« JOTWELL: Leong on Mika on gender disparity before SCOTUS | Main | Ideological Diversity and Party Affiliation »

Saturday, June 17, 2017

SCOTUS Symposium: Patent Law & Legal Process

This past Monday (6/12/17), the Supreme Court issued its sixth (!) patent case of the term, Sandoz v. Amgen. The lone copyright case for the Term, Star Athletica v. Varsity Brands, came down in late March. The lone trademark case for the term, Lee v. Tam, has yet to issue. The Court in Sandoz reversed the Federal Circuit in part, and affirmed it in part (though on a different rationale). Relative to the other five cases, that’s an improvement for the Federal Circuit; in those other five, the Court simply reversed the Federal Circuit outright. Across all six cases, there were dissents in only two (Impression Products, and SCA Hygiene), and both were lone dissents. The Court largely agrees that the Federal Circuit is largely wrong.

Professor John Duffy, an especially astute observer of the Supreme Court’s return to patent law in the mid-1990s (see, e.g., The Federal Circuit in the Shadow of the Solicitor General, 78 Geo. Wash. L. Rev. 518 (2010)), has—at SCOTUSblog—summarized the Court’s decision in Sandoz with his customary clarity. That’s no small feat, given the complexity of the biologics/biosimilars statute at issue in the case. More important than the summary, though, are Professor Duffy’s observations about the broader relationship between the Supreme Court and the Federal Circuit’s patent law decisions. First, observes Duffy, the Court is deciding more patent cases: “Now the court has been averaging over three cases per term for several years, and two more are already slated for argument next term. The lesson to patent lawyers is clear: Every significant issue in patent law could end up at the Supreme Court.” Second, the two courts are at odds in patent law: “the justices have not—to put it mildly—provided a ringing endorsement of the Federal Circuit’s patent jurisprudence. That provides an especially hard lesson for lawyers: They constantly face the challenge of preparing their patent cases for two audiences of appellate judges who often see the law in systematically different ways.”

The October 2016 Term parallels, in these respects, the October 2013 Term, in which the Supreme Court heard six other patent law cases on review from the Federal Circuit. None of those case drew a dissent, and only one (Alice Corp v. CLS Bank) drew a concurrence (and a brief one at that, comprising a single 3-sentence paragraph plus citations). The Supreme Court reversed, 9-0, in five of the six cases; it affirmed, 9-0, in the sixth.

I began looking at the Supreme Court’s post-1993 patent law decisions more systematically after the 2013 Term. There are 46 such cases, starting with Asgrow Seed v. Winterboer, 513 US 179 (1995), and ending with Sandoz v. Amgen. The 5-year running average of patent cases per Term, now at 4, is at a level not seen since the early 1940s (as I read Fig. 2 in Duffy’s 2010 paper). Below is a graph of the running average. For comparison, I include the 5-year running average for the sum of the copyright and trademark cases, which—even together—run well behind.


There are many things to say about this run of patent cases. But I think it’s fair to summarize the overall pattern as follows: 2/3rds of the Supreme Court’s patent law decisions in this period are unanimous (and 91% are 9/0, 8/1, or 7/2); 2/3rds of them are flat reversals of the Federal Circuit; and authorship of the opinions is spread across the Court’s full ideological spectrum. Justice Thomas has written 11 of the majority opinions, and Justices Breyer and Sotomayor have each written 5 of the majority opinions. In short, there is full Court agreement on largely unanimous rejection of Federal Circuit patent law decisions, even as it reviews those decisions more often.

What’s happening? It’s not just boredom, which would explain unanimity but not an increase in cases. It’s not an “IP is important!” vibe, which would show an uptick not only in patent but in copyright and trademark cases too (not seen) and would not explain the unanimity or the reversals. It’s not even a “patent law is important!” vibe, which—again—would explain neither the high unanimity rate nor the high reversal rate. There’s a growing literature that explores the issue, and I’m working on a draft paper myself. The full answer must wait. But, my short answer is, the Legal Process—i.e., Hart & Sack’s The Legal Process (1958). The strongest tell, in this vein, is this Term’s decision on the patent venue statute, TC Heartland v. Kraft Foods Group. The opinion, 8-0, is an exemplar of Legal Process reasoning ~ heavy reliance on stare decisis, and a keen appreciation for the different institutional roles that Congress and the courts play in providing settled ‘rules of the road.’ And this is so even though it makes patent law very much unlike other federal cases, which are controlled by the general federal venue statute. (The June 12 episode of the First Mondays podcast astutely notes this facet of TC Heartland.)

Perhaps the Supreme Court has found, in patent law (which is exclusively federal, and which has a weak agency and a governing statute that includes at least some text dating back to 1793), a domain in which it can express core Legal Process values that all the justices endorse, and that all can equally well articulate for a frequently unanimous court. In many areas, the justices go at each other hammer and tong, with many 5-4 and 6-3 decisions. In patent law, by contrast, the justices calmly reflect the whole Court’s commitment to Legal Process norms (weathered a bit, of course, by law & econ critiques from the right and CLS critiques from the left). The Court’s citation patterns in these case are certainly consistent with a prominent role for stare decisis. I’ve been tallying, from these 46 Supreme Court patent cases, all the other Supreme Court cases cited at least once in a given case, as well as the Courts of Appeals cases cited (from both the Federal Circuit and the regional circuits).

A few observations about the citations, across all 46 cases . . .

The Court cited a total of 669 other Supreme Court cases, 326 patent cases and 343 nonpatent cases. The median number of other S. Ct. cases cited, in any given case, is 12.0 (and the average is 14.5). The Court cited a total of 216 Court of Appeals cases (less than a third as many as S. Ct. cases), 98 from the Federal Circuit and 118 from the regional circuits. (Of the 118 cites to regional circuit cases, 25 were to the 2nd Circuit and 20 were to the 7th. The next highest was the 3rd Circuit, with 12.) The median number of Courts of Appeals cases cited, in any given case, is 2.5 (and the average is 4.7)—about 1/5th the median number of S. Ct. cases cited per decision.

The age range of the other S. Ct. cases the Court cites (in the 46 patent cases) is large, stretching from 1803 to 2016. And though more recent cases are cited more frequently, the cites to older cases are by no means infrequent. The first graph below groups cited cases by decade, marking frequency of citation to cases in that decade and the cumulative percentage of the 669 cases contributed by that decade. Nearly 50% of the cases cited are from before 1977. The second graph below groups cases by how old they were when cited, again showing frequency and cumulative percentage. Nearly 50% of the cases were 50 or more years old when cited.



If one separates the 326 cited S. Ct. patent cases from the 343 nonpatent cases, the greater age of the cited patent precedents is even more apparent. Among the cited S. Ct. patent cases, the median year of decision is 1942 and the median age of the cases when cited is 65 years. By contrast, among the cited S. Ct. nonpatent cases, the median year of decision is 1986 and the median age of the cases when cited is 25 years.

Of course, as is common in studies of citation rates, a smaller number of cases garners a larger share of the citations. Among the cited S. Ct. nonpatent cases, only one—Franchise Tax Board of California v. Construction Laborers Vacation Trust, 463 US 1 (1983)—is cited in three different cases; the remainder are cited in two cases or, far more often, in only one case. Among the cited S. Ct. patent cases, 34 cases are cited in 3 or more cases. One (Bonito Boats v. Thunder Craft Boats, 489 US 141 (1989)) is cited in 10 of the 46 decisions, one (Diamond v. Chakrabarty (1980)) is cited in 6 of the 46, 7 are cited in 5 of the 46, 13 are cited in 4 of the 46, 12 are cited in 3 of the 46, and the remaining are cited in two or one. Among these same 34 cases, although 8 (23.5%) were decided after 2000, 7 (20.5%) were decided in the 1800s.

The challenge for patent litigators is even greater, perhaps, than Professor Duffy suggests. In the Supreme Court’s opinions explaining its approach to patent law, a 19th-century precedent is as likely to prove germane as a 21st-century precedent. This is certainly consistent with strong-form stare decisis, and the Legal Process jurisprudence that underwrites it. But is it the felt sense of most patent litigators about how to handle the legal infrastructure of their cases? I doubt it.

Posted by Joe Miller on June 17, 2017 at 02:50 PM in 2016-17 End of Term | Permalink


Post a comment