On Tuesday, the Supreme Court decided Bittner v. United States. The question presented was whether a “violation” under the Bank Secrecy Act is the failure to file an annual form (no matter the number of foreign accounts), or whether there is a separate violation for each individual account that was not properly reported. The majority held that the penalty can only be assessed once per-report. The dissent held that the penalty can be assessed for each report. Justice Gorsuch wrote the majority opinion, which was joined in full by Justice Jackson. Justice Gorsuch wrote a section on the rule of lenity, which only Justice Jackson signed onto. Chief Justice Roberts and Justices Alito and Kavanaugh joined the rest of the majority opinion. Justice Barrett wrote the dissent, which was joined by Justices Thomas, Sotomayor, and Kagan. Jon Adler already pointed out the unusual split. Here, I’ll try to make sense of the votes.
As a threshold matter, the identity of the authoring Justices should not come as a surprise. Justice Gorsuch consistently reads federal statutes in a narrow way that inures to the benefit of defendants. His separate analysis about the rule of lenity, which only Justice Jackson joined, undergirds Gorsuch’s approach to statutory construction. On the other hand, Justice Barrett seem to consistently disagree with Justice Gorsuch on how to read a statute. (I am reasonably confident she would have dissented in Bostock.) Jon pointed out how often they disagree. I am also not surprised that Justice Thomas joined Barrett. He seems to vote consistently with Barrett on criminal-law related cases.
The other votes require some more thinking. More often than not, Chief Justice Roberts and Justice Kavanaugh vote in lockstep. (Cruz v. Arizona was an outlier.) Perhaps they were persuaded by Justice Gorsuch’s textualism. Or perhaps this join was a way to reduce the number of conservative 5-4 decisions for the term, and pad the heterodox splits. Those statistics do add up! What about Justice Alito? It is a very rare case indeed where he votes for a criminal defendant. I can only think of a handful. But Alito may have felt sympathetic to the defendant here who made a non-willful–that is a good-faith–error. Indeed, many accountants claimed they weren’t even aware this filing requirement exists.
What do we make, then of the Court’s three progressives? As a general rule, I presume they will vote in a bloc, unless there are strategic advantages to breaking up. I am willing to abandon this presumption if they consistently vote out of formation. But if last term is any indication, like the three musketeer, they are all for one and one for all! What then was the strategery in Bittner? Why would Justices Kagan and Sotomayor rule against the criminal law defendant in what is ostensibly an unfair ruling. Well, perhaps they, like Justice Thomas, were persuaded by South Bend’s finest. Or, perhaps, they saw value in joining Justice Barrett in this case, as a way of signaling that they are reasonable free agents, and are not locked in a formation. Maybe the conference vote was 7-2, and Kagan and Sotomayor jumped ship after Barrett circulated her dissent. In the appropriate case, Sotomayor and Kagan can cash in on that credibility to bring Justice Barrett over to their side. (The student loan cases may fit the bill.) And as the cost of the join, Part III of Barrett’s dissent flags an alternate way in which future defendants can prevail, notwithstanding the Court’s ruling.
On the flipside, Justice Jackson had a free pass to form a majority, and the defendant would prevail. She was also able to give Justice Gorsuch a +1 for his rule of lenity section. In other cases, Justice Sotomayor has joined Gorsuch’s dissentals when it affect a criminal justice issue. The three progressives did not need to vote as a bloc here–a five-member majority reached the right outcome–and they did not vote as a bloc.
Of course, I have no inside information, and you are free to disregard everything I write here. But when it comes to the Court’s votes, I try to see patterns where none may exist.