A person celebrates the Supreme Court’s ruling on gay marriage outside the court on June 26. (Jacquelyn Martin/Associated Press)

Danielle Allen is a political theorist at the Institute of Advanced Study and a contributing columnist for The Post.

In a strange feature of his dissent in the recent Supreme Court decision on gay marriage, Obergefell v. Hodges, Chief Justice John G. Roberts Jr. actually agrees with a core plank of Justice Anthony M. Kennedy’s argument for the majority, without seeming fully to realize it. Even stranger, this agreement comes out most sharply in his comparison of Kennedy’s opinion to the notorious 1857 Dred Scott decision, which extended the rights of slave-holders across state lines. The points of agreement between Kennedy and Roberts have to do, first, with the relationship between fundamental rights and property and, second, with the historic relationship between marriage and property.

Paradoxically, Roberts introduces Dred Scott in his dissent to criticize Kennedy. Kennedy, like the Dred Scott court, he argues, has handed down a “substantive due process” decision based on his own view of the right policy on gay marriage, just as the court in 1857 “relied on its own conception of liberty and property” to make its decision. Although this is an obviously provocative comparison, we should look through the provocation to the substance, which is revealing.

What matters about the Dred Scott comparison is less the technical legal point about “substantive due process” than the logic of the underlying question of property rights that it taps into. As with the issue of slavery, the issue at the heart of the dispute over marriage involves property.

As Kennedy argues, marital status determines property rights, parental rights and a host of other benefits. This is to the point. Speaking as a historian, I would say that if there is any universal and distinguishing feature of marriage across millennia, it is its connection to property, not procreation. Consider only the importance throughout the ages of concubinage, which was often just as central as marriage to intimacy and reproduction but did nothing to confer property rights. Marriage has always been set apart by the property rights it generates.

An important due process question, then, is on what basis people can be prevented from distributing their property, or claiming particular benefits, by blocking them from marriage. And if one state permits people to activate property rights through marriage, do those property rights disappear when a person crosses a state line? The comparison of Obergefell to Dred Scott is apt because of this connection between marriage and property.

To see that comparison clearly, however, is also to see a fundamental difference. The property claims in the Dred Scott case — of slave-owners over fellow human beings — were morally illegitimate, even if they were legal in some states. In contrast, the property claims involved in gay marriage are morally legitimate. These involve the decision of two individuals to voluntarily merge their property, to share parental rights and to embark on the joint responsibilities thereby entailed, to which various further rights and privileges also adhere.

Roberts concedes that an argument more single-mindedly based on the specific property rights and privileges connected to marriage might have made more headway with him. “Although [the petitioners] discuss some of the ancillary legal benefits that accompany marriage, such as hospital visitation rights and recognition of spousal status on official documents,” he writes, “petitioners’ lawsuits target the laws defining marriage generally rather than those allocating benefits specifically.”

“The equal protection analysis might be different, in my view,” he continues — implying by that that it would have been more successful — “if we were confronted with a more focused challenge to the denial of certain tangible benefits.”

In this passage, Roberts seems to agree that the denial to gay couples of the property rights and benefits triggered by marriage was indeed open to an equal-protection challenge, but he also suggests that this issue ought to have been addressed in a piecemeal fashion, targeting specific tangible benefits, not the basic definition of marriage.

But this is a strange thing to wish for because marriage is at core a set of property rights, with attendant responsibilities and privileges. To change who can access those rights, even if you do it piecemeal, is to change the definition of marriage. In this regard, the majority has offered the more clear-eyed view of the long history of marriage and what it amounts to: a structure of property rights whose purpose is to control inheritance and define rights and responsibilities in relation to those property claims and the relationships in which they are embedded.

Now, you might wonder, if marriage is just a mode of organizing property rights and attendant duties and privileges, why couldn’t gay couples be given access to those rights via another organizational form, for instance, civil union? But this would have been to permit the re-emergence of a world of separate but equal, where there is no justification for the separate vehicles other than animus and discrimination.

With his arguments about dignity, Kennedy reminds us that this world of “separate but equal” is firmly closed off. That’s why, in his invocation of the problem of harms to the children of gay couples prohibited from marrying, we also hear hints of Brown v. Board of Education. Kennedy’s jurisprudential two-step — connecting a due-process property claim to an equal-protection dignity claim — has a structure that emerges out of the complex and technical histories of legal reasoning. But it also tracks the basic logic that links property rights and equality.