Last month, we briefly discussed an article that proposed that “baseball-style” arbitration be used to resolve FRAND licensing disputes. The following guest post about this article was authored by David Balto, a former Policy Director of the Bureau of Competition for the Federal Trade Commission who currently has his own public interest antitrust practice here in Washington, DC. David’s views expressed below are his own, and do not necessarily reflect the views of the authors of The Essential Patent Blog, Dow Lohnes PLLC, or Dow Lohnes’s clients.
Baseball has the Best Rules: Using Arbitration to Solve FRAND Disputes
How to calculate Fair, Reasonable and Nondiscriminatory (FRAND) royalties seems like one of the most intractable problems firms, standard setting organizations and the courts are grappling with. No wonder, there is sparse authority and relatively few litigated cases.
Two of the most thoughtful scholars on antitrust intellectual property issues — Professors Mark Lemley and Carl Shapiro — have weighed in and issued a paper outlining an interesting solution to the FRAND licensing problem and it provides a clarion call for how to grapple with the problem.
The FRAND licensing problem seems complex: When a standard setting organization (SSO) creates a technology standard that includes patented technologies they will require the patentees to agree to license those patents on FRAND terms. The problem is that FRAND is poorly defined. This can result in abuses by both patentees and potentially licensees, and ultimately result in drawn out court battles where judges have to determine what the FRAND rate actually is. The stakes for these court battles are high, creating an incentive for firms to battle from the extremes — to build a case for an extremely high rate or an extremely low rate knowing that a judge will pick something in between.
Lemley and Shapiro provide a simple and thoughtful approach by suggesting that SSOs require binding baseball-style arbitration as a part of FRAND commitments. In baseball-style arbitration each party presents one number and the arbitrator is only tasked with choosing which number is a more accurate representation of the FRAND licensing rate.
This proposal is so ingenious because it uses game theory (and common sense) to create an incentive for the parties to choose reasonable rates instead of trying to substantiate an extreme rate. Each arbitration is like a game with one winner. A player seeking to maximize her odds of winning will have to balance the reward of a better number against the risk that that number won’t be picked. To demonstrate this I will walk through a hypothetical example where most evidence points to a real FRAND rate between 5 and 8.
It immediately becomes apparent that it would be hard for the parties to pick a number lower than 5 or higher than 8. If a licensee wanted to pick an unreasonably low number like 1 it will have to deal with the risk that the patentee will pick 8. With 8 the patentee maximizes their revenue while still having the evidence to make a plausible case. The licensee does not have a plausible case with 1 and will almost surely lose. Therefore the licensee will want to minimize their costs by picking 5. The outer bounds of the numbers picked will be restrained by what the evidence is the most likely to show is reasonable. If the licensee thought that their chance of winning is only 50% with 5 against 8 but 90% with 6 against 8 then they will have the incentive to offer even more. This is because over repeated games the average cost will be 6.5 in the first instance but only 6.2 in the second instance. The patentee, knowing this, will also be motivated to offer a lower number and eventually an equilibrium will be reached.
But what if there are outside forces that motivate a party to strategically pick an extreme number? For example, the negotiating parties are fierce competitors and the patentee thinks it has more to gain by taking the (small) chance on a crippling high rate than it has to lose on getting stuck with a low rate. The Lemley/Shapiro proposal addresses this by requiring disclosure of all previous license agreements dealing with the patents in question. If the patentee loses this game then it risks lower FRAND rates in the future because any future arbitrators would look to that agreement as evidence of the real FRAND rate. This risk also applies to the opposite set of circumstances. For example, a patentee is dealing with a licensee that could be a major threat to one of the patentee’s competitors. Here the patentee might have the incentive to license her patents at a nominal fee under the assumption the licensee uses the lower cost advantage to destroy the patentee’s competitor. This kind of agreement would be forced into the open because as extremely favorable rate for one licensee can become the basis of deciding FRAND rates for all licensees.
The Lemley/Shapiro proposal also deals with the problem of unwilling licensees. A patentee subject to FRAND would have to agree to submit to binding arbitration, but any potential licensees are under no such commitment. This could potentially allow these implementers to refuse to participate in arbitration. Under these circumstances the proposal allows the patentee to sue in court for damages and injunctive relief. Giving the patentee this power is essential — as I have written in comments to the FTC, this problem of licensee hold out can be significant.
The proposal does not bar courts deciding non-FRAND rate matters. Courts can be called on to determine whether a patent is valid or infringed. Evidence that a patent is not valid or not infringed can also be used by arbitrators in choosing between the rates. The more evidence a licensee can present that a patent is invalid the more likely the arbitrator will be to pick a lower number as the FRAND rate.
The Lemley/Shapiro proposal deserves careful consideration and it should be required reading for all standards setting organizations.
David Balto is an antitrust lawyer in Washington, D.C. and is the former Policy Director of the Federal Trade Commission. He represents standard setting organizations, retailers and others in IP issues. He can be reached at firstname.lastname@example.org. He is grateful for the assistance of Matthew Lane (University of Florida Law School) in the preparation of this article.