Wednesday, July 9, 2014

Another View of FRAND: ALJ Essex's Initial Determination in InterDigital

As reported last week on the Essential Patents Blog, Administrative Law Judge (ALJ) Theodore Essex recently released the June 26, 2014 public version of his initial determination in In the Matter of Certain Wireless Devices with 3G and/or 4G Capabilities and Components Thereof,  United States International Trade Commission Investigation No. 337-TA-868.  InterDigital accused Nokia and ZTE of importing wireless devices that infringed certain U.S. patents.  The ALJ concluded that the accused products did not infringe (and that one asserted claim of one patent was invalid for indefiniteness).  Nevertheless, beginning at page 108 (Part VII) of the initial determination, the ALJ addressed the respondents' arguments that InterDigital violated its obligations to license the patents in suit under FRAND terms, in accordance with ETSI policy.  (The ALJ reasoned that he should address this issue, even though his finding of no infringement meant that the patents were not essential, and therefore not subject to InterDigital's FRAND commitment, in view of the possibility that the Commission would reverse on some or all of the infringement findings.)  Here are some highlights of ALJ Essex's analysis:

1.  The ETSI Rule of Procedure that obligates members to license SEPs on FRAND terms is not a contract.  Moreover, the patent owner "does not agree to license the intellectual property owned under FRAND terms, but only agrees to do so under certain conditions," in this case subject to a reciprocity condition (pp. 110-11).  Readers may recall that in the Microsoft, Innovatio, and Apple v. Motorola district court cases (the one before Judge Crabb), the courts concluded that the relevant FRAND commitments did create binding agreements.  As I have discussed previously (see this paper at pp.5-8, 37-38 and this blog post), courts in other countries have tended to disagree, and commentators have expressed a range of views.

2.  The ALJ also noted that parties or third parties have the option of complaining to ETSI if they believe an SEP owner "is refusing to grant a license on FRAND terms," and that the respondents here did not take advantage of this option (pp. 112-13).  Here then is the most significant part of the ALJ's FRAND analysis (pp. 113-14, 117-18):
These Respondents chose take the actions that led to the allegation of infringement rather than follow ETSI policy for obtaining a license. (They only potentially infringe, because while so called “Standard essential patients” must be declared to the organization in a timely fashion (CX-2555C, paragraph 4.1), there is no check by ETSI or any organization that the patents actually do read on the standard.) The Respondents create, outside of the framework of the ETSI agreement a situation where they use the technology that may be covered by the patent, without having licensed it. This puts pressure on the IPR owner to settle, as the owner is not compensated during a period of exploitation of the IP by the unlicensed parties. The ETSI IPR Policy requires companies that wish to use the IPR covered by the agreements to contact the owner of the IP, and take a license. (CX-3860C.) By skipping this step, the companies that use the IPR in violation of the policy are able to exert a pressure on the negotiations with the IPR holder to try to make the agreement in the lower range of FRAND, or perhaps even lower than a reasonable FRAND rate. They also are able to shift the risk involved in patent negotiation to the patent holder. By not paying for a FRAND license and negotiating in advance of the use of the IPR, they force the patent holder to take legal action. In this action, the patent owner can lose the IPR they believe they have, but if the patent holder wins they gets no more than a FRAND solution, that is, what they should have gotten under the agreement in the first place. There is no risk to the exploiter of the technology in not taking a license before they exhaust their litigation options if the only risk to them for violating the agreement is to pay a FRAND based royalty or fee. This puts the risks of loss entirely on the side of the patent holder, and encourages patent hold-out, which is as unsettling to a fair solution as any patent hold up might be. This has happened with patents from one of the patent families in this case, in 337-TA-800, where the patents that were declared SEP were found not to be essential.
Using the “patented” technology prior to negotiating with InterDigital for a license is a violation of the ETSI Rules of Procedure as well. (CX-2555C; CX-3860C.) While this section of the ETSI rules requires the IPR holder to be prepared to offer a license, it also requires the companies that would use the technology to seek a license as well (The above undertaking may be made subject to the condition that those who seek licenses agree to reciprocate) (CX-2555C.) Within the four corners of the agreement, there appears to be no provision made for companies that simply choose to infringe, and then demand FRAND status when caught. The requirement to negotiate a license rests not just on the IPR owner, but on those companies that would use technology prior to engaging in the [potentially] infringing activities. . . .
While the respondents followed none of the provisions of the ETSI policy that would give them standing to complain about the behavior of complainant, their arguments that complainant refused to negotiate in good faith is baseless. First, it is reasonable to note that InterDigital is not engaged in manufacturing handsets or any cell phone equipment. Their business model requires them to license their portfolio to generate revenue. If they should refuse to license their portfolio, or license it at a rate that put their licensee(s) at a competitive disadvantage, the threat to their business would be both immediate and real. In as much as a FRAND license is one that allows for a profit for InterDigital, and allows the licensee to compete and thrive in the market, InterDigital must attempt to make certain all of its licenses are granted on FRAND terms. Respondents, however, have no such need to see to the health of InterDigital. If InterDigital failed and was no longer in business, each respondent would be able to continue at least as profitably as before, and perhaps more so. Thus, in looking to the interests of the parties, holding out meets the interest of the respondents, but if the complainant should “hold up” the respondents, they will suffer losses along with the licensee. The negotiations of the license agreements are complex business dealings, and, in this case, they are being conducted with the backdrop that each day they are not concluded, the Respondents have not had to pay anything for a license they were by ETSI policy to obtain prior to adopting the potentially infringing technology. (CX-3860C.) While the possibility or existence of an exclusion order may benefit InterDigital in negotiating a license, and move the license fee in the upper direction on the FRAND scale, there are hundreds of other economic factors that go into the parties finding a royalty or flat amount both can agree on.
The ALJ then proceeded to analyze the course of the parties' negotiations, most of which is redacted, before concluding (p.123):
The obligation that InterDigital has taken has been fulfilled, and the ETSI agreement anticipates that the parties if necessary will fall back on the national law involved. The Respondents have not taken the steps provided by ETSI to address a failure to license, and so have not done what they ought to if they believe InterDigital has failed to negotiate in good faith. Finally, they have not followed the ETSI process for procuring a license, and have engaged in holdup by making the products that are alleged to infringe before taking a license. Under these facts there is no FRAND duty.
Finally, on the issue of patent holdup, the ALJ asserts (pp. 123, 126):
The essence of the arguments made by the FTC and PTO/DOJ against the availability of exclusion orders is that by having an exclusion order hanging over the negotiations, there is a risk of “Patent holdup”: that is the owner of the IPR may obtain remuneration beyond the value of the IP, because it is a standard. There is no evidence that is the case here. . . .
. . . While there may be a hypothetical case of holdup, we have evidence that it is not a threat in this case, or in this industry. . . .
For the Commission to adopt a policy that would favor a speculative and unproven position held by other government agencies, without proof that the harm they considered exists or that the risk of such harm was so great that the Commission should violate its statutory duty would damage the Commission's reputation for integrity, and violate its duties under the law. We should and must determine the public interest, and the correct outcome of each matter based on the facts presented, and by applying the law to those facts. To take a pre-set position, without hearing evidence, would violate every concept of justice we are tasked to enforce. Within the facts of this case, the Letters of the DOJ/USPTO and FTC do not assist or enlighten. 
3.  The ALJ also considered the "public interest" factors, and concluded that "the public interest does not support using 'FRAND' to deny an exclusion order" (pp. 173-80).

I appreciate the difficulties faced by the Commission in such cases, given that the only remedy it can offer is injunctive relief, not damages--a matter that, in my opinion, should be legislatively changed, however.  (For that matter, I'm on record suggesting the abolition of the ITC, which provides an unnecessary parallel patent litigation forum of a type that, to my knowledge, is found in no other country save South Korea.)  Nevertheless, I don't find the ALJ's positions on "holdup" and "reverse holdup" persuasive.  To state that an implementer must conclude an agreement before launching a potentially infringing product, or else face the risk of injunctive relief, seems to me (in the SEP context) to provide patent owners with far too much leverage to demand terms that will reflect not only the value of the patented technology but also the implementers' cost of switching (which is the definition of patent holdup as I understand it).  Put another way, if I am reading the opinion correctly, the ALJ's view means that, if an implementer believes in good faith that a patent is not infringed or invalid and therefore goes ahead and engages in allegedly infringing behavior, the patent owner (if it winds up prevailing on infringement and validity) can get an injunction, which seems to me to create a serious risk of holdup.  To be sure, it's not holdup in the stronger sense of keeping new technology off the market--but as I have argued before the argument that that is the only kind of holdup that matters is a little like the argument that the Sherman Antitrust Act of 1890 was an ill-advised innovation because the price of kerosene offered by Standard Oil back then was falling, not rising.  (See my posts of June 11 and June 13.)   Moreover, I'm not convinced by arguments that patent owners will be over a barrel because implementers can use their technology for a time without compensating them.  That's often true in patent litigation generally, and the response is that if the implementer loses it has to pay compensatory damages, plus (in most of the world's major patent systems, see my book pp. 147, 209, 276-77, 328, 370) prejudgment interest, and its own attorneys' fees; plus (potentially, depending on the jurisdiction and the underlying facts) the other side's attorneys' fees and enhanced damages.

The case does raise an interesting question, though, about what it means for an implementer to be a willing licensee.  Even I agree that if an implementer utterly refuses to pay a license for a clearly valid and infringed patent, injunctive relief may be warranted.  The question of when injunctive relief may be appropriate due to the implementer's intransigence is one that has been addressed, in accordance with somewhat varying viewpoints, by agencies in the U.S. (see USDOJ and USPTO, Policy Statement on Remedies for Standards-Essential Patents Subject to Voluntary F/RAND Commitments (Jan. 8, 2013), available at; Decision and Order, In the Matter of Motorola Mobility LLC and Google Inc. (FTC Jan. 3, 2013), available at; Decision and Order, In the Matter of Robert Bosch GmbH (FTC Nov. 26, 2012), available at; see also Florian Mueller's posts on these matters of Jan. 4, 2013 and Jan. 8, 2013) and in Europe (see European Commission, MEMO/14/322, Antitrust decisions on standard essential patents (SEPs)-Motorola Mobility and Samsung Electronics-Frequently asked questions (Apr. 29, 2014), available at; LG Düsseldorf, Mar. 21, 2013, 4 b O 104/12, GRUR-RR 2013, 196 (referring the German Orange-Book-Standard procedure to the CJEU; see sources cited at my paper pp. 29-30 n.141).        


  1. It's funny how two people can have such different takeaways from the same language. I thought that this was for the most part very persuasive and reflective of how standards licensing is supposed to work. I explore the one way ratchet a bit in my patent challenges article - if the patentee is faced with zero on one side, FRAND on the other, and litigation no matter what, that's hardly a fair deal.

    I do agree that the ETSI rules should be binding, though.

    The problem I have with the alternate view is that the "they held us up" argument doesn't work if the licensee doesn't even try. And believing in noninfringement/invalidity can't be an excuse, because everyone believes that and thus no one would use the system.

    If there are standards body rules about licensing, and third party beneficiaries want to benefit from them, then they must be bound by them. If ETSI has a procedure for determining infringement, it should be used. If there is a real debate about infringement or invalidity, the license negotiation should reflect that ex ante, and THEN maybe a court can award a FRAND rate if it finds the patent invalid. But that requires affirmative steps by the potential licensee to actually deal with the issue up front, declaratory relief, etc., and that's not happening. If the product is too valuable, and can't wait to resolve that, then risking non-FRAND damages are part of the mix.

  2. I agree, and have written (at length) that contract law is a flawed theory (at least in many cases) for the enforcement of FRAND and related commitments.

    I'm also not sure that Antitrust is a better alternative. I've proposed a reliance-based promissory estoppel theory that I think works in many cases, but it remains untested (but hopefully somebody will try soon!)

    PS -- Tom, the link to your going "on record" to abolish the ITC seems to be inoperative. You may want to check it, as I'd love to see your comments on this point (I've taken a similar position here:

  3. I find the ETSI policy requiring the user to take out a licence before implementation to be interesting. I had understood that even after the standard has been adopted, it is generally impractical for a user to obtain licenses to all SEPs prior to user implementation and general market adoption – Jorge, you make this point at length in your pseudo-pool article. For example, one point among several that you make is that only about a quarter of the patents declared essential to ETSI’s WCDMA standard are actually essential, so licensing the non-essential “SEPs” prior to user implementation is a waste of money. Another is the sheer time and money it takes to engage in prior licensing of dozens of standards in a fast moving market - the pace of the market shouldn’t be determined by limited legal resources. So I had imaged that a large part of the point of the FRAND commitment was to allow users to implement without prior licensing, secure in the knowledge that they would only be obliged to pay a reasonable royalty. The reason for delaying isn’t “heads-I-win, tails-you-lose,” as suggested by the ALJ, but simply that ex post licensing is more efficient. If that is right, then I am surprised that parties engaged in the standardization process would agree to a procedure in which they are required to license prior to implementation or face injunctive relief. Is the ALJ really interpreting the procedure correctly? I am not familiar with the terms of the policy. Is it possible that the term in question doesn’t mean “license before implementation or face injunctive relief,” but only “you have to license from the patentee, not from us” or something else similarly innocuous?