Memory Lapse in Litigation
2008 Legal Times Online
Page printed from: http://www.legaltimes.com
The U.S. Court of Appeals for the D.C. Circuit has just issued an opinion that, coupled with a related jury verdict in California this March, ought to end the Federal Trade Commission’s much ballyhooed but misguided antitrust case against Rambus Inc. Anyone concerned that the law may trample blindly through the fields of technological innovation should read the April 22 decision in Rambus v. FTC.
The history of the case also merits attention. It speaks to the fundamental fairness we ought to require from administrative agencies. Severe flaws in the FTC’s enforcement process permitted its antitrust powers to be used to advance the anti-competitive interests of Rambus’ rivals. And those flaws may also have distracted the commission while the ground shifted out from under its 2002 complaint. In short, the FTC should have dismissed this case long ago.
I write both as a past participant (on Rambus’ side) and as a citizen concerned with the process that led us here today.
Beginning in June 2002, the FTC sought to impose antitrust sanctions for Rambus’ alleged “deception” in an industry standard-setting process in which Rambus participated from late 1991 to late 1995. The FTC charged that the organization now known as the JEDEC Solid State Technology Association was tricked by Rambus’ silence into standardizing computer memory technologies over which Rambus had current or future patent claims. In asserting these charges, the FTC went beyond JEDEC’s written rules, relying also on allegedly “shared expectations” of disclosure among JEDEC members.
At a four-month-long hearing before the FTC’s chief administrative law judge in 2003, the commission lost. Chief Judge Stephen McGuire recommended dismissal, issuing more than 1,600 highly detailed factual findings, most against the FTC. The judge made numerous explicit and implicit credibility determinations concerning, among other things, the scope of those “shared expectations.”
In July 2006, the FTC commissioners, sitting as an appellate court, vacated this result, along with the judge’s fact findings. Instead, the full commission found against Rambus, relying heavily upon the trial testimony of industry witnesses.
Now the D.C. Circuit has reversed the full commission. Its ruling focuses on the FTC’s failure to prove that JEDEC would have standardized other memory technologies but for the alleged deception. The D.C. Circuit did not, however, stop there. The court was broadly critical of the FTC’s “aggressive interpretation of rather weak evidence.” Its main example involved the now-vacated remedy.
The FTC had imposed severe limits on what Rambus could charge for use of its intellectual property. The remedy extended beyond memory technologies standardized while Rambus participated in the process. It also reached patents that covered parts of a technology called DDR DRAM, which had recently come to dominate the market but was not (except perhaps in the most preliminary way) part of any standardization process while Rambus was a JEDEC member.
The D.C. Circuit criticized the reach of this remedy. It was, the court concluded, highly unlikely that the industry would impose on itself such a “sweeping and early” duty to disclose — i.e., one that would cover patents, pending applications, and plans for possible future patent amendments. Nor would such a duty likely be “triggered by the mere chance that a technology might someday (in this case, more than two years later) be formally proposed for standardization.” JEDEC’s written disclosure rules were, at best, vague (a conclusion reached in 2003 by the Federal Circuit in the related case of Rambus v. Infineon Technologies). Even the most helpful contemporaneous document, the D.C. Circuit noted, did not support a “shared expectation” of disclosure as broad as the FTC posited.
Harsh as it was, the D.C. Circuit opinion reveals only the tip of the iceberg. Other problems in the FTC’s case included the following:
• In the early 1990s, two prominent JEDEC members (IBM and Hewlett-Packard) had formally announced, without rebuke from the association, their refusal to disclose any further patents or patent applications.
• While at JEDEC, Rambus itself indicated that it did not care to comment on its intellectual property and stated in writing that nothing about potentially relevant IP rights should be inferred from its silence.
• JEDEC’s leadership wrote to its members in 1997 that the memory standardization efforts were running behind industry needs and that it therefore had to “take the work” of Rambus and issue that work as JEDEC’s “own” standards.
WHY IT STUMBLES
Flaws in the Rambus case highlight far-reaching problems in how the FTC handles antitrust enforcement. Even if the FTC has jurisdiction to issue market-altering remedies of any kind (an open issue), such remedies are unwise. They rest upon an enforcement process too infirm to support them.
• The FTC’s ability to act as both prosecutor and appellate judge inevitably taints its cases. There is an obvious conflict of interest here.
The full commission votes out complaints in the first instance and invests considerable resources in having its staff prosecute those complaints. Clearly, the commission has a vested interest in the outcomes of the cases it elects to bring. And yet when an administrative law judge’s decision is appealed to the full commission, the FTC contends that it may make all fact findings de novo — even on issues of witness credibility. It is small wonder that, according to one amicus brief in the D.C. Circuit case, the full commission has apparently never upheld an administrative law judge’s decision ordering dismissal of an antitrust enforcement complaint.
• The FTC’s conflicts of interest go even further. The commission can be motivated to bring or to keep pushing a weak case for reasons having little to do with the specific facts — e.g., to expand its regulatory reach, to “send a message,” or to address unanswered doctrinal issues. Each of these motives existed in the Rambus case.
Moreover, any prosecutor faces a natural temptation to frame charges broadly in order to gain leverage over a defendant or to attract public attention. But with the FTC, the situation is worse in two ways: First, the FTC has the power to rule on its charges and to impose broad remedies. Second, FTC budget requests to Congress point to the “value” of the remedies it secures. This gives the commission an incentive to expand that value.
There is thus at least the appearance of a conflict of interest in the fact that the FTC’s remedy in the Rambus case went well beyond the older technologies being standardized in the early 1990s. Those technologies had already lost almost all their market share by 2003. How much more impressive (to Congress, the public, and industry) to limit royalties over the newer dominant technology, DDR DRAM.
• The FTC is vulnerable to misdirection from industry lobbying. The FTC staff can become hostage to segments of an industry, beguiled by business witnesses or relying too heavily on donated legal help, and may bring cases it never should have or stick with cases long after discovery has revealed weaknesses. The multiheaded leadership of the FTC compounds the problems: Cases acquire an apparent life of their own, while no part of the organization takes responsibility to admit error.
In the Rambus case, major corporations lobbied the FTC to help them in their multipronged litigation attack on a much smaller rival. One question is whether participating in such a “pile on” is a good use of government resources.
Those in industry who urged the FTC to move against Rambus had strong financial reasons to do so. They were (and are) using Rambus’ patented inventions in billions of dollars’ worth of semiconductors and semiconductor systems shipped each year. Keeping DDR DRAM free of Rambus royalties (or subject to minimal royalties through the overbroad FTC remedy) also improved that product’s competitive position against Rambus’ own designs.
The length of the litigation alone has been useful to its proponents. Even if paying royalties to Rambus has long been inevitable, the delay has transformed many years of those royalties from ongoing costs into one-time, after-the-fact charges. So profits are helped in the short run, and even in the long run, such onetime charges typically raise little concern on Wall Street.
That there was industry misdirection in the Rambus case now appears obvious. McGuire apparently strongly disbelieved certain key industry witnesses. In March, a U.S. district court jury in San Jose felt likewise in Hynix Semiconductor v. Rambus, unanimously rejecting antitrust claims against Rambus based on essentially the same “shared expectations” theory asserted by the FTC and its industry witnesses.
Certainly some of this has surprised the FTC staff. Presumably the staff did not know, for example, that at the same time one CEO was testifying in the FTC’s 2003 administrative proceeding (discussing, among other things, industry competitiveness), his company was about to work out a leniency deal with the Justice Department concerning its participation in the second-largest criminal price-fixing case in Justice Department history. Later-discovered e-mail shows that the price fixing overlapped with the 2001-2002 period when the industry was lobbying the FTC to go after Rambus and that the price fixing was apparently intended, at least in part, to lower DDR DRAM prices in order to drive Rambus products from the market.
So where does the Rambus case leave us? It points to a few specific reforms.
First, the FTC ought to follow the Justice Department’s lead and litigate antitrust issues in federal court from the outset — at least when market-altering remedies are at stake. Reliance on an independent fact-finder and federal court procedures would resolve many of the conflict-of-interest issues raised by the commission’s own review of administrative proceedings. It also might lead to better FTC scrutiny of its own cases, both before they are brought and along the way, when — as the facts develop — settlement may start to seem wise.
Second, the FTC ought to look hard at the extent to which it relies upon industry to help gather evidence and frame cases.
And last but not least, the FTC ought to weigh the timeliness of its enforcement efforts. Patents have a finite life, while technology moves quickly. Back-due royalties will not compensate Rambus for the six years during which the FTC placed a cloud over its intellectual property — thus, among other things, disadvantaging Rambus’ own memory designs in the market. Even an extension of the life of key patents (assuming current law allows that) will not make Rambus whole.
The FTC should heed the lessons of Rambus. Bring antitrust enforcement over standard-setting efforts with caution, lest the prosecution do more harm to competition and innovation than the alleged misdeeds. And whenever and wherever such cases are brought, find a way to resolve them much more quickly.
John Danforth was the general counsel of Rambus from October 2001 to June 2006. From June 2006 to October 2007, he served as Rambus’ senior legal adviser. This commentary has not been approved or reviewed by Rambus or its counsel. The opinions — and any errors — are solely those of the author, who retains a significant financial interest in Rambus.