The core of Mark Tushnet’s three part explanation for John Roberts voting to strike the ACA at oral argument and then voting to uphold it is provocative from a different direction than one might expect, at least if you know Tushnet’s work. Tushnet speculates that Roberts simply wasn’t persuaded by the legal arguments made by the eventual dissenters:
As he and his law clerks worked on the tax issue, the Chief Justice discovered that the opinion “wouldn’t write.” He couldn’t come up with an account of the tax power — persuasive to him — that would invalidate the ACA. Perhaps his difficulty in getting an opinion that “would write” was influenced by institutional concerns about the long-term standing of the Court, or somewhat different institutional concerns about his ability to pursue a conservative legal agenda on matters more important to him over the long run. And, though less plausibly, perhaps these institutional concerns came to have more salience as liberals mounted their post-argument challenges to an anticipated decision invalidating the ACA. But, I would guess, those things operated — if they did, and I’m skeptical — well below the surface. Mostly, the Chief Justice found that the opinion wouldn’t write.
On this view, the Chief Justice didn’t “switch” his vote, because he hadn’t been firmly committed to the view that the ACA couldn’t be defended under the tax power. Once he indicated his view on the tax issue, the rest of the opinions fell into place. The four now-dissenters assembled their separate contributions into their joint dissent and tacked on some discussion of the Chief Justice’s new position.
There are variants based on whether Roberts assigned the original opinion to himself or intended it to be a Cooper/Gregg/Casey-style joint opinion, but the bottom line is the same: the law mattered more than you might think. One crucial point he makes is particularly powerful. To those of us who wrote off the idea that the potential criticism of liberals a significant factor because it was 1)expected , and 2)had never stopped him before, Tushnet argues that the same logic applies to institutional pressures more broadly. Whatever concern Roberts had for the legitimacy of the Court, these concerns should have been the same at the conference vote as they were a month later.
Of course, this doesn’t mean that “the law” was the only variable that mattered; as Tushnet suggests, you can’t separate out the fact that he found out that the opinion “wouldn’t write” from the pressures on the court. More cynically, one could note that one logical place he could have taken his opinion — holding the mandate unconstitutional but rejecting the self-refuting severance argument made by the dissenters — would have been the worst possible outcome for Republican interests. But I find myself surprisingly persuaded that the most straightforward legal realist explanation won’t quite write.
The real lesson here, I think, is once again Graber’s point that it’s impossible to scientifically separate “legal” and “political” and “strategic” factors. Accepting the such categories are oversimplified, however, I think that NFIB actually contains illustrations of how other factors explain Supreme Court votes as well:
- Attitudinal (“High Politics” Division): Breyer, Ginsburg, and Thomas. Breyer is the Court’s most consistent advocate of broad construction of federal power, Thomas its most consistent advocate of a neoconfederate one. All three, in Raich, were willing to apply their principles against what one would assume are their immediate policy and political preferences. It’s possible that Sotomayor, Kagan, and Alito will fall into this category as well, but we lack the necessary information. This isn’t “apolitical” judging. For obvious reasons, given that state governments have to run balanced budgets and are more easily bought off by powerful interests conservatives are more likely to benefit from decentralized power. But there’s a difference between applying a politically-derived principle when it conflicts with your immediate policy preference and just refusing to apply the principle when there’s an exception.
- Attitudinal (“Low Politics” Division): Kennedy and Scalia. As Raich demonstrates, their theory of federal power would seem to be “a law passed pursuant to the commerce clause is constitutional if Justices Kennedy and Scalia like the law and unconstitutional if they do not.” As Ginsburg so beautifully noted, Scalia’s arguments about the necessary and proper clause in Raich are a particular embarrassment for his self-image as the Last Principled Judge in America. Scalia’s performance at oral argument lends further credence to slotting him in this category.
- Strategic: Breyer and Kagan joining Roberts for the Medicaid expansion argument. It’s pretty obvious to me that Obama’s former solicitor general and the Court’s most consistent advocate of federal power would, if voting their sincere preferences, have joined Ginsburg and Sotomayor in upholding the Medicaid expansion in its entirety. But there were already five votes to rule at least some of the exercise of federal spending power unconstitutional. By joining Roberts, Breyer and Kagan presumably 1)helped secure what was by far the better of the two remaining possible outcomes, and perhaps 2)also cemented Roberts’s decision to join the more liberal faction to uphold the rest of the law. (This, at least, is what Stevens and Blackmun did in Casey, immediately joining every part of the Kennedy/O’Connor/Souter plurality that they could stomach.)
If we assume that Roberts switched his vote largely because of legal factors, we can see examples of major factor that influences judicial decision-making in this case.