In my earlier post predicting the outcomes of the Term's final opinions, I had anticipated that in Bilski v. Kappos, Justice Stevens would write an opinion for the Court very narrowly construing the patent eligibility of business methods, a ruling that could call into question the patentability of various modern innovations including software.  Instead, the Court did almost the reverse:  it decided the case in an opinion by Justice Kennedy that affirmed the patentability of business methods and said almost nothing about the applicable test beyond rejecting the Federal Circuit's exclusive reliance on a "machine or transformation" standard.

Like any good lawyer who sees such a solid prediction not come to pass, my first reaction was that the Court had made an error when it released the decision.  Assured that was not so, my second reaction was to ignore my mistake and distract people with the predictions that I had made that were correct.  But my third and final reaction was to dig a little deeper to show that I was almost right.  The available evidence strongly suggests that the Court very nearly came out the other way.

Bilski was argued in the Court's November sitting.  In that sitting, Justice Kennedy wrote the majority opinion in two cases:  Bilski and Graham v. Sullivan.  Justice Stevens wrote no majority opinions.

That is highly anomalous.  Eleven cases were argued in November, so that each Justice would ordinarily be assigned at least one majority opinion; two Justices would write two opinions.  But instead three Justices "“ the Chief Justice, Kennedy, and Ginsburg "“ each wrote two majorities, and Stevens none.

Justice Kennedy also was perhaps the least likely member of the Court to be assigned a second majority opinion in November.  He was at the time writing the principal opinion in Citizens United, the massive campaign finance case argued in September.  He received an October writing assignment.  And his other November opinion, Graham, was a very significant ruling on the constitutionality of life sentences without parole for juveniles.

Furthermore, Justice Stevens was not assigned a second majority opinion "“ i.e., another opinion that would balance out the Court's workload "“ at the soonest available opportunity in February.  There were only nine signed opinions in December and January.  (Justice Ginsburg had two opinions in December (and Justice Kennedy had none) but I think that can be explained for the reasons I discuss below.)  As the Court's second most senior Justice, it's very unlikely that he would not have received an opinion in November and then would also not have been assigned a second opinion in any of the first three months in which it was available:  October, November, and February.

Bilski also lines up perfectly for the scenario of a lost majority.  It is a five-to-four decision.  Justices Kennedy and Stevens are the authors of the principal opinions.  The ruling took a long time to issue "“ it was argued in November but not decided until the last day of the Term "“ suggesting that something may have been amiss.

The Stevens opinion also reads as if it were originally an opinion for the Court "“ in particular, it starts with a statement of the facts (whereas the Kennedy opinion has only a brief procedural history); and it is also much longer than the Kennedy opinion.  The first substantive part of the Stevens opinion "“ Part II, which is a response to the majority; in that section Stevens’ opinion refers to a subsequent section "“ reads as if it was added after Part III was written, and perhaps was originally prepared as a response to a dissent.  And there is a separate section responding to the majority that comes later (Part V) in a place that reads as if it were a response to an argument of the losing side.

Further, the muddled vote of the majority's fifth vote, Justice Scalia "“ who joined the majority in part and Justice Breyer's concurrence in substantial part "“ suggests that he had significant difficulty in resolving the case.  Of note, at oral argument Justice Scalia closely questioned the patentee's lawyer on the history of business method patents.  History is a focus of the Stevens' opinion, and Scalia declines to join the historical sections of the Kennedy opinion.  (Somewhere, a Stevens clerk who did the historical research is quietly weeping.)  And Scalia asked almost no questions of the PTO's lawyer.

Another piece of evidence of rough compromise in Bilski is the Kennedy opinion itself.  While it does decide the case, it is surpassingly narrow.  It rejects the Federal Circuit's "machine or transformation" test as the exclusive test of patentability and stops there.  The Court does not attempt to provide further guidance, which is some evidence that five Justices could not agree on how to articulate an appropriate test.

For all those reasons, it seems quite likely to me that Justice Stevens was originally going to author the Court's opinion in Bilski but subsequently lost his majority to Justice Kennedy.

[UPDATE, 3 p.m.:  An astute observer notes that my post ignored an important possible answer to the question of the "missing" Stevens opinion from November.  In that sitting, the Court heard oral arguments in Pottawattamie County v. McGhee -- a prosecutorial misconduct case -- in which the liberal side was almost certain to prevail, but which settled in late-December before the case could be decided.  There was a substantial chance that Stevens was assigned that opinion.  If so, the glaring inference from Stevens' failure to author an opinion in November would fade away.  I nonetheless think that the other factors I identify above still point to the idea that Stevens was assigned Bilski but lost his majority.  Among other things, if Pottawattamie had been his only November opinion, then I think he would have been assigned a second majority in February.  The fact that Stevens authored only 6 majority opinions for the Term -- while every other member of the Court authored 8 or 9 -- also suggests that maybe Stevens had the Court's opinions in both Pottawattamie and Bilski.  That would be consistent with the general pattern of assigning 2 majority opinions to the more senior members of the Court earlier in the Term.]

A quick aside about the puzzle of the December assignments, which I mentioned briefly above.  In that sitting, Justice Ginsburg had two majority opinions:  Florida v. Powell (a 7-2 opinion regarding Miranda rights) and Black v. United States (which became the second "honest services" ruling).  Justice Kennedy did not have a December opinion.  But I don't think that Kennedy lost a majority that sitting.  Instead, it appears that the two "honest services" cases from November (Black and Weyhrauch) were put on hold pending the expedited February argument in Skilling.  When Justice Ginsburg was given that February assignment, she also seems to have picked up the ancillary opinion in Black.  (No opinion was issued in Weyhrauch.)

Assuming I'm right about what happened in Bilski, three questions remain:  what would the Stevens opinion have said; when did Stevens lose the majority; and what is the significance of the lost majority for the development of patent law.

We cannot simply assume that had Justice Stevens kept his majority his opinion would have taken the categorical position that business methods are not patentable subject matter.  His opinion might originally have been somewhat narrower, because he needed to keep Scalia's fifth vote (and perhaps Breyer's as well).  Or the opinion originally may have been a plurality.

Though we cannot know, it seems most likely to me that Justice Stevens did originally write an opinion taking an absolute or near-absolute position against business method patents, but Justice Scalia concluded that he did not want to go so far and instead joined part of the Kennedy opinion.

We also cannot know at what point in time Justice Stevens lost his fifth vote.  But my best guess is that it was relatively late in the process.  In February, Stevens was not assigned a second majority opinion but Kennedy was, which would be a little unusual in terms of balancing the Court's workload.  More important, as I mention above, Justice Scalia's position in the case "“ joining part of the majority opinion and part of the Breyer concurrence "“ has the feel of an awkward, late compromise.  It suggests that there was not time for Scalia to better articulate his position.

Finally, there remains the "so what" question.  Does all this Supreme Court Kremlinology have any practical consequence?  On the broadest level of whether the Court might revisit Bilski in a later case when Justice Scalia has the opportunity to consider the matter further, I think the answer is clearly no.  In statutory cases like this one (even those in which the statute's construction is informed by a constitutional provision "“ here, the Patent Clause), the Justices try to adhere to stare decisis.  The decision in Bilski settles the question that business methods are patentable subject matter until Congress decides otherwise.

If that issue were to return to the Court soon, the apparent switch in Bilski may inform the Court's approach to the major issue left open by the decision:  just how broadly are business methods patentable.  Scalia seemingly is a fifth vote to construe patent eligibility very narrowly.

But in reality, it is impossible to know how the Court would rule.  Justice Stevens has retired, and Elena Kagan's views on the issue are unknowable.  Other changes in the Court's membership are of course inevitable as well.

Posted in Bilski v. Kappos, Everything Else