Editor's Note: This Essay is the last in a four-part series on United States v. Windsor, the Supreme Court decision that struck down Section 3 of the Defense of Marriage Act (DOMA) last summer.
Justice Antonin Scalia claims that when the Supreme Court invalidated Section 3 of the Defense of Marriage Act (DOMA) in United States v. Windsor, it improperly impugned Congress’s motives. The Court held that the statute, which withheld federal recognition from same-sex marriages for all purposes throughout the U.S. Code, reflected a “bare congressional desire to harm a politically unpopular group.” Scalia is right that opposition to same-sex marriage is not the same as hatred of gays.
Yet Scalia’s own methods of statutory interpretation support what the Court did in Windsor. If one infers the statute’s purpose from its language and interaction with other statutes, with no attention to the legislative history or the subjective intentions of the law’s authors, the result the Court reached is inescapable. The statute may not precisely reflect a bare desire to harm, but it reflects an extreme indifference to the welfare of gay citizens that violates equal protection.
Part I of this Essay describes the debate among the Justices about the purpose of DOMA. Part II uses Scalia’s theory of statutory interpretation to discern DOMA’s purpose. Part III explains why DOMA, as understood through that theory, violates the Equal Protection Clause. The last Part concludes.
I. Bare Desire to Harm?
DOMA declares, in pertinent part, that the word marriage, wherever it appears in the U.S. Code, “means only a legal union between one man and one woman as husband and wife.” Justice Kennedy, writing for the Court, held that the law was unconstitutional because it pursued an impermissible purpose. “The Constitution’s guarantee of equality ‘must at the very least mean that a bare congressional desire to harm a politically unpopular group cannot’ justify disparate treatment of that group.”
How does Kennedy know that such a desire is present? He argued that “[t]he history of DOMA’s enactment and its own text demonstrate that interference with the equal dignity of same-sex marriages, a dignity conferred by the States in the exercise of their sovereign power, was more than an incidental effect of the federal statute. It was its essence.” He concluded that “[t]he avowed purpose and practical effect of the law here in question are to impose a disadvantage, a separate status, and so a stigma upon all who enter into same-sex marriages made lawful by the unquestioned authority of the States.”
Justice Scalia had a powerful answer: “to defend traditional marriage is not to condemn, demean, or humiliate those who would prefer other arrangements.” Justice Alito similarly challenged Kennedy’s attempt to “cast all those who cling to traditional beliefs about the nature of marriage in the role of bigots or superstitious fools.” The Act, Scalia argued, “did no more than codify an aspect of marriage that had been unquestioned in our society for most of its existence—indeed, had been unquestioned in virtually all societies for virtually all of human history.” Kennedy’s logic conflates the desire to harm a group, which is impermissible, with the desire to disapprove of certain choices. The equal dignity of gay people is constitutionally protected, but it is not necessarily identical with “the equal dignity of same-sex marriages.” One can disapprove of same-sex sexual relations or marriage without despising gay people as such.
To see whether Kennedy’s argument can be rehabilitated, we need to analyze what it means to discern the purpose of a statute. In this task, we will rely on a perhaps surprising ally: the noted theorist of interpretation, Antonin Scalia.
II. How to Determine DOMA’s Purpose
A. Discerning Purpose
Justice Scalia is an original-meaning textualist. He “ascribe[s] to th[e] text the meaning that it has borne from its inception, and reject[s] judicial speculation about both the drafters’ extratextually derived purposes and the desirability of the fair readings’s anticipated consequences.” He has offered a well-developed account of how to determine legislative purpose. He argues that “discerning the subjective motivation of those enacting the statute is, to be honest, almost always an impossible task,” but “it is possible to discern the objective ‘purpose’ of a statute (i.e., the public good at which its provisions appear to be directed).” Scalia states:
The evidence suggests that [in statutory interpretation], despite frequent statements to the contrary, we do not really look for subjective legislative intent. We look for a sort of “objectified” intent—the intent that a reasonable person would gather from the text of the law, placed alongside the remainder of the corpus juris. . . . It is the law that governs, not the intent of the lawgiver.
Purpose, thus understood, “is to be gathered only from the text itself, consistently with the other aspects of its context.” “Normally, finding a purpose in text is a straightforward matter requiring no feats of subtle deduction. Generally the purpose is unmistakable.” For example, “[a] statute provides that anyone with three or more convictions for DUI must have his driver’s license permanently revoked: The purpose is to keep those so convicted permanently off the road.” Scalia is opposed to “purposivism” in interpretation, but only when it relies on a purpose that is different from that manifest in the words of the law. “Purpose sheds light only on deciding which of various textually permissible meanings should be adopted.”
This argument is potentially misleading if it is taken to suggest that a text can convey meaning when it is read in isolation, with no reference to extratextual, cultural context. Stanley Fish observes that if this is what Scalia is claiming, then he exaggerates the autonomy of text. No text can tell you what its context is. Without context, a text has no meaning. Thus Scalia may be read charitably as making the more modest claims that legislative history should not be consulted in interpretation, and that interpreters should not impose their own notions of sensible legislation onto statutes.
Scalia sometimes does not even notice how his own knowledge of cultural context shapes his reading of texts. Consider the case of interracial marriage. In an argument over the protection of abortion in Planned Parenthood of Southeastern Pennsylvania v. Casey, the plurality opinion claimed that Scalia’s narrow reading of the Constitution would permit legislation prohibiting interracial marriage. He responded that this “is entirely wrong,” because such laws are “contradicted by a text—an Equal Protection Clause that explicitly establishes racial equality as a constitutional value.” The Clause, however, merely declares that “No State shall . . . deny to any person within its jurisdiction the equal protection of the laws.” No explicit mention of race there. So how can Scalia get to the result he clearly wants?
We all—Scalia included—rely on cultural context to make sense of both the Equal Protection Clause and the miscegenation laws. Everyone knows that the Fourteenth Amendment was enacted with the specific purpose of invalidating laws that imposed unequal legal status on African-Americans. Everyone also knows that the miscegenation laws, along with the rest of the regime of racial segregation, were intended to maintain that unequal status.
In Loving v. Virginia, the Court declared that laws prohibiting interracial marriage were “measures designed to maintain White Supremacy.” It so stated in the teeth of a perfectly innocent alternate purpose that had been offered by the state.
Virginia, in its brief, had cited numerous scientific authorities (since debunked) claiming that racial intermarriage tended to produce physically defective and socially maladjusted children. It argued that, if the Court were to undertake an inquiry into the wisdom of the challenged legislation, “it would quickly find itself mired in a veritable Serbonian bog of conflicting scientific opinion upon the effects of interracial marriage, and the desirability of preventing such alliances, from the physical, biological, genetic, anthropological, cultural, psychological and sociological point of view.” The decision not to allow such marriages, it argued, rested on the acceptance of scientific arguments put forth by respectable authorities. If an innocent explanation was all that was needed, here it was. Moreover, it is nearly certain that at least some of Virginia’s leaders had managed to persuade themselves that these scientific claims were true. The appellants responded that “there is not a single anthropologist teaching at a major university in the United States who subscribes to the theory that Negro–white matings cause biologically deleterious results,” but the Court was certainly not competent to adjudicate this dispute. Moreover, even if the law rested on bogus science, this would not necessarily have impugned the legislators’ motives. Innocent mistakes are not invidious.
Instead, the Court emphasized that “Virginia’s miscegenation statutes rest solely upon distinctions drawn according to race.” Such distinctions are “odious to a free people whose institutions are founded upon the doctrine of equality.” The Court supported its attribution of invidious purpose by noting that “Virginia prohibits only interracial marriages involving white persons,” but it also indicated that Virginia would not be able to cure the difficulty by enacting a more broadly worded statute.
Why was the Court entitled to make that move? The answer is that the Court was aware of the cultural context of miscegenation laws—a context in which those laws were inseparable from racist ideology. It read the laws in light of that context. That justified the attribution of a racist purpose. It was the most reasonable reading of the laws’ public meaning. If a law’s words “mean what they conveyed to reasonable people at the time they were written,” then the purpose conveyed by the miscegenation law is promoting white supremacy. Alternate readings are available. But in light of everything we know about the society in which the laws were enacted, they are not believable. Scalia evidently has something like this line of reasoning in mind when he concludes that these laws run afoul of the Equal Protection Clause.
Scalia has held that objective reading of statutes can yield an invidious purpose that renders a law invalid. In the process, he has revealed his unstated willingness to consider cultural context. For example, in Church of the Lukumi Babalu Aye, Inc. v. City of Hialeah, the Court struck down four ordinances that a city had enacted with the avowed purpose of preventing a Santeria church from practicing animal sacrifice. The laws, the Court held, violated the Free Exercise Clause of the First Amendment because their object was the suppression of a religious practice. This impermissible purpose was evident on the face of three of the statutes because they only prohibited the “sacrifice” of animals as part of a “ritual.”
Justice Kennedy, who wrote the majority opinion, was able to find lurid statements by Hialeah city officials urging the city “not to permit this Church to exist” and declaring that Santeria was the worship of “demons” and therefore was “an abomination to the Lord.” But Kennedy lost his majority in the section of the opinion that cited this history; only Justice Stevens joined it. The majority held that “suppression of the central element of the Santeria worship service was the object of the ordinances,” and cited the language and operation of the statutes, as well as the fact that almost all nonreligious killings of animals were expressly exempted. That majority included Scalia.
If an innocent-sounding explanation will save a statute, even when it is part of a legislative scheme that impermissibly targets a group, then this law should have been upheld. In oral argument, Scalia flirted with this possibility. He suggested that a facially nondiscriminatory law, even one that specifically refers to “rituals,” should be upheld as not targeting religion because the law would also ban nonreligious ritual killing of animals, for example in fraternity initiations. Attorney Douglas Laycock responded: “If that’s a valid argument, you really have repealed the free exercise clause. Any lawyer in the country with that standard of drafting can draft an ordinance to get any church that happens to be at crosswise with the city council.” In the end, Scalia concluded that the Court should invalidate “those laws which, though neutral in their terms, through their design, construction, or enforcement target the practices of a particular religion for discriminatory treatment.” Thus he evidently concedes Fish’s point that meaning depends on cultural context.
The last bit of context that matters is not cultural, but textual. Any statute is embedded in a larger body of codified law. Scalia has argued that, when reading statutes, we should presume that the overall code is coherent and reflects a unified set of purposes and that any specific provision is enacted with a full understanding of preexisting law. Knowledge of the interaction of the new statute with that preexisting law is appropriately imputed to the legislature. This canon of interpretation, he concedes, counterfactually assumes an implausible legislative knowledge of related legislation in the past and an impossible legislative knowledge of related legislation yet to be enacted. The canon is, however, based upon a realistic assessment of what the legislature ought to have meant. “It rests on two sound principles: (1) that the body of the law should make sense, and (2) that it is the responsibility of the courts, within the permissible meanings of the text, to make it so.” Any interaction with other statutes is irrebuttably presumed to be intended, a part of the purpose of the statute.
B. The Purpose of DOMA
Is the most reasonable reading of DOMA one that attributes an invidious purpose to the statute?
Justice Scalia thinks that the law’s “many perfectly valid—indeed, downright boring—justifying rationales” make it hard “to maintain the illusion of the Act’s supporters as unhinged members of a wild-eyed lynch mob.” These purposes “give the lie to the Court’s conclusion that only those with hateful hearts could have voted ‘aye’ on this Act.” Chief Justice John Roberts similarly writes: “At least without some more convincing evidence that the Act’s principal purpose was to codify malice, and that it furthered no legitimate government interests, I would not tar the political branches with the brush of bigotry.”
Scalia offers two innocent purposes.
1. Does DOMA Avoid Difficult Choice of Law Issues?—
Justice Scalia’s most plausible argument is that “DOMA avoids difficult choice-of-law issues that will now arise absent a uniform federal definition of marriage.” What happens, he asks, to the marriage of a couple who wed in New York and move to Alabama? “DOMA avoided all of this uncertainty by specifying which marriages would be recognized for federal purposes.”
But federal laws and regulations already deal with those questions, which will still arise with underage marriages, cousin marriages, commonlaw marriages, and the like. Federal agencies have routinely addressed these situations for more than a century. Scalia does not explain why same-sex marriage is any different.
More important is the question of proportionality. Conflict of laws problems are rare. That is why, for a long time, almost no one has noticed that federal law is often unclear about which state’s law to apply to determine marriage for federal purposes. In a country with thousands of same-sex marriages, remarkably few conflict of law issues have actually arisen.
DOMA’s blunderbuss response was to withhold recognition from all of these marriages, even that overwhelming majority in which the couple never changes their home, and in which no choice of law problem arises.
The consequences are far-reaching. Same-sex spouses could not file joint tax returns. Pretax dollars could not be used to pay for health insurance or healthcare expenses for a same-sex spouse or that spouse’s dependent children. Same-sex spouses’ debts incurred under divorce decrees or separation agreements were dischargeable in bankruptcy. Same-sex spouses of federal employees were excluded from the Federal Employees Health Benefits Program, the Federal Employees’ Group Life Insurance Program, and the Federal Employees’ Compensation Act, which compensates the widow or widower of an employee killed in the performance of duty. Same-sex spouses were the only surviving widows and widowers who would not have automatic ownership rights in a copyrighted work after the author’s death. Same-sex spouses were denied preferential treatment under immigration law and, therefore, were the only legally married spouses of American citizens who faced deportation. It is a federal crime to assault, kidnap, or kill a member of the immediate family of a federal official in order to influence or retaliate against that official—but it was not if you did that to a same-sex spouse. With the end of the exclusion of gay people from the military, DOMA made it official policy to withhold any survivor’s benefits from the surviving spouse of a soldier killed in the line of duty. And so on.
Once every few years, DOMA may simplify some federal bureaucrat’s job. Everyone else is saddled with enormous burdens. Thousands of employers in states that recognize same-sex marriage must maintain two separate administrative regimes for benefits—one that handles federal benefits affected by DOMA, and another to comply with state law. Mapping the border between those two regimes is so complex that the task has begotten a small industry of compliance specialists—expensive professionals whose work is a deadweight loss for the American economy. A group of 278 employers and organizations, in an amicus brief to the Court, explained:
Some amici have had to pay vendors to reprogram benefits and payroll systems, to add coding to reconcile different tax and benefit treatments, to reconfigure at every benefit and coverage level, and to revisit all of these modifications with every change in tax or ERISA laws for potential DOMA impact. . . . Benefits and human resources departments, facing questions from employees with same-sex spouses regarding workplace benefit selections and coverage, must be adequately trained and prepared to explain the disparate treatment to employees who may later realize (perhaps too late) that their benefits choices and decisions carried unanticipated and significant financial implications.
Even large employers can be overwhelmed. Yale University had to tell its employees that, because of a programming error, it had failed to withhold taxes for the imputed value of health coverage for same-sex spouses in 2010, resulting in extra deductions in 2011. Because state antidiscrimination laws protect many gay employees, the employer must determine, at its own risk, where DOMA does and does not supersede state law.
DOMA does not even “eas[e] administrative burdens” for government actors. For example, it creates fiendishly complex problems for bankruptcy courts, which must deal with property rights created by marriages recognized under state law (which, of course, DOMA could not annul without running afoul of the takings clause) while somehow withhold recognition for federal purposes.
DOMA deems the interest in nonrecognition of same-sex couples to be so overridingly urgent as to justify sacrificing a huge range of other government interests, some of the highest order. Bankruptcy courts cannot accomplish efficient and predictable adjudication. Government employees cannot insure their dependents. The Social Security survivor benefits program, “the primary purpose [of which] is to pay benefits in accordance with the probable needs of the beneficiaries,” cannot accomplish this goal. Retirees cannot provide for the security of their dependent spouses. Income taxation cannot take account of family obligations. National safety itself is compromised because the military cannot provide its members’ families with healthcare, housing, and survivorship benefits that are essential to military effectiveness.
Other conflict of law solutions were available. The most obvious ones are those based on the place of celebration or the domicile of the couple (which, for many couples, are the same state, so the choice between these rules makes no difference). Blanket nonrecognition is such an extraordinary overreaction that even the Jim Crow South did not adopt it with respect to interracial marriage.
Scalia had one other argument.
2. Does DOMA Preserve Prior Legislation?—
Justice Scalia also argued that DOMA just preserved the intended effects of prior legislation:
When Congress provided (for example) that a special estate-tax exemption would exist for spouses, this exemption reached only opposite-sex spouses—those being the only sort that were recognized in any State at the time of DOMA’s passage. When it became clear that changes in state law might one day alter that balance, DOMA’s definitional section was enacted to ensure that state-level experimentation did not automatically alter the basic operation of federal law, unless and until Congress made the further judgment to do so on its own. That is not animus—just stabilizing prudence.
This is circular. Just how is “the basic operation of federal law” altered when the law recognizes a new set of marriages? That operation changes in just the same way—incorporating new couples into the existing structure—whenever a marriage takes place, regardless of the gender of the spouses.
3. To Disparage and to Injure.—
Note what DOMA does not do. Justice Kennedy claimed that Congress’s purpose was “to influence or interfere with state sovereign choices.”Justice Alito sensibly responded that DOMA “does not prevent any State from recognizing same-sex marriage or from extending to same-sex couples any right, privilege, benefit, or obligation stemming from state law.” There is no evidence that DOMA has influenced any state’s decision whether to adopt same-sex marriage, and it is hard to imagine how it could.
But Congress’s helplessness cuts against the law’s reasonableness. DOMA is often defended as manifesting moral opposition to same-sex marriage—opposition, we have stipulated, that is a constitutionally permissible purpose. But that purpose cannot justify a statute that does not promote that end in any significant way. The question before Congress was not whether same-sex marriages would exist. It was undisputed that Congress had no power to answer that question, which is reserved to the states. The real issue was whether to have a set of second-class marriages, denied recognition for all federal purposes, even in contexts in which the whole purpose of the federal classification is defeated by not recognizing the marriage.
The creation of such a set of marriages does not do much to serve either of Justice Scalia’s enumerated purposes. A law will fail even rational basis review if the “purported justifications” make “no sense in light of how the [government has] treated other groups similarly situated in relevant respects.” DOMA does, however, very effectively “tell [same-sex] couples, and all the world, that their otherwise valid marriages are unworthy of federal recognition.”
Here we return to cultural context. The most pertinent context is the fact that when the law was enacted, gay people were still the objects of pervasive and, more importantly, unquestioned prejudice. A few years before, when newly elected President Bill Clinton tried to end their exclusion from the military, the public reaction was so negative that he had to settle for the lousy, and now abandoned, “don’t ask, don’t tell” compromise. That prejudice is easily inferred from “the text itself, consistently with the other aspects of its context.” Read in light of the entire U.S. Code that it amends and the culture in which it was enacted, the law’s central purpose is “to disparage and to injure” gay people at every opportunity, by any available means, heedless of the cost.
III. The Equal Protection Violation
A. Extreme Indifference
The impact on gay people was far from Congress’s mind when the law was enacted. DOMA was enacted by huge, veto-proof majorities in 1996, after the Hawaii Supreme Court made it seem likely that that state was about to adopt same-sex marriage. It was enacted, not to address any existing problem, but because Republicans hoped to place Clinton in a difficult spot: either ineffectually veto the law, and be branded as favoring same-sex marriage (which only a third of Americans then supported), or sign it and alienate his gay supporters, in either case hurting his chances of reelection. He chose the second option and was reelected anyway.
Congress was not thinking about solving a policy problem at all, and it certainly was not thinking about the actual human beings whom this law was going to injure. It lashed out at gay people for the sake of pure political posturing.
When DOMA was enacted, there were no same-sex marriages recognized under state law. Today, on the other hand, there are more than 130,000 such marriages. About half of the U.S. population lives in a jurisdiction in which same-sex relationships are legally recognized. When Congress passed the law, there were no actual people in same-sex marriages for the law to hurt and insult. Now there are plenty of them.
It is imprecise to characterize DOMA as reflecting a bare desire to harm a politically unpopular group. The offense against equal protection is different. Paul Brest observes that one way in which equal protection can be violated is for state actions to reflect “racially selective sympathy and indifference,” meaning “the unconscious failure to extend to a minority the same recognition of humanity, and hence the same sympathy and care, given as a matter of course to one’s own group.” John Hart Ely argues that in order for legislation to be legitimate, the citizens must all “be represented in the sense that their interests are not to be left out of account or valued negatively in the lawmaking process.”
In DOMA, Congress did not mean to hurt anybody. It was merely oblivious. In 1996, otherwise reasonable people thought it a pointless waste of taxpayer dollars to look after the basic needs of gay couples and their families. They also wanted to express moral disapproval of homosexuality, and the desire to make that symbolic point overwhelmed any concern about the concrete consequences of their actions. The statute reflects the fantasy, unfortunately quite common, that gay people can be wished out of existence. Aquinas observed that ignorance is not an excuse “when somebody chooses not to be informed, in order to find some excuse for sin or for not avoiding it,” or “when a person does not actually attend to what he could and should consider.”
More than any other field of law, criminal law parses out culpable states of mind with great precision. One such state of mind is “depraved-heart murder,” in which the defendant does not intend to kill, but causes death when he “engages in conduct which manifests a depraved indifference to the value of human life.” The Model Penal Code uses more precise phrasing, referring to homicide “committed recklessly under circumstances manifesting extreme indifference to the value of human life.” In such cases, death is caused by “the intentional doing of an uncalled-for act in callous disregard of its likely harmful effects on others.” One element in determining extreme indifference is the social utility of the conduct: speeding through crowded streets for the thrill of it is a paradigmatic case, while doing that to carry a passenger who needs emergency surgery to the hospital may not be a crime at all. (That, I will shortly show, is relevant to assessing the wildly disproportionate approach codified in DOMA.)
Of course, given Justice Scalia’s method, Congress’s actual state of mind is irrelevant to the interpretation of DOMA. The issue is “the original meaning of the text, not what the original draftsmen intended.” But the idea that gay people’s interests have so little weight that the burden imposed on them is justified by the trivial interests DOMA promotes is part of “what the text would reasonably be understood to mean.” It is the law’s objective meaning. Given the weakness of the justifications for DOMA, the pertinent legislative purpose was closer to extreme indifference than to anything else.
B. Justice Scalia and Equal Protection
Does Justice Scalia think that a bare desire to harm a politically unpopular group violates equal protection? He has not said so. In the cases in which the question arose, Scalia attacked the majority’s minor premise, claiming that it had misconstrued the purpose of the law in question. More generally, although Scalia has elaborated at length on the theory of interpretation of legal texts—that is what we have relied upon here—it is surprisingly difficult to reconstruct his theory of the Equal Protection Clause. With the data that exists, we can conclude that his theory of interpretation is inconsistent with his restrictive view of equal protection and that a less restrictive view supports the result in Windsor.
The core of equal protection, in Scalia’s view, is a ban on racial classification: “In my view the Fourteenth Amendment’s requirement of ‘equal protection of the laws,’ combined with the Thirteenth Amendment’s abolition of the institution of black slavery, leaves no room for doubt that laws treating people differently because of their race are invalid.” Scalia would also invalidate facially neutral laws that have a racist purpose. On the other hand, he does not think that the courts can remedy unconscious racism, even if it demonstrably influences life and death decisions. He also disagrees with the Court’s imposition of heightened scrutiny for sex-based classifications, and thinks that sex discrimination raises no equal protection issue.
There may be a limit to his minimalism. He has declared that what keeps government action within “reasonable and humane limits” is “the Equal Protection Clause, which requires the democratic majority to accept for themselves and their loved ones what they impose on you and me.” Perhaps for this reason, “[t]he Equal Protection Clause epitomizes justice more than any other provision of the Constitution.” On the other hand, he has never endorsed any doctrinal device for operationalizing this requirement: if equal protection prohibits racial classifications and nothing else, then the majority can impose nearly anything it likes on you and me.
Scalia is committed to reading the Fourteenth Amendment in accordance with its original public meaning. That is why he rejects its application to sex discrimination: “Nobody thought it was directed against sex discrimination” when the Amendment was enacted. But then we must apply the same test to the following rule, which is what Scalia’s position amounts to in practice: The Fourteenth Amendment prohibits classifications and purposive government actions based on race and national origin, and that is all that it prohibits.
At the time of the framing, no one would have understood the Amendment to mean that. Equal protection was understood at the time to prohibit class legislation—legislation that unjustifiably singled out a particular group for special benefits or burdens. The Amendment enacted a “doctrine against ‘partial’ or ‘special’ laws, which forbade the state to single out any person or group of persons for special benefits or burdens without an adequate ‘public purpose’ justification.” The Black Codes, which did not always formally classify based on race, were prime examples. Scalia has shown remarkably little interest in any historical evidence of the original meaning of the Amendment. There are, of course, difficult questions about how to turn these into judicially administrable rules. If courts get to evaluate the justification of every law that singles someone out for a benefit or burden, then they effectively become another house of the legislature. But Scalia has not explored the original meaning carefully enough to even encounter those questions.
An approach to equal protection based on his general theories of interpretation would look very different from the approach he actually takes. It would have to begin with the original public meaning of the text and then attempt somehow to operationalize it.
Begin with his requirement that people be valued as individuals. Before he joined the Court, he argued that affirmative action would inevitably produce injustice because it is “based upon concepts of racial indebtedness and racial entitlement rather than individual worth and individual need; that is to say, because it is racist.” The core claim is that laws should respond to “individual worth and individual need.”
You cannot respond to individual worth if you regard some people as having no worth. You cannot respond to individual need if you think that some people’s needs do not matter. Scalia’s logic thus supports the familiar conclusion that equal protection prohibits laws based on selective sympathy and indifference. That is why the majority must “accept for themselves and their loved ones what they impose on you and me.” If they must do that, then a fortiori a law is not valid if it reflects a bare desire to harm a politically unpopular group—or extreme indifference toward a group. The Moreno Court was right: “[I]f the constitutional conception of ‘equal protection of the laws’ means anything, it must at the very least mean that a bare . . . desire to harm a politically unpopular group cannot constitute a legitimate governmental interest.” It must also mean that the legislature cannot single a group out for enormous burdens in order to produce an insignificant benefit. The notion of class legislation may, as a general matter, involve too many discretionary judgments to be appropriately administrable by an unelected judiciary, but that is not to say that it imposes no limits at all on the legislature. Perhaps the courts should intervene only in the rare case where the inadequacy of the justification is flagrant—“a classification whose relationship to an asserted goal is so attenuated as to render the distinction arbitrary or irrational.” DOMA presents such a case. Scalia ought to support the major premise of Windsor.
And, in fact, he never denies it. The “bare desire to harm” formulation was used in an earlier gay rights case, and he did not dispute the major premise there either. Romer v. Evans struck down an amendment to the Colorado constitution—referred to on the ballot as “Amendment 2”—declaring that neither the state nor any of its subdivisions could prohibit discrimination on the basis of “homosexual, lesbian or bisexual orientation, conduct, practices or relationships.” Justice Kennedy’s opinion for the Court observed that the Amendment “has the peculiar property of imposing a broad and undifferentiated disability on a single named group.” The Court concluded that, “Amendment 2 classifies homosexuals not to further a proper legislative end but to make them unequal to everyone else.”
Scalia dissented in Romer. Given his own interpretive method, he should not have done that. His dissent is based on a misreading of the law. He thought that the law was merely “a modest attempt by seemingly tolerant Coloradans to preserve traditional sexual mores against the efforts of a politically powerful minority to revise those mores through use of the laws.” Even on the narrow reading of Amendment 2 that Scalia proposed, which would have left general guarantees of fair treatment undisturbed and prohibited only antidiscrimination protection of gay people, it was only gays who were singled out for nonprotection. The amendment left undisturbed the protection of heterosexuals against discrimination on the basis of sexual orientation. Scalia neither admitted nor defended this singling out of an unpopular group. And we have already seen the weakness of his defense of DOMA.
Romer and Windsor invalidated laws for lacking a rational basis, but any statute’s terms suggest a purpose that the statute rationally serves. A law that bans the driving of blue Volkswagens on Tuesdays is rationally—indeed, perfectly—related to the purpose of preventing blue Volkswagens from being driven on Tuesdays. The real issue is whether some goals are impermissible or too costly to be worth pursuing, a question that cannot be answered on the basis of “rationality.” David Hume famously wrote: “’Tis not contrary to reason to prefer the destruction of the whole world to the scratching of my finger.” But that is not the sense of rationality that the Court relies on, or should rely on. Innocent explanations did not save the statutes challenged in Loving and Lukumi.
Return to the analogy with extreme-indifference murder. Authorities disagree about whether a defendant can be guilty of extreme indifference if he is unaware of the risk of his conduct, for instance if he is “too absent-minded or to[o] feeble-minded (though not insane) or too drunk to realize the seriousness of the risk.” The better view is that negligence in such cases is not murder. Were one looking for actual legislative intent, it might be difficult to impute to DOMA any desire to harm anyone, or even conscious disregard of a known risk, because Congress’s mental state was not so much depravity of heart as feebleness of mind. Nowhere in the legislative history do the law’s proponents evince any awareness of the law’s actual human consequences. But, once more, Scalia emphatically rejects the “false notion that the purpose of interpretation is to discover intent.” As we have already seen, he thinks that courts should irrebuttably presume Congress’s awareness of any new law’s effect on existing law. That effect, then, must be presumed to be intended.
Recall that a finding of extreme-indifference murder depends on an assessment of the utility of the defendant’s conduct. The exact same behavior will or will not be criminally negligent depending on the value of its object. To trigger liability, it is not necessary that the conduct have no value whatsoever. It really can be thrilling to speed down a crowded street, and pleasure is not valueless. But anyone who thinks that the pleasure is worth it—who endangers others “for his diversion merely”—manifests a “depraved mind, regardless of human life.”
Windsor’s conclusion that DOMA is irrational implicitly relies on a similar proportionality analysis. That analysis is only a minor theme in strict scrutiny, though it has some role: the interest in question has to be a truly compelling one. The analysis balances cost and benefit. Stephen Siegel has shown that cost justification was the original point of strict scrutiny, which was later transformed into a device for discerning illicit motive. That shift was led by John Hart Ely, who thought it inappropriate for judges to second-guess legislatures’ policy decisions. Ely’s caution about judicial policymaking is sensible, and the modern Court evidently shares it. But deference is not necessarily unlimited. If the benefit is trivial by comparison with the cost, then it is appropriate to infer that the decision has an improper purpose.
DOMA’s purpose is to convey a message of disdain for gay couples, with extreme indifference to the human costs. As with the “heart void of social duty” characteristic of extreme-indifference murder, this takes DOMA outside the range of reasonable disagreement. To say that the tiny administrative inconvenience cited by Scalia is sufficient to justify the enormous burden on gay couples and their employers is crazy. It is like strangling the baby because you cannot decide what to name it.
Justice Scalia’s idea of objective statutory purpose is useful. The fault, in his treatment of gay rights claims, is his inability to grasp the minor premise. He persistently fails to confront what the law before him actually does.
We already saw how he relied on an indefensible reading of the law in Romer. In Windsor, by contrast, he attempted no narrowing construction of DOMA. Nor could he have, given its plain language. But neither did he say a word about the way in which the law actually functions. Had he done so, his own canons of interpretation would have precluded him from attributing to it a benign purpose.
Scalia’s most important claim is that Justice Kennedy’s opinion unfairly conflates opposition to same-sex marriage with hatred of gay people. Here Scalia is right. Opposition to same-sex marriage sometimes has nothing to do with devaluation of gays and lesbians.
For many opponents of same-sex marriage, gay people are marginal to their view of the world. Justice Alito nicely summarizes the position: “marriage is essentially the solemnizing of a comprehensive, exclusive, permanent union that is intrinsically ordered to producing new life, even if it does not always do so.” Whatever the merits of this notion, it is not about gay people. It is focused on the value of a certain kind of heterosexual union. The existence of gay people is a side issue. The function of marriage law, on this view, is to protect a human good that gay people happen to be unable to realize: marriage laws do not discriminate against them any more than art museums discriminate against blind people. The Court, Scalia wrote in Romer, should not “verbally disparage[e] as bigotry adherence to traditional attitudes.”
Scalia is also right that the Court should not “take sides in the culture wars.” Disagreement about the value of same-sex relationships is not going to disappear soon. It is fundamentally a disagreement about final ends whose value cannot be demonstrated by argument. It is as intractable as (and, as it happens, largely coextensive with) religious disagreement. The Constitution should be agnostic about unresolvable, contested questions of ultimate value.
But it is not necessary to resolve the ultimate value question in order to conclude that DOMA is unconstitutional. Opponents of same-sex marriage are entitled to their views. But one can hold those views without believing that they justify DOMA’s extraordinary mistreatment of gay couples. Scalia’s own interpretive theories help show why such mistreatment violates equal protection.
Congress passed DOMA without ever focusing on its human costs. DOMA’s contemporary defenders do not have that excuse. They have a weird blind spot. They do not seem to be able to perceive the reality of what they are doing. The need to make a symbolic point overwhelms all other considerations. That self-indulgent blindness has moral dimensions. The notion that you may pursue your goals with no regard at all for the impact on other people is precisely what constitutes extreme indifference.
* John Paul Stevens Professor of Law and Professor of Political Science, Northwestern University School of Law. Thanks to Bob Bennett, Tom Berg, Steve Calabresi, Maggie Gallagher, Jackie Gardina, Rick Garnett, Doug NeJaime, Valerie Quinn, Jonathan Rauch, Marty Redish, Louis Michael Seidman, and Tobias Barrington Wolff for helpful comments on an earlier draft, and to Marcia Lehr for research assistance.
 Windsor, 133 S. Ct. at 2693 (quoting Moreno, 413 U.S. at 534).
 Id. at 2708 (Scalia, J., dissenting). He elaborated the point in his earlier dissent in Romer v. Evans:
Of course it is our moral heritage that one should not hate any human being or class of human beings. But I had thought that one could consider certain conduct reprehensible—murder, for example, or polygamy, or cruelty to animals—and could exhibit even “animus” toward such conduct. Surely that is the only sort of “animus” at issue here: moral disapproval of homosexual conduct . . . .
517 U.S. 620, 644 (1996) (Scalia, J., dissenting) (link).
 Windsor, 133 S. Ct. at 2718 (Alito, J., dissenting).
 Id. at 2709 (Scalia, J., dissenting).
 Id. at 2693 (majority opinion).
 See Andrew Koppelman, The Gay Rights Question in Contemporary American Law 17–19 (2002).
 Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts xxvii (2012).
 I am not here endorsing Scalia’s whole theory of interpretation, which has severe flaws. See generally Andrew Koppelman, Passive Aggressive: Scalia and Garner on Interpretation, 41 Boundary 2 (forthcoming 2014) (book review).
 Antonin Scalia, Common-Law Courts in a Civil-Law System: The Role of United States Federal Courts in Interpreting the Constitution and Laws, in A Matter of Interpretation: Federal Courts and the Law 3, 17 (Amy Gutmann ed., 1997).
 Scalia & Garner, supra note 12, at 33.
 Id. at 34.
 Id. at 57.
 See Stanley Fish, Intention and the Canons of Legal Interpretation, N.Y. Times Opinionator (July 16, 2012, 9:00 PM), http://opinionator.blogs.nytimes.com/2012/07/16/intention-and-the-canons-of-legal-interpretation (link).
 Even those claims are dubious. See William N. Eskridge, Jr., The New Textualism and Normative Canons, 113 Colum. L. Rev. 531 (2013) (book review) (link); Koppelman, supra note 13. They, however, do not involve the radical error of imagining contextless interpretation. They also are not relevant here. DOMA has some pretty nasty legislative history, but it is not necessary to rely upon it to find an equal protection violation.
 Id. at 980 n.1 (Scalia, J., concurring in the judgment in part and dissenting in part).
 In a 1990 opinion, he tried to supplement his reading by reference to another text: “In my view the Fourteenth Amendment’s requirement of ‘equal protection of the laws,’ combined with the Thirteenth Amendment’s abolition of the institution of black slavery, leaves no room for doubt that laws treating people differently because of their race are invalid.” Rutan v. Republican Party of Ill., 497 U.S. 62, 95 n.1 (1990) (Scalia, J., dissenting) (link). But of course the Thirteenth Amendment does not mention race either.
 See generally Theodore Brantner Wilson, The Black Codes of the South (1965). A purely semantic originalism that ignored this context would be oblivious to all considerations of race: in interpreting the meaning of “equal protection of the laws” in the Fourteenth Amendment, we would look up each of the words to discern its meaning in 1868, but we would not be allowed to notice that those words had anything to do with the mistreatment of the former slaves, since that is not part of their dictionary meaning. See Andrew Koppelman, Originalism, Abortion, and the Thirteenth Amendment, 112 Colum. L. Rev. 1917, 1924–25 (2012) (link). That obviously is not Scalia’s position.
 Brief and Appendix on Behalf of Appellee at 41, Loving, 388 U.S. 1 (No. 395).
 Brief for Appellants at 37, Loving, 388 U.S. 1 (No. 395).
 Loving, 388 U.S. at 11.
 Id. at 11 n.11.
 Scalia & Garner, supra note 12, at 16.
 Id. at 527. A more complex line of reasoning, not pertinent here, is necessary to invalidate the fourth statute. That same line provides an independent basis for DOMA’s invalidity. See Andrew Koppelman, Dumb and DOMA: Why the Defense of Marriage Act Is Unconstitutional, 83 Iowa L. Rev. 1, 31 (1997).
 Church of the Lukumi Babalu Aye, 508 U.S. at 541–42 (internal quotation marks omitted).
 Id. at 534.
 Transcript of Oral Argument at 20–21, Church of the Lukumi Babalu Aye, 508 U.S. 520 (No. 91-948). (Thanks to Professor Laycock for confirming that the questioner was Scalia.)
 Id. at 21.
 Church of the Lukumi Babalu Aye, 508 U.S. at 557 (Scalia, J., concurring in part and concurring in the judgment).
 Scalia & Garner, supra note 12, at 252.
 United States v. Windsor, 133 S. Ct. 2675, 2707–08 (2013) (Scalia, J., dissenting).
 Id. at 2707.
 Id. at 2696 (Roberts, C.J., dissenting).
 Id. at 2708 (Scalia, J., dissenting).
 Id. Roberts likewise cites “[i]nterests in uniformity and stability.” Id. at 2696 (Roberts, C.J., dissenting).
 For example, the Social Security Act states that “[a]n applicant is the wife, husband, widow, or widower . . . if the courts of the State in which such insured individual is domiciled . . . would find that such applicant and such insured individual were validly married.” 42 U.S.C. § 416(h)(1)(A)(i) (2006) (link). The Veterans’ Benefits Act directs that “[i]n determining whether or not a person is or was the spouse of a veteran, their marriage shall be proven as valid . . . according to the law of the place where the parties resided at the time of the marriage or the law of the place where the parties resided when the right to benefits accrued.” 38 U.S.C. § 103(c) (2006) (link); see also 29 C.F.R. §§ 825.122(a), 825.800 (2010) (defining spouse under Family and Medical Leave Act as “a husband or wife as defined or recognized under State law for purposes of marriage in the State where the employee resides”) (link) (link).
 I have been on the lookout for them since I wrote Same Sex, Different States: When Same-Sex Marriages Cross State Lines (2006). Because same-sex couples’ strongest claims will be based on unfair surprise, this is not a kind of test case that can be planned. See Andrew Koppelman, The Limits of Strategic Litigation, 17 Law & Sexuality 1 (2008) (link).
 See 17 U.S.C. § 304(a)(C)(ii) (2012) (link); see also Brad A. Greenberg, DOMA’s Ghost and Copyright Reversionary Interests, 108 Nw. U. L. Rev. Colloquy 102 (2013), http://www.law.
 See, e.g., Julia Preston, Bill Proposes Immigration Rights for Gay Couples, N.Y. Times (June 2, 2009), http://www.nytimes.com/2009/06/03/us/politics/03immig.html (link); Editorial, Reunite This Family, Boston Globe (Aug. 27, 2007), http://www.boston.com/news/globe/editorial_opinion/
 Burdens on third parties, other than those facially targeted by a statute, are not normally a part of equal protection analysis. They are, however, relevant to an assessment of a law’s rationality. Thanks to Bob Bennett for helpful conversation on this issue.
 Tara Siegel Bernard, Yale Payroll Error Gives Gay Employees a New Year Surprise, N.Y. Times Bucks Blog (Jan. 11, 2011, 1:45 PM), http://bucks.blogs.nytimes.com/2011/01/11/yale-payroll-error-costs-gay-employees-thousands (link).
 See Koppelman, Same Sex, Different States, supra note 52, at 82–96.
 See id. at 28–50.
 Windsor, 133 S. Ct. at 2708 (Scalia, J., dissenting).
 Id. at 2693 (majority opinion).
 Id. at 2720 (Alito, J., dissenting).
 It is unclear from the legislative history whether Congress intended to influence states’ deliberations on whether to recognize same-sex marriages. The House Judiciary Committee report says repeatedly that each state will remain free to decide this policy issue for itself. See, e.g., H.R. Rep. No. 104–664, at 24–30 (1996), reprinted in 1996 U.S.C.C.A.N. 2905, 2928–34 (link). The report also indicates, however, that the legislation will provide “assistance” to those states that have no declared public policy against recognition of same-sex marriage. See id. at 10 & n.33, reprinted in 1996 U.S.C.C.A.N. at 2914.
 That is one reason why Paul Clement’s argument, on behalf of the House Republicans—nonrecognition is rational because it somehow makes heterosexual couples more likely to marry when unexpected pregnancy occurs—makes so little sense that it was ignored even by Justices Scalia and Alito. See Brief on the Merits for Respondent the Bipartisan Legal Advisory Group of the U.S. House of Representatives, supra note 65, at 44–47. Even if this convoluted causal chain is accepted, DOMA does not prevent any same-sex marriages from occurring. For the same reason, even if one accepts Alito’s view that there is a legitimate moral view that objects to same-sex marriage as such, that premise cannot justify DOMA.
 Windsor, 133 S. Ct. at 2694.
 See Koppelman, supra note 11, at 21–25; Andrew Koppelman, Romer v. Evans and Invidious Intent, 6 Wm. & Mary Bill Rts. J. 89, 123–130 (1997) (link). The law also impermissibly discriminates on the basis of sex. Andrew Koppelman, Beyond Levels of Scrutiny: Windsor and “Bare Desire to Harm,” 64 Case W. Res. L. Rev.(forthcoming 2014).
 See Andrew Koppelman, Gaze in the Military: A Response to Professor Woodruff, 64 UMKC L. Rev. 179 (1995) (criticizing “don’t ask, don’t tell”).
 Scalia & Garner, supra note 12, at 33.
 Windsor, 133 S. Ct. at 2696.
 The bill passed the House by a vote of 342–67 on July 12, 1996. 142 Cong. Rec. 17,094 (1996) (link). It passed the Senate by a vote of 85–14 on September 10, 1996. 142 Cong. Rec. 22,467 (1996) (link).
 See Baehr v. Lewin, 852 P.2d 44, 68 (Haw. 1993) (remanding for trial on whether state can rebut presumption of unconstitutionality under strict scrutiny).
 See Michael J. Klarman, From the Closet to the Altar: Courts, Backlash, and the Struggle for Same-Sex Marriage 61–63 (2013). Representative Barney Frank observed at a hearing on DOMA that the concern of its sponsors “is not what the States decide to do with regard to marriage; they’re worried about how the State decides to allocate its electoral votes, and this is an effort to influence not marriage in the States, but whether the Democratic or Republican tickets win.” Defense of Marriage Act: Hearing Before the Subcomm. on the Constitution of the Comm. on the Judiciary, H.R, 104th Cong., 2d Sess. 6 (1996).
 The President signed the bill at 12:50 AM on September 21, 1996. Peter Baker, President Quietly Signs Law Aimed at Gay Marriages, Wash. Post, Sept. 22, 1996, at A21.
 That lashing out is even more evident in the clumsily drafted choice of law provision, which produces no changes in existing laws that are not unconstitutional. See Koppelman, supra note 38, at 10–24.
 Sarah Trumble & Lanae Erickson Hatalsky, Commitment Campaign, Third Way, The State of Relationship Recognition in 2013, at 1 (2013), available at http://www.thirdway.org/
 On the basis of 2010 census data, the Williams Institute estimates that 131,729 same-sex couples in the United States were married under the law of their states. Gary J. Gates & Abigail M. Cooke, Williams Inst., United States Census Snapshot: 2010, at 1, available at http://williamsinstitute.
 Trumble & Hatalsky, supra note 87, at 1. Of these, some are in a city or county that has domestic partnerships, with no statewide protections. Id. at 2.
 Paul Brest, Foreword: In Defense of the Antidiscrimination Principle, 90 Harv. L. Rev. 1, 7–8 (1976).
 John Hart Ely, Democracy and Distrust: A Theory of Judicial Review 223 n.33 (1980). The theory of equal respect in decisionmaking that these authors rely upon is elaborated on and defended in Andrew Koppelman, Antidiscrimination Law and Social Equality 13–56 (1996), and Koppelman, Romer v. Evans and Invidious Intent, supra note 78, at 101–03.
 Some members, of both political parties, offered delusional estimates of the costs. Senator Phil Gramm warned that the “failure to pass this bill . . . will create . . . a whole group of new beneficiaries—no one knows what the number would be—tens of thousands, hundreds of thousands, potentially more—who will be beneficiaries of newly created survivor benefits under Social Security, Federal retirement plans, and military retirement plans.” 142 Cong. Rec. S10106 (daily ed. Sept. 10, 1996) (link). Senator Robert Byrd said that he did “not think . . . that it is inconceivable that the costs associated with such a change could amount to hundreds of millions of dollars, if not billions—if not billions—of Federal taxpayer dollars.” Id. at S10111. As it turns out, the fiscal consideration cuts the other way: federal recognition of same-sex marriage would produce a modest increase in federal revenue, amounting to a bit less than $400 million annually from 2005 to 2010. See Letter and Report from Douglas Holtz-Eakin, Dir., Cong. Budget Office, to Honorable Steve Chabot, Chairman, Subcomm. on the Constitution, Comm. on the Judiciary 2 (June 21, 2004), available at http://www.cbo.gov/sites/default/files/cbofiles/
 Because linguistic meaning is contingent on background assumptions, symbolism can rationalize anything. See Koppelman, supra note 69, at 66–68. For many years, the criminalization of homosexual sex in the United States was justified on this basis: if the law stopped hunting down gays’ private sex acts, it was argued, this would implicitly send a message of approval. See William N. Eskridge, Jr., No Promo Homo: The Sedimentation of Antigay Discourse and the Channeling Effect of Judicial Review, 75 N.Y.U. L. Rev. 1327, 1341–44 (2000) (link).
 17 Thomas Aquinas, Summa Theologiæ Ia2æ, Q. 6, art. 8, at 33 (Blackfriars 1970).
 Model Penal Code § 210.2(1)(b) (1981).
 Commonwealth v. Malone, 47 A.2d 445, 449 (Pa. 1946).
 2 Wayne R. LaFave, Substantive Criminal Law § 14.4(a), at 439 (2d ed. 2003).
 Scalia, supra note 15, at 38.
 Antonin Scalia, Response, in A Matter of Interpretation: Federal Courts and the Law, supra note 15, at 129, 144.
 See supra and infra text accompanying notes 8, 45–46, 48–49, 71, and 121.
 For attempts to comprehensively survey his equal protection jurisprudence, see Richard A. Brisbin, Jr., Justice Antonin Scalia and the Conservative Revival 150–80 (1997); Ralph A. Rossum, Antonin Scalia’s Jurisprudence: Text and Tradition 157–64 (2006); David A. Schultz & Christopher E. Smith, The Jurisprudential Vision of Justice Antonin Scalia 41–48 (1996).
 Rutan v. Republican Party of Ill., 497 U.S. 62, 95 n.1 (1990) (Scalia, J., dissenting).
 “A racially discriminatory purpose is always sufficient to subject a law to strict scrutiny, even a facially neutral law that makes no mention of race.” Lawrence v. Texas, 539 U.S. 558, 600 (2003) (Scalia, J., dissenting) (link); see also Washington v. Davis, 426 U.S. 229, 241–242 (1976) (link).
 This view is made clear in a short unpublished memo circulated during the deliberations in McCleskey v. Kemp, 481 U.S. 279 (1987) (link), in which a prisoner challenged his death sentence by citing a massive study showing that in Georgia, where the crime occurred, murderers of whites were four times as likely to be sentenced to death as murderers of blacks. The Court, in which Scalia joined, rejected the challenge, because the prisoner had not shown that “the decisionmakers in his case acted with discriminatory purpose.” Id. at 292. Scalia wrote:
I disagree with the argument that the inferences that can be drawn from the Baldus study are weakened by the fact that each jury and each trial is unique, or by the large number of variables at issue. And I do not share the view, implicit in the opinion, that an effect of racial factors upon sentencing, if it could only be shown by sufficiently strong statistical evidence, would require reversal. Since it is my view that the unconscious operation of irrational sympathies and antipathies, including racial, upon jury decisions and (hence) prosecutorial decisions is real, acknowledged in the decisions of this court, and ineradicable, I cannot honestly say that all I need is more proof.
Memorandum from Justice Antonin Scalia to the Conference, McCleskey v. Kemp (No. 84-6811) (Jan. 6, 1987), available at http://law.wlu.edu/deptimages/powell%20archives/McCleskeyKempBasic.pdf (p. 147) (link). Scalia’s equal protection clause is one in which even racially biased imposition of the death penalty is acceptable, so long as it does not overtly or intentionally classify on the basis of race. This indifference to unconscious racism is hard to reconcile with any strain of equal protection theory. See Charles R. Lawrence III, The Id, the Ego, and Equal Protection: Reckoning with Unconscious Racism, 39 Stan. L. Rev. 317 (1987). Scalia indicated in his memo that he would write separately, but never did so, for reasons he later said he could not remember. Joan Biskupic, American Original: The Life and Constitution of Supreme Court Justice Antonin Scalia 153 (2009).
 Cruzan v. Dir., Mo. Dep’t of Health, 497 U.S. 261, 300 (1990) (Scalia, J., concurring) (link). Laurence Tribe’s amicus brief in Romer quoted this passage. Brief of Laurence H. Tribe et al. as Amici Curiae in Support of Respondents at 5, Romer v. Evans, 517 U.S. 620 (1996) (No. 94-1039). Scalia ignored it.
I give little weight to Scalia’s joinder of the per curiam opinion in Bush v. Gore, 531 U.S. 98 (2000) (link), which held that the Equal Protection Clause guarantees to individuals that their ballots cannot be devalued by “later arbitrary and disparate treatment.” Id.at 104. He also wrote separately, and may have joined the equal protection holding only because it was important that there be a majority opinion. Biskupic, supra note 106, at 243.
 Antonin Scalia, The Rule of Law as a Law of Rules,56 U. Chi. L. Rev. 1175, 1178 (1989). He has expanded on the point:
Consistency is the very foundation of the rule of law. If you go through our Bill of Rights, I daresay it does not contain a single provision that various cultures, in various ages, have not in principle rejected . . . . But you will search long and hard to find anyone, in any age, who would reject the fundamental principle underlying the equal protection clause: that persons similarly situated should be similarly treated—that is to say, the principle that the law must be consistent.
Antonin Scalia, Assorted Canards of Contemporary Legal Analysis,40 Case W. Res. L. Rev. 581, 588 (1989–90).
 An exception is his approval of Williams v. Illinois, 399 U.S. 235 (1970) (link), which invalidated the centuries-old practice of permitting convicted criminals to reduce their prison sentences by paying fines. Scalia argued that no practice with a long historical pedigree could violate due process, and distinguished Williams thus:
The basis of that invalidation was not denial of due process but denial to indigent prisoners of equal protection of the laws. The Equal Protection Clause and other provisions of the Constitution, unlike the Due Process Clause, are not an explicit invocation of the “law of the land,” and might be thought to have some counterhistorical content.
Pac. Mut. Life Ins. Co. v. Haslip, 499 U.S. 1, 38 (1991) (Scalia, J., concurring in the judgment) (link). The general theory of equal protection that entails Williams was not further specified.
 Bob Egelko, Constitution Does Not Ban Sex Bias, Scalia Says, S.F. Chron. (Sept. 18, 2010, 4:00 AM), http://www.sfgate.com/news/article/Constitution-does-not-ban-sex-bias-Scalia-says-3252922.php (internal quotation marks omitted) (link).
 Melissa L. Saunders, Equal Protection, Class Legislation, and Colorblindness, 96 Mich. L. Rev. 245, 247–48 (1997). The idea of class legislation was forgotten in the twentieth century, but it is the original meaning. Jack Balkin notes this, in a defense of Windsor, in Windsor and the Constitutional Prohibition Against Class Legislation, Balkinization Blog (June 26, 2013, 1:05 PM), http://balkin.blogspot.com/2013/06/windsor-and-constitutional-prohibition.html (link). Balkin elaborates on his historical evidence in Living Originalism 220–55 (2011). For more such evidence, see Steven G. Calabresi & Julia T. Rickert, Originalism and Sex Discrimination, 90 Tex. L. Rev. 1, 29–46 (2011) (link).
 See Wilson, supra note 26, at 69–70, 76–79, 97–98, 100–03, 105, 107–11, 113.
 Cruzan v. Dir., Mo. Dep’t of Health, 497 U.S. 261, 300 (1990) (Scalia, J., concurring).
 U.S. Dep’t of Agric. v. Moreno, 413 U.S. 528, 534 (1973). I develop this analysis in greater detail in Beyond Levels of Scrutiny: Windsor and “Bare Desire to Harm,” supra note 78.
 Id. at 632.
 Id. at 635.
 Id. at 636 (Scalia, J., dissenting).
 See Robert W. Bennett, “Mere” Rationality in Constitutional Law: Judicial Review and Democratic Theory, 67 Calif. L. Rev. 1049, 1056–57 (1979) (link); Joseph Tussman & Jacobus tenBroek, The Equal Protection of the Laws, 37 Calif. L. Rev. 341, 345–47 (1949) (link); Note, Legislative Purpose, Rationality, and Equal Protection, 82 Yale L.J. 123, 124 (1972) (written by Robert F. Nagel).
 See Bennett, supra note 122, at 1078.
 David Hume, A Treatise of Human Nature 416 (L.A. Selby-Bigge ed., 2d ed. 1978).
 LaFave, supra note 99, § 14.4(b), at 441.
 Scalia & Garner, supra note 12, at 391.
 See supra text accompanying note 99.
 Brown v. Commonwealth, 17 S.W. 220, 221 (Ky. 1891).
 The Court explains:
[W]henever the government treats any person unequally because of his or her race, that person has suffered an injury that falls squarely within the language and spirit of the Constitution’s guarantee of equal protection. . . . The application of strict scrutiny . . . determines whether a compelling governmental interest justifies the infliction of that injury.
Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 229–30 (1995) (link).
 Id. at 397–98.
 Most of the time, anyway. Its hostility to racial classifications, when these could not possibly reflect biased decisionmaking, frankly rests on policy objections about the bad consequences of using such classifications. See Parents Involved in Cmty. Schs. v. Seattle Sch. Dist. No. 1, 551 U.S. 701, 746 (2007) (link).
 Even Ely conceded this. Ely, supra note 91, at 147–48.
 State v. Nueslein, 25 Mo. 111, 121 (1857).
 Yet another gay rights case in which this occurred is Boy Scouts of America v. Dale, 530 U.S. 640 (2000) (link), which invalidated the application to the Scouts of an antidiscrimination statute that made no specific reference to speech. For many years, Scalia has championed the view that laws of general applicability, not directed at First Amendment interests, raise no First Amendment issue even if in some applications they happen to restrict First Amendment activity. Scalia joined the Dale majority without explanation. See generally Stephen Clark, Judicially Straight? Boy Scouts v. Dale and the Missing Scalia Dissent, 76 S. Cal. L. Rev. 521 (2003) (showing that Scalia’s vote in the majority in Boy Scouts was inconsistent with his other jurisprudence) (link).
 United States v. Windsor, 133 S. Ct. 2675, 2718 (2013) (Alito, J., dissenting).
 For critique, see Andrew Koppelman, Judging the Case Against Same-Sex Marriage, 2014 U. Ill. L. Rev. (forthcoming); Andrew Koppelman, More Intuition than Argument, Commonweal, May 3, 2013, at 23 (reviewing Sherif Girgis et al., What is Marriage? Man and Woman: A Defense (2012)) (link).
 See Girgis et al., supra note 138, at 10–12, 86–93.
 Romer v. Evans, 517 U.S. 620, 652 (1996) (Scalia, J., dissenting).
 See Koppelman, supra note 11, at 89–90.
 This, I argue elsewhere, is the core commitment of American religious liberty. See generally Andrew Koppelman, Defending American Religious Neutrality (2013).
Copyright 2013 by Andrew Koppelman
Cite as: 108 Nw. U. L. Rev. Colloquy 131 (2013), http://www.law.northwestern.edu/lawreview/colloquy/2013/12/LRColl2013n12Koppelman.pdf.