POSNER, Circuit Judge.
Indiana and Wisconsin are among the shrinking majority of states that do not recognize the validity of same-s** marriages, whether contracted in these states or in states (or foreign countries) where they are lawful. The states have appealed from district court decisions invalidating the states' laws that ordain such refusal.
Formally these cases are about discrimination against the small h*mos**ual minority in the United States. But at a deeper level, as we shall see, they are about the welfare of American children. The argument that the states press hardest in defense of their prohibition of same-s** marriage is that the only reason government encourages marriage is to induce heteros**uals to marry so that there will be fewer “accidental births,” which when they occur outside of marriage often lead to abandonment of the child to the mother (unaided by the father) or to foster care. Overlooked by this argument is that many of those abandoned children are adopted by h*mos**ual couples, and those children would be better off both emotionally and economically if their adoptive parents were married.
We are mindful of the Supreme Court's insistence that “whether embodied in the Fourteenth Amendment or inferred from the Fifth, equal protection is not a license for courts to judge the wisdom, fairness, or logic of legislative choices. In areas of social and economic policy, a statutory cla**ification that neither proceeds along suspect lines nor infringes fundamental constitutional rights must be upheld against equal protection challenge if there is any reasonably conceivable state of facts that could provide a rational basis for the cla**ification.” FCC v. Beach Communications, Inc., 508 U.S. 307, 313 (1993) (emphasis added). The phrase we've italicized is the exception applicable to this pair of cases.
We hasten to add that even when the group discriminated against is not a “suspect cla**,” courts examine, and sometimes reject, the rationale offered by government for the challenged discrimination. See, e.g., Village of Willowbrook v. Olech, 528 U.S. 562 (2000) (per curiam); City of Cleburne v. Cleburne Living Center, 473 U.S. 432, 448–50 (1985). In Vance v. Bradley, 440 U.S. 93, 111 (1979), an illustrative case in which the Supreme Court accepted the government's rationale for discriminating on the basis of age, the majority opinion devoted 17 pages to an*lyzing whether Congress had had a “reasonable basis” for the challenged discrimination (requiring foreign service officers but not ordinary civil servants to retire at the age of 60), before concluding that it did.
We'll see that the governments of Indiana and Wisconsin have given us no reason to think they have a “reasonable basis” for forbidding same-s** marriage. And more than a reasonable basis is required because this is a case in which the challenged discrimination is, in the formula from the Beach case, “along suspect lines.” Discrimination by a state or the federal government against a minority, when based on an immutable characteristic of the members of that minority (most familiarly skin color and gender), and occurring against an historical background of discrimination against the persons who have that characteristic, makes the discriminatory law or policy constitutionally suspect. See, e.g., Bow- en v. Gilliard, 483 U.S. 587, 602–03 (1987); Regents of University of California v. Bakke, 438 U.S. 265, 360–62 (1978); St. John's United Church of Christ v. City of Chicago, 502 F.3d 616, 638 (7th Cir. 2007); Wilkins v. Gaddy, 734 F.3d 344, 348 (4th Cir. 2013); Gallagher v. City of Clayton, 699 F.3d 1013, 1018–19 (8th Cir. 2012). These circumstances create a presumption that the discrimination is a denial of the equal protection of the laws (it may violate other provisions of the Constitution as well, but we won't have to consider that possibility). The presumption is rebu*table, if at all, only by a compelling showing that the benefits of the discrimination to society as a whole clearly outweigh the harms to its victims. See, e.g., Grutter v. Bollinger, 539 U.S. 306, 326–27 (2003); United States v. Virginia, 518 U.S. 515, 531–33 (1996).
The approach is straightforward but comes wrapped, in many of the decisions applying it, in a formidable doctrinal terminology—the terminology of rational basis, of strict, heightened, and intermediate scrutiny, of narrow tailoring, fundamental rights, and the rest. We'll be invoking in places the conceptual apparatus that has grown up around this terminology, but our main focus will be on the states' arguments, which are based largely on the a**ertion that banning same-s** marriage is justified by the state's interest in channeling procreative s** into (necessarily heteros**ual) marriage. We will engage the states' arguments on their own terms, enabling us to decide our brace of cases on the basis of a sequence of four questions:
1. Does the challenged practice involve discrimination, rooted in a history of prejudice, against some identifiable group of persons, resulting in unequal treatment harmful to them?
2. Is the unequal treatment based on some immutable or at least tenacious characteristic of the people discriminated against (biological, such as skin color, or a deep psychological commitment, as religious belief often is, both types being distinct from characteristics that are easy for a person to change, such as the length of his or her fingernails)? The characteristic must be one that isn't relevant to a person's ability to participate in society. Intellect, for example, has a large immutable component but also a direct and substantial bearing on qualifications for certain types of employment and for legal privileges such as entitlement to a driver's li- cense, and there may be no reason to be particularly suspicious of a statute that cla**ifies on that basis.
3. Does the discrimination, even if based on an immutable characteristic, nevertheless confer an important offsetting benefit on society as a whole? Age is an immutable characteristic, but a rule prohibiting persons over 70 to pilot airliners might reasonably be thought to confer an essential benefit in the form of improved airline safety.
4. Though it does confer an offsetting benefit, is the dis- criminatory policy overinclusive because the benefit it confers on society could be achieved in a way less harmful to the discriminated-against group, or underinclusive because the government's purported rationale for the policy implies that it should equally apply to other groups as well? One way to decide whether a policy is overinclusive is to ask whether unequal treatment is essential to attaining the de- sired benefit. Imagine a statute that imposes a $2 tax on women but not men. The proceeds from that tax are, let's a**ume, essential to the efficient operation of government. The tax is therefore socially efficient, and the benefits clearly outweigh the costs. But that's not the end of the inquiry. Still to be determined is whether the benefits from imposing the tax only on women outweigh the costs. And likewise in a same-s** marriage case the issue is not whether heteros**ual marriage is a socially beneficial institution but whether the benefits to the state from discriminating against same-s** couples clearly outweigh the harms that this discrimination imposes.
Our questions go to the heart of equal protection doc- trine. Questions 1 and 2 are consistent with the various formulas for what entitles a discriminated-against group to heightened scrutiny of the discrimination, and questions 3 and 4 capture the essence of the Supreme Court's approach in heightened-scrutiny cases: “To succeed, the defender of the challenged action must show ‘at least that the cla**ification serves important governmental objectives and that the discriminatory means employed are substantially related to the achievement of those objectives.'” United States v. Virginia, supra, 518 U.S. at 524 (1996), quoting Mississippi University for Women v. Hogan, 458 U.S. 718, 724 (1982).
The difference between the approach we take in these two cases and the more conventional approach is semantic rather than substantive. The conventional approach doesn't purport to balance the costs and benefits of the challenged discriminatory law. Instead it evaluates the importance of the state's objective in enacting the law and the extent to which the law is suited (“tailored”) to achieving that objective. It asks whether the statute actually furthers the interest that the state a**erts and whether there might be some less burdensome alternative. The an*lysis thus focuses not on “costs” and “benefits” as such, but on “fit.” That is why the briefs in these two cases overflow with debate over whether prohibiting same-s** marriage is “over- or underinclu- sive”—for example, overinclusive in ignoring the effect of the ban on the children adopted by same-s** couples, under- inclusive in extending marriage rights to other non- procreative couples. But to say that a discriminatory policy is overinclusive is to say that the policy does more harm to the members of the discriminated-against group than necessary to attain the legitimate goals of the policy, and to say that the policy is underinclusive is to say that its exclusion of other, very similar groups is indicative of arbitrariness.
Although the cases discuss, as we shall be doing in this opinion, the harms that a challenged statute may visit upon the discriminated-against group, those harms don't formally enter into the conventional an*lysis. When a statute discriminates against a protected cla** (as defined for example in our question 2), it doesn't matter whether the harm inflicted by the discrimination is a grave harm. As we said, a statute that imposed a $2 tax on women but not men would be struck down unless there were a compelling reason for the discrimination. It wouldn't matter that the harm to each per- son discriminated against was slight if the benefit of imposing the tax only on women was even slighter.
Our pair of cases is rich in detail but ultimately straight- forward to decide. The challenged laws discriminate against a minority defined by an immutable characteristic, and the only rationale that the states put forth with any conviction— that same-s** couples and their children don't need marriage because same-s** couples can't produce children, intended or unintended—is so full of holes that it cannot be taken seriously. To the extent that children are better off in families in which the parents are married, they are better off whether they are raised by their biological parents or by adoptive parents. The discrimination against same-s** couples is irrational, and therefore unconstitutional even if the discrimination is not subjected to heightened scrutiny, which is why we can largely elide the more complex an*lysis found in more closely balanced equal-protection cases.
It is also why we can avoid engaging with the plaintiffs' further argument that the states' prohibition of same-s** marriage violates a fundamental right protected by the due process clause of the Fourteenth Amendment. The plaintiffs rely on cases such as Hodgson v. Minnesota, 497 U.S. 417, 435 (1990), and Zablocki v. Redhail, 434 U.S. 374, 383–86 (1978), that hold that the right to choose whom to marry is indeed a fundamental right. The states reply that the right recognized in such cases is the right to choose from within the cla** of persons eligible to marry, thus excluding children, close relatives, and persons already married—and, the states contend, persons of the same s**. The plaintiffs riposte that there are good reasons for ineligibility to marry children, close relatives, and the already married, but not for ineligibility to marry persons of the same s**. In light of the compelling alternative grounds that we'll be exploring for allowing same- s** marriage, we won't have to engage with the parties' “fundamental right” debate; we can confine our attention to equal protection.
We begin our detailed an*lysis of whether prohibiting same-s** marriage denies equal protection of the laws by noting that Indiana and Wisconsin, in refusing to authorize such marriage or (with limited exceptions discussed later) to recognize such marriages made in other states by residents of Indiana or Wisconsin, are discriminating against h*mos**uals by denying them a right that these states grant to heteros**uals, namely the right to marry an unmarried adult of their choice. And there is little doubt that s**ual orientation, the ground of the discrimination, is an immutable (and probably an innate, in the sense of in-born) characteristic rather than a choice. Wisely, neither Indiana nor Wisconsin argues otherwise. The American Psychological Association has said that “most people experience little or no sense of choice about their s**ual orientation.” APA, “Answers to Your Questions: For a Better Understanding of Sexual Orientation & h*mos**uality” 2 (2008), (visited Sept. 2, 2014, as were the other websites cited in this opinion); see also Gregory M. Herek et al., “Demographic, Psychological, and Social Characteristics of Self- Identified Lesbian, Gay, and Bis**ual Adults in a US Probability Sample,” 7 Sexuality Research and Social Policy 176, 188 (2010) ("combining respondents who said they'd had a small amount of choice with those reporting no choice, 95% of gay men and 84% of lesbians could be characterized as perceiving that they had little or no choice about their s**ual orientation"). that h*mos**ual orientation is not a choice is further suggested by the absence of evidence (despite extensive efforts to find it) that psychotherapy is effective in altering s**ual orientation in general and h*mos**ual orientation in particular. APA, “Answers to Your Questions,” supra, at 3; Report of the American Psychological Association Task Force on Appropriate Therapeutic Responses to Sexual Orientation 35–41 (2009).
The leading scientific theories of the causes of h*mos**uality are genetic and neuroendocrine theories, the latter be- ing theories that s**ual orientation is shaped by a fetus's exposure to certain hormones. See, e.g., J. Michael Bailey, “Bio- logical Perspectives on Sexual Orientation,” in Lesbian, Gay, and Bis**ual Identities Over the Lifespan: Psychological Perspec- tives 102–30 (Anthony R. D'Augelli and Charlotte J. Patter- son eds. 1995); Barbara L. Frankowski, “Sexual Orientation and Adolescents,” 113 Pediatrics 1827, 1828 (2004). Although it seems paradoxical to suggest that h*mos**uality could have a genetic origin, given that h*mos**ual s** is non- procreative, h*mos**uality may, like menopause, by reducing procreation by some members of society free them to provide child-caring a**istance to their procreative relatives, thus increasing the survival and hence procreative prospects of these relatives. This is called the “kin selection hypothesis” or the “helper in the nest theory.” See, e.g., Association for Psychological Science, ”Study Reveals Potential Evolutionary Role for Same-Sex Attraction,” Feb. 4, 2010, www.psychologicalscience.org/media/releases/2010/vasey.cf m. There are other genetic theories of such attraction as well. See, e.g., Nathan W. Bailey and Marlene Zuk, “Same-Sex Sexual Behavior and Evolution,” forthcoming in Trends in Ecology and Evolution,. For a responsible popular treatment of the subject see William Kremer, “The Evolutionary Puzzle of h*mos**uality,” BBC News Magazine, Feb. 17, 2014,.
The harm to h*mos**uals (and, as we'll emphasize, to their adopted children) of being denied the right to marry is considerable. Marriage confers respectability on a s**ual relationship; to exclude a couple from marriage is thus to deny it a coveted status. Because h*mos**uality is not a voluntary condition and h*mos**uals are among the most stigmatized, misunderstood, and discriminated-against minorities in the history of the world, the disparagement of their s**ual orientation, implicit in the denial of marriage rights to same-s** couples, is a source of continuing pain to the h*mos**ual community. Not that allowing same-s** marriage will change in the short run the negative views that many Americans hold of same-s** marriage. But it will enhance the status of these marriages in the eyes of other Americans, and in the long run it may convert some of the opponents of such marriage by demonstrating that h*mos**ual married couples are in essential respects, notably in the care of their adopted children, like other married couples.
The tangible as distinct from the psychological benefits of marriage, which (along with the psychological benefits) enure directly or indirectly to the children of the marriage, whether biological or adopted, are also considerable. In Indiana they include the right to file state tax returns jointly, Ind. Code § 6-3-4-2(d); the marital testimonial privilege, § 34- 46-3-1(4); spousal-support obligations, § 35-46-1-6(a); survivor benefits for the spouse of a public safety officer k**ed in the line of duty, § 36-8-8-13.8(c); the right to inherit when a spouse dies intestate, § 29-1-2-1(b), (c); custodial rights to and child support obligations for children of the marriage, and protections for marital property upon the d**h of a spouse. §§ 12-15-8.5-3(1); 12-20-27-1(a)(2)(A). Because Wisconsin allows domestic partnerships, some spousal benefits are available to same-s** couples in that state. But others are not, such as the right to adopt children jointly, Wis. Stat. § 48.82(1); spousal-support obligations, §§ 765.001(2), 766.15(1), 766.55; the presumption that all property of married couples is marital property, § 766.31(2); and state-mandated access to enrollment in a spouse's health insurance plan, § 632.746(7).
Of great importance are the extensive federal benefits to which married couples are entitled: the right to file income taxes jointly, 26 U.S.C. § 6013; social security spousal and surviving-spouse benefits, 42 U.S.C. § 402; d**h benefits for surviving spouse of a military veteran, 38 U.S.C. § 1311; the right to transfer a**ets to one's spouse during marriage or at divorce without additional tax liability, 26 U.S.C. § 1041; ex- emption from federal estate tax of property that pa**es to the surviving spouse, 26 U.S.C. § 2056(a); the tax exemption for employer-provided healthcare to a spouse, 26 U.S.C. § 106; Treas. Reg. § 1.106–1; and healthcare benefits for spouses of federal employees, 5 U.S.C. §§ 8901(5), 8905.
The denial of these federal benefits to same-s** couples brings to mind the Supreme Court's opinion in United States v. Windsor, 133 S. Ct. 2675, 2694–95 (2013), which held un- constitutional the denial of all federal marital benefits to same-s** marriages recognized by state law. The Court's criticisms of such denial apply with even greater force to Indiana's law. The denial “tells those couples, and all the world, that their otherwise valid marriages are unworthy of federal recognition. [No same-s** marriages are valid in Indiana.] This places same-s** couples in an unstable position of being in a second-tier marriage [in Indiana, in the lowest—the unmarried—tier]. The differentiation demeans the couple ... [and] humiliates tens of thousands of children now being raised by same-s** couples. The law ... makes it even more difficult for the children to understand the integrity and closeness of their own family and its concord with other families in their community and in their daily lives.” Id. at 2694.
The Court went on to describe at length the federal marital benefits denied by the Defense of Marriage Act to married same-s** couples. Of particular relevance to our two cases is the Court's finding that denial of those benefits causes economic harm to children of same-s** couples. “It raises the cost of health care for families by taxing health benefits provided by employers to their workers' same-s** spouses. And it denies or reduces benefits allowed to families upon the loss of a spouse and parent, benefits that are an integral part of family security. [The Act also] divests married same-s** couples of the duties and responsibilities that are an essential part of married life and that they in most cases would be honored to accept.” Id. at 2695 (citations omitted).
Of course there are costs to marriage as well as benefits, not only the trivial cost of the marriage license but also the obligations, such as alimony, that a divorcing spouse may be forced to bear. But those are among “the duties and responsibilities that are an essential part of married life and that [the spouses] in most cases would be honored to accept.” That marriage continues to predominate over cohabitation as a choice of couples indicates that on average the sum of the tangible and intangible benefits of marriage outweighs the costs.
In light of the foregoing an*lysis it is apparent that groundless rejection of same-s** marriage by government must be a denial of equal protection of the laws, and there- fore that Indiana and Wisconsin must to prevail establish a clearly offsetting governmental interest in that rejection. Whether they have done so is really the only issue before us, and the balance of this opinion is devoted to it—except that before addressing it we must address the states' argument that whatever the merits of the plaintiffs' claims, we are bound by Baker v. Nelson, 409 U.S. 810 (1972) (mem.), to reject them. For there the Supreme Court, without issuing an opinion, dismissed “for want of a substantial federal question” an appeal from a state court that had held that prohibiting same-s** marriage did not violate the Constitution. Alt- hough even a decision without opinion is on the merits and so binds lower courts, the Supreme Court carved an exception to this principle of judicial hierarchy in Hicks v. Miranda, 422 U.S. 332, 344 (1975), for “when doctrinal developments indicate otherwise”; see also United States v. Blaine County, 363 F.3d 897, 904 (9th Cir. 2004); Soto-Lopez v. New York City Civil Service Commission, 755 F.2d 266, 272 (2d Cir. 1985). Baker was decided in 1972—42 years ago and the dark ages so far as litigation over discrimination against h*mos**uals is concerned. Subsequent decisions such as Romer v. Evans, 517 U.S. 620, 634–36 (1996); Lawrence v. Texas, 539 U.S. 558, 577– 79 (2003), and United States v. Windsor are distinguishable from the present two cases but make clear that Baker is no longer authoritative. At least we think they're distinguishable. But Justice Scalia, in a dissenting opinion in Lawrence, 539 U.S. at 586, joined by Chief Justice Rehnquist and Justice Thomas, thought not. He wrote that “principle and logic” would require the Court, given its decision in Lawrence, to hold that there is a constitutional right to same-s** marriage. Id. at 605.
First up to bat is Indiana, which defends its refusal to al- low same-s** marriage on a single ground, namely that government's sole purpose (or at least Indiana's sole purpose) in making marriage a legal relation (unlike cohabitation, which is purely contractual) is to enhance child welfare. Notably the state does not argue that recognizing same-s** marriage undermines conventional marriage.
When a child is conceived intentionally, the parents normally intend to raise the child together. But pregnancy, and the resulting birth (in the absence of abortion), are some- times accidental, unintended; and often in such circumstances the mother is stuck with the baby—the father, not having wanted to become a father, refuses to take any responsibility for the child's welfare. The sole reason for Indiana's marriage law, the state's argument continues, is to try to channel unintentionally procreative s** into a legal regime in which the biological father is required to a**ume parental responsibility. The state recognizes that some or even many h*mos**uals want to enter into same-s** marriages, but points out that many people want to enter into relations that government refuses to enforce or protect (friendship being a no- table example). Government has no interest in recognizing and protecting same-s** marriage, Indiana argues, because h*mos**ual s** cannot result in unintended births.
As for the considerable benefits that marriage confers on the married couple, these in the state's view are a part of the regulatory regime: the carrot supplementing the stick. Marital benefits for h*mos**ual couples would not serve the regulatory purpose of marital benefits for heteros**ual couples because h*mos**ual couples don't produce babies.
The state's argument can be an*logized to requiring drivers' licenses for drivers of motor vehicles but not for bi- cyclists. Motor vehicles are more dangerous to other users of the roads than bicycles are, and therefore a driver's license is required to drive the former but not to pedal the latter. Bicyclists do not and cannot complain about not having to have a license to pedal, because obtaining, renewing, etc., the li- cense would involve a cost in time and money. The an*logy is not perfect (if it were, it would be an identity not an an*logy) because marriage confers benefits as well as imposing costs, as we have emphasized (indeed it confers on most couples benefits greater than the costs). But those benefits, in Indiana's view, would serve no state interest if extended to h*mos**ual couples, who should therefore be content with the benefits they derive from being excluded from the marriage-licensing regime: the cost of the license and the burden of marital duties, such as support, and the costs a**ociated with divorce. Moreover, even if possession of a driver's li- cense conferred benefits not available to bicyclists (dis- counts, or tax credits, perhaps), the state could argue that it offered these benefits only to induce drivers to obtain a li- cense (the carrot supplementing the stick), and that bicyclists don't create the same regulatory concern and so don't de- serve a carrot.
Another an*logy: The federal government extends a $2000 “saver's credit” to low- and middle-income workers who contribute to a retirement account. Although everyone would like a $2000 credit, only lower-income workers are entitled to it. Should higher-income workers complain about being left out of the program, the government could reply that only lower-income workers create a regulatory con- cern—the concern that they'd be unable to support them- selves in retirement without government encouragement to save while they're young.
In short, Indiana argues that h*mos**ual relationships are created and dissolved without legal consequences be- cause they don't create family-related regulatory concerns. Yet encouraging marriage is less about forcing fathers to take responsibility for their unintended children—state law has mechanisms for determining paternity and requiring the father to contribute to the support of his children—than about enhancing child welfare by encouraging parents to commit to a stable relationship in which they will be raising the child together. Moreover, if channeling procreative s** into marriage were the only reason that Indiana recognizes marriage, the state would not allow an infertile person to marry. Indeed it would make marriage licenses expire when one of the spouses (fertile upon marriage) became infertile because of age or disease. The state treats married h*mos**uals as would-be “free riders” on heteros**ual marriage, un- reasonably reaping benefits intended by the state for fertile couples. But infertile couples are free riders too. Why are they allowed to reap the benefits accorded marriages of fertile couples, and h*mos**uals are not?
The state offers an involuted pair of answers, neither of which answers the charge that its policy toward same-s** marriage is underinclusive. It points out that in the case of most infertile heteros**ual couples, only one spouse is infer- tile, and it argues that if these couples were forbidden to marry there would be a risk of the fertile spouse's seeking a fertile person of the other s** to breed with and the result would be “multiple relationships that might yield unintentional babies.” True, though the fertile member of an infertile couple might decide instead to produce a child for the couple by surrogacy or (if the fertile member is the woman) a s**m bank, or to adopt, or to divorce. But what is most unlikely is that the fertile member, though desiring a biological child, would have procreative s** with another person and then abandon the child—which is the state's professed fear.
The state tells us that “non-procreating opposite-s** couples who marry model the optimal, socially expected behavior for other opposite-s** couples whose s**ual intercourse may well produce children.” That's a strange argument; fertile couples don't learn about child-rearing from infertile couples. And why wouldn't same-s** marriage send the same message that the state thinks marriage of infertile heteros**uals sends—that marriage is a desirable state?
It's true that infertile or otherwise non-procreative heteros**ual couples (some fertile couples decide not to have children) differ from same-s** couples in that it is easier for the state to determine whether a couple is infertile by reason of being of the same s**. It would be considered an invasion of privacy to condition the eligibility of a heteros**ual couple to marry on whether both prospective spouses were fertile (although later we'll see Wisconsin flirting with such an approach with respect to another cla** of infertile couples). And often the couple wouldn't know in advance of marriage whether they were fertile. But then how to explain Indiana's decision to carve an exception to its prohibition against marriage of close relatives for first cousins 65 or older—a population guaranteed to be infertile because women can't conceive at that age? Ind. Code § 31-11-1-2. If the state's only interest in allowing marriage is to protect children, why has it gone out of its way to permit marriage of first cousins only after they are provably infertile? The state must think marriage valuable for something other than just procreation— that even non-procreative couples benefit from marriage. And among non-procreative couples, those that raise children, such as same-s** couples with adopted children, gain more from marriage than those who do not raise children, such as elderly cousins; elderly persons rarely adopt.
Indiana has thus invented an insidious form of discrimination: favoring first cousins, provided they are not of the same s**, over h*mos**uals. Elderly first cousins are permit- ted to marry because they can't produce children; h*mos**uals are forbidden to marry because they can't produce children. The state's argument that a marriage of first cousins who are past child-bearing age provides a “model [of] family life for younger, potentially procreative men and women” is impossible to take seriously.
At oral argument the state‘s lawyer was asked whether “Indiana's law is about successfully raising children,” and since “you agree same-s** couples can successfully raise children, why shouldn't the ban be lifted as to them?” The lawyer answered that “the a**umption is that with opposite- s** couples there is very little thought given during the s**ual act, sometimes, to whether babies may be a consequence.” In other words, Indiana's government thinks that straight couples tend to be s**ually irresponsible, producing unwanted children by the carload, and so must be pressured (in the form of governmental encouragement of marriage through a combination of sticks and carrots) to marry, but that gay couples, unable as they are to produce children wanted or unwanted, are model parents—model citizens re- ally—so have no need for marriage. Heteros**uals get drunk and pregnant, producing unwanted children; their reward is to be allowed to marry. h*mos**ual couples do not produce unwanted children; their reward is to be denied the right to marry. Go figure.
Which brings us to Indiana's weakest defense of its distinction among different types of infertile couple: its a**umption that same-s** marriage cannot contribute to alleviating the problem of “accidental births,” which the state contends is the sole governmental interest in marriage. Sup- pose the consequences of accidental births are indeed the state's sole reason for giving marriage a legal status. In advancing this as the reason to forbid same-s** marriage, Indiana has ignored adoption—an extraordinary oversight. Un- intentional offspring are the children most likely to be put up for adoption, and if not adopted, to end up in a foster home. Accidental pregnancies are the major source of un- wanted children, and unwanted children are a major problem for society, which is doubtless the reason h*mos**uals are permitted to adopt in most states—including Indiana and Wisconsin.
It's been estimated that more than 200,000 American children (some 3000 in Indiana and about the same number in Wisconsin) are being raised by h*mos**uals, mainly h*mos**ual couples. Gary J. Gates, “LGBT Parenting in the United States” 3 (Williams Institute, UCLA School of Law, Feb. 2013),; Gates, “Same-Sex Couples in Indiana: A Demographic Summary” (Williams Institute, UCLA School of Law, 2014),; Gates, “Same-Sex Couples in Wisconsin: A Demo- graphic Survey” (Williams Institute, UCLA School of Law, Aug. 2014),. Gary Gates's demographic surveys find that among couples who have children, h*mos**ual couples are five times as likely to be raising an adopted child as heteros**ual couples in Indiana, and two and a half times as likely as heteros**ual couples in Wisconsin.
If the fact that a child's parents are married enhances the child's prospects for a happy and successful life, as Indiana believes not without reason, this should be true whether the child's parents are natural or adoptive. The state's lawyers tell us that “the point of marriage's a**ociated benefits and protections is to encourage child-rearing environments where parents care for their biological children in tandem.” Why the qualifier “biological”? The state recognizes that family is about raising children and not just about producing them. It does not explain why the “point of marriage's a**ociated benefits and protections” is inapplicable to a couple's adopted as distinct from biological children.
Married h*mos**uals are more likely to want to adopt than unmarried ones if only because of the many state and federal benefits to which married people are entitled. And so same-s** marriage improves the prospects of unintended children by increasing the number and resources of prospective adopters. Notably, same-s** couples are more likely to adopt foster children than opposite-s** couples are. Gates, “LGBT Parenting in the United States,” supra, at 3. As of 2011, there were some 400,000 American children in foster care, of whom 10,800 were in Indiana and about 6500 in Wisconsin. U.S. Dept. of Health & Human Services, Children's Bureau, “How Many Children Are in Foster Care in the U.S.? In My State?”.
Also, the more willing adopters there are, not only the fewer children there will be in foster care or being raised by single mothers but also the fewer abortions there will be. Carrying a baby to term and putting the baby up for adoption is an alternative to abortion for a pregnant woman who thinks that as a single mother she could not cope with the baby. The pro-life community recognizes this. See, e.g., Students for Life of America, “Adoption, Another Option,”: “There may be times when a mother facing an un- planned pregnancy may feel completely unequipped to parent her child. She may feel her only option is to k** her pre- born child. Pro-life individuals touch lives by helping women place their baby or child for adoption. It is important to show women on your campus that adoption can be the answer to all of her fears” (emphasis in original).
Consider now the emotional comfort that having married parents is likely to provide to children adopted by same-s** couples. Suppose such a child comes home from school one day and reports to his parents that all his cla**mates have a mom and a dad, while he has two moms (or two dads, as the case may be). Children, being natural conformists, tend to be upset upon discovering that they're not in step with their peers. If a child's same-s** parents are married, however, the parents can tell the child truthfully that an adult is permitted to marry a person of the opposite s**, or if the adult prefers as some do a person of his or her own s**, but that either way the parents are married and therefore the child can feel secure in being the child of a married couple. Conversely, imagine the parents having to tell their child that same-s** couples can't marry, and so the child is not the child of a married couple, unlike his cla**mates.
Indiana permits joint adoption by h*mos**uals (Wisconsin does not). But an unmarried h*mos**ual couple is less stable than a married one, or so at least the state's insistence that marriage is better for children implies. If marriage is better for children who are being brought up by their biological parents, it must be better for children who are being brought up by their adoptive parents. The state should want h*mos**ual couples who adopt children—as, to repeat, they are permitted to do—to be married, if it is serious in arguing that the only governmental interest in marriage derives from the problem of accidental births. (We doubt that it is serious.)
The state's claim that conventional marriage is the solution to that problem is belied by the state's experience with births out of wedlock. Accidental pregnancies are found among married couples as well as unmarried couples, and among individuals who are not in a committed relationship and have s**ual intercourse that results in an unintended pregnancy. But the state believes that married couples are less likely to abandon a child of the marriage even if the child's birth was unintended. So if the state's policy of trying to channel procreative s** into marriage were succeeding, we would expect a drop in the percentage of children born to an unmarried woman, or at least not an increase in that percentage. Yet in fact that percentage has been rising even since Indiana in 1997 reenacted its prohibition of same-s** marriage (thus underscoring its determined opposition to such marriage) and for the first time declared that it would not recognize same-s** marriages contracted in other states or abroad. The legislature was fearful that Hoosier h*mos**uals would flock to Hawaii to get married, for in 1996 the Hawaii courts appeared to be moving toward invalidating the state's ban on same-s** marriage, though as things turned out Hawaii did not authorize such marriage until 2013.
In 1997, the year of the enactment, 33 percent of births in Indiana were to unmarried women; in 2012 (the latest year for which we have statistics) the percentage was 43 percent. The corresponding figures for Wisconsin are 28 percent and 37 percent and for the nation as a whole 32 percent and 41 percent. (The source of all these data is Kids Count Data Center, “Births to Unmarried Women,”.) There is no indication that these states' laws, ostensibly aimed at channeling procreation into marriage, have had any such effect.
A degree of arbitrariness is inherent in government regulation, but when there is no justification for government's treating a traditionally discriminated-against group significantly worse than the dominant group in the society, doing so denies equal protection of the laws. One wouldn't know, reading Wisconsin's brief, that there is or ever has been dis- crimination against h*mos**uals anywhere in the United States. The state either is oblivious to, or thinks irrelevant, that until quite recently h*mos**uality was anathematized by the vast majority of heteros**uals (which means, the vast majority of the American people), including by most Americans who were otherwise quite liberal. h*mos**uals had, as h*mos**uals, no rights; h*mos**ual s** was criminal (though rarely prosecuted); h*mos**uals were formally banned from the armed forces and many other types of government work (though again enforcement was sporadic); and there were no laws prohibiting employment discrimination against h*mos**uals. Because h*mos**uality is more easily concealed than race, h*mos**uals did not experience the same economic and educational discrimination, and pub- lic humiliation, that African-Americans experienced. But to avoid discrimination and ostracism they had to conceal their h*mos**uality and so were reluctant to participate openly in h*mos**ual relationships or reveal their h*mos**uality to the heteros**uals with whom they a**ociated. Most of them stayed “in the closet.” Same-s** marriage was out of the question, even though interracial marriage was legal in most states. Although discrimination against h*mos**uals has diminished greatly, it remains widespread. It persists in statutory form in Indiana and in Wisconsin's constitution.
At the very least, “a [discriminatory] law must bear a rational relationship to a legitimate governmental purpose.” Romer v. Evans, supra, 517 U.S. at 635. Indiana's ban flunks this undemanding test.
Wisconsin's prohibition of same-s** marriage, to which we now turn, is found in a 2006 amendment to the state's constitution. The amendment, Article XIII, § 13, provides: “Only a marriage between one man and one woman shall be valid or recognized as a marriage in this state. A legal status identical or substantially similar to that of marriage for un- married individuals shall not be valid or recognized in this state.” Opponents of same-s** marriage in Indiana have tried for a number of years to insert a prohibition of such marriages into the state's constitution, as yet without success. A number of large businesses in Indiana oppose such a constitutional amendment. With 19 states having authorized same-s** marriage, the businesses may feel that it's only a matter of time before Indiana joins the bandwagon, and that a constitutional amendment would impede the process— and also would signal to Indiana's gay and lesbian citizens, some of whom are employees of these businesses, that they are in a very unwelcoming environment, with statutory re- form blocked. (On the attitude of business in Indiana and Wisconsin to same-s** marriage, see, e.g., Nick Halter, “Target Files Court Papers Supporting Same-Sex Marriage in Wisconsin and Indiana,” Aug. 5, 2014,)
Wisconsin's brief in defense of its prohibition of same-s** marriage adopts Indiana's ground (“accidental births”) but does not amplify it. Its “accidental births” rationale for prohibiting same-s** marriage is, like Indiana's, undermined by a “first cousin” exemption—but, as a statutory matter at least, an even broader one: “No marriage shall be contracted ... between persons who are nearer of kin than 2nd cousins except that marriage may be contracted between first cousins where the female has attained the age of 55 years or where either party, at the time of application for a marriage license, submits an affidavit signed by a physician stating that either party is permanently sterile.” Wis. Stat. § 65.03(1). Indiana's marriage law, as we know, authorizes first-cousin marriages if both cousins are at least 65 years old. But—and here's the kicker—Indiana apparently will as a matter of comity recognize any marriage lawful where contracted, including therefore (as an Indiana court has held) marriages of first cousins contracted in Tennessee, a state that places no restrictions on such marriages. See Tenn. Code Ann. § 36-3- 101; Mason v. Mason, 775 N.E.2d 706, 709 (Ind. App. 2002). Indiana has not tried to explain to us the logic of recognizing marriages of fertile first cousins (prohibited in Indiana) that happen to be contracted in states that permit such marriages, but of refusing, by virtue of the 1997 amendment, to recognize same-s** marriages (also prohibited in Indiana) contracted in states that permit them. This suggests animus against same-s** marriage, as is further suggested by the state's inability to make a plausible argument for its refusal to recognize same-s** marriage.
But back to Wisconsin, which makes four arguments of its own against such marriage: First, limiting marriage to heteros**uals is traditional and tradition is a valid basis for limiting legal rights. Second, the consequences of allowing same-s** marriage cannot be foreseen and therefore a state should be permitted to move cautiously—that is, to do nothing, for Wisconsin does not suggest that it plans to take any steps in the direction of eventually authorizing such marriage. Third, the decision whether to permit or forbid same- s** marriage should be left to the democratic process, that is, to the legislature and the electorate. And fourth, same-s** marriage is an*logous in its effects to no-fault divorce, which, the state argues, makes marriage fragile and unreliable—though of course Wisconsin has no-fault divorce, and it's surprising that the state's a**istant attorney general, who argued the state's appeal, would trash his own state's law. The contention, built on the an*logy to no-fault divorce and sensibly dropped in the state's briefs in this court—but the a**istant attorney general could not resist resuscitating it at the oral argument—is that, as the state had put it in submissions to the district court, allowing same-s** marriage creates a danger of “shifting the public understanding of marriage away from a largely child-centric institution to an adult-centric institution focused on emotion.” No evidence is presented that same-s** marriage is on average less “child- centric” and more emotional than an infertile marriage of heteros**uals, or for that matter that no-fault divorce has rendered marriage less “child-centric.”
The state's argument from tradition runs head on into Loving v. Virginia, 388 U.S. 1 (1967), since the limitation of marriage to persons of the same race was traditional in a number of states when the Supreme Court invalidated it. Laws forbidding black-white marriage dated back to colonial times and were found in northern as well as southern colonies and states. See Peggy Pascoe, What Comes Naturally: Miscegenation Law and the Making of Race in America (2009). Tradition per se has no positive or negative significance. There are good traditions, bad traditions pilloried in such famous literary stories as Franz Kafka's “In the Penal Colony” and Shirley Jackson's “The Lottery,” bad traditions that are historical realities such as cannibalism, foot-binding, and suttee, and traditions that from a public-policy standpoint are neither good nor bad (such as trick-or-treating on Halloween). Tradition per se therefore cannot be a lawful ground for discrimination—regardless of the age of the tradition. Holmes thought it “revolting to have no better reason for a rule of law than that so it was laid down in the time of Henry IV.” Oliver Wendell Holmes, Jr., “The Path of the Law,” 10 Harv. L. Rev. 457, 469 (1897). Henry IV (the English Henry IV, not the French one—Holmes presumably was referring to the former) died in 1413. Criticism of h*mos**uality is far older. In Leviticus 18:22 we read that “thou shalt not lie with mankind, as with womankind: it is abomination.”
The limitation on interracial marriage invalidated in Loving was in one respect less severe than Wisconsin's law. It did not forbid members of any racial group to marry, just to marry a member of a different race. Members of different races had in 1967, as before and since, abundant possibilities for finding a suitable marriage partner of the same race. In contrast, Wisconsin's law, like Indiana's, prevents a h*mo- s**ual from marrying any person with the same s**ual orientation, which is to say (with occasional exceptions) any person a h*mos**ual would want or be willing to marry.
Wisconsin points out that many venerable customs appear to rest on nothing more than tradition—one might even say on mindless tradition. Why do men wear ties? Why do people shake hands (thus spreading germs) or give a peck on the cheek (ditto) when greeting a friend? Why does the President at Thanksgiving spare a brace of turkeys (two out of the more than 40 million turkeys k**ed for Thanksgiving dinners) from the butcher's knife? But these traditions, while to the fastidious they may seem silly, are at least harmless. If no social benefit is conferred by a tradition and it is written into law and it discriminates against a number of people and does them harm beyond just offending them, it is not just a harmless anachronism; it is a violation of the equal protection clause, as in Loving. See 388 U.S. at 8–12.
Against this the state argues in its opening brief that Loving “should be read as recognizing the constitutional restrictions on the government's ability to infringe the freedom of individuals to decide for themselves how to arrange their own private and domestic affairs.” But that sounds just like what the government of Wisconsin has done: told h*mos**uals that they are forbidden to decide for themselves how to arrange their private and domestic affairs. If they want to marry, they have to marry a person of the opposite s**.
The state elaborates its argument from the wonders of tradition by a**erting, again in its opening brief, that “thou- sands of years of collective experience has [sic] established traditional marriage, between one man and one woman, as optimal for the family, society, and civilization.” No evidence in support of the claim of optimality is offered, and there is no acknowledgment that a number of countries permit polygamy—Syria, Yemen, Iraq, Iran, Egypt, Sudan, Morocco, and Algeria—and that it flourishes in many African countries that do not actually authorize it, as well as in parts of Utah. (Indeed it's been said that “polygyny, where- by a man can have multiple wives, is the marriage form found in more places and at more times than any other.” Stephanie Coontz, Marriage, a History: How Love Conquered Marriage 10 (2006).) But suppose the a**ertion is correct. How does that bear on same-s** marriage? Does Wisconsin want to push h*mos**uals to marry persons of the opposite s** because opposite-s** marriage is “optimal”? Does it think that allowing same-s** marriage will cause heteros**uals to convert to h*mos**uality? Efforts to convert h*mos**uals to heteros**uality have been a bust; is the opposite conversion more feasible?
Arguments from tradition must be distinguished from arguments based on morals. Many unquestioned laws are founded on moral principles that cannot be reduced to cost- benefit an*lysis. Laws forbidding gratuitous cruelty to animals, and laws providing public a**istance for poor and dis- abled persons, are examples. There is widespread moral op- position to h*mos**uality. The opponents are entitled to their opinion. But neither Indiana nor Wisconsin make a moral argument against permitting same-s** marriage.
The state's second argument is: “go slow”: maintaining the prohibition of same-s** marriage is the “prudent, cautious approach,” and the state should therefore be allowed “to act deliberately and with prudence—or, at the very least, to gather sufficient information—before transforming this cornerstone of civilization and society.” There is no suggestion that the state has any interest in gathering information, for notice the a**umption in the quoted pa**age that the state already knows that allowing same-s** marriage would transform a “cornerstone of civilization and society,” namely monogamous heteros**ual marriage. One would expect the state to have provided some evidence, some reason to believe, however speculative and tenuous, that allowing same-s** marriage will or may “transform” marriage. At the oral argument the state's lawyer conceded that he had no knowledge of any study underway to determine the possible effects on heteros**ual marriage in Wisconsin of allowing same-s** marriage. He did say that same-s** marriage might somehow devalue marriage, thus making it less attractive to opposite-s** couples. But he quickly acknowledged that he hadn't studied how same-s** marriage might harm marriage for heteros**uals and wasn't prepared to argue the point. Ma**achusetts, the first state to legalize same-s** marriage, did so a decade ago. Has heteros**ual marriage in Ma**achusetts been “transformed”? Wisconsin's lawyer didn't suggest it has been.
He may have been gesturing toward the concern ex- pressed by some that same-s** marriage is likely to cause the heteros**ual marriage rate to decline because heteros**uals who are hostile to h*mos**uals, or who whether hostile to them or not think that allowing them to marry degrades the institution of marriage (as might happen if people were allowed to marry their pets or their sports cars), might decide not to marry. Yet the only study that we've discovered, a reputable statistical study, finds that allowing same-s** marriage has no effect on the heteros**ual marriage rate. Marcus Dillender, “The d**h of Marriage? The Effects of New Forms of Legal Recognition on Marriage Rates in the United States,” 51 Demography 563 (2014). No doubt there are more persons more violently opposed to same-s** marriage in states that have not yet permitted it than in states that have, yet in all states there are opponents of same-s** marriage. But they would tend also to be the citizens of the state who were most committed to heteros**ual marriage (devout Catholics, for example).
No one knows exactly how many Americans are h*mo- s**ual. Estimates vary from about 1.5 percent to about 4 per- cent. The estimate for Wisconsin is 2.8 percent, which includes bis**ual and transgendered persons. Gary J. Gates & Frank Newport, “LGBT Percentage Highest in D.C., Lowest in North Dakota,” Gallup (Feb. 15, 2013),. Given how small the percentage is, it is sufficiently implausible that allowing same-s** marriage would cause palpable harm to family, society, or civilization to require the state to tender evidence justifying its fears; it has provided none.
The state falls back on Justice Alito's statement in dissent in United States v. Windsor, supra, 133 S. Ct. at 2716, that “at present, no one—including social scientists, philosophers, and historians—can predict with any certainty what the long-term ramifications of widespread acceptance of same- s** marriage will be. And judges are certainly not equipped to make such an a**essment.” What follows, if prediction is impossible? Justice Alito thought what follows is that the Supreme Court should not interfere with Congress's determination in the Defense of Marriage Act that “marriage,” for purposes of entitlement to federal marital benefits, excludes same-s** marriage even if lawful under state law. But can the “long-term ramifications” of any constitutional decision be predicted with certainty at the time the decision is rendered?
The state does not mention Justice Alito's invocation of a moral case against same-s** marriage, when he states in his dissent that “others explain the basis for the institution in more philosophical terms. They argue that 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.” Id. at 2718. That is a moral argument for limiting marriage to heteros**uals. The state does not mention the argument because as we said it mounts no moral arguments against same-s** marriage.
We know that many people want to enter into a same-s** marriage (there are millions of h*mos**ual Americans, though of course not all of them want to marry), and that forbidding them to do so imposes a heavy cost, financial and emotional, on them and their children. What Wisconsin has not told us is whether any heteros**uals have been harmed by same-s** marriage. Obviously many people are dis- tressed by the idea or reality of such marriage; otherwise these two cases wouldn't be here. But there is a difference, famously emphasized by John Stuart Mill in On Liberty (1869), between the distress that is caused by an a**ault, or a theft of property, or an invasion of privacy, or for that matter discrimination, and the distress that is caused by behavior that disgusts some people but does no (other) harm to them. Mill argued that neither law (government regulation) nor morality (condemnation by public opinion) has any proper concern with acts that, unlike a punch in the nose, inflict no temporal harm on another person without consent or justification. The qualification temporal is key. To be the basis of legal or moral concern, Mill argued, the harm must be tangible, secular, material—physical or financial, or, if emotional, focused and direct—rather than moral or spiritual. Mill illustrated nontemporal harm with revulsion against polygamy in Utah (he was writing before Utah agreed, as a condition of being admitted to the union as a state, to amend its constitution to prohibit polygamy). The English people were fiercely critical of polygamy wherever it occurred. As they were entitled to be. But there was no way polygamy in Utah could have adverse effects in England, 4000 miles away. Mill didn't think that polygamy, however offensive, was a proper political concern of England.
Similarly, while many heteros**uals (though in America a rapidly diminishing number) disapprove of same-s** marriage, there is no way they are going to be hurt by it in a way that the law would take cognizance of. Wisconsin doesn't argue otherwise. Many people strongly disapproved of interracial marriage, and, more to the point, many people strongly disapproved (and still strongly disapprove) of h*mos**ual s**, yet Loving v. Virginia invalidated state laws banning interracial marriage, and Lawrence v. Texas invalidated state laws banning h*mos**ual s** acts.
Though these decisions are in the spirit of Mill, Mill is not the last word on public morality. But Wisconsin like Indiana does not base its prohibition of same-s** marriage on morality, perhaps because it believes plausibly that Lawrence rules out moral objections to h*mos**uality as legitimate grounds for discrimination.
In pa**ing, Wisconsin in its opening brief notes that it “recogniz[es] domestic partnerships.” Indeed it does: Wis. Stat. ch. 770. And the domestic partners must be of the same s**. Id., § 770.05(5). But the preamble to the statute states: “The legislature ... finds that the legal status of domestic partnership as established in this chapter is not substantially similar to that of marriage,” § 770.001, citing for this proposition a decision by a Wisconsin intermediate appellate court. Appling v. Doyle, 826 N.W.2d 666 (Wis. App. 2012), affirmed, 2014 WI 96 (Wis. July 31, 2014). Indeed that is what the court held. It pointed out that chapter 770 doesn't specify the rights and obligations of the parties to a domestic partner- ship. Rather you must go to provisions specifying the rights and obligations of married persons and see whether a provision that you're concerned with is made expressly applicable to domestic partnerships, as is for example the provision that gives a surviving spouse the deceased spouse's interest in their home. 826 N.W.2d at 668. But as the court further ex- plained, the rights and obligations of domestic partners are far more limited than those of married persons. See id. at 682–86. (For example, only spouses may jointly adopt a child. Id. at 685.) They have to be far more limited, because of the state's constitutional provision quoted above that “a legal status identical or substantially similar to that of marriage for unmarried individuals shall not be valid or recognized.” Wis. Const. Art. XIII, § 13. Domestic partnership in Wisconsin is not and cannot be marriage by another name.
It is true that because the state does not regard same-s** marriages contracted in other states as wholly void (though they are not “recognized” in Wisconsin), citizens of Wisconsin who contract same-s** marriages in states in which such marriages are legal are not debarred from receiving some of the federal benefits to which legally married persons (including parties to a same-s** marriage) are entitled. Not to all those benefits, however, because a number of them are limited by federal law to persons who reside in a state in which their marriages are recognized. These include benefits under the Family & Medical Leave Act, see 29 C.F.R. § 825.122(b), and access to a spouse's social security benefits. See 42 U.S.C. § 416(h)(1)(A)(i).
So look what the state has done: it has thrown a crumb to same-s** couples, denying them not only many of the rights and many of the benefits of marriage but also of course the name. Imagine if in the 1960s the states that forbade interracial marriage had said to interracial couples: “you can have domestic partnerships that create the identical rights and ob- ligations of marriage, but you can call them only ‘civil un- ions' or ‘domestic partnerships.' The term ‘marriage' is re- served for same-race unions.” This would give interracial couples much more than Wisconsin's domestic partnership statute gives same-s** couples. Yet withholding the term “marriage” would be considered deeply offensive, and, having no justification other than bigotry, would be invalidated as a denial of equal protection.
The most arbitrary feature of Wisconsin's treatment of same-s** couples is its refusal to allow couples in domestic partnerships to adopt jointly, as married heteros**ual couples are allowed to do (and in Indiana, even unmarried ones). The refusal harms the children, by telling them they don't have two parents, like other children, and harms the parent who is not the adoptive parent by depriving him or her of the legal status of a parent. The state offers no justification.
Wisconsin's remaining argument is that the ban on same- s** marriage is the outcome of a democratic process—the enactment of a constitutional ban by popular vote. But h*mos**uals are only a small part of the state's population— 2.8 percent, we said, grouping transgendered and bis**ual persons with h*mos**uals. Minorities trampled on by the democratic process have recourse to the courts; the recourse is called constitutional law.
In its reply brief Indiana adopts Wisconsin's democracy argument, adding that “h*mos**uals are politically powerful out of proportion to their numbers.” No evidence is presented by the state to support this contention. It is true that an increasing number of heteros**uals support same-s** marriage; otherwise 11 states would not have changed their laws to permit such marriage (the other 8 states that allow same-s** marriage do so as a result of judicial decisions in- validating the states' bans). No inference of manipulation of the democratic process by h*mos**uals can be drawn, how- ever, any more than it could be inferred from the enactment of civil rights laws that African-Americans “are politically powerful out of proportion to their numbers.” It is to the credit of American voters that they do not support only laws that are in their palpable self-interest. They support laws punishing cruelty to animals, even though not a single animal has a vote.
To return to where we started in this opinion, more than unsupported conjecture that same-s** marriage will harm heteros**ual marriage or children or any other valid and important interest of a state is necessary to justify discrimination on the basis of s**ual orientation. As we have been at pains to explain, the grounds advanced by Indiana and Wisconsin for their discriminatory policies are not only conjectural; they are totally implausible.
For completeness we note the ultimate convergence of our simplified four-step an*lysis with the more familiar, but also more complex, approach found in many cases. In SmithKline Beecham Corp. v. Abbott Laboratories, 740 F.3d 471, 483 (9th Cir. 2014), the Ninth Circuit concluded, based on a reading of the Supreme Court's decisions in Lawrence and Windsor, that statutes that discriminate on the basis of s**ual orientation are subject to “heightened scrutiny”—and in do- ing so noted that Windsor, in invalidating the Defense of Marriage Act, had balanced the Act's harms and offsetting benefits: “Notably absent from Windsor's review of DOMA are the ‘strong presumption' in favor of the constitutionality of laws and the ‘extremely deferential' posture toward government action that are the marks of rational basis review. ... In its parting sentences, Windsor explicitly announces its balancing of the government's interest against the harm or injury to gays and lesbians: ‘The federal statute is invalid, for no legitimate purpose overcomes the purpose and effect to disparage and injure those whom the State, by its marriage laws, sought to protect in personhood and dignity.' 133 S. Ct. at 2696 (emphasis added). Windsor's balancing is not the work of rational basis review.”
The Supreme Court also said in Windsor that “the Act's demonstrated purpose is to ensure that if any State decides to recognize same-s** marriages, those unions will be treat- ed as second-cla** marriages for purposes of federal law.” 133 S. Ct. at 2693–94. A second-cla** marriage would be a lot better than the cohabitation to which Indiana and Wisconsin have consigned same-s** couples.
The states' concern with the problem of unwanted children is valid and important, but their solution is not “tailored” to the problem, because by denying marital rights to same-s** couples it reduces the incentive of such couples to adopt unwanted children and impairs the welfare of those children who are adopted by such couples. The states' solution is thus, in the familiar terminology of constitutional dis- crimination law, “overinclusive.” It is also underinclusive, in allowing infertile heteros**ual couples to marry, but not same-s** couples.
Before ending this long opinion we need to address, though only very briefly, Wisconsin's complaint about the wording of the injunction entered by the district judge. Its lawyers claim to fear the state's being held in contempt be- cause it doesn't know what measures would satisfy the in- junction's command that all relevant state officials “treat same-s** couples the same as different s** couples in the context of processing a marriage license or determining the rights, protections, obligations or benefits of marriage.” If the state's lawyers really find this command unclear, they should ask the district judge for clarification. (They should have done so already; they haven't.) Better yet, they should draw up a plan of compliance and submit it to the judge for approval.
The district court judgments invalidating and enjoining these two states' prohibitions of same-s** marriage are
AFFIRMED.