Mises Wire

Home | Wire | Patent Exhaustion

Patent Exhaustion

Okay, this post is a bit inside baseball. But some lawyers and law students may at least find this of interest. In an upcoming case, Quanta v. LG Electronics, the Supreme Court may refine the "patent exhaustion" doctrine to make it more difficult for patentees to extract royalties from multiple parties for the same device. I find this case and the arguments presented absolutely fascinating (a bit morbidy, since I am opposed to the notion of patent law, and also dislike the positivistic and utilitarian assumptions that permeate these arguments; but still). It's a Supreme Court case that was recently argued; we patent lawyers await the decision, probably coming in July or so.The basic issue is patent exhaustion and the ability to override it contractually. Patent exhaustion means that if you authorize someone to sell an item covered by your patent (or if you sell it yourself), and the only reasonable use of the product is covered by the patent, then you can't sue any more under patent rights; you've "exhausted" them. As the Court has said, "[W]here one has sold an uncompleted article which, because it embodies essential features of his patented invention, is within the protection of his patent, and has destined the article to be finished by the purchaser in conformity to the patent," the patentee "has sold his invention so far as it is or may be embodied in that particular article." So, if I sell you my patented mousetrap, I can't sue your customers for patent infringement. This case involves LG's license to Intel to sell microprocessors covered by LG's patents. LG wants to sue Intel's customers who used those chips in their computers. Here's a good summary of the LG case. Fascinating excerpts from oral argument can be found here; full argument transcript. There's also a huge collection of amicus briefs (showing how significant this case is, and also how much patent law affects the economy). It's not clear who has the "better" argument. It seems to me that Quanta has the "better" argument, but on the other hand, it's hard to be objective since I want Quanta to win, so that patent law and the power of patentees to license is weakened. In any event, the legal arguments here are extremely interesting (to those of the right bent). I find Quanta's arguments compelling, as well as those of IBM and Dell. The EFF's brief is also interesting. For those interested take a look at the summary linked above; then read the "Summary" section" which summarizes each amici/party's argument. My guess is that the court won't want to let LG have 2 bites at the apple (getting a royalty from Intel, then another from its customers), but because this case is impossible to solve "correctly" (because the patent statute is simply incoherent and leads to impossible-to-resolve cases), it will probably do something along the lines suggested by IBM (a sort of compromise position, "between two extremes"). When you read the arguments by LG and its supporters (including Richard Epstein, by the way, on the wrong side of this) you see that this is an example of why the Crits are right; there is no "right" answer here, because the justifications of the various laws are incoherent and their various "goals" incompatible (since this is all artificial law). However, as noted, I favor the Quanta side, think their argument (at least the moderate version of IBM) is more legally persuasive, and also more policy-persuasive, and also predict the Court will at least partially reverse the CAFC and side with Quanta. But who knows. One problem with Quanta/IBM winning (and this illlustrates that the law is really not determinative or even coherent) is that IBM argues that LG, say, could make a deal with Intel to license to Intel the right to sell Intel chips that are covered by LG's patents, and authorize the sale by Intel to third parties; or it can specify that it's not authorized, unless Intel makes its customers agree to pay royalties to LG. I.e., they want it to be a contract action only, not a property action (hence Breyer's seemingly-strange comment about "equitable servitudes on chattel"). So what IBM is saying is LG can only sue Intel's customers, if the sale is not authorized. But if the sale is not authorized, then Intel itself would be infringing and could be sued too. Now I can see one way the patentee might try to draft around this. A license to Intel only means LG promises not to sue them. So let's say that LG does what IBM proposes: it does a license with Intel that authorizes Intel to sell chips ONLY to customers who first contact LG and sign a separate license with LG. This would mean that if intel just sells chips to Quanta, without first making sure Quanta has a license, Intel is using patented technology of LG without permission, and may be sued. And so may Quanta. Fine. But suppose LG sues Intel and Quanta, then immediately settles with Intel for a dollar. See where I'm heading? Or, suppose that, instead, it inserts clause 31.3 in the license agreement where it promises not to sue Intel, if Intel pays it a dollar. Etc. So now we have creative drafting that provides (as LG wants now) that the sale to the customer is not authorized, so patent exhaustion is not triggered (and the customer, Quanta, may be sued for infringement). But Intel can't be sued because LG agreed not to sue them. So this seems a way parties could get around a ruling in Quanta/IBM's favor. Unless ... the court were to hold in this case that because Intel can't be sued, the sale is authorized, and therefore there is patent exhaustion. Sheesh. It's hard to predict.

Stephan Kinsella

Stephan Kinsella is an attorney in Houston, director of the Center for the Study of Innovative Freedom, and editor of Libertarian Papers.

Shield icon wire