The U.S. Court of Appeals for the Ninth Circuit in San Francisco will decide America’s 5G future when it considers Qualcomm’s appeal of an antitrust remedy urged by the Federal Trade Commission.

The FTC’s economically dubious victory against the American standard bearer in mobile phone technologies has resulted, in my view, in a myopic and improvident lower-court order apt to disincent technology development and create confusion in the mobile phone technology marketplace. The Ninth Circuit ought to stay, and ultimately overturn, U.S. District Judge Lucy Koh’s injunction lest it hurt competition and transfer American leadership in 5G wireless technology to Huawei.

A wider view of the public interest must be countenanced, as also expressed by the U.S. Department of Justice in its statements of interest filed before both courts. An unbounded antitrust remedy can endanger innovation, competitiveness and even national security. How has this debacle, with our two federal antitrust agencies at war with each other, come about?

Standards development organizations—notably the European Telecommunication Standards Institute—require technology contributors to agree to license in exchange for adequate reward on fair, reasonable, and nondiscriminatory terms. This arrangement is fragile because patents are not self-enforcing: Technology users are not obliged to seek a license.

Ultimately, the technology contributor must assert their patents, sometimes through expensive litigation, which Qualcomm has demonstrably sought to avoid. Judge Koh’s decision in FTC v. Qualcomm does not provide adequate, fair, and structurally sound compensation for Qualcomm’s technology within the open-innovation FRAND licensing regime that ETSI has established.

Understanding dynamic technology markets populated with licensing arrangements is difficult. My reading of Judge Koh’s opinion indicates that neither she nor the FTC properly understood innovation and licensing in the mobile wireless industry. The conclusory confidence of the remedial order exhibits none of the caution that industry analysts and innovation and licensing scholars would exhibit when facing a complex reality.

Qualcomm makes two main products: wireless communications technology, often incorporated into standards, that proximately benefits chip manufacturers and handset manufacturers; and industry-leading chipsets for sale to those handset manufacturers. To avoid exhaustion of its patent portfolio by licensing at multiple levels of the value chain, Qualcomm has customarily licensed only handset manufacturers. To purchase Qualcomm’s chipsets, those handset manufacturers (and not chipset manufacturers) tend to license the underlying patents.

Capturing value through a reasonable royalty is a concern for a business seeking to profit from its risky investment in breakthrough technology. Technology licensing accounts for reportedly over two-thirds of the company’s value, reflecting Qualcomm’s business model designed to capture value to fund further innovation. Qualcomm is one of America’s most research-intensive companies, spending 25% of revenue on R&D.

Heavy investment in R&D and concomitant pioneering technological development creating huge consumer benefits did not earn Qualcomm any points with Judge Koh. None of Qualcomm’s executives and scientists escaped being tarred as “not credible,” partly due to their “practiced narratives” of in-court testimony.

Ironically, and surprisingly given concerns raised elsewhere by governments and the private sector, the judge found Huawei’s witnesses credible enough to rely upon and cite repeatedly. But the judge did not even dare to name the FTC’s economic expert whose “bargaining leverage” theory she tacitly adopted despite its rejection by District Judge Richard J. Leon in last year’s AT&T-TimeWarner ruling for its lack of “reliability and factual credibility.”

Judge Koh’s first main conclusion was that Qualcomm had monopolized two carefully tailored modem chip markets, seemingly only through the force of licensing practices she deemed anticompetitive. That Qualcomm produced a superior product found no place in the court’s decision. Her finding that “Qualcomm had two decades head start on Intel” would seem to me to demonstrate Qualcomm’s technological leadership (and perhaps Intel’s failure to develop a better modem) but was used by the court only as evidence of anti-competitive behavior.

Repeated testimony that original equipment manufacturers, or OEMs, had “no choice” but to accede to Qualcomm’s licensing so as to gain access to Qualcomm’s chips invites the question as to what made Qualcomm’s chips so special. The judge’s silence on Qualcomm’s technological advantage is noteworthy.

One possible explanation was stated by an Apple executive in 2015: “iPhone cellular: we’ve been using QCOM, and engineering wise, they have been the best.” Judge Koh did not merely decline to acknowledge this explanation but struck it from the record in the court’s post-trial denial of a stay pending appeal.

The judge’s second main conclusion, drawn from the FTC’s theory of the case, held that this alleged monopoly power was leveraged to obtain royalty rates for patented technology that were not just high, but unreasonably high. She attributed every chipmaker’s failure to Qualcomm’s licensing practice. That is just wrong. Evidence was presented in court that Intel and others had exercised poor foresight of the mobile chip market.

In my research, I call that a failure of sensing and sensemaking, reflecting what I call weak dynamic capabilities. Not surprisingly, poor execution was also in evidence, but that did not find its way into the decision. Qualcomm’s expert testimony analyzing rivals’ decisions was dismissed as “not reliable” for the reason that it did not analyze Qualcomm’s behavior, thereby failing to reject the judge’s null hypothesis that Qualcomm’s licensing was anticompetitive.

“Guilty until proven innocent” is a novel approach in antitrust law and not consistent with American jurisprudence. This approach enabled the judge to slide easily from the uniqueness of Qualcomm’s licensing practice to its unreasonableness (hence illegality). Qualcomm’s business model is unique, and its dynamic capabilities and engineering achievements are unparalleled.

Qualcomm’s technological prowess was evidenced in events after the close of trial, while Judge Koh was writing her opinion. On April 16, Intel “threw in the towel” on 5G modem chips despite being Apple’s sole supplier of iPhone modem chips for the last couple of years. Plainly, Qualcomm’s market power is not the issue. Mobile wireless technologies are difficult to master; success is not the norm.

Intel was not handicapped because Qualcomm had in prior years entered an alleged exclusive deal with Apple, as the judge found in her opinion. When Intel itself had the same supposed benefit of sole supply to Apple, Intel was not able to leverage that into the “market power” that the judge has imputed to Qualcomm under similar arrangements.

This “natural experiment” indicates that Qualcomm’s market position stems from its technological capability. It’s a formidable competitor and does exactly what we want our competitors to do. It out-innovates others and is a powerful driver of improvement and competition—the very stuff the FTC is supposed to favor.

So why is this such a mess? Are our antitrust laws out of date? Has the judge’s ruling failed to follow the law?

In my view, it’s a bit of each, but mainly the latter. As I understand it, the law requires the FTC to show consumer harm in a Section 5 “unfair method of competition” case. Indeed, my understanding of the burden of proof set by the U.S. Supreme Court in Ohio vs. American Express is that the mere tending of harm to competition isn’t enough for a violation of the antitrust laws; The burden of proof is to show actual harm to competition based on economic evidence.

It does not appear that any evidence of actual (end-user) consumer harm ever came before the judge. More revealing is that the “consumers” mentioned in industry testimony are largely the OEMs as “consumers” of Qualcomm’s technology, not the end users. Clearly, the American Express standard hasn’t been met and couldn’t be met at the time of trial for nonexistent markets such as the just-now-emerging 5G mobile phone technology.

Leaning heavily on mere intuition, not evidence, that “Qualcomm is likely to replicate its market dominance during the transition to 5G,” the court’s remedy seeks to regulate a market that did not exist at trial. 5G will likely take decades to fully develop, but the ruling discloses little understanding of the evolutionary development of the 5G standard. This proceeds by periodic updates every 18 months or so. Much associated development in this field is long, is sometimes slow, and is yet to happen.

It is also remarkable that Judge Koh’s ruling requires Qualcomm to license its IP for standardized cellular technology to chip makers as well as handset makers, rather than solely to handset makers (as is the industry norm). In my view, the outcome of the court’s “remedy” will likely be chaotic.

Not all Qualcomm technology is embedded exclusively in modem chips. Some is embedded in the handsets, and the patent exhaustion doctrine may not absolve devices from infringement, even after a chip license, requiring licensing at the device level, too. Concern about the “additional uncertainty and confusion” wrought by the remedy is one reason that cellular base-station manufacturer (and Qualcomm customer and rival) Ericsson filed an amicus curiae brief in support of the Ninth Circuit’s staying that remedy.

U.S. Chief District Judge Rodney Gilstrap in the Eastern District of Texas recognized in a recent ruling that device-level licensing is the industry norm and satisfies the FRAND commitment. If Judge Koh heard evidence that chip-level licensing was better for innovation, efficiency or competition, she makes no mention of it. She simply assumed its public benefit and has endeavored to force a cockeyed licensing regime on not just on Qualcomm but Qualcomm’s licensees. The soundness of split-level regulation of the industry has not been supported by evidence. I am skeptical that Judge Koh is even aware of these implications.

The DOJ under Assistant Attorney General Makan Delrahim knows the IP licensing world better than to be this incautious—no doubt one reason the DOJ pleaded for hearings to have these complicated issues carefully reviewed. However, Judge Koh rejected DOJ’s caution in favor of a more reckless approach.

The court’s remedies will disrupt industry practices such that Qualcomm will license variously at the device and chip (or component) levels, while other licensors will continue to license at the device level. This will be a feast for patent and antitrust lawyers, confusion for licensees, and a likely disaster for Qualcomm or any innovator, because it’s now clear you can get singled out simply for being the technology leader.

Should Judge Koh’s remedies be upheld, every Qualcomm licensee (except Apple, who has settled similar litigation with Qualcomm) will likely stop paying Qualcomm its royalties right away. A Qualcomm executive declared to the judge in June that Huawei had stopped paying royalties on an active license agreement, resulting in significant lost royalty revenue. Despite Qualcomm’s strong balance sheet and cash reserves, this interruption writ large would be traumatic and could lead to the layoff of thousands of engineers, mostly in the U.S.

Harm would be inflicted on the American economy with no benefit for consumers. Harming of competition is the natural corollary. U.S. innovators, large and small, beware. If the ruling stands, American leadership in cellphone technology will be the loser, and Huawei will be the clear winner. American consumers and society will be worse off.

Device makers around the world, especially abroad, may (for a short while) be delighted by the judge’s decision, but her decision was a sad day for American inventors, entrepreneurs and consumers—and for competition.

Broader issues are at stake, too. Judge Koh doesn’t seem to realize that the whole cooperative innovation enterprise in wireless (such as embodied in ETSI) envisions companies that contribute technology in exchange for royalties as the primary way to recoup investments in R&D. If there is no such mechanism, or if recoupment is made even harder, the whole global endeavor fails, particularly for the U.S. companies that cannot rely on the government to fund their R&D. While Huawei’s ownership is murky, it appears to be effectively a state-owned enterprise and can find other ways to fund its investment positions.

I am not the only one to conclude that the court has overreached so brashly. In addition to the DOJ, FTC Commissioner Christine Wilson opined in the Wall Street Journal with dismay that the judge took the opportunity to “create new legal obligations, undermine intellectual-property rights, and expand the application of our antitrust laws beyond US borders.”

She rightly pointed out that Judge Koh has given foreign rivals a club against U.S. technology developers. “Foreign competitors will no longer have to steal American technology; henceforth, they can simply cite this decision for the proposition that US innovators must share it. The expropriation of American technology will continue, but now under the auspices of US law.”

I’ve spent much of my career trying to help policymakers and executives of pioneering companies figure out how to capture a fair and reasonable share of the rewards from innovation in order to keep on innovating. Their battle is almost always uphill, unless they have compelling network externalities on their side—which Qualcomm does not.

I’ve rarely seen such a poorly supported FTC antitrust case coupled with an erroneous court decision (so confidently put forward), contrary to caution from a sister government agency and scholarship, that at minimum scopes the complexities of the issues and signals only that “fools rush in where angels fear to tread.”

Postscript. On August 23, 2019, the Ninth Circuit granted Qualcomm’s request for the stay of certain portions of Judge Koh’s remedial injunction until Qualcomm’s appeal could be heard and decided. The Ninth Circuit panel, unlike Judge Koh, acknowledged the DOJ’s warning “that the injunction has the effect of harming rather than benefiting consumers.” “[T]his case is unique, as the government itself is divided about the propriety of the judgment and its impact on the public interest.”

On the chip-level licensing that the remedial order would have imposed, the Ninth Circuit noted, “Qualcomm likewise has made the requisite showing that its practice of charging OEMs royalties for its patents on a per-handset basis does not violate the antitrust laws.”

I strongly agree with the Ninth Circuit’s recognition that “The fundamental business changes that the injunction imposes cannot be easily undone should Qualcomm prevail on appeal.” Those business changes should not be wrought.