California Law Review
88 Calif. L. Rev. 65
LENGTH: 5176 words
1998-99 Brennan Center Symposium Lecture: Cover-Blindness
NAME: Thomas C. Grey +
+ Nathan Bowman Sweitzer and Marie B. Sweitzer Professor of Law, Stanford Law School. Thanks to Barbara Babcock for her editorial suggestions.
... Robert Post's original and challenging Brennan Lecture starts with a puzzle: Why don't the champions of civil rights line up against "lookism" as they have against racism and sexism? The rhetoric of civil rights law denounces, in general and unqualified terms, discrimination based on stereotyping - which is to say, judging the worth of human beings on the basis of superficial characteristics like skin color. ... I will also suggest that certain of the more abstract and formal aspects of antidiscrimination law, which might be thought targets of Professor Post's critique, actually are consistent with a contextualist approach, once we take account of law's own political context, and its specificity as a set of social practices. ... This point should persuade even those who do not agree with each distinction drawn by the courts in Professor Post's examples, as long as they are viewed as the kinds of distinctions that judges must make to decide discrimination cases sensibly. ... I think my amplification of the contextual approach, which draws attention to the two stages of civil rights law - the classical and the modern - can offer some help in explaining, for example, the differences to which Professor Post points between the law of gender and race discrimination. ... Recall that race and sex discrimination were embodied in well-defined and legally sanctioned social hierarchies, which liberal reformers began to attack long ago. ... In these latter cases, contemporary law makers cannot identify invidious discrimination by looking for those aspects of a social hierarchy that mirror a past legal hierarchy. ...
Robert Post's original and challenging Brennan Lecture 1 starts with a puzzle: Why don't the champions of civil rights line up against "lookism" as they have against racism and sexism? The rhetoric of civil rights law denounces, in general and unqualified terms, discrimination based on stereotyping - which is to say, judging the worth of human beings on the basis of superficial characteristics like skin color. You can't judge a book by its cover, or so we say when we talk about race and gender. Does not the same principle make us say that beauty is only skin deep?
And yet even many ardent civil rights advocates share the intuition that using law to prohibit discrimination on the basis of appearance goes too far. Professor Post thinks this intuition is no mere reflex of prejudice; he invokes it as the basis of an argument reductio ad absurdum, meant to put in question some standard assumptions of antidiscrimination law.
The maxim about books and their covers implies that a meritocratic and just society looks beyond mere appearances, and considers only the real person underneath when allocating important benefits and burdens. That would make discrimination against the physically ugly a paradigmatic form of bigotry. And if it is bigotry, it is a damaging kind; studies show that less attractive people do tend to get fewer of the socially valued goods. 2 So why don't we rally around laws aimed to combat appearance discrimination?
Professor Post wants to plant some doubts about the underlying "book and cover" distinction between mere appearance and underlying reality, with its notion of a core (but unapparent) "real individual" behind the facade, in whom all genuine merit resides. This concept is too unhistorical, acontextual, and utopian - in a word, too abstract - to serve as a plausible underpinning for antidiscrimination law. 3 In the actual life-world, not only do we judge and get judged by appearances, but we sense that it could not [*66] be otherwise - and that is why the anti-lookism laws strike even many egalitarians as overreaching.
The more general point is that a stark contrast between "mere physical appearance" and "true human reality" does not provide a satisfactory basis even for the core of civil-rights law, its prohibition of race and gender discrimination. 4 We cannot live behind a veil effacing individuals' appearances as women or men, black, brown, or white. The social categories of race and gender are so much constituted by our conventional associations of appearance with identity that to disestablish those conventions across the board is surely beyond the capacity of the law. And if judges take seriously a rhetoric that purports to require this impossible degree of "cover-blindness," they will invariably produce dishonest and misguided doctrine.
To replace the abstracted approach that is now dominant, Professor Post proposes a "sociological" (that is, more contextual) conception of antidiscrimination law. 5 On this conception, the links between physical appearance and the status-ascriptions we make when we identify someone's race or gender are not simply a set of harmful stereotypes, obstacles to individual liberation, to be extirpated by law insofar as possible. Rather, our conventions for attributing group identity to persons are embedded in social practices that partly constitute our very selves, collective and individual.
We pass antidiscrimination laws, then, not to abolish all stereotypes, but to reform selected aspects of the practices that mold our notions of gender and race. And those reforms are to be implemented through yet another set of embedded social practices: legislation, regulation, adjudication, and the informal dissemination of legal norms.
Professor Post supports his critique and proposal mainly by reference to the law of gender discrimination, where the gap between the case law and the rhetoric of stereotype-extirpation is particularly striking. 6 But his scattered references to the law of race, and particularly his praise of Justice Brennan's opinion in United Steelworkers of America v. Weber, 7 show how his argument applies throughout the domain of antidiscrimination law.
I agree with the general thrust of Professor Post's approach, and in this Response, I mainly want to further situate his position in the familiar territory of American civil rights history, theory, and doctrine. I will also suggest that certain of the more abstract and formal aspects of antidiscrimination law, which might be thought targets of Professor Post's [*67] critique, actually are consistent with a contextualist approach, once we take account of law's own political context, and its specificity as a set of social practices.
Classical liberalism is suspicious of status hierarchies. Indeed, the equality aspect of the liberal ideal of a society of free and equal individuals seems originally to have meant simply that law should not define or maintain any caste-like arrangements. Thus, nineteenth-century liberals sought the abolition of slavery and supported the rights of non-Christians, non-whites, women, servants, paupers, and immigrants to own property, make contracts, engage in lawful occupations, travel, and change residences on equal legal terms with members of the most favored class. The idea was that every competent adult should fit within a single legal category, that of "free person," or simply, "person." The Roman and medieval "law of persons," with its complex status hierarchies, was to wither away.
Modern antidiscrimination law is in a sense a natural extension of that classical liberal ideal of equality under law. Slavery was to be abolished, not just the law of slavery repealed. The removal of formal legal status hierarchies left behind entrenched social norms that often effectively denied equal opportunity in much the same way as the old legal restraints. Customary norms still, for example, kept qualified blacks and women from holding many jobs, using some public accommodations, and owning or renting real property.
At the same time, the modern administrative state put in place the kind of legal machinery that could be directed against such nominally private structures of power. In the age of the welfare state, it came to be seen as a new requirement of the old liberal ideal of legal equality that government must combat discriminatory social practices that serve to maintain the functional effects of the old legal restrictions.
The first step, the abolition of formal legal status hierarchies, could be undertaken without choosing a theoretical framework for the liberal civil rights project. A sociological/contextual argument might well have been made for the classical liberal conclusion that every competent adult should have the single legal status of person. But this conclusion followed equally well from a more abstract and philosophical justification - the principle that justice always and everywhere requires blindness to the particularities by which we separate individuals into social and cultural groups. The law of status could then be conceived as reduced to a single principle: Each free rational actor (which is to say, competent adult) has, by virtue of that fact alone, the sole legal status of person. It is not who you are but what you do that should be the sole concern of the law, and this principle is to be understood abstractly and universally.
[*68] However, the second or modern step cannot be taken so readily without attention to the choice between an abstract and a contextual account. First, modern antidiscrimination law requires identifying those social practices that operate as functional equivalents of the old formal restrictions. There is no simple formal test that can do this; many social practices work to the practical disadvantage of identifiable social groups, but do so legitimately. We do not believe that equality of opportunity guarantees equality of result. We have to separate invidiously discriminatory practices from neutral and legitimate social sorting mechanisms, and this requires a good deal of sociological knowledge, whether systematic or intuitive. Why were the racially segregated bathrooms of Jim Crow so clearly invidious, while bathrooms segregated by sex equally clearly are not?
Further, modern civil rights law requires deliberate strategic intervention into social life, with an eye to achieving the distribution of benefits and burdens that would have obtained had the old formal restraints not been continued in force through their surrogate discriminatory customs and social norms. Social engineering like this requires understanding how norm structures work, and how law can effectively counter their caste-preserving effects. By contrast, the first or formal stage of civil rights law required nothing more than the identification and repeal of laws that openly imposed unequal status hierarchies.
One solution is simply to deny that the social intricacies exist, or at least refuse to deal with them through law. Some contemporary classical liberals would just carry forward the assumptions (e.g., the true inner self; you can't tell a book by its cover) that worked well in implementing the first stage of the law of equal opportunity. But as Professor Post shows, these assumptions run into trouble when applied within modern antidiscrimination law.
When the rhetorical distinction between books and covers gets translated into legal doctrine, it can become a principle condemning all socially constructed race and gender differences as superficial, and hence to be disregarded in the law's allocation of burdens and benefits. Only biologically based differences can be accepted as real, and hence only proven biological necessity justifies differential treatment of the members of those old status groups. Because there are few if any provable and legally relevant biological differences between the races, we should strictly scrutinize all racially differential practices. By contrast, while many gender-based distinctions are socially constructed and must be abolished, some are justified by genuine biological differences - hence, intermediate scrutiny, and the BFOQ exception, in gender discrimination law.
But as Professor Post shows, these explanations will not hold up when they are examined in detail. Particularly in the area of gender discrimination, judges cannot sensibly and honestly decide cases unless they [*69] distinguish forms of gender-differential treatment on grounds that have nothing to do with biological necessity. Thus men can be rejected as nursing home attendants for female patients, 8 but women must be allowed to guard male prisoners. 9 This is so even though gendered privacy conventions involving nudity in the presence of the opposite sex are the difficulty in both cases. Employers can require Playboy waitresses to be female, but airlines cannot require the same of flight attendants, even though the same "sex appeal for male customers" business rationale applies in both cases. 10 Employers can require that men, but not women, have short hair and wear ties, and they can require women, but not men, to wear skirts 11 - although the justification for the differential treatment is not biology, but gendered hair and dress conventions. Even in the area of race, few believe that casting actors in roles to match for race is the kind of injustice that employment discrimination laws were meant to prohibit - though, again, the practice is rooted not in biology but in the cultural and historic conventions that govern theatrical portrayal.
In general, the creation and enforcement of modern antidiscrimination law involves the attempt to identify and, insofar as it is practicable, eliminate forms of differential treatment that are, to invoke a conclusory but still useful doctrinal term, invidious. In the examples noted above, gender (or race) differentiations are acceptable when they are based on conventions that are regarded as neutral - not part of the racial or gender caste system.
The contextual approach, if accepted, would influence legislators, judges, and commentators to discuss these matters openly, and confront the difficulties of drawing lines between invidious and innocent differentiation. By contrast, the abstracted approach, with its rhetorical attachment to treating all socially constructed differentiation as suspect, hides these questions - while still requiring that we answer them. This point should persuade even those who do not agree with each distinction drawn by the courts in Professor Post's examples, as long as they are viewed as the kinds of distinctions that judges must make to decide discrimination cases sensibly. No such distinctions could be made if, as the official rhetoric suggests, civil rights law were really committed to requiring blindness to all gender or race differentiations derived from convention or practice.
I think my amplification of the contextual approach, which draws attention to the two stages of civil rights law - the classical and the modern - can offer some help in explaining, for example, the differences to which Professor Post points between the law of gender and race discrimination. Recall that the abstract approach can justify differential treatment [*70] only on the basis of biological necessity, and must say (however inconsistently with the actual course of decisions) that it condemns all conventional or socially constructed grounds for differentiation.
The social-biological distinction cannot explain the law's quite different treatment of racial and sexual separate-but-equal segregation. It is a pillar of civil rights law that mandated racial segregation, even with materially equal facilities, is invidiously discriminatory. By contrast, sex segregation has been readily accepted with respect to bathroom, dressing, and sleeping facilities, and found debatable on a case-by-case basis with respect to separate educational facilities for males and females.
We cannot plausibly explain these differences in treatment on the ground that biological differences justify the sex but not the race segregation. They make much more sense when one considers contemporary civil rights law as a strategic response to the social norms left in the wake of dismantled status hierarchies that valued whites over blacks, and men over women.
A strategic response must take account of the nature of the enemy, and the different histories of racism and patriarchy help explain the varied responses to "separate-but-equal" segregation. Racism and its legal manifestations have always rested squarely on an ideology of white supremacy. Attempts to justify segregation as a path to the separate development of different but equally valuable human groups have never been more than transparent rationalizations devised by white supremacists in response to external restraints on the simpler forms of discrimination.
By contrast, the evolution of the ideology governing relations between the sexes has been more complex. From an earlier normative order based on the conception of women as clearly inferior, placed below men on the Great Chain of Being, the nineteenth century developed an ideology of separate spheres - that men and women were of equal worth, though different in their natures. Men were stronger, more logical, and more decisive, and thus more fit for the public spheres of marketplace, forum, and battlefield. Women were more flexible, intuitive, nurturing, and pure, and thus more suited for the private realms of home, church, and salon.
Simple rationalization of male domination played a major role in this ideology, but there was more to it than that. The separate spheres theory was, in certain respects, a genuine first step on the road to legal and social equality for women. Indeed, the notion that women are importantly different from men, but equal to them in worth, still survives in the influential "difference" strand of contemporary feminist theory. 12 The ambiguous legacy of separate spheres ideology means that contemporary sex discrimination law must make a nuanced distinction between the manifestations of [*71] that ideology that are simply ruses for male supremacy and those that are reflections of egalitarianism.
This brief discussion of separate-but-equal segregation is meant to suggest how conceptual division of civil rights law into its two stages - classical and modern - might be helpful in making the contextual judgments that Professor Post's approach requires. It works most clearly in the cases of race and sex discrimination, but modern law also prohibits discrimination on other grounds - age, disability, sexual orientation, among others. How can it apply to these?
Recall that race and sex discrimination were embodied in well-defined and legally sanctioned social hierarchies, which liberal reformers began to attack long ago. That assault on the legally entrenched caste system was the "first stage" I have referred to in my account. But no similar first stage can be identified with respect to contemporary laws prohibiting discrimination on the basis of age, sexual orientation, and disability, which are not generally enacted in response to a history of legally formalized status hierarchy. Still, the two-stage analysis can, with appropriate modification, be of some guidance in formulating a contextual version of these bodies of law.
The basic idea of the contextual approach is to identify and articulate the terms of the hierarchy that the civil rights laws in question seek to dismantle. Sometimes, as with race and gender, these hierarchies will have been given formal legal definition. In other cases, as with age and the like, the present structure of discriminatory social norms may lack clear legal antecedents in a defined status system.
In these latter cases, contemporary law makers cannot identify invidious discrimination by looking for those aspects of a social hierarchy that mirror a past legal hierarchy. They must ascertain the content of the social hierarchy through systematic social investigation, or (by default) through a kind of intuitive ethnography - like native speakers feeling their way to the grammatical patterns that structure their own language. Having found "the enemy" - that fragment of a social practice which sufficiently resembles an invidious legal status hierarchy - the law makers must then formulate a strategy to dismantle this structure, or at least remedy its unjust effects, within the boundaries of legal institutions and practices.
Of course the same mix of intuitive and systematic ethnography supplements the formal use of legal history in the development of modern race and gender law. The difference is just that, in the case of age, disability, and sexual orientation, the guidance from formal legal history is absent or less helpful.
Does a sociological account like Professor Post's simply mean that antidiscrimination law must proceed case-by-case, identifying and attempting to reform group-based practices that are found, taking all factors into account, to be unacceptably invidious? It might seem so at first glance, but a second look at the implications of a contextual approach should correct any first impression that sociological jurisprudence is inevitably so informal or undisciplined.
Professor Post's approach does suggest a flexible and situational set of responses to the various categories of discrimination. The process of identifying invidious social practices depends on (often intuitive) sociological insight; designing a legal campaign to eradicate those norms calls for strategic acumen. There are no handbooks or recipes that make either of these tasks routine or mechanical, so it can be expected that quite variable judgments will be reached about what practices are unacceptably invidious, and what remedies are appropriate and effective in dealing with them.
And indeed in his specific examples, Professor Post does call for judges in discrimination cases to make subtler distinctions than the rhetoric of the dominant abstract approach encourages. He suggests that a judge distinguishing the nursing home attendant case 13 from the prison guard case 14 is considering the differing weights conventionally attributed to the privacy interests of elderly female patients and male prison inmates, and should make any such consideration explicit. 15
But this example should make us notice a set of factors that will work to limit the degree of flexibility and nuance we might otherwise expect to find in a contextualized body of antidiscrimination law. As Professor Post emphasizes, civil rights law is the effort to reform a unique set of cultural and historic practices, those structuring caste-like social hierarchies. As he mentions, but does not emphasize, the effort to reform those customs is channeled through another set of social practices - our legal institutions. And within those legal practices is a complex of traditions that constitute the political ideal we call "the rule of law," the notion that coercive state power should be limited so that we may have a government of laws and not of individuals.
It should be noted that when the set of traditions constituting the rule of law create barriers to our most worthy strivings for social change, we are likely to condemn them as "legalism." Legalism and the rule of law are not easily separated from each other, perhaps because they are two sides of the same coin. But a constitutive part of our legal institutions is the [*73] aspiration that law should be relatively objective and neutral, in the sense that it can be administered to roughly the same effect by individuals with different views on the partisan conflicts of the day.
Now let's return to the contrast of the prison guard case (women must be allowed to guard male prisoners) and the case of the nursing home (which need not hire male attendants for female patients). One aspect of the rule of law - indeed, an aspect deeply implicated in civil rights law itself - is a preference for weighing the interests of individuals separately from their transitory legal status. Though it would not necessarily be decisive, this norm cuts against the practical and contextual tendency Professor Post invokes in his explanation of the distinction made between the two cases. As a matter of social mores, we do tend to find prisoners' expectations of privacy less compelling than those of elderly female nursing home patients. But when we "think like lawyers," our suspicion of such status-based distinctions makes us resist them, even if only momentarily. A judge would be more likely than a politician or a social worker to reject the offered distinction, precisely because of the judge's acculturation in the ways of thinking that are distinctive to legal training and legal institutions.
If you think this effect of the judge's legal training is a salutary check in this case, you will speak reverently of the rule of law and its value in promoting equality. But if instead you think it is a rigid refusal to make a commonsense distinction, you will speak ruefully of the legalist tendency to insist on treating genuinely disparate phenomena the same. In either case, the point remains that the law, even when understood as a set of socially embedded practices and not an abstract structure of norms and concepts, is an instrument that reinterprets and often redirects the force of the social impulses that drive its content, and usually in a relatively formal direction.
More generally, we can say that rule-of-law values would operate to restrain the tendency of the sociological approach to produce a highly situational law of discrimination. The rule of law and its bad twin, legalism, prefer objective rules to flexible standards. Thus, many of the doctrines that Professor Post criticizes as historically crude and sociologically insensitive can nonetheless be defended as more impersonal and detached than the doctrinal alternatives that an unadulterated contextual approach would otherwise suggest.
Commentators inculcated in the legalist tradition will regard an unduly situational body of civil rights law not as a properly lawful structure of neutral rules and principles, but as a reform project that imposes an unacceptably managerial task on judges. Egalitarian lawyers will worry that if judges (who tend to be creatures of the establishment) are left alone to decide civil rights cases on the basis of particularized judgments of invidiousness, they will be too inclined to find familiar practices lawful. [*74] Conservative lawyers will worry that if liberal judges follow their ethnographic intuitions and strategic instincts, they will tend to find invidiousness everywhere and then impose a draconian regime of political correctness to stamp it out. Judges themselves, faced with decisions that engage burning social passions on both sides, will look for solutions that do not force them to take sides too obviously.
A practical response to these interacting concerns might well be a compromise on a relatively (though not absolutely) formal version of civil rights law. All parties, legalist and judicial, liberal and conservative, fear the consequences of too much judicial discretion. All might then converge on the substitution of relatively crude, yet objective and thus stable rules for the sociologically nuanced and flexibly strategic responses that one would expect if Professor Post's contextual approach were generally accepted. In practice, of course, any workable body of rules will leave considerable room for the exercise of situational judgment. But when we take account of the additional context supplied by politics, and by the acculturation of lawyers and judges in a legalist tradition, we may reintroduce a considerable amount of the formalism we were originally inclined to displace.
The resulting law of civil rights might look something like the law we have today. And yet such a body of law would not rest to any degree on any lawyer's or judge's failure to understand the cultural and historical specificity of the social norms through which we construct race, gender, and so on. Its formalistic and legalistic features would be grounded not in philosophical error but in the practices and practicalities, themselves quite culturally and historically specific, that make up the institutions and environment of a particular legal system at a particular time.
My point is not that a sociologically grounded critique of the kind represented by this remarkable Brennan Lecture devours itself, and so is futile and should be avoided. Rather, I agree with the thrust of Professor Post's argument, and have tried to amplify it in what I hope is a useful way. I do not believe that making the critique, and then qualifying it by reference to our legal culture's devotion to the rule of law (and embroilment in legalism), simply leaves things where they were.
Emphasis on the fact that the rule-of-law/legalism complex is part of the context of law, and so of antidiscrimination law, does not take away the force of the contextualist critique. No doubt, in discussing issues of discrimination and prejudice, we will continue to recite to each other the maxim that you can't tell a book by its cover. But we should not confuse this sometimes useful nostrum with an inflexible axiom or a philosophical principle. We make a lot of judgments about what to read on the basis of what we know about books from their covers alone, and we could scarcely do otherwise.
[*75] In the same vein, the notion that the controversy over affirmative action can be resolved by simple application of a slogan about color blindness cannot survive Professor Post's critique. Not only is there an important distinction between an abstract and a contextual approach to antidiscrimination law, but the contextual approach is better.
For related research and practice materials, see the following legal topics:
Education LawDiscriminationRacial DiscriminationDesegregationFaculty & StaffHealthcare LawBusiness Administration & OrganizationPatient ConfidentialityGeneral OverviewLabor & Employment LawDiscriminationGender & Sex DiscriminationDefenses & ExceptionsBona Fide Occupational Qualifications
n1. See Robert Post, Prejudicial Appearances: The Logic of American Antidiscrimination Law, 88 Calif. L. Rev. 1 (2000).
n2. See id. at 7.
n3. See id. at 2-16.
n4. See id. at 16-30.
n5. See id.
n6. See id. at 18-30.
n7. 443 U.S. 193, 208-09 (1979) (holding that an unqualified statutory prohibition of employment discrimination on the basis of race should not be read to prohibit a race-based affirmative action hiring program).
n8. See Post, supra note 1, at 25-26.
n9. See id. at 27.
n10. See id. at 23-24.
n11. See id. at 24-25, 29.
n12. See, e.g., Deborah L. Rhode, Justice and Gender: Sex Discrimination and the Law 306-13 (1989).
n13. See Post, supra note 1, at 25-26.
n14. See id. at 27.
n15. See id.