Printer Friendly

"The very stereotype the law condemns": constitutional sex discrimination; law as a quest for perfect proxies.

In his lone dissent from the Supreme Court's decision holding the Virginia Military Institute's exclusion of women unconstitutional, Justice Scalia worked up to yet another fever pitch of outrage at what he perceived as the majority's unprincipled departure from established doctrine. According to an indignant Justice Scalia:
   [T] he Court proceeds to interpret "exceedingly persuasive justification"
   in a fashion that contradicts the reasoning of ... other precedents.

      That is essential to the Court's result, which can only be achieved by
   establishing that intermediate scrutiny is not survived if there are some
   women interested in attending VMI, capable of undertaking its activities,
   and able to meet its physical demands.


      Only the amorphous "exceedingly persuasive justification" phrase, and
   not the standard ,elaboration of intermediate scrutiny, can be made to
   yield this conclusion that VMI's single-sex composition is unconstitutional
   because there exist several women (or, one would have to conclude under the
   Court's reasoning, a single woman) willing and able to undertake VMI's
   program.... There is simply no support in our cases for the notion that a
   sex-based classification is invalid unless it relates to characteristics
   that hold true in every instance.(1)

The view that U.S. v. Virginia ("VMI") is a departure from precedent and that the standard it articulates is "amorphous" is not unique to Justice Scalia, or even to critics of the decision. A variety of commentators have seen new promise in the decision. For example, Cass Sunstein, who speaks well of the VMI decision, also characterizes it as unsettling the constitutional law of sex discrimination, changing the test, "heighten[ing] the level of scrutiny and bring[ing] it closer to ... `strict scrutiny.'"(2) Sunstein puts VMI in a category of "narrow," "deeply reasoned" decisions he calls "the most rule-free" of all the decision categories he describes.(3)

This Article will demonstrate that Justice Scalia is quite right in his articulation of the standard applied by the majority in VMI, but wrong to assert that it is a departure of any kind from precedent.(4) Cass Sunstein is also wrong to suggest that "[a]fter United States v. Virginia, it is not simple to describe the appropriate standard of review"(5) and to view the VMI decision as in any way "rule-free" or an example of even a limited move "in the direction of open-ended balancing."(6)

VMI instead marks yet another application of the rule that has governed constitutional sex discrimination cases since the early 1970s, a rule quite "simple to articulate" and applied in a way that might satisfy the most rigid of formalists, but one often lurking, as common law rules(7) sometimes do, just below the surface of the decisions applying it. It is my contention that the components of the intermediate scrutiny standard---a practice "substantially related to an important governmental objective"--have rarely been the moving parts(8) in a Supreme Court sex discrimination decision.(9) Rather, the bulk of the work in these decisions has been done by what readers of the opinions may be tempted to treat as mere decorative rhetorical flourish--the proposition that there are constitutional objections to "gross, stereotyped distinctions between the sexes,"(10) that is to say, to "classifications based on sex ... premised on overbroad generalizations."(11) To determine whether there is unconstitutional sex discrimination, one need generally ask only two questions: 1) Is the rule or practice at issue sex-respecting, that is to say, does it distinguish on its face between males and females?(12) and 2) Does the sex-respecting rule rely on a stereotype?

In the constitutional, just as in the statutory, law of sex discrimination, "stereotype" has become a term of art by which is simply meant any imperfect proxy, any overbroad generalization. For a sex-respecting rule to withstand constitutional scrutiny by the Court, it seems to be at least necessary and usually sufficient that it embody some perfect proxy.(13) That is to say, the assumption at the root of the sex-respecting rule must be true of either all women or no women or all men or no men; there must be a zero or a hundred on one side of the sex equation or the other.(14) Even a generalization demonstrably true of an overwhelming majority of one sex or the other does not suffice to overcome the presumption of unconstitutionality the Court has attached to sex-respecting rules: virtually every sex-respecting rule struck down by the Court in the last quarter century embodied a proxy that was overwhelmingly, though not perfectly, accurate.(15) Moreover, overbreadth alone seems to be enough to doom a sex-respecting rule. This is so even though many of the generalizations embodied in sex-respecting rules struck down by the Court are not only overbroad but also "archaic." That is to say, that as well as being descriptively less than perfectly accurate, these generalizations also embody outdated normative stereotypes (i.e., "fixed notions concerning the roles and abilities of males and females"(16) or "the accidental byproduct of a traditional way of thinking about females"(17).

On this view of the law, as I shall discuss, VMI is an extremely easy case, the logical culmination of a long line of cases rather than any sort of new beginning. Justice Ruth Bader Ginsburg, author of the majority opinion, stands at both ends of this line, now able to affirm as a Justice what she first argued as an advocate. All of the moving parts of the present law are fully articulated in her brief for the appellant in Reed v. Reed,(18) although it took until the second modern constitutional sex discrimination case, Frontiero v. Richardson,(19) which Ginsburg argued for the ACLU as amicus curiae, for members of the Court explicitly to adopt them.

I do not mean to suggest that modern constitutional sex discrimination law sprang full grown from the head of Ruth Bader Ginsburg like Athena from the head of Zeus.(20) Not only were many other advocates and theorists involved in shaping the arguments presented to the Court and to the nation, the Court's view of sex discrimination was strongly shaped by what went "[b]efore, behind, between, above, below."(21) Three main influences-the developing law of sex discrimination under Title VII of the Civil Rights Act of 1964;(22) the Court's own past decisions in early, discredited sex and race cases; and its contemporaneous obsession with closely examining all irrebuttable presumptions--primed the Court to formulate the law as it did. As I shall attempt to demonstrate, constitutional sex discrimination law is in many ways path dependent on Title VII, which since 1964 has outlawed discrimination in employment on the basis of sex. Title VII made an explicit textual commitment to a "sameness," or anti-discrimination, approach.(23) And, under the rubric of bona fide occupational qualification ("BFOQ"), which is the single justification under the statute for discrimination on the basis of sex, there developed an anti-stereotyping jurisprudence that centered on a search for perfect proxies.(24) As the Court began seriously to consider questions of sex discrimination under the Constitution, it saw that the very sort of stereotyping now prohibited by Title VII was the basis for earlier constitutional decisions such as Justice Bradley's concurrence in Bradwell v. Illinois,(25) which denied employment opportunities to women on the basis of sex;(26) it rejected stereotypes in rejecting Justice Bradley's opinion.(27) Finally, the current constitutional law of sex discrimination crystallized in that narrow window of time during which a majority of the Court seemed disposed to take seriously the notion that there was something constitutionally problematic about any irrebuttable presumption, any failure to make room for individual exceptions to general rules, whatever the characteristic on which such rules were based.(28) Thus the justices to whom Ginsburg first presented her arguments were a receptive audience to the extent they were impelled by Title VII's prohibition of sex-stereotyping, repelled by their predecessors' reliance on stereotypes, and surrounded by doubts about all reliance on any overbroad generalizations.

Nor do I mean to suggest that VMI is merely a dead end for the law. Rather, as I shall discuss, taking its anti-stereotyping message seriously would lead the laws governing men and women, particularly as sex and marriage partners and as mothers and fathers, in interesting and radical directions. The decision also highlights the growing disparity between the Court's treatment of race and sex in matters involving the relationship between stereotyping and affirmative action.

Before hypothesizing further about where the constitutional rule against sex-stereotyping came from and what it may mean for the future adjudication of constitutional questions of sex discrimination and sex equality, let me try to explain how the rule differs from a more conventional application of heightened scrutiny and how the rule is articulated and applied in Supreme Court cases over the past quarter century.



Perhaps the best way of illustrating the difference in the requirements of the prohibition on stereotyping on the one hand and conventional heightened scrutiny on the other is by applying both tests to the facts of the case generally recognized as the first major step toward the construction of today's tiers of scrutiny, Korematsu v. United States.(29) Infamously, the Court in Korematsu subjected an order banning those of Japanese ancestry from their homes on the West Coast to "the most rigid scrutiny"(30) and nevertheless upheld the order.(31) Korematsu is particularly useful as an illustration because it is one of the few race cases in which race is clearly being used as a proxy, specifically, for loyalty to the United States.(32)

If the perfect proxy test is applied to the order in Korematsu, it clearly fails the test. The order at issue is clearly both over- and under-inclusive. On neither side of the equation does it embody a perfect proxy since there are concededly both loyal Japanese-Americans and disloyal non-Japanese. The difference between this test and the conventional test of strict scrutiny--namely, whether the order is narrowly tailored to meet a compelling governmental interest--is the difference between attention to false positives and false negatives. Whereas the perfect proxy rule asks whether there are any non-traitorous Japanese-Americans and would use a "yes" answer to strike down the rule, avoiding false positives, conventional strict scrutiny, having established that the exclusion of Japanese traitors during a war with Japan is a compelling interest, would instead ask whether there is any rule more narrowly tailored than the exclusion order at issue that would exclude all traitorous Japanese. It seeks to avoid false negatives, but by the least restrictive means. Since there may well be no more narrowly tailored rule that would serve, some rules that might fail the perfect proxy test could nevertheless survive strict scrutiny.(33) This is in fact what happened in Korematsu, where the majority acknowledged that, "[i] t was because we could not reject the finding of the military authorities that it was impossible to bring about an immediate segregation of the disloyal from the loyal that we sustained the validity of the ... order as applying to the whole group."(34)

Among many other things, Korematsu may demonstrate how very malleable conventional heightened scrutiny is to judicial manipulation. More importantly, the comparison of the two standards suggests that those commentators who see the standard applied in VMI as more exacting than that set forth in the conventional formulation of intermediate scrutiny may be on to something, but this more exacting standard is nothing new. The perfect proxy test has always had the capacity to be more strict even than strict scrutiny. One reason this may not have been obvious in the past may be that, for decades, very few of the race cases to which strict scrutiny was applied involved the use of race as any sort of proxy.(35) Instead, perhaps the most pernicious aspect of the racial classifications at issue in cases from Brown to Loving v. Virginia(36) was that race was being used for its own sake.(37)

Now we are entering an era in which the bulk of the race cases reaching the Court can be characterized as using race as a proxy--for disadvantage or diversity in the affirmative action cases,(38) or for community of interest in the voting rights cases--and once again the language of perfect proxy is asserting itself. This could lead to a convergence of standards for race and sex, even though there is disagreement on the Court about the form that convergence should take. Thus, though the view of liberal justices that intermediate scrutiny should govern "benign" racial classifications has been rejected,(39) language in the voting rights cases suggests that the conservatives on the court may insist that race be used as a proxy for community of interest only when it is a perfect proxy.(40)

Interestingly, there is a sense in which VMI could have been a much more difficult case to fit under the perfect proxy, anti-stereotyping rubric: VMI was not really using sex as a proxy for anything; it was maleness itself in which the school was interested. What really mattered to VMI was its cult of masculinity in a world sealed from the presence of women who might either meet or undermine the masculine standard, in each case threatening male privilege. In some respects, this makes VMI an even easier case, because it then begins to look more like the many race cases from Plessy v. Ferguson(41) through Brown and Loving, in which race also was not really used as a proxy for anything, it was the thing itself' that the legislators were after. In that event, it is far more difficult to argue that a legitimate governmental interest could be served thereby: just as it was a case of preserving white supremacy in Loving, it was a case of preserving male supremacy in VMI; and Virgina's lack of concern for its daughters in VMI is paralleled by its lack of concern about race mixtures among the "inferior" races in Loving.(42)

The analogy between VMI and the early race cases extends beyond doctrine to history and sociology. What is going on at VMI is very close to what happened in the South more generally in the era of segregation and massive resistance--the fear of contamination by a group. VMI's student population is by and large a hard scrabble group, fairly poor, and with low SAT scores; they are clinging to male privilege because it's all they have.(43) The admission of women to VMI can also be predicted to have some of the same effects as the admission of blacks. For example, hierarchy works differently in a cross-race or cross-sex context. Whether it be a black man giving orders to a white man in the 60s South, or a white man insulting a black man in the way superior officers are meant to insult entering freshman at VMI, the introduction of culturally coded racial differences disturbs the illusion of equality in sameness on which VMI claims to rely.(44) As Justice Breyer indicated at the oral argument of VMI, the exclusion of blacks could be justified in much the same terms as VMI used to try to justify the exclusion of women.(45)

In choosing to litigate the case as involving maleness as a proxy for both interest and ability,(46) VMI made a decision that, whatever its temporary success in the benighted Fourth Circuit, was bound to fail when set squarely against Supreme Court precedent. According to the District Court, its findings of fact were proof that the establishment of an all-male and an all-female VWIL(47) rested on real differences between the sexes, and not on stereotypes. But precisely what the District Court found as a matter of fact not to be a stereotype, conclusively proved as a matter of law that stereotypes were the basis of the sex distinctions in the case. None of the many facts as to ability or inclination of males and females was categorical. In each case the most that was claimed was that the findings were true of the vast majority of one or the other sex. But, as Justice Ginsburg noted, what the constitutional law of sex discrimination teaches unequivocally is that "generalizations about `the way women are,' ... no longer justify denying opportunity to women whose talent and capacity place them outside the average description."(48)



In arguing that the rule of decision in constitutional sex discrimination cases has been that of the perfect proxy as described above, I do not mean to indicate endorsement of this rule. Not only do I question whether our concerns about sex equality are appropriately circumscribed by an inquiry only into the closeness of fit of sex-respecting rules, it also seems to me, as it has to many other commentators, that at least some of the perfect proxies found by the Court in upholding sex-respecting rules since the early 1970s are specious.(49) My claim that post-Frontiero constitutional sex discrimination cases can be explained by reference to such a rule is a descriptive one, although I address some of its normative implications. Descriptively, while I do not claim that the perfect proxies the Court found are good ones, I do claim (1) that the majority found a perfect proxy in every sex-respecting rule the Court upheld since Frontiero, and (2) for every sex-respecting rule struck down since Frontiero, no perfect proxy, even a farfetched one, occurred to the court, or to me.(50)

In the years since Frontiero, the case in which the Court first expressly noted constitutional opposition to sex-stereotyping, the Court has examined approximately two dozen sex-respecting rules for constitutionality and upheld only about a half dozen. In some of these, the perfect proxy identified was created by another sex-respecting rule whose constitutionality was not before the court; another group involved enduring "[p]hysical differences between men and women"(51) and, in the third group, "[s]ex classifications [were] ... used to compensate women `for particular economic disabilities [they have] suffered.'"(52)

Perhaps most notorious is the perfect proxy identified by the plurality in Michael M. v. Superior Court.(53) upholding California's statutory rape law, which criminalized exclusively sexual intercourse with females under the age of eighteen. The statute at issue in Michael M. was in several ways a sex-respecting rule: as the California Supreme Court held, it "discriminates on the basis of sex because only females may be victims, and only males may violate the section."(54) For the California Supreme Court, as for the Supreme Court plurality, both halves of this sex distinction rested on a perfect proxy, and hence, the sex distinction embodied in the law was constitutionally permissible. As Justice Rehnquist noted in summarizing the lower court opinion with approval, "the classification was `supported not by mere social convention but by the immutable physiological fact that it is the female exclusively who can become pregnant.'"(55) Moreover, "males alone can `physiologically cause, the result which the law properly seeks to avoid,' ... [so] the gender classification was readily justified as a means of identifying offender and victim."(56) Whatever one may think about the archaic nature of Justice Rehnquist's foray into reproductive biology, which conjures up Aristotelian images of homunculi in sperm as well as reinforcing stereotypical notions of female passivity and male activity in sex, he got from it the prefect proxy he seemed to think he needed.(57)

The second group of perfect proxies are no more normatively satisfying. Although, as I have noted, Justice Bradley's concurrence in Bradwell has become the Court's favorite example of what was wrong with earlier views of relations between the sexes, the one aspect of Justice Bradley's opinion that has not been rejected by the modern Court is the notion that one sex discriminatory law can be justified by reference to others not before the court. Just as the legal disabilities of a regime of coverture were used by Justice Bradley to justify Bradwell's exclusion from the bar, so women's exclusion from combat justified both the extra time given women to achieve promotion,(58) and women's exclusion from military registration.(58) And, notwithstanding that the number of women eligible for veterans' preferences in Massachusetts was kept infinitessimally low(60) by "the variety of federal statutes, regulations, and policies that have restricted the number of women who could enlist in the United States Armed Forces, and ... the simple fact that women have never been subject to a military draft,"(61) the Court in Personnel Administrator v. Feeney upheld an absolute preference for veterans whose effect was admittedly to keep qualified women out of the upper echelons of the Massachusetts Civil Service.(62) Thus, the combat exclusion is used to justify both a "benefit" to women (longer time), and women's exemption from the "burden" of draft registration. But, note, of course, that the net result to women includes the resulting deprivation, not only of concrete opportunity as in Feeney, but also of citizenship value, inclusion, and respect.

The most interesting group of cases in which a sex-respecting rule was upheld are those in which the purpose of the rule is seen to be to compensate women for discrimination against them as a sex. This group includes Ballard, Califano v. Webster(63) and Kahn v. Shevin.(64) To see how sex can serve as a proxy for discrimination in these cases, it is best to begin with Ballard. In this case women were given longer than men to demonstrate their promotability under an "up or out" military regime that mandated discharge for those not promoted within a set time. The justification given was, once again, women's exclusion from combat positions, in which promotability could be demonstrated far more readily. The fact that, notwithstanding the combat exclusion, some women could and did get promoted in the time allowed men did not mean that the proxy was imperfect, because the exclusion still affected all women, even those who by extraordinary effort, luck, or skill managed to overcome its handicap. Similarly, if a rule says a woman must be twice as good as a man to be promoted, and she is twice as good, the discriminatory rule will not have changed the outcome in her case, but she will still not have been exempt from the application of the rule.(65) With this structure of analysis in mind, it becomes easier to see how even the rule in Kahn,(66) which gave a $500 annual property tax exemption to widows, but not to widowers, can rely on the perfect proxy of societal discrimination against women, notwithstanding that many female beneficiaries of the exemption may not have suffered identifiable harms from discrimination. As women, they are still surrounded by the ambient level of discrimination against them; just like the military women in Ballard, they may be excluded from jobs they neither need nor want, they may succeed despite the exclusion, but they will have been excluded nevertheless. The interesting question posed by these cases and by Justice Ginsburg's reaffirmation in VMI that sex-respecting rules will be upheld if they are "used to compensate women `for particular economic disabilities they have suffered,' ... `promote equal employment opportunity,' [or] advance full development of the talent and capacities of our Nation's people,"(67) is what limits, if any, there are on the use of discrimination as a proxy to justify compensatory or affirmative action schemes for women.(68) Must there be some sort of narrow tailoring, or will any scheme do so long as there is across the board discrimination to serve as a perfect proxy?



The current constitutional law of sex discrimination is an excellent example of what David Strauss has called "common law constitutional interpretation."(69) Not only is the notion that discrimination on the basis of sex presumptively violates the Equal Protection Clause a "settled principle[]"(70) of law with no clear source in the text or original understanding of the Constitution, not only does it "rely ... on the elaborate body of law that has developed, mostly through judicial decisions, over" the past quarter century,(71) it is also a fine illustration of the "fact [that] rules, as well as case-by-case decision making, are an important part of the common law."(72) While Justice Ginsburg's majority opinion exemplifies the application of common-law-style rules, Justice Rehnquist's concurrence, like so much of Justice Rehnquist's jurisprudence, is common law constitutional interpretation of a far more primitive sort, "encompassing case-by-case method that emphasizes analogy, context, and `situation sense.'"(73)

Justice Rehnquist concurrence in VMI is a splendid argument for what, with a shudder at being unable to find a more graceful term, I am tempted to call pro-active constitutionalism. This is the notion that constitutional actors such as the state of Virginia and VMI are under an obligation to take seriously the commands of the Constitution as interpreted by the precedents of the Court and apply them even in the absence of litigation prodding them to do so. In faulting Virginia for not seeing the handwriting on the wall, and following its command even in advance of litigation, Justice Rehnquist seemed to begin with a premise more associated with the civil law than the common law tradition, to wit that only a consistent line of cases (what the Germans call "staendige Rechtsprechung" and the French "jurisprudence constante)," rather than a single case, has any strong precedential force. He recounts the history of modern sex discrimination law, beginning with Reed v. Reed.(74) That case, he said, involved a fact pattern quite different from VMI and hence did not put the school "on notice that its holding would be extended across the constitutional board. They were entitled to believe that `one swallow doesn't make a summer' and await further developments."(75) Interestingly, however, it is not the long line of cases after Reed that, for Justice Rehnquist, constitute the relevant developments. For him, "[t]hose developments were 11 years in coming.... Mississippi University for Women v. Hogan ... actually involv[ed] a single-sex admissions policy in higher education.... This holding did place Virginia on notice that VMI's men-only admissions policy was open to serious question."(76) This is common law reasoning with a vengeance--only a precedent on all fours as to the facts, not a developed role of decision, is seen as binding.

Several different lines of precedent led to modern constitutional sex discrimination law.

A. Title VII

Among the most important is Title VII, with its strong textual commitment to anti-discrimination, and the subsequent interpretation of the BFOQ requirement to mandate a perfect proxy, that is to say, to be limited in applicability to circumstances where no person of the opposite sex could perform the job.(77)

Anti-stereotyping doctrine developed in a series of lower court cases,(78) some of which combined constitutional and statutory questions, and hence impelled courts to apply the same standards to both. Notable among these latter are (1) cases covering government employment, particularly those encompassing periods of time both before and after such employment was subject to Title VII, and (2) cases involving state laws limiting women's employment, which embody or are based on stereotypes.(79) such as the California law prohibiting the employment of women bartenders at issue in Sail'er Inn v. Kirby,(80) a case cited with approval by the Supreme Court. To the extent the statute at issue in Kirby was applied to employers large enough to be covered by Title VII, the, California Supreme Court struck it down as conflicting with federal law, because it rested on impermissible stereotypes rather than anything amounting to a BFOQ. As to employers not covered by Title VII, the California court reached the same result under the state and federal constitutions by applying strict scrutiny, which it held applicable, "first, because the statute limits the fundamental rights of one class of persons to pursue a lawful profession, and, second, because classifications based upon sex should be treated as suspect."(81) The earliest Supreme Court reference to sex stereotyping appears during 1970 injustice Marshall's concurrence in Phillips v. Martin Marietta.(82) a Title VII case. The term enters the Court's constitutional vocabulary in Frontiero three years later.

Approaching the early constitutional sex discrimination cases with Title VII standards in mind may have been encouraged by the fact that a disproportionate number of such cases before the Court, among them Frontiero, involved employment related issues such as the terms and conditions of government employment and pension benefits. Even Reed's petition to be appointed estate administrator is in essence an application for a job. Moreover, Hogan, VMI, and Stanton involve education explicitly seen as vocational; and many of the most notorious cases of the pre-Reed era, for example, Bradwell, Muller, and Goesshart, restricted women's job opportunities. In examining the extent to which constitutional sex discrimination law is path dependent on Title VII, it is important to note where the paths diverge. Of course, the most dramatic divergence is in the failure of the Court to adopt a constitutional disparate impact standard as expansive as that imposed under Title VII. But it is also worth considering if the question of whether pregnancy-based rules are sex-respecting, and hence problematic, might have been answered differently had a statutory case, in which women could have relied more strongly on EEOC Guidelines, had reached the Court before the purely constitutional case of Geduldig.

B. The ERA

At the other extreme of the path dependency question, it is also worth asking how, if at all, the constitutional law of sex discrimination might have evolved differently if the text of reference had not been Title VII, which explicitly prohibits "discriminat[ion] ... because of ... sex,"(83) but instead the ERA, which, in its 1970s textual incarnation provided instead that "[e]quality of rights under the law shall not be denied or abridged ... on account of sex."(84) Some evidence of a different path can be seen in interpretations of state ERAs. Thus, for example, the Washington Supreme Court, interpreting that state's ERA, upheld a requirement that "the two members of the State Democratic Committee elected by the county central committees be of the opposite sex"(85) and the chair and vice-chair also be of the opposite sex on the grounds that these rules "assure[d] women actual as well as theoretical equality of rights."(86)

On the whole, however, there are numerous persuasive reasons to suspect that a change in textual basis from the Fourteenth Amendment to an Equal Rights Amendment would have made little difference. Among them are, first, that the relevant portion of the Fourteenth Amendment, like the ERA, makes a textual commitment to equality rather than against discrimination.(87) Second, all sides in the ERA debate, but particularly the most vocal and authoritative proponents of the Amendment, were in agreement that its effect would be to prohibit sex distinctions in the law--whatever else it might have done, the Amendment would not, according to participants in ERA debates, have changed the result in those cases in which the Court struck down sex-respecting rules. Indeed, as noted above, the current constitutional doctrine on sex discrimination bears a striking resemblance to that set forth by the authors of an influential Yale Law Journal article endorsing and interpreting the ERA.(88)

C. The Court's Contemporaneous Campaign Against Irrebuttable Presumptions and in Favor of Individualized Adjudication

Most interestingly, the Court's rejection of overbroad sex stereotypes is part of the jurisprudential climate of the time, during which, across a wide spectrum of subject areas and predicate constitutional provisions, the Court insisted on individualized adjudication rather than reliance on overbroad presumptions. A sea of concern about the constitutional difficulties with overbroad rules and imperfect proxies swept over the Court in precisely that time period during which the current constitutional law of sex discrimination was developing. At the time of its development, sex discrimination doctrine was one of many related, flourishing life forms in this sea, but, by the time the sea receded toward the end of the 1970s, only sex discrimination doctrine had developed the hard skeleton necessary to insure both a clear fossil record and the continuing viability of the organism.(89)

The most closely related and well known of the group of doctrines I have in mind is the notion that irrebuttable presumptions are constitutionally problematic. In addition to being the background assumption in sex discrimination cases like Reed,(90) this doctrine also was decisive in cases involving issues related to sex, but not analyzed by the Court as involving sex-respecting rules. Chief among the latter group of cases is Cleveland Board of Education v. LaFleur,(91) involving a requirement that pregnant teachers stop work at a fixed point fairly early in their pregnancy. Although a lower court saw such a rule as sex discriminatory, the Supreme Court struck it down as instead involving an impermissible irrebuttable presumption of incapacity to work.

Other cases at the intersection of sex discrimination and irrebuttable presumptions involve rules governing the rights of illegitimate children and their parents(92) such as Stanley v. Illinois,(93) the second case after Reed in which the Court struck down a sex-respecting rule.(94) The rule in Stanley, struck as a violation of due process as well as equal protection, would have deprived a father of the custody of his illegitimate, motherless children without a hearing. In arguing unsuccessfully against an individualized hearing for Stanley's claim, Illinois relied on stereotypes, on overbroad generalizations, on irrebuttable presumptions--three terms of art for the same practice--about fathers in general and non-marital fathers in particular.(95) Illinois "argued that unmarried fathers are so seldom fit that [it] ... need not undergo the administrative inconvenience of inquiry in any case.(96) Citing, inter alia, Reed,(97) the Court held that, though Illinois's wish for "prompt efficacious procedures" was a "proper state interest," it was insufficient: "[T]he Constitution recognizes higher values than speed and efficiency."(98)

But numerous cases in this line had nothing whatever to do with sex. In the years surrounding the development of current sex discrimination law, the Court also held it would be unconstitutional to presume conclusively that soldiers were not bona fide residents of the town in which they were stationed;(99) that only property holders and parents were interested in school board elections;(100) that Communist party members would engage in subversive acts if given passports(101) or public employment(102) and that recent migrants from one state to another were not entitled to voting rights,(103) welfare(104) or in-state tuition benefits.(105) As with sex-respecting rules so with the presumptions in these cases--a single exception was enough to doom the rule. Only presumptions universally true withstood scrutiny and individual exceptions to the rule were entitled to individual adjudication.

In addition to the line of causes explicitly objecting to irrebuttable presumptions, the Court during the decade in which constitutional sex discrimination doctrine developed emphasized individualized adjudication and opposed overbroad rules in a variety of other contexts. For example, the early 1970s were the high water mark of First Amendment overbreadth doctrine.(106) By the mid-1970s, the Court insisted that mandatory death sentences for certain offenses were over-broad and consideration of individual circumstances was constitutionally required.(107) And an emphasis on the benefits and corresponding lack of concern for the costs of individualized adjudication extended at the time beyond the area of irrebuttable presumptions into other due process cases.(108)

Finally, in his opinion in Weinberger v. Salfi,(109) which is generally considered to mark the end of the era of heightened scrutiny for most irrebuttable presumptions, Justice Rehnquist, at the time still not convinced that special attention to laws discriminating on the basis of sex was a good idea, indicated that he understood quite well the relationship between irrebuttable presumptions and the evolving law of sex discrimination when he cited with approval, as examples of permissible presumptions, the restrictions on women at issue not only in Geduldig, but also in Goesaert v. Cleary.(110)

It is in no small part because Justice Rehnquist's view of the continuing validity of precedents such as Goesart, was distinctly a minority opinion on the Court by the mid-1970s that heightened scrutiny for sex-respecting rules survived the demise of the irrebuttable presumption doctrine. By that time the Court had adopted Ruth Bader Ginsburg's view that all pre-Reed cases involving sex distinctions were "Precedent in Need of Re-evaluation."(111) The Court's re-evaluation of this precedent was itself a major influence on the development of the current law.

D. Justice Bradley as the Bogeyman

The embodiment of all that modern constitutional sex discrimination law sets its face against is the concurring opinion of Justice Bradley in Bradwell v. Illinois,(112) an opinion repeatedly cited with strong disapproval by the modern Court and commentators. The majority in Bradwell held that, a license to practice law not being a privilege or immunity of citizenship, the state of Illinois had not violated the Privileges or Immunities clause by denying Myra Bradwell's petition for admission to the bar solely on grounds of her sex.(113) Justice Bradley, as one of the dissenters in the Slaughter-House Cases, had a more expansive view of privileges and immunities; indeed, as further discussed below, his dissent opined that "a law which prohibits a large class of citizens from adopting a lawful employment ... [deprives] them of liberty as well as property, without due process of law."(114) In Bradwell, he exhibits no such qualms about depriving the large class of female citizens of such an opportunity, even in the absence of express legislation.(115) Passages from his concurrence have been quoted in so many opinions and commentaries in the last quarter century that scholars of sex discrimination can practically recite them by heart.(116) It was these passages that Justice Brennan quoted just before he said in Frontiero that, "as a result of notions such as this, our statute books became laden with gross, stereotyped distinctions between the sexes and, indeed, throughout much of the 19th century the position of women in our society was, in many respects, comparable to that of blacks under the pre-Civil War slave codes."(117) Justice Bradley's opinion is chockablock with stereotypes. He regularly paired normative and descriptive stereotypes, talking of what is "natural and proper" or what "belongs to the female sex."(118)

Although Justice Bradley recognized that there are exceptions to all his male and female stereotypes, for example, there are unmarried women not subject to coverture, he left no space for them. A similar recognition that there are individual exceptions, coupled with a resolute refusal to make space for them in the law, infected other earlier Supreme Court sex discrimination cases that the modern Court views as negative precedent. Thus, for example, Justice Brewer in Muller v. Oregon(119) justified the paternalistic protection of women in employment contracts by enumerating the various physical, educational, and temperamental disadvantages that a woman faced when asserting her rights.(120) He then admitted that "[d]oubtless there are individual exceptions," but took no account of the effect of the rule the Court upheld on such exceptions.(121) The modern constitutional and statutory law of sex discrimination makes a place for these exceptional women. For, in rejecting Justice Bradley's opinion as precedent, the Court has concluded that "rules of civil society" not only can but "must be adapted to ... exceptional cases."(122)

E. Analogies from Race Cases

In insisting on a place for exceptional individuals, modern sex discrimination doctrine not only rejects earlier cases involving sex discrimination, but, furthering the parallels Justice Brennan observed between race and sex discrimination, draws implicit support from race cases interpreting the requirements of equality in a time of separate spheres for race as well as sex, the time of "separate but equal."(123) Given that what all these lines of precedent leading up to the constitutional law of sex discrimination have in common is a focus on the individual, it should hardly have surprised Justice Scalia that the existence of even one woman interested in and capable of attending VMI would require an end to its all-male status.(124)

Having set forth what I believe to be the central features of the current constitutional law of sex discrimination, its parameters and its origins, I should now like to consider the implications of that law for several open issues likely to be presented to courts in the near future.



As Ruth Colker has observed, "[t]wo sometimes conflicting principles, anti-differentiation and anti-subordination, underlie equal protection jurisprudence."(125) At present, our constitutional standard with respect to sex is not, as Colker would have it, "anti-subordination above all," but rather "anti-stereotyping" above all. Notwithstanding the heroic efforts of Colker and others to argue to the contrary, I remain of the view that with respect to the sexes, anti-differentiation is also normatively (i.e., in terms not only of what the law now does, but what it should do), an attractive principle, unfashionable though that view may now be.(126)

There are two main ways of formulating the principle behind the constitutional norm against the denial of equal protection on grounds of sex. The first is that women Should not be subordinated by the law or, more broadly, by men. The second is that sex should be irrelevant to an individual's treatment by the law, and, more broadly, to his or her life chances. On the latter view, "fixed notions concerning the roles and abilities of males and females"(127) are problematic when embodied in law, even in law that does not in any articulable way subordinate women to men. Note, despite superficial similarity leading to a temptation to conflate them, it is not the case that the latter principle necessarily reduces to formal equality and the former to substantive equality, as those are conventionally defined. Even if one confines one's inquiry exclusively to laws that discriminate on their face on the basis of sex (i.e., to sex-respecting rules, of the sort at issue in the vast majority of modern Supreme Court cases on the sexes and equal protection), one might still care whether the purpose or effect of a particular sex-respecting rule was to subordinate women and be inclined to strike down only those that can be demonstrated to do so. For example, this led then-Justice Rehnquist to object to striking down Oklahoma's higher beer-purchasing age for men.(128) By analogy to race cases,(129) it may lead to an easier defense of affirmative action measures for women. It leads Andrew Koppelman, among others who claim that prohibiting same sex marriage impermissibly discriminates on the basis of sex, to demonstrate at some length that restricting entry into marriage to two persons of different sexes has the intent and effect of subordinating women,(130) On the other hand, those who think sex should be irrelevant need not confine themselves to urging the abolition of sex-respecting rifles. Thus, for example, Wendy Williams, who was accused of seeking only "formal equality" in her call for "equal treatment," insisted that her "larger strategy" was to "get the law out of the business of reinforcing traditional, sex-based family roles and to alter the workplace so as to keep it in step with the increased participation of women."(131) Not just any sex-neutral rule would accomplish this end, according to Williams; instead "[a]n androgynous prototype requires sex neutral schemes that take into account the normal range of human characteristics--including pregnancy."(132)

It is no accident that we now refer to the law of sex discrimination rather than say, of sex equality. The Court has notoriously failed to consider anything that is not a sex-respecting rule to violate the constitutional norm against the denial of equal protection on grounds of sex. What it has required is not that the protection be equal, but that the rule be the same.(133) Not only has it failed ever to find disparate impact by sex to rise to the level of unconstitutional discrimination, it has occasionally been blind even to the disparity of the impact(134) and for long did not even see sex discrimination implicated in abortion questions. (135)

It may be worth noting that the main proponent on the Court of the anti-subordination strand in the current constitutional law of sex discrimination is Chief Justice Rehnquist.(136) Not only has he objected in the past to giving heightened scrutiny to laws disfavoring men, his VMI concurrence insisted:
   [I] t is not the "exclusion of women' that violates the Equal Protection
   Clause, but the maintenance of an all-men school without prodding any--much
   less a comparable--institution for women.... An adequate remedy in my
   opinion might be a demonstration by Virginia that its interest in educating
   men in a single-sex environment is matched by its interest in educating
   women in a single-sex institution. To demonstrate such, the Commonwealth
   does not need to create two institutions with the same number of faculty
   Ph.D.'s, similar SAT scores, or comparable athletic fields.... Nor would it
   necessarily require that the women's institution offer the same curriculum
   as the men's; one could be strong in computer science, the other could be
   strong in liberal arts. It would be a sufficient remedy, I think, if the
   two institutions offered the same quality of education and were of the same
   overall caliber.(137)

This opinion sounds suspiciously like a vision of separate (but equal) spheres, a vision of equality in sexual difference often announced, but never, to my mind, realized.(138) As with the separate but equal racial spheres categorically rejected in Brown and its progeny, much of the problem with constitutionally endorsed, state-enforced, separate spheres for the sexes may be a practical one. As Justice Souter noted at oral argument in VMI, because we do not stand "on the world's first morning" with respect to sex distinctions, but rather at the close of millenia of subordination, continued separation of the sexes along the remedial lines suggested by Chief Justice Rehnquist cannot be free of a subordinating taint.(139)

I would argue, however, that the objections go beyond impracticability. The constitutional principle that "[t]here is no caste here"(140) is not cashed out by "[t] here is no subordination here." The Constitution "neither knows nor tolerates classes among citizens," not even separate but equal classes.(141) Imagine, for example, a society with two castes, not upper and lower, not Brahmin and untouchable, but priest and warrior.(142) The two castes are equal in status, but radically different in role. Those born into the priest caste are limited to the role of priest even if they would rather fight than pray, and vice versa. Is such a division consistent with the American Constitution? I do not think so.(143) As Justice Harlan's dissent in Plessy v. Ferguson makes clear, the problems with equality of separate spheres exceed the practical: the Constitution guarantees liberty as well as equality.(144) Indeed, the constitutional equality norm itself has regularly been interpreted to guarantee equal liberty.(145) I would say of Justice Rehnquist's remedy for VMI what Justice Harlan said of the legislation at issue in Plessy: "Indeed, such legislation, as that here in question, is inconsistent not only with that equality of rights which pertains to citizenship, National and State, but with the personal liberty enjoyed by every one within the United States."(146) My contention is that, under the Constitution, no less than under Title VII, "[a]s for the legal relevance of sex stereotyping, we are beyond the day when [individuals of either sex can be] evaluate[d] by assuming or insisting that they match[] the stereo-type associated with their group."(147) This is so whether or not that stereotype is itself subordinating or demeaning.



I began this paper with the assertion that VMI marked the end, not the beginning, of a line of constitutional cases. This is because, within the narrow range the Court has identified as raising the constitutional problem of denial of equal protection on grounds of sex, few open questions remain. With all but a few exceptions, several of which I will discuss below, sex-respecting rules have vanished from the statute books and practices of state actors. Without exception, no practice that does not involve a sex-respecting rule has yet been seen by the Court as denying equal protection on grounds of sex and the VMI opinion certainly does not signal a change in this regard.

Many commentators, most of them critics, see the progress of the constitutional law of sex discrimination as, for better and for worse, the triumph of "sameness" feminism. A "sameness feminist," by my definition, is one who, if asked when, if ever, there should be sex-respecting rules, that is to say, rules that distinguish on their face between males and females, would give an answer straight out of Gilbert and Sullivan: When? "[n]ever ... What, never? No, never! ... Hardly ever!"(148)

In part because of the limited success constitutional sex discrimination law has had in promoting sex equality, "sameness" feminism is out of favor. Nevertheless, I, for one, still find it appealing. I by no means believe that all the problems posed by feminist jurisprudence can be solved by abolishing sex-respecting rules. But I think very few can be solved by instituting or affirming such rules. And, although many sensible criticisms have been made of the oppressive downside of abolishing such rules,(149) I think the radical possibilities of abolition have not been fully explored.

This Article is part of a larger project examining the tag ends of "sameness" feminist jurisprudence, including an inquiry, from both a legal and a theoretical perspective, into the few sex-respecting rules we have left.(150) A complete consideration of the appropriate treatment of each of these rules under the current structure of the constitutional law of sex discrimination, while certainly more manageable than any broader consideration of the appropriate scope of claims for sex equality under the Constitution, is still beyond the scope of this Article. Rather, I will limit myself to considering briefly three broad categories of remaining sex-respecting rules, those involving males and females as mothers and fathers, as participants in marriage, and as students in public schools.

A. Miller v. Albright and the Future of the Stereotyping of Mothers and Fathers

One piece of evidence that VMI changes nothing is furnished by the only constitutional sex discrimination case subsequently decided by the Supreme Court, Miller v. Albright.(151) That case involved a complex network of sex-respecting rules on citizenship whose net effect in the case before the Court was to deny citizenship to Lorelyn Miller. Miller was born in the Phillippines to a U.S. citizen and a Phillippine woman not his wife, but only acknowledged by her American father after she had grown to adulthood.(152) The laws at issue, because they granted citizenship to the nonmarital children of citizen fathers and alien mothers on different and less generous terms than they did to the nonmarital children of citizen mothers and alien fathers, facially discriminated on the basis of sex.(153) In a welter of opinions,(154) seven justices explicitly agreed that if the sex-respecting rules at issue were based on stereotypes, they would be unconstitutional. Five of those justices also explicitly stated their opinion that the rules were indeed based on impermissible stereotypes. Thus, if Congress and the Immigration and Naturalization Service heed Chief Justice Rehnquist's admonition to take precedent to heart, even before the commencement of litigation, they should read the collection of opinions as an injunction to alter the laws at issue, notwithstanding that Miller herself, for procedural reasons, lost her case.

But what alteration in the law would produce equal protection in this case? The answer is far from simple, either as a practical or as a theoretical matter. As a practical matter, the laws bestowing U.S. citizenship at birth to children only one of whose parents is a U.S. citizen are quite complex, with the requirements depending, not only on whether the citizen parent is the mother or the father as in Miller, but on whether the parents are married to one another. Even if a majority of the Court were to find standing in a Miller-like case, and also to disagree with Justices Scalia and Thomas that granting citizenship in such a case is within the Court's power, it would still be quite difficult for a court surgically to alter the laws at issue so as to eliminate any discrimination, without distorting the congressional scheme.(155) Perhaps this is one reason why Justices O'Connor and Kennedy, though denying Miller herself standing, went on in dicta to express their doubts as to constitutionality. They may have been signalling Congress and the State Department to get busy drafting before a plaintiff with standing came to court.

Congress could, of course, solve the Miller problem by requiring of all offspring of only one citizen parent of either sex or marital stares, merely proof of parental relationship, coupled with some uniform and minimal period of parental residence in the United States.(156) This would be a generous, straightforward and perhaps even just result. Rather than responding with aversion to the spectre of invasion by the scattered seed of American service men, which seemed to haunt at least Justice Stevens,(157) perhaps we as a nation should welcome and take responsibility for them.(158) If Congress remains disinclined to be so generous with citizenship, solving the problems raised by Miller will not be easy.

Squarely facing the question of remedy in Miller might cause the Court and legislators who respond to it to realize how poorly each has thought through the issues presented by earlier cases concerning men and women as mothers and fathers, in particular of illegitimate children. Commentators who have tried to make sense of the Court's pronouncements in this area and the legislative responses thereto admit defeat.(159) It would be beyond the scope of this Article for me to make another such attempt. A few observations should be made, however. First, of the two main branches of the Court's illegitimacy cases, one involving sex discrimination in the treatment of mothers and fathers of illegitimates, the other involving distinctions made on the basis of illegitimacy, the issue in Miller comes closer to those in the latter branch. Citizenship bears a certain resemblance to an inheritance: the question is not one of zero-sum competition between males and females as mothers and fathers, but of the ability of each to transmit benefits to and to gain them through their children. As in cases of inheritance and wrongful death, where, mutatis mutandis, both parents and children could claim benefits through one another, and rights claimed through a child could redound to the benefit of the other parent, so with citizenship. Just as the worker's compensation benefits in Weber v. Aetna Casualty,(160) the disability benefits in Jimenez v. Weinberger,(161) and the inheritance in Trimble v. Gordon(162)--claimed by illegitimate minor children through their deceased father--would have assisted their mother financially, so the U.S. citizenship a child claims through one parent may assist her other, alien parent, to obtain preferential U.S. immigration status through the child. Also, there is a similar willingness to allow some choice about transmission: for both citizenship and inheritance, shirking fathers can disqualify their offspring out of benefits by failing to formalize relationships by legitimation, will, or written agreement. The question is also generally more straightforwardly one of biological and legal relationship, not of attachment.(163) As Justice Harlan insisted in his dissent in Glona v. American Guarantee,(164) the statutory scheme, there of wrongful death, here of citizenship transmission, "generally defined classes of proper plaintiffs by highly arbitrary lines based on family relationships, excluding issues concerning the actual effect" of family members on each others' lives.(165) Justice Harlan added: "[I]t does not matter who loved or depended on whom, ... mak[ing] everything the Court says about affection and nurture and dependence altogether irrelevant. The only question in any case is whether the plaintiff falls within the class of persons to whom the State has accorded a right of action."(166)

It is unhelpful for the Miller dissenters to insist that the statute's problem can be solved by "distinguish[ing] between caretaker and non-caretaker parents, rather than between men and women."(167) What is called for is an easily administered bright line requirement, and again, the requirement need not be competitive between parents. What does Justice Breyer envision, that immigration officials take testimony about who changed a child's diapers?(168) At what point in time, if citizenship can be claimed at any point from birth through adulthood? And how often? Should divorced citizen parents no longer be able to transmit citizenship if they do not get custody or pay child support? Should citizenship transmitted to an infant through a caretaker parent be revocable in the event the bulk of the caretaking shifts to the alien parent? Can a parent who does not live day-to-day with the child ever qualify as a caretaker for citizenship-transmittal purposes? Can mere financial contribution, however substantial and reliable, qualify as caretaking? How about financial support combined with shared residence? Or combined with regular visits, letters, and phone calls? These questions are far too complicated to administer, and unnecessary for citizenship, because generosity to fathers does not hurt mothers and probably not the polity either.

A different set of questions is raised in the main context in which the Court earlier faced sex distinctions between nonmarital fathers and mothers--adoption and custody of children. If the rule in Miller is unconstitutional for the reasons suggested by the dissenters, then there is reason to think, for example, that most states have not adequately solved the constitutional problems with their adoption statutes, because, to me at least, the sex distinctions a majority of the Court found problematic in Miller look very close to a state's good faith response to the Stanley/Caban/Quilloin/Lehr line of cases, holding that the rights nonmarital fathers must receive are a function of the extent to which those fathers have seized their unique opportunity to invest in their children.(169)

One way of analyzing the remedy in Miller and other cases involving men and women as mothers and fathers is analogous to affirmative action cases. Assuming that, as Justice Ginsburg stated in VMI, both "[p]hysical differences between men and women ... [that] are enduring"(170) and affirmative action(171) can be acceptable reasons for sex-respecting rules, it does not seem to follow that this justifies any and every sex-respecting rule claimed to be premised on either of these bases. Just as, therefore, a claim of affirmative action was not enough to justify an all-female nursing school at Mississippi University, so physical difference may not be enough to justify the diverse requirements imposed on citizen mothers and fathers by the statutes in Miller. Note, I am not arguing that any preference for mothers is affirmative action, just that there are similar questions as to fit once the threshhold question of difference has been answered.

Another way of looking at the remedy phase in Miller is as a necessary application of the doctrine of separate but equal. If men and women are not quite similarly situated with respect to their children, at least at birth, how do you treat "unalikes" equally? An insistence that pregnancy is unique is no help here, since, that granted, the questions begin--some regime must be established for both mothers and fathers, and, unless pregnancy is to count for nothing (as it would in a scheme based simply on the fact of blood relationship), or to count for everything (with no father's rights at all), its uniqueness is question begging. What then should be set aside pregnancy in establishing requirements for rights flowing from fatherhood? Note that the normative framework looks very different if we are talking about fathers' rights, fathers' obligations (e.g., children's rights vis a vis their fathers), or third party rights flowing from the fact of fatherhood to both fathers and children. It also matters whether the statutory scheme is zero sum, as it most clearly is when fathers are in competition with mothers, for example, for custody or control over adoption, as in cases like Caban v. Mohammed.(172) At least as applied, most of the benefit scheme cases also have, a zero sum potential. After all, if the illegitimate child (or its father or mother) does not inherit or bring a wrongful death action, or collect workers' compensation, someone else likely will; even if no one does, the state gets the estate, and the tortfeasor or the employer or the fund keeps the money. Citizenship is really the only infinitely expandable benefit at issue in these cases.

B. Sex Discrimination in Education: VMI, The Citadel, and the Harlem Girls' School

Many of the most interesting open questions in the constitutional law of sex discrimination involve education. Consider three concrete questions. The first two involve the application of the VMI decision to the parties at whom it was directed, VMI and the Citadel, each of which has chosen a different path to implementing the integration of women. The Citadel, in line with the predictions of all parties and courts in the VMI litigation, made modifications for women in its grooming codes and physical fitness requirements, as well as some of its hazing rituals and restrictions on personal privacy.(173) All of these modifications are sex-respecting rules and hence subject to heightened scrutiny if challenged, for example, by a male cadet who wishes to grow longer hair, have more privacy, exhibit less upper body strength, and be treated more civilly. Indeed, the rules are also subject to challenge by a female cadet who--if reports about the treatment of two women who withdrew are accurate(174)--could claim that the modifications stigmatized her as a second class citizen-soldier and impeded her chances for success.

VMI, by contrast, modified none of its fitness standards for women. It modified its grooming standards for new cadets or "rats,"(175) only to the extent of adding a uniform skirt and giving the women a haircut slightly longer than their male classmates, but in line with those of upperclass men.(176) To accommodate personal privacy for female cadets, VMI made minimal changes in bathroom architecture.(177) Except for these changes in grooming and architecture,(178) VMI's treatment of cadets will not be sex respecting, and can be challenged as discriminatory only to the extent that it has a disparate impact on women and was adopted or maintained because of, and not merely in spite of, this disparate impact. Given that VMI spent years in litigation denying that women could be admitted without massive changes to the requirements imposed on cadets, and that it announced its decision on unchanged standards only after the start of an academic year for which, notwithstanding the Court's order, it had taken no visible steps to admit women,(179) there is some chance the requirements of unconstitutional disparate impact have been met. The litigation history might be evidence that VMI's motive is an exclusionary, not an egalitarian one.(180) Nevertheless, there is no blinking the fact that, of the two potential lawsuits, the one challenging the Citadel's effort to accommodate women has a far better chance of success. This is further evidence of the law's troubling blindness to discrimination caused by anything other than a sex-respecting rule.

As for the Harlem Girls' School, officially known as the Young Women's Leadership Academy, as I understand the facts, its exclusion of boys is de facto, not de jure. The school has announced its willingness to accept applications from boys, but no boy has yet applied, nor has any organization interested in filing a legal challenge to the school yet found a boy willing to serve as the named plaintiff. The school's case is therefore interesting to me less as a case of sex discrimination than of gender discrimination. The lack of interest on the part of boys is yet one more example of the phenomenon of the devaluing of women and all that is associated with them, i.e., as coded feminine.(181) Will the presence of even a single boy cause a tipping point in the Harlem Girls' School? I leave that to the educational theorists, but I predict that as a legal matter, so long as his exclusion is not de jure, explicitly sex-respecting rules may not be at issue, and courts will be correspondingly hesitant to interfere.(182)

C. Same Sex Marriage and the Defense of Marriage Act as a Sex-Respecting Rule

Bans on same sex marriage, and indeed most other discrimination against homosexuals, including criminal laws proscribing only homosexual sodomy, have all the formal structure of a discrimination on the basis of' sex in that, but for a gay person's sex, his or her treatment by the law would be different. This has been clear to gay rights litigators for decades,(183) but has only recently gained acceptance by a court: the Hawaii Supreme Court in Baehr v. Lewin(184) held it to be discrimination on the basis of sex of the sort that, under the Hawaii Constitution, could only be upheld if narrowly tailored to serve a compelling state interest, for Hawaii to deny same-sex couples entry into marriage.(185) Inevitably after Baehr, further challenges to sex-respecting rules of marriage, under the federal as well as state constitutions, will follow. The only purportedly compelling justifications offered on remand in Baehr were based on stereotypes and were, as such, rejected by the lower court,(186) Though that court did not use the language of stereotyping or imperfect proxy, but rather the more conventional strict scrutiny language of narrow tailoring to a compelling state interest,(187) the fact remains that the lower court relied on exceptions to the state's assumptions in rejecting them as an acceptable basis for the law.(188) The state insisted that denying marriage to gays and lesbians promoted the "compelling interest in protecting the health and welfare of children."(189) In rejecting any "causal link between allowing same-sex marriage and adverse effects upon the optimal development of children,"(190) the trial court relied on evidence from even Hawaii's own expert witnesses that there were gay and lesbian parents who provided good homes for children; heterosexual, married, biological parents who did not; as well as, more broadly, a diversity in functioning family structure.(191) In other words, being a male-female couple was an imperfect proxy for being a good parent.

More broadly, prohibitions on homosexuality rely on stereotypes in the sense that they are based on "fixed notions concerning the roles and abilities of men and women."(192) These notions have some subordinating taint, in that among their normative premises are that women should not be free of men and that men should not behave sexually as women do--receptively in anal intercourse or fellatio.(193) Thus, both for the social and the sexual role division in marriage, requiring one of each sex for a marriage is stereotypical.(194) Legally enforced role differentiation in marriage was ruled out of bounds in cases like Orr v. Orr(195) and Stanton v. Stanton.(196) Since it is no longer the case that in marriage he pays, she serves, why not take Sylvia Law's point about the anti-stereotyping impact of having two men or two women divide up these roles--either equally and symmetrically, or, at least for one of them in a same-sex couple, counter to the predictive and normative stereotype(197) associated with his or her sex?(198)

One way of making the law of marriage neither sex-respecting nor useable by same sex couples might be to make the ability to engage in vaginal intercourse with one's marital partner a condition for entry into marriage. Marriage can thus be seen as a license to engage in vaginal intercourse. For centuries this was the law of marriage in most of the Western world, with impotence in men and vaginal obstruction in women, if existing at the time of marriage, a ground for annullment. Capacity for vaginal intercourse in marriage law is like capacity for pregnancy in Geduldig, it is strongly, indeed almost definitionally, sex-linked,(199) but on tire other hand, not, as a formal matter, dividing the world into classes by sex. Ultimately, however, the difficulty with any such attempt to avoid a sex-discriminatory law of marriage, while still excluding same-sex couples, is that state laws no longer confine access to marriage to those capable of vaginal intercourse, as they do not confine it to fertile couples,(200) and it is quite doubtful that such a restriction would survive a substantive due process challenge by a man and woman.(201)

As the Hawaii Supreme Court in its Baehr v. Lewin opinion may well have realized, to strike down a ban on same-sex marriage on equal protection grounds as violative of norms against sex discrimination rather than on substantive due process grounds as violative of guarantees of associational privacy and family autonomy, is in many respects the more conservative, more easily limited decision. Not that a sex discrimination holding could not have quite radical implications, but unisex bathrooms and women on football teams are both less frightening, and more easily avoided, through distinction and manipulation of existing doctrine, than are polygamy or adult incest, the proximate entrants in the parade of horribles posed by a privacy/substantive due process holding. A ruling that the constitutional basis for same-sex marriages is a prohibition on sex discrimination gets Hawaii same-sex marriages, but puts very little else at risk.

A federal constitutional challenge to sex-respecting rules governing marriage is inevitable, and one likely source is the Defense of Marriage Act.(202) Much attention has been paid to the question of the constitutionality of the first portion of the Defense of Marriage Act, concerning the full faith and credit implications of same-sex marriage.(203) Substantially less attention has been paid to the second half of the Act, which provides that, "[i]n determining the meaning of any [federal statute, regulation, or administrative ruling], the word `marriage' means only a legal union between one man and one woman as husband and wife, and the word `spouse' refers only to a person of the opposite sex who is a husband or wife."(204) This is, of course, a sex-respecting rule, and hence subject to heightened scrutiny under Supreme Court precedent.(205) The fact pattern I see as presenting the strongest challenge to the Act involves a heterosexual couple disadvantaged by the federal definition of marriage. To see how they could be, it is important to remember that, under federal law, it is sometimes advantageous and often disadvantageous to be treated as married. Not only does the Internal Revenue Code impose a so-called "marriage penalty" on two-earner married couples, other federal statutes often terminate benefits to those who marry (for example the benefits due as the result of the death of a previous spouse). Thus, especially if they are residents of a state that permits same-sex marriage, a married couple who conforms to the federal definition may be disadvantaged by comparison with a married same-sex couple from their state. For the same-sex couple will be eligible for all the privileges of marriage under state law and none of its burdens under federal law. Thus, a remarrying widow who loses her federal pension when she acquires her new husband can claim that, but for her sex, or that of her partner, she would be better off.

Questing for proxies, perfect or otherwise, strikes me as quite difficult in this scenario. On the other hand, it also seems to me difficult to imagine what important governmental interest the sex-respecting rule in the Defense of Marriage Act can be closely related to. I wonder how many sex-respecting rules, other than that in the Defense of Marriage Act, are left in the United States Code. And I'm hard pressed to imagine what a reasoned opinion upholding the Act could possibly say.


I began this Article by asserting that, notwithstanding the fears of its critics and the hopes of its fans, the VMI case was more like the end of the line than it was some sort of new beginning for the constitutional law of sex discrimination. If this branch of law is indeed to have a new beginning, most of the plausible routes would involve what the VMI case does not--a departure from precedent. One such new route might be to apply an anti-subordination approach to the facts of cases like Lanier. Still another approach would be to revitalize constitutional disparate-impact doctrine, for example, by forcing a shift in support from the masculinist practices of VMI to those of VWIL. A third approach, advocated by Chief Justice Rehnquist in his VMI concurrence, but also by many committed feminists, might be to pay increased attention to the possibilities of equality in difference, something required by the parenting cases and imaginable in a wide variety of other contexts. What these routes have in common is a shift in emphasis from the problem of sex discrimination to the problem of sex equality. By contrast, the Hawaii Supreme Court's decision to take restrictions on same-sex marriage seriously, as a problem of sex discrimination, shows that a new beginning is also possible for a court prepared to follow the old anti-stereotyping route past its comfort zone to its logical conclusion.

(1) United States v. Virginia, 518 U.S. 515, 572-74 (1996) (Scalia, J., dissenting).

(2) Cass R. Sunstein, Foreword: Leaving Things Undecided, 110 HARV. L. REV. 4, 73 (1996).

(3) Id. at 23-26.

(4) Fortunately, not all interpreters of the decision have made Scalia's mistake. See Cohen v. Brown Univ., 101 F.3d 155, 179, 183 & n.22 (1st Cir. 1996) (correctly describing the current law as focused on the condemnation of "discrimination based upon 'archaic and overbroad generalizations' about women" and noting that VMI "adds nothing to the analysis of equal protection challenges to gender-based classifications that has not been part of that analysis since 1979" (citations omitted)).

(5) Sunstein, supra note 2, at 75.

(6) Id. at 78.

(7) See David A. Strauss, Common Law Constitutional Interpretation, 63 U. CHI. L. REv. 877, 879 (1996); infra text accompanying notes 134-46.

(8) As Sylvia Law put it in commentary on this paper at an N.Y.U. workshop, the "emperor" of equal protection law is generally seen to have three garments: red for stop, yellow for caution, and green for go--equivalent to strict, intermediate, and rational basis scrutiny. In questioning this conventional wisdom, I am not suggesting that the emperor has no clothes, instead I am suggesting that what may generally have been seen as a fluid yellow garment is instead more akin to a rigid suit of armor--it is a rule, not a standard.

(9) The single exception may be Heckler v. Mathews, 465 U.S. 728 (1984), discussed infra note 59. Additionally, Justice Brennan's majority opinion in Davis v. Passman, 442 U.S. 228 (1979), directed that, on remand, "inquiry ... be undertaken into what 'important governmental objectives,' if any, are served by the gender-based employment of congressional staff" but neither undertook ,such an inquiry nor "express[ed any] views as to [its] outcome." 442 U.S. at 235 n.9 (citations omitted).

(10) Frontiero v. Richardson, 411 U.S. 677, 685 (1973).

(11) Schlesinger v. Ballard, 419 U.S. 498, 507 (1975).

(12) Of course, in theory, a "No" answer does not end the inquiry. Instead, it triggers inquiry into constitutional disparate impact, with the next set of questions being the following: (a) Does the practice have a disparate impact on one sex?, and (b) If yes, was the practice adopted "because of, and not in spite of" this diparate impact? See Personnel Adm'r v. Feeney, 442 U.S. 256 (1979). Notoriously, however, as will be discussed in detail below, in practice the Court has seen only sex-respecting rules as denying equality on grounds of sex.

(13) I do not claim that this is a startlingly original discovery which has completely escaped the notice of the many able commentators on the constitutional dimensions of sex equality. But, especially given reactions like those above cited to VMI, I do think it worthwhile to set forth the parameters of this rule more explicitly and fully than has yet been done by either courts or commentators.

(14) This should have been the test for sex-respecting rules under the Equal Rights Amendment, according to an influential 1971 article on the meaning of the ERA. See Barbara A. Brown, Thomas I. Emerson, Gail Falk & Ann E. Freedman, The Equal Rights Amendment: A Constitutional Basis for Equal Rights for Women, 80 YALE L.J. 871,893 (1971).

(15) Thus, for example, in Craig v. Boren, 429 U.S. 190 (1976), though Justice Brennan tried to play games with the math in his :majority opinion, more than 90% of those teenagers arrested in Oklahoma for driving while intoxicated were male. See id. at 200 n.8. In Califano v. Goldfarb, 430 U.S. 199 (1977), the evidence indicated that approximately 90% of women and 3% of men in the relevant pool were dependent on their spouses. See id. at 238-39 n.7 (Rehnquist, J., dissenting). These figures are representative, not extraordinary. The exceptions involve proxies, such as those behind sex-based peremptory challenges to jurors, whose accuracy cannot readily be measured.

(16) Mississippi Univ. for Women v. Hogan, 458 U.S. 718, 725 (1982).

(17) Califano, 431) U.S. at 223 (Stevens, J., concurring).

(18) 404 U.S. 71 (1971).

(19) 411 U.S. 677 (1973).

(20) If it had, of course, the law would be importantly different, and much more expansive. For example, Ginsburg has long argued that laws concerning abortion and pregnancy should be seen as implicating sex equality, although the Court expressly rejected the latter view in Geduldig v. Aiello, 417 U.S. 484, 496 n.20 (1974), and has largely ignored it in abortion cases. In addition to briefing the losing side in Geduldig, Ginsburg argued the losing side in Kahn v. Shevin, ,[16 U.S. 351 (1974). She might well also be generally more sympathetic to claims of disparate impact by sex.

(21) JOHN DONNE, ELEGY XIX, To HIS MISTRESS GOING TO BED (1669), reprinted in JOHN DONNE'S POETRY 62 L26 (Arthur L. Clements ed., W.W. Norton 1992).

(22) 42 U.S.C. [subsections] 2000e-2000e17 (1994).

(23) Id.

(24) The outlines of Title VII stereotyping jurisprudence are set forth in Case, supra note [double dagger], at 3641.

(25) 83 U.S. 16 Wall. 130 (1872) (Bradley, J., concurring).

(26) See id. at 135 (Bradley, J., concurring).

(27) See id. at 14042.

(28) As I will discuss in detail below, the heyday of the Court's willingness to see constitutional problems with irrebuttable presumptions of all kinds extended from the mid 1960s to 1975. See discussion infra Part III.C By the time Justice Rehnquist's opinion for the majority in Weinberger v. Salfi, 422 U.S. 749 (1975), killed the move toward mandating individual determinations or perfect accuracy in all sorts of classification, the current constitutional law of sex discrimination, with its anti-stereotyping core, was already firmly in place. See infra note 109 and accompanying text.

(29) 323 U.S. 214 (1944).

(30) Id. at 216,

(31) See id. at 219.

(32) Indeed, the very fact that race is being used as a proxy, however imperfect, is what may save its use in the opinion of the Korematsu majority, because this suggests to them that the rule is not the result of "antagonism" or "racial prejudice." Id. at 223. In the bulk of the race cases, perhaps the most pernicious thing of all about the classification is that race itself is the characteristic at which the law aims. Of this, more later.

(33) Far from viewing this result as anomalous, many of the litigators and commentators who paved the way for today's constitutional law of sex discrimination have affirmatively endorsed it. For example, the authors of an influential Yale Law Journal article on the Equal Rights Amendment argued that strict scrutiny for sex classifications, even if available in court, would not be enough to guarantee equality: "[B]ecause this doctrine allows the government to justify even a suspect classification by 'compelling reasons,' it would permit some classifications based on sex to survive. Thus this standard too would not guarantee an effective system of equality which, as we shall argue, demands the elimination of all such classifications." Brown et al., supra note 14, at 880-81 (footnote omitted).

(34) Korematsu, 323 U.S. at 219 (discussing Hirabayashi v. United States, 320 U.S. 81 (1943), which upheld a similar order imposing a curfew on those of Japanese ancestry).

(35) Note, however, that one form of massive resistance to Brown v. Board of Education, 347 U.S. 483 (1954), was the attempt to reimpose segregated schools by alleging race as a proxy for intellectual ability. See Singleton v. Jackson Mun. Separate Sch. Dist., 355 F.2d 865, 866 (5th Cir. 1966) (insisting on nonracial assignment of children notwithstanding "voluminous testimony [offered] to show that allegedly innate racial differences furnish a reasonable basis for classifying school children according to race and therefore justify continued segregation ... based on [inter alia] the disparity in intelligence and achievement between Negro and white pupils"). As, later, with sex, the attempt to use imperfect proxies to bolster race prejudice was found unconstitutional. In arguments identical to those made by opponents of sex discrimination, Thurgood Marshall insisted to the Supreme Court during the Brown remedy argument:
      They give tests to grade children so what do we think is the solution?
   Simple. Put the dumb colored children in with the dumb white children, and
   put the smart colored children with the smart white children .... [T]here
   are geniuses in both groups and there are lower ones in both groups, and it
   has no beating. No right of an individual can be conditioned as to any
   average of other people in his racial group or any other group.

MARK V. TUSHNET, MAKING CIVIL RIGHTS LAW: THURGOOD MARSHALL AND THE SUPREME COURT, 1936-1961 925 (1994) (quoting Thurgood Marshall's oral argument before the Supreme Court for the remedy portion of Brown).

(36) 388 U.S. 1 (1967).

(37) Cf. Romer v. Evans, 517 U.S. 620, 635 (1996) (holding Colorado's Amendment Two unconstitutional because "it is a classification of persons undertaken for its own sake, something the Equal Protection Clause does not permit").

(38) Note that even the partial dissent in Regents of the University of California v. Bakke, 438 U.S. 265 (1978), specifically urged the adoption of "considerations ... developed in gender-discrimination cases" for allegedly benign race-based classifications. Id. at 359430 (Brennan, J., dissenting in part).

(39) See Adarand Constr., Inc. v. Pena, 515 U.S. 200, 222-27 (1995). The Adarand decision results in what some, see, e.g., id. at 242 (Stevens,J., dissenting), consider an anomaly--if racial affirmative action is judged under strict scrutiny and sex-based affirmative action under intermediate scrutiny, sex-based programs may be upheld when identical race-based programs will be struck down. It is my guess that this potential anomaly, together with the Court's simultaneous assertion that "strict in theory" should no longer be taken to mean "fatal in fact," see id. at 237, was one of the factors that led the United States in VMI to switch, at the last minute, to a claim that strict scrutiny should govern sex classifications. Another factor ma), well be the presence on the Court of Justice Ginsburg, who has supported strict scrutiny for sex since her brief in Reed. See supra notes 18-20 and accompanying text. Of course, not all would see this situation as anomalous or troubling. Counterposed to the view that the Fourteenth Amendment should be interpreted above all to help blacks, including by preferential programs, is the view that the Amendment should be interpreted to see all racial classifications, but not necessarily any other form of classification, as inherently pernicious.

As I shall explain below, to date the Court's application of the perfect proxy rule is consistent with easy approval of many sex-based affirmative action programs, if, as the Court seems to have done, one views all women, even those not demonstrably materially affected by it, as subject to ambient discrimination on the basis of their sex. This leads to the paradox that the quest for perfect proxies, while animating both race and sex based affirmative action analysis, has led to diametrically opposed results in the two areas, because in the sex cases, but emphatically not in the race cases, the Court has been willing to accept that all members of the disadvantaged group were subject to discrimination. What may help account for the persistence of this paradox is that, because the sex-based affirmative action blessed by the Court is closely associated with de jury exclusion of women from job opportunities (for example, federal regulations excluding women from combat positions in the military), the perfection of the proxy (i.e., the absence of an exception to the exclusion) may be more easily shown for sex than it can be in the case of race, where, at least since the fall of Jim Crow, exclusion based on race is not de jury and thus not as demonstrably categorical.

(40) See, e.g., Bush v. Vera, 517 U.S. 9!52, 967439 (1996) (plurality opinion) (condemning as "unjustified racial stereotyping" by government actors the use of race as "a proxy for political characteristics," notwithstanding evidence that 97% of black voters in the district were Democratic voters).

(41) 163 U.S. 537 (1896), overruled by, Brown v. Board of Educ., 347 U.S. 483 (1954).

(42) The statute at issue in Loving did not involve a perfectly parallel treatment of the races. Instead, the statute only prohibited "whites from marrying any nonwhite (subject to the exception for the descendants of Pocahontas), [while] Negroes, Orientals, and any other racial class may intermarry without statutory interference." 388 U.S. 1, 11 n.11 (1967). This made white supremacy, rather than racial purity more generally, the obvious purpose of the statute.

(43) Cf Richard H. McAdams, Cooperation and Conflict: The Economics of Group Status Production and Race Discrimination, 108 FLay. L. REv. 1003, 1008 (1995) (presenting the contention that "because of concern for status, cooperation arises within groups and conflict occurs between groups").

(44) See generally Dianne Avery, Institutional Myths, Historical Narratives and Social Science Evidence: Reading the "Record" in the Virginia Military Institute Case, 5 S. CAL. REv. L. & WoMEN'S STUD. 189 (1996) (exploring the history of the Virginia Military Institute and the litigation concerning it).

(45) See Oral Argument at 50-52, United States v. Virginia, 518 U.S. 515 (1996) (No. 941941). Although the transcript does not identify the speaker, I was personally present at the oral argument and observed Justice Breyer make this comment.

(46) What at first glance may seem a stronger argument--that the exclusion of women is a proxy for the exclusion of sexual tension between cadets, is strongly undercut by an amazing and little-known fact about VMI, one that, to the best of my knowledge, never made it into the record of the case: according to its officials, VMI does not discriminate on the basis of sexual orientation. VMI's officials repeatedly assured me over a period of years that they not only did not ask, they would not care if applicants or cadets told, about homosexual inclinations. (They added that homosexual conduct in the barracks was prohibited and that there were, as far as they knew, no homosexuals at VMI.) Given that VMI's announced purpose is to train "citizen-soldiers" and that the U.S. military still does discriminate against homosexuals, VMI's policy may seem particularly bizarre. But to me, it is conclusive evidence that what VMI fears is not sexual tension, but women, pure and simple. In remarking how sad he was to see VMI change, the Superintendent of VMI recently spoke of having the strains of Gustav Mahler's score to the movie adaptation of Thomas Mann's Death in Venice in the back of his mind as he looked out over the campus in its final all-male days. This anecdote adds a peculiar flavor to the homoerotic overtones of VMI, given that, of course, Death in Venice is the story of an aging repressed homosexual in a dying city lusting after beautiful young men. See Jeffrey Rosen, Like Race, Like Gender?, NEW REPUBLIC, Feb. 19, 1996, at 21, 24.

(47) The Virginia Women in Leadership Program ("VWIL") at Mary Baldwin College, a private, all-female college near VMI, was VMI's response to the Fourth Circuit's order that it either admit women, go private, or establish an alternative program for women. See United States v. Virginia, 976 F.2d 890, 900 (4th Cir. 1992), vacating, 766 F. Supp. 1407 (W.D. Va. 1991).

(48) United States v. Virginia, 518 U.S. 515, 517 (1996).

(49) Most obviously, both the law at issue in Michael M. v. Superior Court, 450 U.S. 464 (1981), and the court opinions upholding it, discussed below, rest on stereotypes, as dissenting justices and commentators have made clear. See, e.g., id. at 489 n.2 (Brennan, J., dissenting); Wendy W. Williams, The Equality Crisis: Some Reflections on Culture, Courts, and Feminism, 7 WOMEN'S RTS. L. REP. 175, 186 (1982).

(50) All this does not answer several questions. For example, even if a perfect proxy is necessary, as I hope to demonstrate it has been and will be, is it always sufficient? No case to date has tested this question.

(51) VMI, 518 U.S. at 533. This group includes not only Michael M., but also Pa(ham v. Hughes, 441 U.S. 347 (1979). In Parham, the sex-respecting rule distinguished between the rights of fathers and mothers of illegitimate children not only on the basis of reproductive biology, but also, as with the second group of cases, on the basis of other statutes not before the Court, such as those providing fathers of illegitimate children the opportunity to seek orders of filiation. See id. Wendy Williams has criticized this as the "`one discrimination justifies another' approach." Williams, supra note 49, at 182 n.50.

(52) VMI, 518 U.S. at 533 (quoting Califano v. Webster, 430 U.S. 313, 320 (1977)). The quotations in the text are taken from that portion of the VMI opinion in which Justice Ginsburg acknowledged that the "heightened review standard our precedent establishes does not make sex a proscribed classification" and set forth the permissible uses of sex as a classification. Id.

(53) 450 U.S. 464 (1981).

(54) Id. at 467 (quoting Michael M. v. Superior Court, 601 P.2d 572, 574 (Cal. 1980), aff'd, 450 U.S. 464 (1981)) (internal quotation marks omitted).

(55) Id. (quoting Michael M., 601 P.2d at 574).

(56) Id.

(57) Justice Stevens, the Justice most alive to the problem of stereotypes that are "archaic" and not merely overbroad, described as an unsupported assumption perhaps reflecting "nothing more than an irrational prejudice," the assumption that "the decision to engage in risk creating conduct is always--or at least typically--a male decision." Id. at 501 (Stevens, J., dissenting). Cf. Territory v. Armstrong, 28 Haw. 88, 95 (1924) (upholding against equal protection challenge higher penalties for male than for female adulterers on grounds, inter alia, that women already run the risk of pregnancy and social ostracism, that men are more often "the agressor, [or] `hunter,'" and that family members should be encouraged to bring charges without fear that women would bear the brunt of punishment).

(58) See Schlesinger v. Ballard, 419 U.S. 498 (1975).

(59) See Rostker v. Goldberg, 453 U.S. 57 (1981). Heckler v. Mathews, 465 U.S. 728 (1984), also involved reference over to another sex-respecting rule, but, unlike the combat exclusions at issue in Ballard and Rostker, the sex-respecting rule behind the one at issue in Heckler not only had been before the Court, but had been held unconstitutional. The law at issue in Heckler makes no mention of sex on its face, it merely incorporates by reference another subsection of the Social Security law "as it was in effect and being administered in January 1977." Id. at 733 (internal quotation marks omitted). Precisely that subsection, requiring proof of dependency from husbands but not from wives, was struck down as discriminatory in Califano v. Goldfarb, 430 U.S. 199 (1977), early in 1977. See id. To protect those persons, most of them women, who had relied on the law struck down in Goldfarb in planning their retirement, Congress enacted a five-year-transition-period exemption from a substitute, sex-neutral rule that would require no proof of dependency, but would require an offset of spousal pensions. See Heckler, 465 U.S. at 732-34. Congress further provided that the transition rule could not be extended to persons, like Mathews, whose retirement took place outside the protected period; instead, if the rule were successfully challenged, those within the window would lose their exemption. See id. Although I am prepared to consider Heckler as that rare case in which, not the perfect proxy rule, but instead conventional intermediate scrutiny, is doing the work, I do note that the opinion by Justice Brennan described the rule as
   distinguish [ing] social security applicants, not according to archaic
   generalizations about the roles and abilities of men and women, but rather
   according to whether they planned their retirement with the expectation,
   created by the law in effect in January 1977, that they would receive both
   full spousal benefits and a government pension.

Id. at 730.

(60) "[O]ver 98% of the veterans in Massachusetts were male." Personnel Adm'r v. Feeney, 442 U.S. 256, 270 (1979). No women (except for those passing as men) were given full military rank until the early twentieth century, and, until 1967, a statutory quota kept women at no more than 2% of enlisted strength. See id. at 269 n.21. Women were less than 5% of the military at the time of the Feeney litigation. See id.

(61) Id. at 269-70.

(62) See id. So-called "women's requisitions" (of the lower echelon, clerical variety) were initially excluded by law from the veterans preference and, by the time of Feeney, were the sort of "lower paying positions for which males traditionally had not applied." Id. at 270 & n.22.

(63) 430 U.S. 313 (1977) (allowing women wage earners to exclude three more low earning years than men in the computation of the base from which Social Security benefits were calculated).

(64) 416 U.S. 351 (1974) (upholding the granting to widows but not widowers of an annual $500 property-tax exemption).

(65) Compare the 1991 Amendments to Title VII, Civil Rights Act of 1991, Pub. L. No. 102-166, 105 Stat. 1071, which for the first time gave a cause of action to victims of discrimination even if no employment outcome in their case was altered by the discrimination. These amendments may provide some acknowledgment that the burden of having to be twice as good is a difficult and unjust one, even to those who can meet it.

(66) Most commentators see Kahn as an outlier, an aberration, an early mistake made on the Court's way to shaping current sex discrimination doctrine. It is also one of Ginsburg's few losses before the Court. In her brief, quoting Sarah Grimke, Ginsburg insisted

that she "asked no favors for her sex." See Mary Anne Case, Of Richard Epstein and Other Radical Feminists, 18 HARV. J.L & PUB. POL'Y 369, 407 (1995).

(67) United States v. Virginia, 518 U.S. 515, 533 (1996).

(68) One limit, of course, is that imposed by the Court in Mississippi University for Women v. Hogan, 458 U.S. 718 (1982), to wit that the compensatory or affirmative action scheme actually have the announced intent and effect. See id. In Hogan, the Court relied on the longstanding numerical prominence of women in nursing and evidence that to keep nursing a predominantly female profession would depress nurses' wages, to reject the notion that a de jure all-female nursing school could be justified by a compensatory purpose. See id. at 729-30. Moreover, for some sex-based affirmative action cases, a stereotype that is archaic alone, even if not overbroad, might suffice to doom a rule. Ginsburg's answers to questions about Kahn suggested that what was constitutionally objectionable about the tax break at issue was that it lumped widows with the blind and infirm. Similarly, Justice O'Connor's opinion in Hogan rejects an affirmative action justification for limiting training in a traditionally female field to women. See Hogan, 458 U.S. at 730.

(69) Strauss, supra note 7, at 879. I give the attention I do in this discussion to Strauss's description of the common law method in constitutional law in part as the preemptive response to what might be characterized as epistemological questions about my methods of analysis.

(70) Id. at 877.

(71) Id.

(72) Id. at 909.

(73) Id. As discussed elsewhere in this paper, Justice Rehnquist was slow to conclude that sex-respecting rules are constitutionally problematic. He dissented in virtually every case striking down a sex-respecting rule. The rare exceptions include Kirchberg v. Feenstra, 450 U.S. 455 (1981), in which he joins Justice Stewart's two paragraph concurrence to the effect that "[s]ince men and women were similarly situated for all relevant purposes with respect to the management and disposition of community property, [the statute at issue], which allowed husbands but not wives to execute mortgages on jointly owned real property without spousal consent, violated the Equal Protection Clause." Id. at 463. But, like the good common law lawyer he is, Justice Rehnquist deployed the very principle from whose use against sex-respecting rules he dissented to write for the majority upholding such a rule in Michael M.

(74) 404 U.S. 71 (1971).

(75) United States v. Virginia, 518 U.S. 515, 561 (1996) (Rehnquist, J., concurring).

(76) Id.

(77) By 1968, the Equal Employment Opportunity Commission (EEOC) set forth a narrow, anti-stereotyping construction of the BFOQ requirement. According to the EEOC:
   [T]he following situations do not warrant the application of the bona fide
   occupational qualification exception ... (ii) The refusal to hire an
   individual based on stereotyped characterizations of the sexes. Such
   stereotypes include, for example, that men are less capable of assembling
   intricate equipment: that women are less capable of aggressive
   salesmanship. The principle of nondiscrimination requires that individuals
   be considered on the basis of individual capacities and not on the basis of
   any characteristics generally attributed to the group.

29 C.F.R. [sections] 1604.2 (1998) (EEOC guidelines on discrimination because of sex).

(78) Not all early lower court cases vindicating claims of unconstitutional sex discrimination used anti-stereotyping reasoning, or indeed any reasoning at all. See, e.g., Owen v. Illinois Baking, 260 F. Supp. 820, 821 (W.D. Mich. 1966) (holding that "to grant a husband the right to sue [for loss of consortium] while denying the wife access to the courts in the assertion of the same right is too clearly a violation of Fourteenth Amendment equal protection guarantees to require citation of authority").

(79) The bulk of such laws concerned sex respecting conditions of employment, such as limitations on women's lifting more than a certain specified (low) weight on the job or on women's working overtime or nights. After an initial period of indecision, the EEOC condemned such laws as incompatible with Title VII.

(80) 485 P.2d 529 (Cal. 1971).

(81) Id. at 539.

(82) 400 U.S. 542 (1971).

(83) Civil Rights Act of 1964, Pub. L. No. 88-352, [sections] 703, 78 Stat. 241,255 (codified at 42 U.S.C. [sections] 2000e-2).
   [T]he statute forbids an employer to `fail or refuse to hire or to
   discharge any individual, or otherwise to discriminate with respect to his
   compensation, terms, conditions, or privileges of employment,' or to
   `limit, segregate, or classify his employees or applicants for employment
   in any way which would deprive or tend to deprive any individual of
   employment opportunities or otherwise adversely affect: his status as an
   employee, because of such individual's ... sex.' ... We take these words to
   mean that gender [i.e., sex] must be irrelevant to employment decisions.

Price Waterhouse v. Hopkins, 490 U.S. 228, 239-40 (1989) (citations omitted) (case superseded by statute but not for the proposition cited).

(84) S.J. Res., 92d Cong. (1971). This is a different question than the one, pressed by Mary Becker, of how, if at all, the law might be different if feminists in the early 1970s had concentrated their quest for equality on the legislatures instead of the courts. See Mary Becker, The Sixties Stiff to Formal Equality and the Courts: An Argument for Pragmatism and Politics, 40 WM. & MARY L. REV. 209, 211 (1998).

(85) Machiaro v. Chaney, 582 P.2d 487, 489 (1978).

(86) Id. at 491 (internal quotations omitted). The court insisted:
   Neither sex may predominate. Neither may discriminate or be discriminated
   against. There is an equality of numbers and an equality of rights to be in
   office and to control the affairs of the state committee. The ironic result
   of the plaintiffs' theory would be to abolish a statute which mandates
   equality by invoking a provision of the constitution passed to guarantee

Id. at 492. It was left to the dissent to articulate principles more closely associated with the anti-stereotyping lines on which the federal constitutional law developed, to the effect that "the majority opinion prevents the Equal Rights Amendment from achieving its purpose of making sex a neutral factor, one to be disregarded in favor of ability and performance." Id. at 497 (Horowitz, J., dissenting).

(87) The phrasing is, of course, not identical, but it is difficult to read into the change in wording a move away from an anti-discrimination toward an equality norm.

(88) See Brown et al., supra note 14.

(89) Cf. RUDOLF A. RAFF, THE SHAPE OF LIFE: GENES, DEVELOPMENT, AND THE EVOLUTION OF ANIMAL FORM 63-102 (1996) (surveying paleontological data to understand the development of animals found on Earth).

(90) As a technical matter, the Court in Reed did not discuss the overbroad generalizations on which the preference for males might have been based, such as what the lower court identified as their greater likelihood of having business experience. Instead the Court treated the preference for males as arbitrary. The rule struck down in Reed is like a coin flip in which the coin comes up heads disproportionately. Cf. TOM STOPPARD, ROSEN-CRANTZ AND GUILDENSTERN ARE DEAn act. 1, sc. 1 (Henry Popkin ed., 1967). A fair coin flip is bad enough--arbitrary, but random. Sex-respecting rules work like flipping a weighted coin--men are disproportionately likely to win and women to lose. This is why courts will not allow stereotypical generalizations, even highly accurate ones (i.e., good predictors) to justify race or sex-respecting rules--because those on the losing side of the generalization or the coin flip almost always lose. Compare alphabetical order by last name as a basis for the state to distribute things that really matter. This system would be troubling whether the first in line gets punished or rewarded.

(91) 414 U.S. 632 (1974).

(92) This line of cases begins with two wrongful death suits involving illegitimate children and their mothers: Levy v. Louisiana, 391 U.S. 68 (1968), and Glona v. American Guarantee, 391 U.S. 73 (1968). Justice Doaglas for the majority insisted that illegitimate children are no less close to their mother, nor she less solicitous of them, than legitimate children. See Glona, 391 U.S. at 75. Justice Harlan in his dissent insisted that the fact of such closeness is irrelevant to the statutory scheme which, both before and after the Court's expansive intervention, "generally defined classes of proper plaintiffs by highly arbitrary lines based on family relationships, excluding issues concerning the actual effect of the death on the plaintiff." Id. at 77 (Harlan, J., dissenting). (93) 405 U.S. 645 (1971).

(94) There is a direct line from Stanley to the most recent Supreme Court case examining the constitutionality of a sex-respecting rule, also a rule premised on a father's lack of connection to his illegitimate children, Miller v. Albright, 523 U.S. 420 (1998), discussed below.

(95) See Stanley, 405 U.S. at 653 n.5 (noting how Illinois's Brief cited "physiological and other studies ... [for] the proposition that men are not naturally inclined to childrearing").

(96) Id. at 656.

(97) The Court also cited irrebuttable presumption cases like Carrington v. Rash, 380 U.S. 89 (1965). See Stanley, 405 U.S. at 655-56.

(98) Stanley, 405 U.S. at 656.

(99) See Carrington, 380 U.S. at 96.

(100) See Kramer v. Union Free Sch. Dist., 395 U.S. 621,632-33 (1969).

(101) See Aptheker v. Secretary of State, 378 U.S. 500, 511-17 (1964).

(102) See Elfbrandt v. Russell, 384 U.S. 11, 16-19 (1966).

(103) See Dunn v. Blumstein, 405 U.S. 330, 359430 (1972).

(104) See Shapiro v. Thompson, 394 U.S. 618, 63842 (1969), overruled in part by, Edelman v. Jordan, 415 U.S. 651 (1974).

(105) See Vlandis v. Kline, 412 U.S. 441,453-54 (1973).

(106) See, e.g., Gooding v. Wilson, 405 U.S. 518, 527-28 (1972).

(107) See Woodson v. North Carolina, 428 U.S. 280, 3044)5 (1976).

(108) See, e.g., Goldberg v. Kelly, 397 U.S. 254, 268-70 (1970).

(109) 422 U.S. 749 (1975) (upholding prohibition on widow's receipt of Social Security survivor's benefits when her relationship with the wage earner had not existed at least nine months prior to his death). Justice Rehnquist had long been of the view, hitherto expressed in dissent, that what his brethren considered impermissible, irrebuttable presumptions were what, in the days since the rejection of the Chancellor's foot as a measure, was prized as law--general rules announced in advance and uniformly applied.

(110) 335 U.S. 464 (1948) (upholding prohibition on women working in bar not owned by her husband or father).

(111) Amicus Brief for the American Civil Liberties Union at 34, Frontiero v. Richardson, 411 U.S. 677 (1973), cited in Sylvia A. Law, Rethinking Sex and the Constitution, 132 U. PA. L. REV. 955, 979 n.88 (1984).

(112) 83 U.S. (16 Wall.) 130 (1872) (Bradley, J., concurring). Bradwell joins Dred Scott v. Sandford, 60 U.S. 393 (1857), and Lochner v. New York, 198 U.S. 45 (1905), in a small group of cases the modern Court uses as the antithesis of a precedent: It is, for the modern Court, an argument against holding a certain way that to do so would be in conformity with the reasoning of the majority in one of these anti-precedents.

(113) See Bradwell, 83 U.S. (16 Wall.) at 139.

(114) The Slaughter-House Cases, 83 U.S. (16 Wall.) 36, 122 (1872) (Bradley, J., dissenting).

(115) See Bradwell, 83 U.S. (16 Wall.) at 139-42 (Bradley, J., concurring).

(116) The passages are found in Bradwell, id. at 141-42 (Bradley, J., concurring).

(117) Frontiero v. Richardson, 411 U.S. 677, 685 (1973).

(118) Bradwell, 83 U.S. at 141. At least one of the stereotypes Justice Bradley relied on is expressly prohibited by the EEOC guidelines cited above. See supra note 77 and accompanying text. Thus, while the EEOC deems it impermissible stereotyping to assume that "women are less capable of aggressive salesmanship," 29 C.F.R. [sections] 1604.2(a)(1)(ii) (1990), Justice Bradley used the alleged need for lawyers to exhibit "that decision and firmness which are presumed to predominate in the sterner sex," Bradwell, 83 U.S. (16 Wall.) at 142 (Bradley, J., concurring), in essence to establish a BFOQ for males as lawyers. Note that nothing in the EEOC's formulation addresses stereotyping of the job, rather than the applicant. Thus, while it constitutes sex discrimination to assume that "women are less capable of aggressive salesmanship," the EEOC does no better than Justice Bradley with the stereotype that success on the job requires qualities such as aggressiveness, "presumed to predominate in the sterner sex." The regulations merely give individual women the opportunity to prove that they can be as aggressive (as masculine?) as is assumed to be required. Judging by the oral argument of the VMI case, we seem, unfortunately, not to have advanced far beyond Justice Bradley in our assumptions that good lawyering requires qualifies gendered masculine. Arguing for the United States, Paul Bender analogized VMI to a traditional Socratic method law school with large classes and issue spotting exams, and VWIL to an all female law school with a nurturing teaching style and seminar papers, on the assumption, apparently shared by a majority of the Court, that only the former would adequately prepare students. See Oral Argument at 22-24, United States v. Virginia, 518 U.S. 515 (1996) (No. 94-1941). For further discussion, see Mary Anne Case, Two Cheers for Cheerleading: The Noisy Integration of VMI and the Quiet Success of Virginia Women in Leadership, 1999 U. CHI. LEGAL F. 347. By contrast to both Justice Bradley and Bender, Matthew Hale Carpenter, Bradwell's lawyer, while he accepted the stereotype of women as gentler than men, argued to the court that traits gendered as feminine can have advantages over those gendered as masculine in the practice of law. According to Carpenter, "It]here may be cases in which a client's rights can only be rescued by an exercise of the rough qualities possessed by men. There are many causes in which the silver voice of woman would accomplish more than the severity and sternness of man could achieve." Bradwell, 83 U.S. (16 Wall.) at 137.

(119) 208 U.S. 412 (1908).

(120) See id. at 421-22.

(121) Id. at 422. The Brandeis brief, relied on by the court in Muller, quotes a certain "Mr. B.--, a foreman of a large printing establishment" to the effect that he "never knew but one woman, and she a strong, vigorous Irishwoman, of unusual height, who could stand at the case like a man." Brief for Defendant in Error at 38, Muller v. Oregon, 208 U.S. 412 (1908) (No. 107).

(122) Bradwell v. Illinois, 83 U.S. (16 Wall.) 130, 142 (1872) (Bradley, J., concurring).

(123) These cases include, for example, McCabe v. Atchison, Topeka & Santa Fe Railway Co., 235 U.S. 151 (1914), which rejected the argument that limited demand by blacks justified providing sleeping cars only for whites, on the grounds that it made the right
   depend upon the number of persons who may be discriminated against, whereas
   the essence of the constitutional fight is that it is a personal one.... It
   is the individual who is entitled to the equal protection of the laws, and
   if he is denied ... a facility or convenience ... which under substantially
   the same circumstances is furnished to another.... he may properly complain
   that his constitutional privilege has been invaded.

Id. at 161-62; see also Missouri ex rel. Gaines v. Canada, 305 U.S. 337, 350 (1938) ("We find it impossible to conclude that what otherwise would be an unconstitutional discrimination ... can be justified by requiring resort to opportunities elsewhere."). The implication of these opinions was clear to the lower court judge deciding a companion case to VMI, Shannon Faulkner's petition to be admitted to the Citadel. Faulkner v. Jones, 858 F. Supp. 552 (D.S.C. 1994), aff'd, United States v. Jones, 136 F.3d 342 (4th Cir. 1998). Responding to the argument that too few women were interested in a Citadel-style education to make it worth the state's while to provide them with one, Judge Houck said, "[t]o suggest that a lack of demand for a certain type of equal protection can somehow justify the denial of another person's constitutional right thereto undermines the express intent of the Fourteenth Amendment." 858 F. Supp. at 564. Justice Ginsburg had the same principle in mind when, in her Miller dissent, she quoted Burnita Sheldon Matthews, the first female federal district judge, to the effect that "[w]hether there are a lot of people who suffer or whether there are a few who suffer, it seems to us that the principle of equal application of the law to men and women ought to receive recognition." Miller v. Albright, 523 U.S. 420, 471 (1998) (Ginsburg, J., dissenting) (quoting testimony of Burnita Shelton Matthews, Hearings on H. R. 3673 and H. R. 77 Before the House Comm. on Immigration and Naturalization, 73d Cong. 36 (1933)) (internal quotation marks omitted).

(124) Moreover, Justice Scalia's own focus on the individual in race cases is in substantial tension with his and other conservatives' willingness to focus on the group in sex cases. Thus, for example, other than the substitution of sex for race, the position on the relationship between remedy and standing Justice Scalia articulates in his J.E.B. dissent, J.E.B. v. Alabama, 511 U.S. 127, 156-63 (1994) (Scalia, J., dissenting), is identical to that of Justice Stevens set forth in Shaw v. Reno, 509 U.S. 630, 676-79 (1993); a tension Justice Scalia nowhere bothers to resolve or even acknowledge. This inconsistency is common among conservatives: Ted Olson, admittedly a hired gun, but one who frequently chooses his clients for the ideological appeal of their position, represented both Virginia in the Supreme Court argument of the VMI case and Cheryl Hopwood in her litigation successfully challenging the University of Texas's affirmative action policies. See Hopwood v. Texas, 78 F.3d 932 (5th Cir. 1996). This put him squarely on both sides of the anti-stereotyping question. For women, Olson argued to the Supreme Court in VMI, individual merit was or should legally be irrelevant--group averages or tendencies could and should shape the law and exceptions be damned. See United States v. Virginia, 518 U.S. 515, 517-18 (1996). But, in Hopwood, he successfully insisted on behalf of plaintiffs that all applicants to the University of Texas Law School be evaluated as individuals and not lumped with their racial group. See Hopwood, 78 F.3d at 944.

(125) Ruth Colker, Anti-subordination Above All: Sex, Race, and Equal Protection, 61 N.Y.U. L. REV. 1003, 1003 (1986). Especially to the great extent to which these principles do not conflict, both are important in constitutional sex-equality jurisprudence. See, e.g., Sylvia A. Law, Rethinking Sex and the Constitution, 132 U. PA. L. REV. 955, 1017 (1984) (proposing scrutiny of laws to ensure that the law "has no significant impact in perpetuating the oppression of women or culturally imposed sex role constraints on individual freedom").

(126) Obviously, I do not endorse the subordination of women, but unlike some feminist theorists, see, e.g., Christine A. Littleton, Does It Still Make Sense to Talk About "Women"?, 1 U.C.L.A. WOMEN'S L.J. 15, 33 n.84 (1991), I see the goal of feminism as seeking for men, as well as women, liberty as well as equality. See generally Case, supra note [double dagger], at 4.

(127) Mississippi Univ. for Women v. Hogan, 458 U.S. 718, 725 (1982).

(128) See Craig v. Boren, 429 U.S. 190, 219-220 (1976) (Rehnquist, J., dissenting). In Craig, Justice Rehnquist observed:
   There is no suggestion in the Court's opinion that males in this age group
   are in any way peculiarly disadvantaged, subject to systematic
   discriminatory treatment, or otherwise in need of special solicitude from
   the courts.... [T]here being no plausible argument that this is a
   discrimination against females, the Court's reliance on our previous sex
   discrimination cases is ill-founded. It treats gender-classification as a
   talisman which--without regard to the rights involved or the persons
   affected--calls into effect a heavier burden of judicial review.

Id. (Rehnquist, J., dissenting) (footnote omitted).

(129) For the difference between a welcome mat and a no trespassing sign, see Adarand Constructors v. Pena, 515 U.S. 200, 245 (Stevens, J., dissenting).

(130) See Andrew Koppelman, Why Discrimination Against Lesbians and Gay Men Is Sex Discrimination, 69 N.Y.U.L. REV. 197, 255-57 (1994). Cf Sylvia A. Law, Homosexuality and the Social Meaning of Gender, 1988 WIS. L. REV. 187, 188, 232 (arguing that legal condemnation of homosexuality "can best be understood as preserving traditional concepts of masculinity and femininity" and as a result injures "everyone who seeks freedom to experience the full range of human emotions, behavior and relationships without gender-defined constraints"). I take Law to be saying far more explicitly than Koppelman that "fixed notions concerning the roles and abilities of males and females" are problematic, quite apart from their subordinating effects. See Mississippi Univ. for Women v. Hogan, 458 U.S. 718, 725 (1982).

(131) Wendy W. Williams, Equality's Riddle: Pregnancy and the Equal Treatment/Special Treatment Debate, 13 N.Y.U. REV. L. & Soc. CHANGE 325, 352 (1985).

(132) Id. at 369. See generally Case, supra note [double dagger], passim (calling for attention to be paid to which sex neutral rules are chosen).

(133) Thus, the Court has not come close to adopting anything like Robin West's "pure protection" approach whose "goal is a community in which all are equally protected by the state against private encroachment of rights." Robin West, Equality Theory, Marital Rape, and the Promise of the Fourteenth Amendment, 42 FLA. L. REV. 45, 62 (1990); see also Case, supra note 66, at 401-05 (1995) (discussing West's "pure protection theory" and citing relevant sources). One of the many reasons for the Court's failure in this regard may be the extent to which it today ranks Muller v. Oregon, 208 U.S. 412 (1908), together with cases like Bradwell as negative precedents. In Muller, the Court upheld differential-hours laws for certain working women on the grounds, inter alia, that, because "there is that in [woman's] disposition and habits which will operate against a full assertion of those rights. She will still be where some legislation to protect her seems necessary to secure a real equality of right." Muller, 208 U.S. at 422.

(134) See, e.g., Geduldig v. Aiello, 417 U.S. 484 (1974).

(135) The true test of the Court's blindness to sex (not to be contused with sex-blindness) might have come with a final resolution in United States v. Lanier, 73 F.3d 1380 (6th Cir. 1996), vacated, 520 U.S. 259 (1997), the federal criminal case against a Tennessee state judge who, inter alia, forcibly raped female employees and litigants "while wearing his judicial robes" and implicitly or explicitly threatened them with retaliation in the form of dismissal or an adverse judgment if they resisted. Id. at 1400 (Keith, J., joining in the dissent). The Sixth Circuit en banc reversed Lanier's conviction for violation of a federal statute, criminalizing willful "`deprivation of any rights ... protected by the Constitution' committed by any person `under color of any law'" and ordered his release from the 25-year prison term to which he had been sentenced. 73 F.3d at 1382-84 (quoting 18 U.S.C. [sections] 242 (1994)). According to a majority of the Sixth Circuit, Lanier's conviction could not stand because no prior case had put Lanier on notice that sexual assault by a judge would constitute a violation of the broadly worded federal statute. See id. at 1384. As several indignant dissenters pointed out, the Sixth Circuit's decision worked a grave injustice. See id. at 1399 (Keith, J., joining in the dissent); id. at 1398 (Nelson, J., concurring in part and dissenting in part). The majority made much of the fact that the United States had charged Lanier only with unconstitutional deprivation of his victims' liberty, not with "a gender-based crime for sexual assault involving discrimination against or oppression of women in violation of the Equal Protection Clause." Id. at 1384. The real equal protection violation will occur, however, not when a judge rapes a woman, but when the law fails to take this violation as seriously as it takes, for example, the nonsexual assaults perpetrated on male prisoners by their guards. See id. at 1414 (Daughtrey, J., dissenting). Whether this violation will occur is not yet clear, given that the Supreme Court, while unanimously reversing the Sixth Circuit, left many crucial issues open on remand and Lanier, by fleeing the country, prevented consideration of those issues by the court below. See United States v. Lanier, 520 U.S. 259 (1997).

(136) I do not mean this observation as a cheap shot against the anti-subordination position, nor do I mean to suggest that other Justices are unconcerned with issues of subordination. See, e.g., Adarand Constructors v. Pena, 515 U.S. 200, 273-74 (1995) (Ginsburg, J., dissenting).

(137) United States w Virginia, 518 U.S. 515, 565 (1996) (Rehnquist, C.J., concurring).

(138) For further discussion, see Mary Anne Case, Unpacking Package Deals: Separate Spheres Are Not the Answer, 75 DENV. L. REV. 1305 (1998).

(139) Oral Argument at 18, United States v. Virginia, 518 U.S. 515 (1996) (No. 94-1941).

(140) Plessy v. Ferguson, 163 U.S. 537, 559 (1896) (Harlan, J., dissenting), overruled by, Brown v. Board of Educ., 347 U.S. 483 (1954). Of course, Brown overruled the majority in Plessy, not Justice Harlan's dissent.

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

(142) This is not all that farfetched a hypothetical, at least in its assumption that there can be equality in difference. Consider, for example, the estates of the clergy and nobility in medieval and early modern France. The relationship between these groups, the First and Second Estates, presents a somewhat different problem than the conventional one of subordination framed by the position of the Third Estate. Both clergy and nobility ran the gamut of wealth and power, from the impoverished country squire and village priest, to the prelates and princes of the royal line. And the two estates were distributed throughout the land. Although nominally the clergy was the premier estate, the nobility was hardly subordinate. Role differentiation, rather than inequality, marked the difference between the two.

(143) Ironically, strong support for the proposition I am here advancing comes from none other than Justice Bradley. See Slaughter-House Cases, 83 U.S. (16 Wall) 36, 111-13 (Bradley, J, dissenting) (arguing that choosing employment is one of a citizen's most valuable rights and using the example of a caste system to support his contention).

(144) 163 U.S. at 555 (Harlan, J., dissenting).

(145) See id.

(146) Id.

(147) Price Waterhouse v. Hopkins, 490 U.S. 228, 251 (1989).

(148) W.S. GILBERT, H.M.S. Pinafore, in THE COMPLETE PLAYS or GILBERT AND SULLIVAN 106 (Random House 1936) (1878).

(149) I take much, but not all, of the quarrel that other feminist theorists have with the opponents of sex-respecting rules to turn on a theory of the second best. If the single, uniform, non-sex respecting form of most rules could reliably be formulated favorably to women's interests or neutrally between men's and women's interests, there might be much less objection to its implementation. But, in a world of law and social practice built largely around male, masculine standards, merely striking down the sex-respecting character of the rule masks, rather than eliminates, the rule's pernicious effects.

(150) The project began with Disaggregating Gender. See Case, supra note [double dagger]. The next article in the series will be on toilets as gendered spaces. This future Article will examine one of the few sex-segregated spaces left in our culture and is intended to confront squarely the use of the prospect of mandatory single-sex public toilets as an argument-stopping reductio for the "sameness" approach. The first part of this future Article will use the all too concrete example of bathrooms to ask what and how we are equalizing in a world of separate but equal; the second will examine what we would lose if toilets ceased to be segregated spaces. After the "toilet paper," I plan papers on sports and prisons.

(151) 523 U.S. 420 (1998).

(152) See id. at 424-26.

(153) See id. at 426.

(154) Justice Stevens, speaking for himself and Chief Justice Rehnquist, opined that the plaintiff had standing, see id. at 433, but lost on the merits because, although "we may assume that if the classification ... were merely the product of an outmoded stereotype, it would be invalid, [t]he biological differences between single men and single women provide a relevant basis for differing rules." Id. at 443-45. Justice O'Connor, for herself and Justice Kennedy, opined that the plaintiff lacked standing to raise a claim of sex discrimination, from which she suffered only indirectly through her father who had failed to appeal the dismissal of his own claim by a lower court, but that, if challenged by a party with standing, the rule at issue would not "withstand[ ] heightened scrutiny." Id. at 450-51 (O'Connor, J., concurring). Justice Scalia, for himself and Justice Thomas, opined that, because only Congress and not the Court had the power to grant the requested remedy of citizenship, scrutiny of any sort of the rules in question was beside the point. See id. at 45253 (Scalia, J., concurring). Thus, of the six justices who ruled against Miller, only two, Justices Stevens and Rehnquist, needed to reach the merits of her claim to do so. In dissent, Justices Breyer and Ginsburg, each of whom wrote an opinion in which the other and Justice Souter joined, opined that the plaintiff had standing and the Court power to aid her, and that the sex-respecting rules that denied her citizenship relied unconstitutionally on stereotypes. See id. at 460 (Ginsburg, J., dissenting); id. at 471-72 (Breyer, J., dissenting).

(155) Only a subset of the sex-distinctions in requirements for the transmission of citizenship to nonmarital offspring were at issue in Miller, but a subsequent case may challenge others that also may be difficult to justify under heightened scrutiny. The differences include: (1) differential U.S. residency requirements for the citizen parent of a child born outside of the United States or its outlying possessions. Compare 8 U.S.C. [sections] 1409(c) (1994) (requiring unmarried mothers to reside in the United States or one of its outlying possessions for one year), and 8 U.S.C. [sections] 1401(g) (1994) (requiring both mothers and fathers married to noncitizens to reside in the United States or one of its outlying possessions for five years); (2) differential standards of proof of blood relationship, see 8 U.S.C. [sections] 1409(a) (1) (1994) (requiring proof by "clear and convincing evidence" for fathers only); and (3) requirements that apply only to fathers, see 8 U.S.C. [sections] 1409 (a) (1994) (requiring that the father agree "in writing to provide financial support for the [child] until the [child] reaches the age of 18 years," and that the paternity of the child is either legitimated, acknowledged "in writing under oath," or adjudciated by a competent court). Eliminating all discrimination between citizen mothers and fathers of foreign born nonmarital children by imposing on citizen mothers the additional requirements imposed on fathers would deprive of citizenship many children on whom Congress intended to bestow citizenship, some of whom would as a result of such deprivation become stateless. Generously reducing the requirements on fathers to those the current law imposes on mothers also poses problems, however. For example, it would categorically favor nonmarital over marital children with one citizen parent, something unlikely to be endorsed by Congress.

(156) Alternatively, a continued distinction between the one year residence required of mothers and five of everyone else could be justified by reference over to another rule, that of certain other countries which, by making citizenship of nonmarital children dependent on the citizenship of their mothers, might leave stateless such children of American mothers to whom the United States denied citizenship. See Cornelia T.L. Pillard & T. Alexander Aleinikoff, Skeptical Scrutiny of Plenary Power: Judicial and Executive Branch Decision Making in Miller v. Albright, 1998 Sue. CT. REV. 1.

(157) See Miller, 523 U.S. at 438-39.

(158) In an intriguing paper on the implications of Miller, Alex Aleinikoff and Cornelia Pillard point out that "[a]llowing U.S. fathers to elect not to convey citizenship [on children they conceive abroad has] ugly class and race implications." Pillard & Aleinikoff, supra note 156, at 24 n.92.

(159) For example, a leading family law treatise, HOMER CLARK, THE LAW OF DOMESTIC RELATIONS IN THE UNITED STATES (2d ed. 1988), says of "the constitutional position of the father of an illegitimate child," id at 855, that "[i]t is difficult if not impossible to arrive at an accurate or useful assessment of the Supreme Gourt's decisions [in this area]," id. at 860, and "[m]any criticisms and questions can be directed to the Stanley opinion," id. at 856, and "[t]he Supreme Court has attempted on four subsequent occasions to clarify some of the issues raised by the Stanley case, but has succeeded only in compounding the confusion." Id. at 856-58.

(160) 406 U.S. 164 (1972).

(161) 417 U.S. 628 (1974).

(162) 430 U.S. 762 (1977).

(163) Consider, for example, how, if at all, the analysis would have been different, if instead of Lorelynn Miller, the plaintiff in the Supreme Court citizenship case had been Luzvisaminda Ablang, also the daughter of a Phillippine woman and a U.S. citizen-soldier-father, but one whose "mother gave her shortly after birth to another family, which raised but never legally adopted her. Ablang was unaware of her true parentage until her natural mother came to see her some time after her graduation from high school. Her mother informed her that her father was a United States citizen...." Ablang v. Reno, 52 F.3d 801, 802 (9th Cir. 1995) (denying Ablang citizenship on the basis of the statute at issue in Miller).

(164) 391 U.S. 73 (1968) (expanding wrongful death actions to include suits by the mother of a deceased illegitimate child).

(165) Id. at 77 (Harlan, J., dissenting).

(166) Id. at 78-79 (Harlan, J., dissenting).

(167) Miller, 523 U.S. at 487 (Breyer, J. dissenting).

(168) Of course, as Professor Martha Fineman pointed out to me, such intrusive questions in aid of amorphous standards would not be new to those controlling access to the United States. Compare the requirement that a marriage between a citizen and an alien, in order to advance the alien's immigration, must be more than legally binding, it must be "valid." See 8 U.S.C. [sections] 1430(a) (1994).

(169) For example, New York's Domestic Relations Law, see N.Y. DOM. REL. LAW [sections] 111 (McKinney 1999), specifies who must consent to adoption of a child, with respect to children born out of wedlock. Like the statute at issues in Miller, this statute now gives rights to mothers categorically and unconditionally which it gives to fathers only to the extent that they have taken one or more specified steps to evince commitment to the child. Compare id. [sections] 111(c) with id. [sections] 111(d). If such a distinction is to be seen as unconstitutional in the context of citizenship, its continuing viability in the adoption context requires further justification.

(170) United States v. Virginia, 518 U.S. 515, 533 (1996).

(171) "Sex classifications may be used to compensate women `for particular economic disabilities [they have] suffered.'" Id.

(172) 441 U.S. 380 (1979).

(173) These modifications are somewhat difficult to square with the phrase officials of the Citadel repeatedly used in conversations with me to describe their approach to the arrival of women: "Assimilation, not integration, and certainly not accommodation." I take this motto to signify the Citadel's desire to change nothing on account of the arrival of women. It is the women who must assimilate to the Citadel's existing masculine culture. Integration would signify a single standard formed through modification of the rules for both women and men. At present, the Citadel seems to exhibit the worst combination of some unsystematic and half-hearted accommodation with repudiation of any commitent to accommodation.

(174) It is reported that the harassment of the women intensified when they were excused from carrying heavy loads because of injuries. Personal communications with Valorie Vojdik, attorney for Shannon Faulkner. The women also, contrary to the regulation for women, had their heads shaved. Id. It is unclear whether this was imposed on them by harassers or if they did it voluntarily in an effort to fit in. Id.

(175) Upperclass women at VMI will be subject to sex-specific makeup, hair, and jewelry regulations designed, according to administrators, to obviate the risk that its female cadets will be mistaken for males. See Peter Finn, Metro Making Room for a New Breed of Rat: VMI Leaders Say School is Ready for First Women, WASH. POST, July 27, 1997, at 88.

(176) See id.

(177) See id. Among the few across the board changes made coincident with the admission of women are privacy fostering rules regarding attire in public spaces in the barracks, screening of windows on cadets' rooms, and restrictions on the previously unhindered ability of upperclassmen to enter rats' rooms unannounced. See id.

(178) I do not mean to minimize the effect of these differences. I have been an accredited observer at VMI for many of the major events in the integration of women, and I can testify from experience that the very slight difference in hair length is the single most reliable way of distinguishing male from female rats. Breasts and hips-to-waist ratio are effectively hidden by fatigues, and many male cadets are as slight and delicate looking as women. If a press photographer's rule is "look for hair" in singling out female rats, a quarter-inch in length makes a world of difference. In my view, VMI was wrong to abandon its intermediate plan of returning the hair length for all rats to one slightly longer than recently permitted, but consistent with uniform rules in past decades.

As for changes to accommodate privacy, these include individualized, curtained shower stalls for female cadets, but none for the men. When I questioned the justification for this distinction, I was told by the male cadet taking me on tour that the women's showers had been so constructed for "health reasons." Neither I, nor the physicians I have consulted, have any idea what these health reasons might be; the cadet did not know either. But the vague look of horror in his eyes as he said this suggested to me that for the men of VMI, the bodies of women still were a contaminant, infested with cooties, somehow mysteriously unclean.

(179) Administrators at VMI insist that behind the scenes they were frantically preparing for women, even as its Board was openly discussing the option of going private instead.

(180) I should disclose that in a debate at UVA sponsored by the Federalist Society with VMI's counsel Ted Olson, while the case was sub judice before the Court, I asked Olson what VMI was afraid of and told him the following:
   What I don't understand is why VMI doesn't simply say nothing will change.
   We will admit women, and any woman who is foolish enough to want to and
   strong enough to be able to do everything we have required of men thus far,
   we'll let in. No privacy--she goes to the bathroom with, she showers with,
   she does push-ups with all those men. If she has to do it, fine. Why not at
   least give her that chance? Why must everything change? Except for archaic
   and stereotypical reasons about the way the sexes ought to behave to one
   another. Possibly few women will be able to do it, I don't know so much.
   It's not clear that many men at the moment do it.... If you're male you get
   in whether you have the physical abilities required or not. The District
   Court established that 15% of women in the applicant pool can pass the
   physical fitness test, 50% of the incoming class at VMI cannot, 2% graduate
   without ever having met those physical requirements. The only physical
   requirement seems to be that you have male genitalia.

Federalist Society Debate, Univ. of Va., April 1996. If VMI had promptly responded to the Hogan decision, as Chief Justice Rehnquist urged, or even to the filing of a lawsuit against them, by admitting women without changing their requirements, I would have supported them, as few other feminists would. I do not think this estops me from claiming that decades after Hogan, years after litigation commenced and was defended by VMI with relentless insistence on the need for a change in standards, and months after the Supreme Court ordered the admission of women having accepted VMI's insistence that change would be required, VMI could be estopped from its plan of minimal change. For just such behavior as VMI's was the constitutional disparate-impact standard developed. Under that standard, the very same rule deemed acceptable when done for acceptable motives becomes unconstitutional when the motives are as pernicious as VMI's could have been seen to be in this instance.

(181) Just as boys do not want to attend a girls' school, they do not, for example, want to watch Saturday morning cartoons or play video games featuring predominantly female characters. They are afraid of cooties. In the case of cartoon characters, as with so much of contemporary culture, males devalue the feminine while females accept the masculine. The result on Saturday mornings has reportedly been a severe limitation on female characters, reminiscent of the avoidance of "tipping points" in racial integration.

(182) An instructive analogy may be the Detroit Male Academy, whose de jure all-male character was struck down in Garrett v. Board of Education, 775 F. Supp. 1004 (E.D. Mich. 1991). Although the Academy's mission was to educate African-American boys, it appears that it never needed to exclude whites de jure, because of the overwhelming percentage of minority youth in the relevant school system. See id. at 1005, 1014. In any event, only its exclusion of women was challenged in court. See id. at 1005-06.

(183) See, e.g., DeSantis v. Pacific Tel. & Tel., 608 F.2d 327, 329, 331 (9th Cir. 1979) (unsuccessfully arguing, inter alia, that (1) to treat, for example, a male who prefers male sexual partners differently from a female who prefers male partners was to use different employment criteria for men and women, something Title VII prohibits and (2) given that discrimination against those who associated with blacks had been held to violate Title VII, "analogously discrimination because of the sex of the employees' sexual partner should constitute discrimination based on sex"); Singer v. Hara, 522 P.2d 1187, 1190 (Wash. Ct. App. 1974), review denied, 84 Wash. 2d 1008 (1974) (unsuccessfully arguing on the basis of Washington's Equal Rights Amendment that "to permit a man to marry a woman but at the same time to deny him the right to marry another man is to construct an unconstitutional classification `on account of sex'").

(184) 852 P.2d 44 (Haw. 1993).

(185) See id. at 63-64.

(186) See Baehr v. Miike, No. CIV. 91-1394, 1996 WL 694235, at *21 (Haw. Cir. Ct., Dec. 3, 1996). The only compelling state interest on which the state offered evidence on remand was the interest in promoting the wall-being of children who, it claimed, were best off when raised by both of their biological parents in a marital household. See id. at *3. The State failed miserably in this proof, with its own witnesses testifying that gay and lesbian couples can make excellent parents. See id. at *7. 187 See id. at *21

(188) See id.

(189) Id. at *3.

(190) Id. at *18.

(191) See id. at *17. The court explained:
   A father and a mother can, and do, provide his or her child with unique
   paternal and maternal contributions which are important, though not
   essential, to the development of a happy, healthy, well-adjusted child....
   However, there is diversity in the structure and configuration of families.
   In Hawaii, and elsewhere, children are being raised by their natural
   parents, single parents, step-parents, grandparents, adopted parents, hanai
   parents, foster parents, gay and lesbian parents, and same-sex couples....
   There also are families in Hawaii, and elsewhere, which do not have
   children as family members.... The evidence presented by Plaintiffs and
   Defendants establishes that the single most important factor in the
   development of a happy, healthy, well-adjusted child is the nurturing
   relationship between parent and child.... The sexual orientation of parents
   is not in and of itself an indicator of parental fitness.... Gay and
   lesbian parents and same-sex couples can be as fit and loving parents, as
   non-gay men and women and different-sex couples.


(192) Mississippi Univ. for Women v. Hogan, 458 U.S. 718, 725 (1982).

(193) To the extent that disfavoring of homosexuality is tied to the disfavoring of masculine behavior by women and effeminacy in men or of sexually aggressive behavior in women and sexually receptive behavior in men, it is based on stereotypes (in the sense of imperfect proxies) of gay life: it is both over and underinclusive in that not all effeminate men are gay or gay men effeminate, not all masculine women are lesbian nor all lesbians butch.

(194) Consider the defense of role modeling for children--some case law suggests in the employment context that this cannot amount to a BFOQ and why, pace Pat Buchanan, can't you have a female role model if you're a male? Nation Briefs: Buchanan's Role Models Are All Men, CHI. SUN-TIMES, Aug. 21, 1995, at 18.

(195) 440 U.S. 268, 279-80 (1979) (striking down Alabama statute providing alimony on divorce only to wives, never husbands, on the grounds that the state's purpose of reinforcing the model of "allocation of family responsibilities under which the wife plays a dependent role" could not "justify a statute that discriminates on the basis of gender").

(196) 421 U.S. 7, 13-15 (1975) (striking down differential ages of majority for child support and education since "[n]o longer is the female destined solely for the home and the rearing of the family, and only the male for the marketplace and the world of ideas").

(197) I use stereotype here in the sense of an archaic normative template for a given sex, as well as an imperfect descriptive proxy.

(198) See Law, supra note 130, at 232.

(199) See General Elec. Co. v. Gilbert, 429 U.S. 125, 161452 (1976) (Brennan, J., dissenting) (superseded by statute).

(200) Note that the requirement of capacity for vaginal intercourse was probably used as a proxy for fertility in eras when more direct evidence was difficult to obtain.

(201) See Turner v. Safley, 482 U.S. 78, 99-100 (1987) (holding "constitutionally infirm" a Missouri regulation denying prisoners the ability to marry without permission from the superintendent of the prison). For prisoners, the obstacle to intercourse is not simply physiological incapacity, but the fact that one of the parties may be scheduled to spend the remainder of his or her life in a prison that allows no full-contact visits. The Court acknowledged the importance of marriages as "expressions of emotional support and public commitment" with "spiritual significance." Id. at 95-96.

(202) Pub. L. No. 104-199, 110 Stat. 2419, 2419 (1996).

(203) See, e.g., Larry Kramer, Same-Sex Marriage, Conflict of Laws, and the Unconstitutional Public Policy Exception, 106 YALE L.J. 1965, 2000-07 (1997) (arguing that this portion of the Act is unconstitutional).

(204) Defense of Marriage Act [sections] 3(a).

(205) It also may have other constitutional problems unrelated to sex discrimination.

(dagger]) J.E.B.v. Alabama, 511 U.S. 127, 138 (quotation marks omitted) (quoting Powers v. Ohio, 499 U.S. 400, 410 (1991)).

([double dagger]) For reasons set forth more fully in Mary Anne C. Case, Disaggregating Gender from Sex and Sexual Orientation: The Effeminate Man in the Law and Feminist Jurisprudence, 105 YALE L.J. 1, 24, 9-13 (1995), I shall throughout this article use the term "sex discrimination" to refer to discrimination between males and females and reserve the term "gender discrimination" to refer to discrimination on the basis of qualities coded as masculine or feminine, whether the person exhibiting those qualifies is male or female. For reasons set forth more fully below, I shall generally use the term "discrimination" in its technical sense of "making a distinction." Thus, not all sex discrimination in the law (i.e., not all distinctions made in the law on the basis of sex) will necessarily be unconstitutional. Nor will every denial of equal protection to persons on account of their sex necessarily result from sex discrimination. As I shall further explain, because the Supreme Court's inquiries into the denial of equal protection on grounds of sex have focused quite narrowly on deprivations caused by rules that, on their face, distinguish between males and females (i.e., sex-respecting rules), it is no accident that I refer to the law of sex discrimination, rather than say, of sex equality.

Mary Anne Case, Professor of Law, University of Chicago Law School. Versions of this Article were presented at the University of Chicago, New York University, Ohio State University, Quinnipiac College, the University of Virginia and Washington and Lee University; and at the UCLA Colloquium on Justice, the New York Area Feminist Law Teachers Works-in-Progress Group, and a Federalist Society debate with Virginia Military Institute's counsel Ted Olson. I am grateful to participants in those events, particularly Sylvia ]Law, Mary Becker, Brian Bix, Jennifer Brown, Jim Brudney, Ruth Colker, Rochelle Dreyfuss, Chris Eisgruber, Martha Fineman, Louis Halper, Frances Kamm, Larry Kramer, Carlin Meyer, David Millon, Butt Neuborne, Eric Posner, Judith Resnick, Joan Shaughnessy, Seana Shiffrin, Peter Swire, Mary Moers Winig, and Diane Zimmerman, as well as to Alex Aleinikoff, Dianne Avery, Ed Baker, Jack Balkin, Emily Buss, John Harrison, Dick Howard, Para Karlan, Mike Klarman, David Martin, Chuck McCurdy, Steve Newhouse, Dan Ortiz, Tod Preuss, Mike Seidman, David Strauss, Cass Sunstein, Adrian Vermeule, Rip Verkerke, Val Vojdik, Ted White; research assistants Toby Heytens, Allyson Newton, David Ravicher, and Liz Tucci; members of my 1996 class in Regulating ]Family, Sex and Gender; and Si Bunting, Mike Strickler, and Chuck Steenbergen of VMI.
COPYRIGHT 2000 Cornell University
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2000 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Case, Mary Anne
Publication:Cornell Law Review
Geographic Code:1USA
Date:Jul 1, 2000
Previous Article:Afterword: critical strategy and the judicial evasion of difference.
Next Article:Respecting commitment: a proposal to prevent legal barriers from obstructing the effectuation of intestate goals.

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters