Yes, Supreme Court Justices Vote Strategically, And No, There’s Nothing Wrong With That

This thread had a discussion of the the Supreme Court damaging the Medicaid funding mechanism in Sebelius,leaving to the inevitable argument that it was a 7-2 vote and hence Both Sides Do It and so on. My position continues to be that had Breyer or Kagan been the median vote on the Supreme Court there’s a roughly 100% chance that the Medicaid expansion would have been untouched. Equally inevitably, this leads to people expressing shock at the very idea that a Supreme Court justice’s vote could ever represent something other than their sincere, optimal preference. It’s sort of like the argument that if Obama puts Chained CPI in his budget with conditions that make it unpassable or a Republican proposes health care reform that will vanish the second a Republican takes the White House, we have no choice but to assume that this reflects their sincere preferences. Only in the case of elected politicians, this argument is virtually always being made in bad faith; nobody really believes it (almost everybody who assumed that Chained CPI in his budget reflected Obama’s passionate commitment to killing Social Security also assumes that Obama spent most of his first term secretly stopping Congress from adopting legislation he publicly supported.) When it comes to the Supreme Court, though, I think many people really do have a weirdly naive belief that justices rarely do and never should vote strategically. So let’s explain again why this assumption is transparently wrong empirically and unpersuasive normatively.

The political science literature is unambiguous in demonstrating the ubiquity of strategic voting. Epstein and Knight is a superb primer, although nonspecialists may want to start with Walter Murphy’s Elements of Judicial Strategy; the future bestselling novelist wrote very well and it’s an extremely smart and elegantly argued text. In addition to the examples in the first link, there’s an excellent recent example: Northwest Austin. John Roberts’s faux outrage notwithstanding, it should have been obvious what was going on in Northwest Austin at the time: Ginsburg, Breyer, Souter and Stevens joined the Court’s opinion not because they were persuaded by Roberts’s bare assertion that the Constitution protects an extratextual Equal Sovereign Dignitude of the states, but because they preferred an epically specious opinion upholding the Voting Rights Act to an epically specious opinion that went ahead and gutted it, and if they didn’t provide Roberts with the appearance of consensus they substantially risked the latter in 2009. Unless you think that Ginsburg and Breyer radically revised their views in the intervening four years, the former’s dissent in Shelby County settles the question. But Ginsburg and Breyer weren’t lying or acting in bad faith in Northwest Austin (although their strategic judgment might have been wrong); voting for the viable position closest to your own is a perfectly ordinary and defensible practice.  What matters is that the Voting Rights Act survives, perhaps until there can be a different median vote on the Court.  Getting the satisfaction of having spoken your mind isn’t much consolation to everyone whose vote would be suppressed.

Indeed, you don’t really need a lot of legal scholars digging through old conference votes and memos or using fancy game theory to see that strategic voting is inevitable in any court where five votes are necessary to get a majority. Just compare, say, a typical Thomas solo dissent with a typical Thomas majority opinion; the latter will be as a rule much blander and less idiosyncratic, because they have to represent what’s minimally acceptable to every member of the majority coalition. We know Brennan wrote an opinion subjecting gender classifications to intermediate scrutiny (which could command a majority) although his preferred position was strict scrutiny (which couldn’t) because we have the relevant memos, but it should be obvious that this kind of thing is going to happen all the time.

Which brings us back to Sebelius. On the Medicaid expansion, there were three options on the table:

• (ArtI) The Spending Clause in Article I means what it says.  There is no precedent for striking down an exercise of conditional federal grants given to states, and this modification to Medicaid is no less constitutional than any of the previous 50 or so.
• (.5NeoCon) The half-neoconfederate: Congress cannot make Medicaid funding contingent on the accepting the new expansion, even though it would be unquestionably constitutional for Congress to create the new Medicaid from scratch on one hand or to repeal it entirely on the other, because…look, John Calhoun!  Equal Sovereign Dignitude! The Cornhusker Kickback!  Congress, however, can offer the new money as long as it doesn’t threaten to withhold the old money.
• (NeoCon) The full neoconfederate: All of the illogical states’ “rights” gibberish of .5NeoCon, only the expansion is struck down entirely.

Everything about Breyer’s history suggests that his ideal preference ordering is ArtI>.5NeoCon>NeoCon.  If you can name one Supreme Court justice in history with a broader conception of federal authority, that’s one more than me.  With Kagan, we have less history to go on, but she’s a typical Democratic nominee and that means that any preference ordering but ArtI>.5NeoCon>NeoCon would be enormously unusual.  (Members of team Democrat potentially disagree about many issues, but national power isn’t one of them in 2014.)  The oral arguments confirmed this, with both Breyer and Kagan expressing notable hostility to the neoconfederate spending power arguments (not surprisingly, since they were an afterthought even to most people who took the argument that the mandate was unconstitutional seriously.)  Taken literally, however, their votes would indicate a preference ordering of .5NeoCon>ArtI>NecCon.  Did Breyer radically revise his views? Was Kagan a disastrous pick after all?

Almost certainly not.  The only two relevant options — the only positions with a chance of getting 5 votes — were .5NeoCon and NeoCon.  Particularly in a context in which Roberts initially voted to strike down the ACA in its entirety, by far the most plausible explanation is that Breyer (who, you’ll remember, was part of a doomed effort to salvage the Florida recounts by trying to form a consensus on 14th Amendment violations with an actual remedy) and Kagan wanted to add their superfluous votes to ensure Roberts wouldn’t move from the .5NeoCon to the NeoCon camp.   Can I prove to an absolute certainty that Breyer hasn’t radically changed the views he’s consistently held for decades and won’t suddenly start casting votes placing arbitrary limits on the congressional spending power?  No, but the overwhelming likelihood is that his vote was anomalous.  And, certainly, the only thing we can safely infer from the Breyer and Kagan votes is that they preferred .5NeoCon to NeoCon.