Reading Quanta Narrowly

David McGowan is a law professor at the University of San Diego School of Law. Much of his research involves the interplay between new technologies and antitrust law. I asked him to comments on the Supreme Court’s recent patent-antitrust decision: Quanta.

By David McGowan

To the extent possible, parties and not courts should set the terms for practicing inventions.  Parties have better information than courts; they are likely to set more efficient terms.  The Supreme Court’s opinion in Quanta Computer, Inc. v. LG Electronics, Inc, does not embrace this principle, though it may not reject it very broadly. The opinion may and should be narrowly construed.

I presume readers of this blog are generally familiar with the ruling. Intel took from LG electronics a license providing Intel a defense to infringement and immunity from contributory infringement by Intel’s customers, who knew Intel’s license did not extend to them.  The Court nevertheless held the Intel license exhausted LG’s rights insofar as Intel chips were concerned.

The Court’s analysis on the main point is too formal. Because the chips embodied substantially all the relevant LG inventions and had no substantial use unrelated to practicing those inventions, the Court thought it irrelevant that Intel customers knew their chips implied no LG license.

But an invention is an intellectual achievement distinct from its embodiment in a product.   There is no logical reason the sale of a product should exhaust rights in an invention.  There is a logical objection: Forcing patentees to load up initial sales with all the terms necessary to tailor rights to potential uses will tend to increase transaction costs of initial licenses and probably result in lumpier, less tailored contracting in general.  Better to let Intel buy its peace at a price that reflects its business and let its customers do the same. 

The statute does not compel the Court’s holding, which rests on history.  The Court should have limited rather than extended the cases it relied on.  These, such as Bauer and Motion Picture Patents, invoked exhaustion to defeat post-sale restrictions on use of a product.  But such restrictions may facilitate price discrimination or make sense for other business reasons.  Intel and LG, and firms like them, know better than courts when this is the case. 

Against this it might be argued that a robust exhaustion doctrine economizes on downstream transaction costs.  Conceptually that is true, and perhaps persuasive at the retail consumer level, but Quanta was not such a case.  Taken to its logical conclusion, this argument implies patentees are better off suing everyone and negotiating settlements than licensing anyone and giving up the ability to tailor terms.  That implication is perverse.

The Court’s devotion to history is odd because it has not shown much regard for history of late.  In eBay v. Mercexchange it was unimpressed by the long history of presumptive injunctive relief for infringement.  In the antitrust context, in Leegin Creative Leather it recently overruled the longstanding per se prohibition on minimum resale price maintenance.  And in Illinois Tool Works the Court reversed the relatively longstanding presumption that patents imply market power in tying cases. 

Quanta is at least in tension with Leegin and Tool Works.  Those cases recognize that post-sale restrictions may be efficient, a point Leegin makes directly and which is implicit in Section 271(d)(5) of the Patent Act, on which the Tool Works Court relied. In placing economic substance over transactional form, Leegin and Tool Works follow Continental TV, Inc. v. GTE Sylvania, which overruled the Court’s per se prohibition on vertical nonprice restraints.  Like exhaustion, that rule had been based on transactional formalism rather than economic analysis.

Both Leegin and Tool Works reject doctrines that trace to the same cluster of cases, such as Bauer and Motion Picture Patents, the Quanta Court cites favorably. Yet the Quanta Court said nothing to explain how its holding relates to antitrust precedents (or Section 271(d)(5)) recognizing that post-sale restrictions may be efficient.  One therefore might try to read Quanta as qualifying even its recent antitrust cases.

On this reading, at least insofar as pass-through restrictions are concerned Quanta implicitly qualifies the Federal Circuit’s Mallinckrodt precedent, which, in the misuse context, upholds patentees’ ability to impose lawful conditions on use.   Professor Ghosh has suggested (http://lawprofessors.typepad.com/antitrustprof_blog/2008/06/the-quandry-of.html) the Court’s footnote seven, which says exhaustion does not necessarily bar breach of contract claims, may point in this direction.

This reading of Quanta is neither compelled nor desirable.  It is possible to reconcile Quanta with the Court’s recent (and sensible) antitrust decisions by focusing on two things not at issue in Quanta. 

First, the initial license to Intel was not conditional.  The cases the Court cited also involved unconditional sales. Quanta therefore should not be read to extend to conditional licenses, and thus not to undercut Mallinckrodt directly.  Second,  Quanta presented no question regarding pass-through conditions and should not be held to govern them.  Intel merely bought immunity from suit for contributory infringement.  The Court did not stress this fact, but it is fair to repay the woodenness of its analysis with a strict reading of its holding.

A weakness of this reading is that it renders Quanta trivial.  One wonders why the Court would take a such a case.  Possibly the Court believes patent law and the Federal Circuit have got out of hand and must be reigned in.  But its majority opinions have not matched the robust commentary cert grants have prompted.  They are fairly narrow and unambitious.  So is Quanta, and it deserves to be read that way. 

Comments

[Mr. Miles]

Thank you Professor McGowan for the interesting analysis of the potential ramifications of the Quanta decision. I find this discussion particularly intriguing in light of a previous article on Patently-O that argued for reversal in Quanta, which the Supreme Court did. http://www.patentlyo.com/patent/2007/11/reestablishing.html

The consensus appears to be that patent owners should still have robust downstream contract rights, even if their downstream patent rights are now attenuated. The upshot is that Mallinckrodt won't protect such robust contract rights from attack under antitrust theories. Whether the benefits of antitrust defenses outweigh the additional transaction costs of antitrust analysis, nobody seems to say.

[AllSeeingEye]

Thank you Professor McGowan for a very well written analysis. Are there additional facts that may weigh in to the overall scenario that may distinguish this case from pure downstream rights protection? Such as concerns over Intel's ability to use downstream licensing power as a way to support or manipulate prices?

[GP]

So Fannie and Freddie are going to be nationalized.

How long do you think it's going to take before the government nationalizes patents?

I'd give it no more than 5 years.

[AllSeeingEye]

I think the fact that the PTO is a government agency is a fairly strong clue that "patents" are already "nationalized."

[Alan McDonald]

I disagree with Professor McGowan's statement that Quanta does not properly follow other recent Supreme Court decisions.

I believe that all of the decisions have a common theme: patent were seen as too powerful in today's marketplace and that power had to be pulled back.

I will leave to others whether that is a good or bad thing.

[Robert]

LG had every opportunity to place conditions on Intel's sale of articles that embodied their patent claims. Instead, LG opted to permit Intel to sell the articles without conditions. Thus, the sale of the articles exhausts the patentee's ability to control postsale use of those articles via patent law.

I agree that this result is unremarkable. Nevertheless, the case is remarkable in overturning the Federal Circuit's twisted logic regarding the application of exhaustion to method claims.

[Anon E. Mouse]

I agree with the vast majority of your very well-thought out piece, Professor. I've always thought Quanta was decided on very narrow grounds: I figured the only point at issue in the case was whether the sale was conditional.

SCOTUS said it was conditional, ergo, exhaustion. Lesson learned: expressly condition the license.

You wrote:
"...weakness of this reading is that it renders Quanta trivial. One wonders why the Court would take a such a case."

Actually, I saw this as a non-trivial case. First, patent exhaustion was solidified to mean what it meant in the earlier cases (unconditional sale of an item embodying the invention exhausts the patent). This limits exhaustion from the over-broad and over-narrow readings for which some pundits were pleading.

Second, it smacked down the utterly bizarre and relatively new Federal Circuit case law that said a method *cannot* be exhausted. This needed to be nipped in the bud -- go back and follow the case law as this notion evolved. The original Fed. Cir. cases were not so broad (holding that the sale of an item not embodying the invention or having non-infringing uses thus did not exhaust a method claim) and perhaps poorly worded if read out of context. Each subsequent case seemed to forget a piece of the last and this bizarre creature (the exhuastion-free method claim) evolved into being.

In my (pro-licensing!) mind, this was the most important point of the Quanta case.

One other minor point: SCOTUS did clear up the lingering question whether exhaustion crossed patent lines (i.e., whether a sale of Item A, which embodies the '123 patent and the '456 patent, under an uncondition license to sell that explicitly lists the '123 patent and does not mention the '456 patent, exhausts the '456 patent). Now we know that it can. Certainty is a good thing.

[Anon E. Mouse]

pardon the above typo.

Delete: "SCOTUS said it was conditional, ergo, exhaustion. Lesson learned: expressly condition the license."

Replace: "SCOTUS said it was unconditional, ergo, exhaustion. Lesson learned: expressly condition the license.

[Paul F. Morgan]

1. "More efficent licensing terms" for WHOM is the more appropriate Quanta question.

2. This is not an antitrust case. Highly disfavoring restraints on the allienation [subsequent purchasors or users] of chattels goes back to old English common law.

[EG]

"Yet the Quanta Court said nothing to explain how its holding relates to antitrust precedents (or Section 271(d)(5)) recognizing that post-sale restrictions may be efficient. One therefore might try to read Quanta as qualifying even its recent antitrust cases."

Professor McGowan,

Thanks for addressing a glaring omission in the Quanta opinion, namely how does it square with Section 271(d) which the patentee-Respondent and all but one Amicus in support of the patentee-Respondent asked the Supreme Court to address, including one Amici who suggested 271(d) might have done away with the patent exhaustion doctrine. In this regard, it looks like the Supreme Court blipped over more than just Dawson Chemical Co. v. Rohm and Haas Co., 448 U.S. 176 (1980) (cited in the AIPLA Amicus brief) where the Supreme Court did discuss the impact of 271(d) as permitting patentee to exercise control over nonstaple articles used in their inventions, at least from the standpoint of contributory infringement. It's one thing for the Supreme Court to say that 271(d) doesn't apply to patent exhaustion or to the licensing arrangement between Intel and LG Electronics; it's another thing for them to not even address this issue after it was raised not only by the patentee-Respondent, but also by multiple Amici in support.

[M. Samardzija]

Excellent article. As an alumnus of the University of San Diego School of Law, I am glad to see that they continue to hire great legal scholars. I just hope that they can retain these great legal scholars.

[Anonymous and Cowardly Patent Attorney]

CONTRACT YOUR WAY AROUND QUANTA:
The SC found exhaustion here on the thinnest of salami slices: there was no explicit prohibition on sales to 3rd part for improper combination (though it was clear such actual combinations were not allowed by 3rd parties -- the implication being that such sales should not be allowed either). To avoid exhaustion, the court is all but explicitly saying, just prohibit the sale by Intel to 3rd parties for combination. They all but give a roadmap to avoiding exhaustion (between contracting parties) in this case. This is indeed a very narrow ruling, and I have no idea why they took it up. I think the supremes had buyers remorse after they took it and gave it to Thomas because he didn't ask any questions during orals.

[Licensor]

Questions for the group:

What would have been sufficient conditions here to save Quanta's downstream patent rights?

Would either of the following limitations on licensed products work:

1) Chips with Quanta's technology combined only with Intel chips?
2) Chips branded solely with Intel marks?

How about this limitation in the grant clause:

The right to make use and sell, but only sell in combination with Intel chips, ...?

[Anonymous Cowardly Patent Guy]

Licensor:

Your question is a bit mis-placed: the question should be, what would have saved the transaction from exhaustion... the question you are asking is broader: it goes to the questions concerning the limits of conditional licenses.

1 and 2 do not seem to cure the defect the SC picked up on -- that the sale from Intel to Quanta was not prohibited.

Your last one is closest -- in the grant clause is great. Bottom line: let nobody that reads that license in any way think the sales for combination w/ non-Intel chips was "authorized" by the license agreement.

[Andrew Dhuey]

ACPA, whether Justice Thomas asks a question at oral argument has nothing to do with who writes the majority opinion. He asks a question once every blue moon -- less than that, actually.

[`bEn]

Why did the SC not discuss the fact that patent exhaustion depends on the country in which the patented method/apparatus is places in circulation?
The prerequisites for exhaustion are that the patented apparatus or the product of a patented method is placed in circulation IN A SPECIFIC AREA by the patent proprietor or a third person acting with the approval of the proprietor.

[ak]

Another case of patent exhaustion with Intel:

http://www.marketwatch.com/news/story/amd-says-intel-threatening-end/story.aspx?guid={C47C1DB9-374F-46B9-981A-A15554FDB6E8}&dist=msr_1