Now that a critical part of the Affordable Care Act is before the Supreme Court in King v. Burwell we can expect to see many efforts to steer the outcome from both sides in popular media, legal blogs, and briefs. A lot of these attempts will be aimed at the interesting and complex legal issues involved, but others will not be.

On Tuesday, Harvard law professor Noah Feldman speculated in a blog post about a possible “linkage” between the pending ACA case and a likely same-sex marriage case this Term. Not that he asserts any linkage in the facts, legal issues, or background principles, which would be a fascinating read. But he sees a linkage by which the Supreme Court may give itself psychological permission to reach a “conservative” policy result in one case (against the ACA) by balancing it against a “liberal” result in the other (for same-sex marriage). The post is entitled, “Obamacare May Die So Gay Marriage Survives.” Feldman may not have written the title but it doesn’t really mislead the reader about the substance.

I have a great deal of respect for Professor Feldman, who’s clearly a star of his generation in the legal academy. Indeed, I assign my Constitutional Law students the casebook he co-authors. But the conjecture in his BloombergView post is unpersuasive on its own terms and reflects a dismaying trend in constitutional-law writing: a hyper-legal-realism that is neither legal nor very real.

One can fully accept that political allegiances and the Supreme Court’s own desire to preserve the perception of its legitimacy, which is itself a form of politics, can influence its decisions. But Feldman’s approach calls on us to examine the deep and hidden psychological motivations of the Justices–which he agrees they would deny publicly and privately–without even considering whether there are perfectly reasonable and defensible legal grounds for the way they may rule in a particular case.

Here’s how the claim is introduced:

How does linkage work at the Supreme Court? To begin with, recognize that today’s justices never speak in terms of horsetrading or logrolling from one case to another. I don’t just mean that, when asked publicly if justices might trade votes to support causes held dear by one-another, they all deny it. I mean that, even within the court itself or in private conversation with clerks and confidants, the justices would almost certainly say that no such linkage exists. Denial of linkage across unrelated cases is a kind of justices’ dogma — it would be sacrilegious, not to say heretical, for court insiders to say otherwise.

Nevertheless, linkage may exist at the level of psychology — and the justices know it.

To support this psychological-linkage speculation, Feldman offers the example of Justice Stevens’s votes in 2008 agreeing with the five conservatives in a death penalty and international-law decision (Medellin v. Texas) while at the same time voting to apply constitutional-rights protections to prisoners at Guantánamo Bay (Boumediene v. Bush).

What explains Stevens’s votes? The cases presented very different factual backgrounds and distinct constitutional questions about the status of international law, executive authority, and federalism. But Feldman analyzes none of this, instead asserting: “Stevens did not want Justice Anthony Kennedy to be subject to the criticism that he had provided the deciding conservative vote in the Medellin decision and the deciding liberal vote in the Guantánamo case.” So he voted to protect Kennedy by giving him added support in a conservative decision, thereby securing his vote for a liberal decision in a different case. Feldman does not tell us what the basis is for thinking Stevens might have had this latent psychological motivation–sacrificing his own judicial integrity and legacy against his better liberal judgment–to protect Kennedy’s right flank.

Feldman then applies this mode of speculation to the prospective votes of Justice Kennedy and Chief Justice Roberts in the coming ACA case, which challenges on statutory grounds the IRS tax credits being made available to those who purchase health insurance on federally created exchanges. Since Feldman believes Justice Kennedy’s vote for gay marriage is almost certain, he asserts that Kennedy may vote to strike down the ACA in order “to try to maintain his conservative bona fides with a corresponding conservative decision.” He adds: “If you believe that Kennedy will decide in favor of gay marriage — and I do — then it seems almost guaranteed that he will vote with the conservatives against the Barack Obama administration in the new Affordable Care Act case.”

Feldman adds that “some speculated” that Kennedy’s liberal vote in United States v. Windsor  in 2013 may have been linked to his conservative vote to strike down the ACA’s individual mandate in NFIB v. Sebelius a year earlier. Apparentlypsychological-linkage speculation can explain not only the Justices’ votes across ideological divides but can enrich our understanding of why a Justice votes one way in one case while other unrelated cases are still “wending their way” to the Court to provide a future balancing vote.

As for Roberts, whichever way he votes in the gay-marriage case, Feldman is ready to pounce on his possible vote to  strike down the ACA tax credit. Roberts, of course, was the saving vote for the ACA’s individual mandate two years ago. Feldman asserts that Roberts’s vote against the ACA now will be an attempt to balance the ledger against a high-profile liberal legal victory for gay marriage with at least one high-profile conservative win this Term. This stratagem will also be part of a cunning attempt by Roberts to discredit potential liberal critics of the Court, according to Feldman:

Imagine that, in the space of a few days at the end of June, the court decides a landmark case in favor of gay rights and then says that the IRS can’t give subsidies to citizens of states that have created their own health-insurance exchanges [sic, the tax credit would be ended for those purchasing on the federal exchange]: What liberal critic would be able to say with a straight face that this was the most conservative activist court in history? The court would be activist, all right, but it would appear almost evenhandedly so.

It never enters the analysis that a Justice could, for reasons entirely apart from personal vanity or institutional strategy, view the cases as presenting quite distinct questions of fact and law. One case presents serious constitutional questions, the other raises multi-faceted statutory-interpretation issues. One challenges state laws, the other attacks a federal law. One involves a claimed need to defer to democratic decisionmaking, the other involves a claimed need to defer to agency interpretation. A Justice’s vote on whether the “equal protection of the laws” requires states to issue marriage licenses to same-sex couples hardly predetermines his or her vote on the question whether a statutory tax credit for individual health-insurance purchases in markets created “by the State” also includes a tax credit for such purchases in a market created by the federal government.

For Feldman, the obvious distinctions don’t appear to matter. On this view, the conservative votes to strike down the ACA tax credit this Term would be the barely concealed psychological by-product of shrewd politics, legacy building, and institutional protection rather than legal principle. The liberal votes for the ACA tax credit and for same-sex marriage will presumably be matters of constitutional principle, justice, and the American way.

There is almost no decision in the Court’s history that could not be explained by reverting to this crude and non-falsifiable psychological-linkage theory about why the Court issues some decisions favoring liberal policy and some decisions favoring conservative policy. Maybe Kennedy voted to uphold Roe v. Wade in Planned Parenthood v. Casey so that he could later feel ok about voting to strike down part of the Voting Rights Act and the Michigan affirmative-action policy. Maybe Justice Breyer supported the Ten Commandments display in Van Orden v. Perry so he wouldn’t appear anti-religious when he later voted against the city council prayer in Town of Greece v. GallowayMaybe Justice Scalia’s votes to strike down the flag-burning law in Texas v. Johnson (against a dissent from the devious Justice Stevens) and to reject the military detentions in Hamdi (joined by the same wily Stevens) helped him feel better about his votes in, well, just about every other case. Perhaps the Court apologized for Bush v. Gore by striking down sodomy laws in Lawrence v. Texas

Law professors have no particular expertise for concocting such speculation, unrooted as it is to any consideration of the actual legal issues presented in the cases. Instead, whatever you think of the merits of the challenge to the ACA tax credit, this sort of critique is a fairly obvious preemptive political strike on any decision overturning it. If one can be allowed speculation about psychology for a moment, maybe the theory is that at least one of the conservatives on the Court can be embarrassed into saving Obamacare.