non-holding I do not discuss.(141) What the Romer Court did hold was that Colorado had imposed on homosexuals a distinctive, degraded legal status. The Court began with a quotation from Justice Harlan's Plessy dissent: "[T]he Constitution 'neither knows nor tolerates classes among citizens.'"(142) The Court continued by stating that Amendment 2 was a "status-based enactment" that "deem[ed] a class of persons a stranger to [Colorado's] laws."(143) In other words, Colorado had in effect deliberately thrown the backing of the state behind the view of homosexuals as unfit to commingle on equal terms with the rest of the community.(144) Thus Romer illustrates not only the purposive equal protection analysis advocated in Part II--an analysis that precedes and may obviate entirely the application of the various standards of review--but the caste-based or untouchability understanding of equal protection principles advocated in Part III.

B. The Civil Rights Initiative

Foes of affirmative action, if there are any who have read to this point, will not have found much they like. What follows may be consoling. California's Civil Rights Initiative, which bans governmental discrimination or preferences on the basis of race in contracting, hiring, and academic admissions,(145) may well be a racial classification under the relevant Supreme Court precedent.(146) If this precedent is adhered to, the Initiative ought to receive strict scrutiny, and so scrutinized, the Initiative should have a difficult time.(147)

But how could this be? Doesn't the Civil Rights Initiative merely codify as a matter of state law what the Supreme Court has come close to laying down as a constitutional requirement? How can it be unconstitutional for a state to ban race preferences in public decisionmaking when the Supreme Court has held that race preferences in public decisionmaking are presumptively unconstitutional? If there is a contradiction here, it lies in the Court's classification-driven doctrine itself For as we have seen, the Court's classificationism creates the spectacle of a constitutional doctrine very nearly unconstitutional under itself.

In any event, current doctrine should, if fairly applied, cause serious constitutional difficulties for California's effort to eradicate affirmative action. The "benign" motivations underlying the Initiative (benign from the Court's point of view) should not excuse it from strict scrutiny, and the singling out of racial (and gender) preferences should therefore be a cause of serious concern to friends of this law.(148)

But on the analysis that I have suggested, the main thrust of the Initiative would not face serious difficulty. The Initiative is not an untouchability law, and its classification would not trigger heightened scrutiny, because it denies no rights to minorities or women that it bestows upon whites or men. The principal question, therefore, would be whether the law's express purpose (eliminating racial and gender considerations while allowing consideration of almost all other factors) is constitutionally permissible.(149) And the answer to this question would probably be yes. The Constitution permits affirmative action because affirmative action does not force a second-class status or citizenship on anyone. But abolishing affirmative action does not thrust a second-class status or citizenship on anyone either. In other words, foes of affirmative action should stop calling on courts to engage in an activism that they supposedly deplore. Instead they should throw their energies behind their Civil Rights Initiatives, the constitutionality of which turns out to go hand in hand with that of affirmative action itself.

V. CONCLUSION

If I had to choose, I would probably vote to scrap the entire patchwork of affirmative action measures in this country in favor of a massive capital infusion into inner-city day care and educational facilities. But this conclusion goes to affirmative action's costs and benefits, not to its constitutionality. Affirmative action, with all its costs and imperfections, is not inconsistent with the commitment made by this nation when it enacted the Fourteenth Amendment. On the contrary, there is a reason why Congress enacted "colored relief" legislation at the same time this commitment was laid down. The mason is justice--constitutional justice.

Consider the phenomenal success of those ethnic groups whose cultures enforce a strong work ethic and emphasize education. Consider how such a culture is carried on: through families, from one generation to the next. Consider how long it would take for such a culture to be obliterated if members of the group were seized, carried to another place, with no records of its former learning, with its language killed, and with its family structure destroyed by separating children from parents. How long would it be before all the cultural strength of that group, acquired and maintained perhaps for thousands of years, was crushed and replaced by its opposite? Two generations, three, four? America still Pays for the crimes of slavery. We may never stop paying for them. We never ought to, until the day comes when black children have the same opportunities that others have. Affirmative action is no cure-all. It is only a small effort to do some good.

All the same, I have two recommendations for those who favor affirmative action. First, institutions with affirmative action plans should be open about them or scrap them. If the burdens that an honest affirmative action program imposes on its beneficiaries are too great to bear, the correct response is not to prevaricate, but to try something new. Second, the pro-affirmative action crowd needs to own up to the weakness of "diversity" as a defense of most affirmative action plans. Everyone knows that in most cases a true diversity of perspectives and backgrounds is not really being pursued. (Why no preferences for fundamentalist Christians or for neo-Nazis?)

The purpose of affirmative action is to bring into our nation's institutions more blacks, more Hispanics, more Native Americans, more women, sometimes more Asians, and so on--period. Pleading diversity of backgrounds merely invites heightened scrutiny into the true objectives behind affirmative action. This heightened scrutiny would quite properly reveal the existence of a race- or group-related purpose, rather than a genuine interest in achieving a representative diversity of perspectives. In fact, the true, core objective of race-based affirmative action is nothing other than helping blacks. Friends of affirmative action, if there are any left, should acknowledge this objective, and they should embrace it--in the name of justice.

(1.) The Supreme Court; Excerpts from News Conference Announcing Court Nominee, N. Y. Times, July 2, 1991, at A14 (quoting President Bush on his announcement of Clarence Thomas's nomination to the United States Supreme Court).

(2.) Act of July 16, 1866, ch. 200, 14 at 173; Act of Mar. 3, 1865. ch. 90. 13 Stat. 507.

(3.) Readers may think that this paraphrase must be exaggerated. It is not. See infra Section II.C.

(4.) See infra notes 81, 83-85 and accompanying text.

(5.) 116 S. Ct. 1620 (1996).

(6.) Cal. Const. art. 1, [sections] 31.

(7.) Senator Orrin G. Hatch, Speech Before the Federalist Society in Washington, D.C. (Nov. 17, 1996); see also Al Kamen, Chairman's Call on the Court, Wash. Post, Nov. 18, 1996, at A19 ("[Senator Hatch] listed a parade of horribles of what could happen if the president gets to nominate even one `liberal activist' to the Supreme Court: `Almost certainly' lower court judges would be allowed to expand use of racial and gender preferences . . . .").

(8.) E.g., United States v. Lopez, 115 S. Ct. 1624, 1642 (1995) (Thomas. J., concurring); Robert H. Bork, The Tempting of America 143-60 (1990),

(9.) Michael H. v. Gerald D., 491 U.S. 110, 128 n.6 (1989) (Scalia, J.).

(10.) Cf., e.g., Lopez, 115 S. Ct. at 1643 (Thomas, J., concurring) (arguing against congressional authority to regulate manufacturing or agriculture because "[a]t the time the original Constitution was ratified . . . the term `commerce' was used in contradistinction to productive activities such as manufacturing and agriculture"); Tony Mauro, Scalia Says There Is No Right To Die. USA Today, Oct. 28, 1996, at 1A ("`It's absolutely plain there is no right to die,' Scalia [said] . . . . Scalia said his view was based on the fact that laws against suicide were universally accepted at the time of the drafting of the Constitution.").

(11.) Bork, supra note 8, at 160.

(12.) Justice Scalia does not embrace the term "strict construction." See Antonin Scalia, Common-Law Courts in a Civil-Law System: The Role of United States Federal Courts in Interpreting the Constitution and Laws, in A Matter of Interpretation: Federal Courts and the Law 3, 23 (Amy Gutmann ed., 1997). I use the term to refer to any constitutional method that (1) insists on strict adherence to a putatively historically fixed meaning (whether said to be fixed by text or by original intent); and (2) on that basis campaigns against all introduction of judicial inquiry into right and wrong or just and unjust.

(13.) City of Richmond v. J.A. Croson Co., 488 U.S. 469, 528 (1989) (Scalia, J., concurring); see also Adarand Constructors, Inc. v. Pena, 115 S. Ct. 2097, 2119 (1995) (Thomas, J., concurring) ("[G]ood intentions cannot provide refuge from the principle that under our Constitution, the government may not make distinctions on the basis of race.").

(14.) Bork, supra note 8, at 107; accord id. at 246.

(15.) See Act of July 28, 1866, ch. 296, 14 Stat. 3 10.

(16.) Id., 14 Stat. at 317 (emphasis added). The act appropriated money for the "National association for the relief of destitute colored women and children," id., an association incorporated by federal statute three years earlier "for the purpose of supporting . . . aged or indigent and destitute colored women and children." Act of Feb. 14, 1863, ch. 33, 12 Stat. 650, 650.

(17.) See Resolution of Mar. 16, 1867, No. 4, 15 Stat. 20.

(18.) Id., 15 Stat. at 20 (emphasis added).

(19.) E.g., Act of Mar. 3, 1873, ch. 227, 17 Stat. 510, 528 (emphasis added), Act of Mar. 3, 1869, ch. 122, 15 Stat. 301, 302 (emphasis added); Resolution of June 15, 1866, No. 46, 14 Stat. 357, 358-59 (emphasis added).

(20.) Act of July 16, 1866, ch. 200, 14 Stat. 173; Act of Mar. 3, 1865, ch. 90, 13 Stat. 507.

(21.) For originalist arguments in favor of affirmative action on the basis of the Freedmen's Bureau Acts, see Louis H. Pollak, "Mr. Chief Justice; May It Please the Court:", in Constitutional Government in America 247, 252 (Ronald K.L. Collins ed., 1980); and Eric Schnapper. Affirmative Action and the Legislative History of rite Fourteenth Amendment, 71 Va. L. Rev. 753 (1985).

(22.) See Charles Fried, Metro Broadcasting, Inc. v. FCC: Two Concepts of Equality. 104 Harv. L. Rev. 107, 111 n.21 (1990) ("[R]eference to the Reconstruction-era Freedmen's Bureaus . . . is not quite apposite: they no more accorded benefits in terms of race than did the Emancipation Proclamation.").

(23.) Schnapper cited a number of these statutes over 10 years ago. See Schnapper, supra note 21, at 775, 778-80. Unfortunately, his references to these laws appear in the middle of a much lengthier discussion of the Freedmen's Bureau Acts as race-conscious legislation. See id. at 754-75, 780-83. As a result, some concluded that Schnapper's essay was "convincingly rebutted" by those who pointed out that the Freedmen's Bureau Acts, at least after 1865, purposefully employed race-neutral language. Jeffrey Rosen, The Color-Blind Court, 45 Am. U. L. Rev. 791, 795 (1996) (citing, for example, Paul Moreno, Racial Classifications and Reconstruction Legislation, 61 J.S. Hist. 271 (1995)). But while the Freedmen's Bureau Acts were arguably race-neutral, the statutes referred to in the text above were not. And these statutes convincingly rebut even the most tortuously phrased conclusions of those who deny the existence of race-based affirmative action in the Reconstruction period: "It is not unreasonable to conclude . . . that the Reconstruction Congress might have allowed race-conscious remedial legislation with preference for blacks, since it permitted the anti-black variety [!]. But in fact it did not . . . ." Paul D. Moreno, From Direct Action To Affirmative Action 11 (1997) (emphasis in original). In fact, the Reconstruction Congress enacted just such legislation, and it did so repeatedly. That Congress also passed non-color-based remedial legislation in this era, such as the Freedmen's Bureau Acts, or provisions for the relief of all destitute persons in the District of Columbia, see, e.g., Resolution of Apr. 17, 1866, No. 25, 14 Stat. 3.47, 353, does not alter this fact. A full list of "race-conscious remedial legislation" of the period would also include the many federal grants of aid to Indians, which represented race- or blood-based remedial legislation on an enormous scale. See, e.g., Resolution of Dec. 21, 1865, No. 1, 14 Stat. 347, 347 (authorizing the expenditure of $500,000 "for the immediate subsistence and clothing of destitute Indians"). It would also be proper to add the special allocations of federal resources for "colored schools." See. e.g., Act of July 28, 1866, ch. 308, 14 Stat. 310, 343 (donating federally owned land in the District of Columbia "for the sole use of schools for colored children").

(24.) Planned Parenthood v. Casey, 505 U.S. 833, 998 (1992) (Scalia, J., dissenting) (stating the belief "that the text of the Constitution, and our traditions, say what they say and there is no fiddling with them").

(25.) 347 U.S. 483 (1954). To be sure, today's originalists constantly discover new ways to save Brown. For example, Bork--after conceding that the original understanding of the Equal Protection Clause countenanced race-segregated schools--argues that "equality and segregation were mutually inconsistent, though the ratifiers did not understand that." Bork, supra note 8, at 82 (emphasis added). Hence originalism dictates the result in Brown after all, because in the choice between "equality and segregation," originalist judges must opt for "equality," the "purpose that brought the fourteenth amendment into being." Id. This is pure result-oriented reasoning. Originalism is no longer the method it claimed to be if judges are free to reject the specific understanding of a constitutional provision in light of a more general putative "purpose" such as "equality." Every view that originalists want to insist on elsewhere--such as a narrow reading of the Commerce Clause or the Due Process Clause--depends on their cutting off the move from specific intentions to general purposes. If achieving "equality" is the relevant intention, it would be equally originalist to say that the Fourteenth Amendment enacted Marxism, on the theory that equality and capitalism were mutually inconsistent, though the ratifiers did not understand that. Certainly affirmative action could be deemed permissible on this brand of originalism.

(26.) See, e.g., Adarand Constructors, Inc. v. Pena, 115 S. Ct. 2097, 2112-14 (1995); Metro Broad., Inc. v. FCC, 497 U.S. 547, 563-66 (1990); City of Richmond v. J.A. Croson Co., 488 U.S. 469, 493-98 (1989) (plurality opinion).

(27.) "Intermediate scrutiny" applies to so-called semi-suspect classifications and requires a law to be "substantially related" to "important" or "significant" governmental interests. Metro Broad., 497 U.S. at 564-65. Gender has been the prime example of a semi-suspect classification. See Mississippi Univ. for Women v. Hogan, 458 U.S. 718, 724 (1982); Craig v. Boren, 429 U.S. 190, 197 (1976). Recent cases have reopened the question of whether sex classifications should be shifted up to full "suspect" status. See, e.g., United States v. Virginia, 116 S. Ct. 2264, 2274-76 (1996); J.E.B. v. Alabama ex rel. T.B., 511 U.S. 127, 137 n.6 (1994).

(28.) 115 S. Ct. 2097 (1995).

(29.) See id. at 2117. Strict scrutiny requires that the state action at issue be "narrowly tailored to further" a "compelling governmental interest." Id.

(30.) See id. at 2111; see also Loving v. Virginia, 388 U.S. 1, 11 (1967) ("At the very least, the Equal Protection Clause demands that racial classifications ... be subjected to the `most rigid scrutiny' ....").

(31.) See Adarand, 115 S. Ct. at 2110; see also Buckley v. Valeo, 424 U.S. 1, 93 (1976) ("Equal protection analysis in the Fifth Amendment area is the same as that under the Fourteenth Amendment."); Boiling v. Sharpe, 347 U.S. 497, 500 (1954) ("[I]t would be unthinkable that the same Constitution would impose a lesser duty on the Federal Government.").

(32.) See City of Richmond v. J.A. Croson Co., 488 U.S. 469 (1989) (plurality opinion).

(33.) See, e.g., Bush v. Vera, 116 S. Ct. 1941 (1996) (plurality opinion); Shaw v. Hunt, 116 S. Ct. 1894 (1996). Race-conscious redistricting raises issues that other, more common affirmative action measures do not. Some scholars have advanced arguments that could support the Court's recent redistricting decisions without embracing the Court's hostile treatment of race preferences in general. See, e.g., Richard H. Pildes & Richard G. Niemi, Expressive Harms, "Bizarre Districts," and Voting Rights: Evaluating Election District Appearances After Shaw v. Reno, 92 Mich. L. Rev. 483 (1993).

(34.) See, e.g., Hopwood v. Texas, 78 F.3d 932 (5th Cir.) (invalidating race preferences in law school admissions), reh'g and reh'g en banc denied, 84 F.3d 720 (5th Cir.), cert. denied, 116 S. Ct. 2580 (1996); see also Taxman v. Board of Educ., 91 F.3d 1547 (3d Cir. 1996) (en banc) (invalidating face preferences in employee layoff decisions under Title VII), cert. granted, 117 S. Ct. 2506 (1997).

(35.) See 115 S. Ct. at 2111.

(36.) Id. at 2111 (quoting Croson, 488 U.S. at 494 (plurality opinion)).

(37.) See Craig v. Boren, 429 U.S. 190, 220 (1976) (Rehnquist, J., dissenting).

(38.) See id. at 191-92.

(39.0 Id. at 197.

(40.) See id. at 204.

(41.) Id. at 217 (Rehnquist, J., dissenting).

(42.) See, e.g., City of Richmond v. J.A. Croson Co., 488 U.S. 469, 492-96 (1989) (plurality opinion).

(43.) Craig, 429 U.S. at 220 (Rehnquist, J., dissenting); see supra text accompanying note 37.

(44.) Adarand Constructors, Inc. v. Pena, 115 S. Ct. 2097, 2119 (1995) (Thomas, J., concurring) (quoting id. at 2120 (Stevens, J., dissenting)).

(45.) Id. at 2119 (Thomas, J., concurring).

(46.) Harris v. McRae, 448 U.S. 297, 323 (1980).

(47.) See Lassiter v. Northampton County Bd. of Elections, 360 U.S. 45, 50-54 (1959). Congress banned the use of some literacy tests in the Voting Rights Act of 1965, 42 U.S.C. [sections] 1973 (1994), the relevant provisions of which were upheld in South Carolina v. Katzenbach, 383 U.S. 301, 308 (1966).

(48.) See John Hart Ely, Democracy and Distrust 146 (1980) ("[F]unctionally, special scrutiny, in particular its demand for an essentially perfect fit, turns out to be a way of `flushing out' unconstitutional motivation ...."). If there is a meaningful difference between motivation, intention, and purpose in this context, I regard intentions and purposes, rather than motivations, as the appropriate object of inquiry. For example, it may be possible to regard what is being smoked out as the intention or purpose that a citizen would reasonably impute to the law (or that which a rational, consistent legislature would have had to enact the law), rather than the motivations actually obtaining. This issue raises difficult questions that I do not try to answer in this Essay.

(49.) Redistributive motives have not been held constitutionally suspect since the demise of Lochner v. New York, 198 U.S. 45 (1905). See, e.g., Hawaii Hous. Auth. v. Midkiff, 467 U.S. 219 (1984) (holding that redistribution of land ownership to remedy "land oligarchy" is a legitimate state purpose)

(50.) For more on the issue of paternalism, see infra Subsection III.A.1

(51.) Craig v. Boren, 29 U.S. 190, 220 (1976) (Rehnquist, J., dissenting).

(52.) Adarand Constructors, Inc. v. Pena, 115 S. Ct. 2097, 2114 (1995) (emphasis added).

(53.) Richmond v. J.A. Croson Co., 488 U.S. 469, 493 (1989) (plurality opinion) (emphasis added).

(54.) Id. (plurality opinion).

(55.) Id. (plurality opinion).

(56.) Richard J. Herrnstein & Charles Murray, The Bell Curve (1994). on the contrary, an adherent of The Bell Curve could be expected to oppose affirmative action. See id. at 447-508

(57.) E.g., Bush v. Vera, 116 S. Ct. 1941, 1963 (1996) (plurality opinion) ("foster harmful and divisive stereotypes"); Adarand, 115 S. Cf. at 2119 (Thomas, J., concurring) ("reinforce ... the way of thinking that produced race slavery, race privilege and race hatred"); see also Croson, 488 U.S. at 520 (Scalia, J., concurring); infra Sections III.A-B.

(58.) Adarand, 115 S. Ct. at 2114.

(59.) See cases cited supra note 27.

(60.) Adarand, 115 S. Ct. at 2114.

(61.) A temporary quarantine is not equivalent to permanent or indefinite apartheid. Specifying exactly what makes a particular state objective such as apartheid impermissible under the Equal Protection Clause is not a trivial task. I return to this problem below. See infra Section III.A. But while a temporary racial separation order in very confined contexts might be constitutional, permanent racial separation laws can never be. Cf., e.g., Lee v. Washington, 390 U.S. 333 (1968) (per curiam) (rejecting permanent race segregation in prison alleged to be necessary to avoid racial violence).

(62.) For example, a cost-benefit version of strict scrutiny appears to have been the one used with dubious success in free exercise doctrine until it was overruled in the well-known Smith case. See, e.g.. Sherbert v. Verner, 374 U.S. 398 (1963), limited by Employment Div. v. Smith, 494 U.S. 872 (1990). Hence the analysis suggested here might provide a strong defense of Smith. The compelling state interest test also plays an important role in free speech doctrine. See, e.g., Austin v. Michigan Chamber of Commerce, 494 U.S. 652, 657 (1990). See generally Eugene Volokh, Freedom of Speech, Permissible Tailoring and Transcending Strict Scrutiny, 144 U. Pa. L. Rev. 2417 (1996). Here, too, useful results might be achieved by distinguishing those cases in which strict scrutiny has been used to smoke out impermissible purposes from the cases in which it has been used as a cost-benefit test.

(63.) See, e.g., Terry Eastland, Ending Affirmative Action: The Case for Colorblind Justice 196 (1996) ("Whoever would have been admitted to a school, or won the promotion or the contract, but for race, has suffered discrimination-and there is no good discrimination."): Charles Murray, Affirmative Racism, in Racial Preference and Racial Justice: The New Affirmative Action Controversy 393, 408 (Russell Nieli ed., 1991) ("There is no such thing as good racial discrimination.").

(64.) See, e.g., Adarand, 115 S. Ct. at 2121 (Stevens, J., dissenting) ("The consistency that the Court espouses would disregard the difference between a `No Trespassing' sign and a welcome mat.").

(65.) See, e.g., Antonin Scalia, The Disease as Cure, 1979 Wash. U. L.Q. 147, 154 ("The affirmative action system now in place . . . is based upon concepts of racial indebtedness and racial entitlement rather than individual worth and individual need; that is to say.... it is racist.").

(66.) See, e.g., Stanley Fish, Reverse Racism or How the Pot Got To Call the Kettle Black, Atlantic Monthly, Nov. 1993, at 128, 130. Fish writes:

[B]lacks have not simply been treated unfairly; they have been subjected

first to decades of slavery, and then to decades of second-class

citizenship, widespread legalized discrimination, economic persecution,

education deprivation, and cultural stigmatization.... When the deck is

stacked against you in more ways than you can even count, it is small

consolation to hear that you are now free to enter the game and take your

chances.

Id.

(67.) See, e.g., Adarand, 115 S. Ct. at 2119 (Thomas, J., concurring); Nathan Glazer, Affirmative Discrimination: Ethnic Inequality and Public Policy 168-95 (1975); Thomas Sowell, Black Education: Myths and Tragedies 292 (19;2); Murray, supra note 63, at 401-06.

(68.) See, e.g., Ronald Dworkin, A Matter of Principle 294-303 (1985); Randall Kennedy, Persuasion and Distrust: A Comment on the Affirmative Action Debate, 99 Harv. L. Rev. 1327, 1330-33 (1987).

(69.) City of Richmond v. J.A. Croson Co., 488 U.S. 469, 493 (1989) (plurality opinion); see also id. at 527 (Scalia, J., concurring) ("[T]hose who believe that racial preferences can help . . . display, and reinforce, a manner of thinking by race . . . .").

(70.) Bush v. Vera, 116 S. Ct. 1941, 1963 (1996) (plurality opinion).

(71.) Adarand, 115 S. Ct. at 2119 (Thomas, J., concurring) (citations omitted).

(72.) Croson, 488 U.S. at 493 (plurality opinion); see supra text accompanying note 69.

(73.) Fullilove v. Klutznick, 448 U.S. 448, 545 (1980) (Stevens, J., dissenting). As Justice Stevens put it,

[E]ven though it is not the actual predicate for this legislation, a

statute of this kind inevitably is perceived by many as resting on an

assumption that those who are granted this special preference are less

qualified in some respect that is identified purely by their race. . . .

[T]hat perception--especially when fostered by the Congress of the United

States--can only exacerbate rather than reduce racial prejudice. . . .

Id. This passage was quoted approvingly in Adarand, 115 S. Ct. at 2113.

(74.) Adarand, 115 S. Ct. at 2119 (Thomas, J., concurring); see supra text accompanying note 71.

(75.) Croson, 488 U.S. at 493 (plurality opinion); see supra text accompanying note 69.

(76.) Lochner v. New York, 198 U.S. 45 (1905).

(77.) The Lochner Court indicated that heightened means-ends scrutiny was justified because of the suspect politics underlying prolabor legislation. "It is impossible for us to shut our eyes to the fact that many of the laws of this character, while passed under what is claimed to be the police power . . . are, in reality, passed from other motives." Id. at 64. To understand what "motives" the Court was referring to, see, for example, id. at 48 (argument of counsel for Lochner) (attacking the statute at issue as "[c]lass legislation").

(78.) Dred Scott v. Sandford, 60 U.S. (19 How.) 393 (1857).

(79.) 347 U.S. 483 (1954).

(80.) See supra text accompanying note 71.

(81.) See, e.g., Personnel Adm'r v. Feeney, 442 U.S. 256, 272 (1979) ("[E]ven if a neutral law has a disproportionately adverse effect upon a racial minority, it is unconstitutional under the Equal Protection Clause only if that impact can be traced to a discriminatory purpose."); Arlington Heights v. Metropolitan Hous. Dev. Corp., 429 U.S. 252, 264-65 (1977) ("Proof of racially discriminatory intent or purpose is required to show a violation of the Equal Protection Clause.").

(82.) See, e.g., Washington v. Seattle Sch. Dist. No. 1, 458 U.S. 457, 470 (1982) (invalidating a referendum targeted at race-integrating busing and stating that heightened equal protection scrutiny was required because the referendum "explicitly us[ed] the racial nature of a decision" to alter the requisite decisionmaking process); Hunter v. Erickson, 393 U.S. 385, 386 (1969) (striking down as an "explicitly racial classification" a law preventing the city council from issuing "any ordinance dealing with racial, religious, or ancestral discrimination without the approval of a majority of the voters")

(83.) On the contrary, they characteristically champion the use of such tests, without acknowledging that the inadvertent stigmatization argument they make against affirmative action applies equally to Standardized testing. Thus, Terry Eastland affirms that "[t]he black and Mexican-American applicants admitted [to the University of Texas Law School] under affirmative action were not unqualified to study law; their academic qualifications were good enough to win admission under non-affirmative action standards at fully two-thirds of the nation's law schools." Eastland, supra note 63, at 9. "Affirmative action thus stigmatizes beneficiaries who could succeed--and be seen to succeed--without it." Id. This reasoning is really quite strange. Affirmative action "stigmatizes" an individual at a top echelon law school by making others think that she "could not have landed the opportunity in open competition." Id. But to be assigned by test score to a bottom-two-thirds law school, which confirms that the individual could not land the top-echelon opportunity in open competition, is to "succeed." In other words, being viewed as "not unqualified to study law" is stigmatizing only when the person is at one of the best law schools. It evidently becomes a badge of success, or merely the truth, when the person has been correctly assigned to the kind of law school where she belongs. I am not an opponent of standardized tests. The point is that one cannot condemn affirmative action because of its inadvertent fostering of notions of racial inferiority without recognizing the same danger as a result of standardized tests.

(84.) Washington v. Davis, 426 U.S. 229, 240 (1976).

(85.) Mobile v. Bolden, 446 U.S. 55, 63 n.10 (1980) (emphasis added) (citing Feeney, 442 U.S. at 272; Arlington Heights, 429 U.S. at 265, 266; Davis, 426 U.S. at 240).

(86.) It might be objected that strictly scrutinizing all racial classifications would also fail strict scrutiny on the smoking-out view, because not all kinds of racial classification are likely to reflect an invidious purpose, and at least some facially neutral state action (although by no means every instance of state action with a racially disparate impact) should be regarded as sufficiently suspicious to trigger strict scrutiny. But this objection does not undermine the smoking-out understanding of strict scrutiny. On the contrary, it precisely explains why smoking out strict scrutiny must tailor more narrowly the class of race-related laws to which it applies. In other words, this objection merely reconfirms what we have insisted upon from the beginning: that the application of heightened scrutiny, to be true to itself, must depend on the particular persons affected and the rights involved.

(87.) 347 U.S. 483 (1954).

(88.) Id. at 494.

(89.) 163 U.S. 537 (1896).

(90.) See id. at 557 (Harlan, J., dissenting) ("Every one knows that [Louisiana's statute] had its origin in the purpose, not so much to exclude white persons from railroad cars occupied by blacks, as to exclude colored people from coaches occupied by or assigned to white persons.")

(91.) Id. at 562 (Harlan, J., dissenting).

(92.) Many scholars have criticized an intent- or purpose-based equal protection jurisprudence, arguing that invidious racial effects should be regarded either as persuasive evidence of discriminatory purpose or as unconstitutional (in the absence of strong justification) in their own right. For an excellent concise review of the literature, see Theodore Eisenberg & Sheri Lynn Johnson, The Effects of Intent: Do We Know How Legal Standards Work?, 76 Cornell L. Rev. 1151, 1160-62 (1991). In the literature, an effects-test reading of Brown often occupies pride of place. See, e.g., David A. Strauss, Discriminatory Intent and the Taming of Brown, 56 U. Chi. L. Rev. 935 (1989).

(93.) See 347 U.S. at 494 & n.11.

(94.) See Brown v. Board of Educ. (Brown II), 349 U.S. 294 (1955) (ordering the implementation of school desegregation measures); Cooper v. Aaron, 358 U.S. 1 (1958) (same); Watson v. Memphis, 373 U.S. 526 (1963) (ordering the desegregation of municipal recreation facilities).

(95.) 388 U.S. 1 (1967).

(96.) Id. at 11.

(97.) Id.

(98.) Washington v. Davis, 426 U.S. 229, 240 (1976).

(99.) See supra notes 84-86 and accompanying text.

(100.) Adarand Constructors, Inc. v. Pena, 115 S. Ct. 2097, 2119 (1995) (Thomas, J., concurring).

(101.) Ely specifically observed that most affirmative action would be constitutional under his thinking. See Ely, supra note 48, at 170-72.

(102.) Ronald Dworkin is by far the most influential partisan of this view. For a recent restatement of Dworkin's thinking on the role of morality in constitutional law, see Ronald Dworkin, Freedom's Law: The Moral Reading of the American Constitution (1997).

(103.) See, e.g., Ronald Dworkin, Law's Empire 395 (1986) ("Candidates are no more responsible for their ability to score well on conventional intelligence tests or for their . . . skill at football than for their race . . . ."); Thomas Nagel, Equal Treatment and Compensatory Discrimination, in Equality and Preferential Treatment 3 (M. Cohen et al. eds., 1977) (arguing that affirmative action is "not seriously unjust" because even though it allocates goods on the basis of ascriptive, morally irrelevant traits, so do the alternative merit-based distributions).

(104.) See, e.g., Dworkin, supra note 103, at 395 ("[I]f race were a banned category because people cannot choose their race, then intelligence, geographical background, and physical ability would have to be banned as well.").

(105.) For an account and defense of the paradigm-case method of constitutional interpretation, see Jed Rubenfeld, Reading the Constitution as Spoken, 104 Yale L.J. 1119, 1169-84 (1995).

(106.) Plessy v. Ferguson, 163 U.S. 537 (1896).

(107.) Loving v. Virginia, 388 U.S. 1 (1967).

(108.) Plessy, 163 U.S. at 559 (Harlan, J., dissenting).

(109.) Loving, 388 U.S. at 11 (quoting Korematsu v. United States. 323 U.S. 214, 216 (1944)).

(110.) Hirabayashi v. United States, 320 U.S. 81, 100 (1943) (upholding a curfew on persons of Japanese ancestry).

(111.) Alexander M. Bickel, The Morality of Consent 133 (1975).

(112.) William Van Alstyne, Rites of Passage: Race, rite Supreme Court and the Constitution, 46 U. Chi. L. Rev. 775, 782, 809 (1979).

(113.) Plessy v. Ferguson, 163 U.S. 537, 551 (1896).

(114.) See supra note 90.

(115.) Id. at 556-57 (Harlan, J., dissenting).

(116.) Id. at 560 (Harlan, J., dissenting).

(117.) Plessy, 163 U.S. at 559 (Harlan, J., dissenting).

(118.) Loving v. Virginia, 388 U.S. 1 (1967).

(119.) Brown v. Board of Educ., 347 U.S. 483, 495 (1954).

(120.) Hirabayashi v. United States, 320 U.S. 81, 100 (1943).

(121.) Plessy, 163 U.S. at 545.

(122.) Id. at 544.

(123.) Id. at 559 (Harlan, J., dissenting).

(124.) See Owen M. Fiss, Groups and the Equal Protection Clause, 5 Phil. & Pub. Aff. 107 (1976).

(125.) See, e.g., Kenneth L. Karst, Belonging to America: Equal Citizenship and the Constitution (1989); Laurence H. Tribe, American Constitutional Law [subsections] 16-13 to 16-16 (2d ed. 1988); Jack Balkin, The Constitution of Status, 106 Yale L.J. 2313 (1997); and Cass Sunstein, The Anticaste Principle, 92 Mich. L. Rev. 241 (1994). Balkin's essay interestingly discusses the link between the anticaste dimension of the Fourteenth Amendment and other constitutional provisions such as the Title of Nobility Clauses. See Balkin, supra, at 2350-51.

(126.) Cf., e.g., Paul R. Dimond, The Anti-Caste Principle--Toward a Constitutional Standard for Review of Race Cases, 30 Wayne L. Rev. 1, 3 (1983) (arguing against governmental action that "perpetuate[s) longstanding stereotypes that stigmatize").

(127.) See, e.g., City of Cleburne v. Cleburne Living Ctr., Inc., 473 U.S. 432, 441 (1985) (quoting Massachusetts Bd. of Retirement v. Murgia, 427 U.S. 307, 313 (1976) (per curiam)).

(128.) 429 U.S. 190 (1976).

(129.) See supra text accompanying note 37.

(130.) Craig, 429 U.S. at 220.

(131.) See City of Richmond v. J.A. Croson Co., 488 U.S. 469, 495-96 (1989) (plurality opinion); John Han Ely, Vie Constitutionality of Reverse Racial Discrimination, 41 U. Chi. L. Rev. 723, 739 n.58 (1974).

(132.) I realize that powerful arguments have been made suggesting that the distinction between not-benefiting and penalizing is impossible to pin down. See, e.g., Kathleen M. Sullivan, Unconstitutional Conditions, 102 Harv. L. Rev. 1413, 1439-43 (1989). I think this impossibility has been overstated. Obviously, to make good on the idea of taking away something a person already has, as opposed to denying him an opportunity, we would have to rule out taking away opportunities as an instance of taking away something a person already has. And then we might need an account of those opportunities in which one already has some son of vested expectation, as distinct from mere applied-for opportunities to which one has, as yet, no entitlement. This is not the place to pursue this tine of thought. I note only that it would sustain standard affirmative action programs while potentially embracing the result in Taxman v. Board of Education, 91 F.3d 1547 (3d Cir. 1996) (en banc), cert. granted, 117 S. Ct. 2506 (1997), which invalidated under Title VII a race-based preference scheme for employee layoffs.

(133.) Plessy v. Ferguson, 163 U.S. 537, 560 (1896) (Harlan, J., dissenting).

(134.) 116 S. Ct. 1620 (1996).

(135.) The amendment barred any state actor from enacting or enforcing "`any statute, regulation, ordinance or policy whereby homosexual, lesbian or bisexual orientation, conduct, practices or relationships shall ... be the basis of or entitle any person or class of persons to have or claim any ... protected status or claim of discrimination.'" Id. at 1623 (quoting COLO. Const. art. II [sections] 30b). In other words, no law could be passed prohibiting discrimination on the basis of homosexuality, and no individual (private or public) could be made legally liable for engaging in such discrimination. An exception in the case of common carriers--by which I mean those charged by law with the duty to offer their services, if at all, to the general public--was by no means clear from the face of the law. But the state argued that common carriers would not be legally immune under the amendment if they denied services to persons on the basis of homosexuality, and the Court did not reach the issue. See Romer, 116 S. Ct. at 1626.

(136.) Id. at 1627.

(137.) Id. at 1628.

(138.) Id. at 1627-28.

(139.) See id. at 1627 ("[T]he amendment seems inexplicable by anything but animus toward the class that it affects; it lacks a rational relationship to legitimate state interests.").

(140.) See, e.g., id. at 1631-33 (Scalia, J., dissenting).

(141.) My views on the constitutionality of laws criminalizing homosexual conduct maybe found in Jed Rubenfeld, The Right to Privacy, 102 Harv. L. Rev. 737, 765-82. 799-802 (1989).

(142.) Romer, 116 S. Ct. at 1623 (quoting Plessy v. Ferguson. 163 U.S. 537, 559 (1896) (Harlan, J., dissenting)).

(143.) See Romer, 116 S. Ct. at 1629.

(144.) See Daniel Farber & Suzanna Sherry, The Panah Principle, 13 Const. Commentary 257, 265-67 (1996). For a related analysis--based, however, on the concept of attainder--see Akhil Reed Amar, Attainder and Amendment 2: Romer's Rightness, 95 Mich. L. Rev 203 (1996).

(145.) See Cal. Const. art. 1, [sections] 31. A preliminary injunction entered against the Initiative was vacated in Coalition for Economic Equality v. Wilson, 110 F.3d 1431 (9th Cir.), reh'g en banc denied, Nos. 97-15030, 9715031, 1997 WL 528335 (9th Cir. Aug. 21, 1997).

(146.) See supra note 82 and accompanying text.

(147.) The difficulty is that the Initiative singles out only race and gender preferences as unlawful. All other preferences that might be used in hiring or admissions decisions (including preferences based on other purely ascriptive traits) remain untouched by the Initiative. Thus the Initiative appears not to he "narrowly tailored" to furthering the state interest in pure-merit or nonarbitrary public decisionmaking.

(148.) The Ninth Circuit held that strict scrutiny did not apply, on the ground that the Initiative is "a law that addresses in neutral fashion race-related and gender-related matters." Coalition for Econ. Equality 110 F.3d at 1444. The difficulty with this proposition is that, under current doctrine, when a law employs a racial classification, there is supposed to be no way of determining whether it is benign except through the application of strict scrutiny. See Adarand Constructors, Inc. v. Pena, 115 S. Ct. 2097, 2111-13 (1995).

(149.) There might be other questions that I do not address here. For example, there might be a question as to whether the Initiative was motivated by racial animosity. There might also be a question as to whether its provisions would unconstitutionally bar race-conscious relief in discrimination actions.

Jed Rubenfeld, Professor, Yale Law School. This Essay profited enormously from a workshop held at Duke Law School. Special thanks to Bruce Ackerman, Akhil Amar, Kate Bartlett, George Christie, Amy Chua, Jim Coleman, Zoe Hilden, Don Horowitz, Trina Jones, and Jeff Powell.
COPYRIGHT 1997 Yale University, School of Law
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1997 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Rubenfeld, Jed
Publication:Yale Law Journal
Date:Nov 1, 1997
Words:22438
Previous Article:On the meaning of congressional silence: using federal common law to fill the gap in Congress's residual statute of limitations.
Next Article:The look within: property, capacity, and suffrage in nineteenth-century America.
Topics:


Related Articles
25 years of affirmative action.
Down but not out: the Court makes it tougher for affirmative action programs, but doesn't eliminate them.
The affirmative action deathwatch.
Affirmative action. (You be the Judge!).
Supreme Court ruling supports affirmative action. (Capitol briefs).
Equality control: a Catholic perspective on affirmative action in the wake of the recent Supreme Court decisions.
Owens straddles affirmative-action line.
Affirmative action for the future.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters