Patent Chaos and "Dorf's Law"

Writing in the NY Times on Tuesday, Adam Liptak reported that, as a result of analysis by GW Law Prof John Duffy, the courts may have to "undo thousands of patent decisions concerning claims worth billions of dollars." Duffy's analysis is excellent, but I want to suggest here that the bigger the problem is, the smaller the problem will be. How's that? Let me explain.

In a little-noticed provision of a 1999 appropriations measure, Congress changed the way that patent judges of the Board of Patent Appeals and Interferences are appointed. As a result, all such administrative judges appointed since 2000 have been named by the Director of the Patent and Trademark Office, but the Director is not the Head of a Department (since he serves under the Commerce Secretary). Thus, these judges were appointed in violation of the Appointments Clause of the Constitution, Art. 2, sec 2, cl. 2. And thus, cases on which they have sat are invalid. Duffy argues that because of the breadth of standing the Supreme Court has permitted in Appointments Clause challenges, and the automatic nature of the remedy, there is no good way for the government to avoid this result---although prospectively, Congress can vest appointment authority of the patent judges in the Commerce Secretary.

I would argue that the government may indeed have a way out, which I'll call "Dorf's Law." Dorf's Law states that courts do not provide a remedy for really really big constitutional problems where doing so would create chaos. One example of Dorf's Law in the United States may be McCleskey v. Kemp, in which the Supreme Court refused to invalidate Georgia's death penalty despite evidence showing with statistical significance that the race of the victim was a determinative factor in whether a defendant was sentenced to death. The Court feared that invalidating the death penalty on this basis would, among other things, require acknowledging that racial considerations pervade the criminal justice system, and the Justices couldn't imagine invalidating the entire criminal justice system. As a result, Justice Powell wrote a majority opinion that is, to put it kindly, innumerate. It says (nonsensically, for anyone who understands statistics) that proof that race infected death penalty decisions in Georgia in general doesn't count as proof that race infected the death penalty decision in any particular case. (This is nonsense because the Court was NOT saying that there is a higher persuasion burden where statistical proof is used, nor was it saying that there was something about McCleskey's case that made it atypical.)

As McCleskey illustrates, it's possible (perhaps likely) that examples of Dorf's Law will not be expressly acknowledged as such. Towards the end of his opinion, Justice Powell does candidly say that the Court can't give McCleskey the relief he wants because doing so would open the floodgates, but he offers this as an "additional concern" that is meant to bolster his decision nominally based on the statistical (non)argument.

My favorite illustration of Dorf's Law is the Supreme Court of Canada's decision in the Manitoba Language Rights Case. Manitoba was obligated to enact all laws in French and English, but from 1890 until 1985, the Province only enacted laws in English. The Supreme Court had little difficulty finding that practice invalid, but balked at the suggestion that as a result, all laws in Manitoba since 1890 were null and void. Such a legal vacuum, the Court said, would itself violate the rule of law. Thus, the Court continued in effect the prior laws for a temporary period while Manitoba translated and re-enacted its laws.

Dorf's Law suggests a legal strategy for the government in the patent imbroglio: Emphasize the utter waste and chaos that would ensue from declaring void all the decisions rendered by the improperly appointed patent judges. The courts (that's the Article III courts that would adjudicate this issue, presumably including, eventually, the U.S. Supreme Court) could either follow the McCleskey route and make up some nonsensical reason why the Appointments Clause isn't really violated, or they could follow the (more honest) Manitoba route and say simply that the remedy is too costly. This is a risky strategy primarily because the chaos that would ensue from invalidating eight years worth of patent decisions isn't quite on the scale of the chaos of having no law at all in an entire Canadian province or no criminal law in the United States.

Still, the Supreme Court did something a lot like this once before in a context that looks quite similar. In Northern Pipeline Co. v. Marathon Pipe Line Co., after finding that bankruptcy courts had been improperly constituted, the Court decided that its ruling only applied prospectively, and gave Congress a few months to fix the problem. In the ensuing years, the Court has ruled out pure prospectivity, but if there's enough at stake, look for something like it to re-emerge in the patent cases.

Posted by Mike Dorf