Remarks as Prepared for Delivery
"The times they are a'changin'": the Nine No-No's in 2019
Thank you, Ron, and thank you to the Licensing Executives Society for inviting me to be with you today to share some thoughts about intellectual property and antitrust law this morning.
At the outset, I must admit I have no personal connection to Arizona – no roots and no family living in the state – but when I think of Arizona, I’m reminded of Justice O’Connor’s description of growing up in a cattle ranch, the Lazy B Ranch, which straddled the Arizona-New Mexico Border. In her memoir, she described her parents negotiating a deal to sell a set number of calves, and sealing the promise with only a handshake. It would be, in her words, “many years before a written contract of sale became commonplace. And for many years sales would be made without a down payment.”
This story merits two observations. First, it is clear that those cowboys are not IP lawyers. Second, even in the idyllic environment of Justice O’Connor’s family ranch, the law nevertheless adapts in response to new teachings. As she explained, “Gradually, … not only a written contract but a down payment were required” to give those same agreements force.
The eventual requirement of a written contract in the Lazy B Ranch is a good segue into our discussion today, which focuses on innovation not only in the economy, but in the law. In response to changing circumstances, the relative scope of IP law and antitrust law has changed over time. One regime or another has been given more sway in various points of our legal history. To reach the modern understanding – that the two bodies of law need not be at odds – was no simple feat.
Today, the U.S. economy is driven by innovation, which in turn is driven by technology licensing and partnerships – topics the LES Conference brings to the forefront. Accordingly, during our time this morning, I will discuss the evolution in thinking over the last couple of decades about the important licensing practices many of you encounter every day. We will, of course, also look to the future to assess when these practices should attract more antitrust scrutiny. In addition, I will share some thoughts on circumstances where antitrust enforcers should act with humility in order to avoid imposing burdens that may end up stalling important innovation.
The prevailing view in the 1970s was that antitrust law and IP law shared no common purpose. One created monopolies and the other sought to prevent them, so the two regimes were seen as not only in tension, but in conflict.
Unsurprisingly, the response by the Antitrust Division of the Department of Justice to many IP arrangements was one of suspicion and formalism. This was expressed most clearly by a list of practices in the 1970s that the Division considered anticompetitive “in virtually all cases”: the “Nine No-No’s.” Some of these No-Nos prohibited forms of vertical control, such as resale price maintenance or restraints on restricted vertical field of use licensing. Others more directly restricted a licensee’s ability to use products outside the scope of the patent. For example, the Division viewed tying of unpatented supplies to the licensing of patented products as anticompetitive.
In 2004, when I was a Deputy Assistant Attorney General in the Antitrust Division, I took a close look at the infamous Nos-Nos and discussed them publicly as examples of how well-intentioned antitrust enforcement, if not carefully done, has the potential to undermine the lawful exercise of IP rights and hence harm innovation. I spoke approvingly of how the bright lines of the No-Nos were increasingly replaced in favor of a flexible effects-based analysis on those same licensing practices. I also hypothesized then that stark antitrust rules and prohibitions would become a thing of the past.
Since then, actions by the courts and by antitrust enforcers have proven my prediction true, especially when it comes to five particular licensing practices I will talk about today:
- First, a licensor’s restrictions on a licensee’s freedom to deal in products or services outside the scope of the patent, which has similarities to current debates about standard setting organizations and FRAND licensing;
- Second, payment of royalties beyond the scope of the product, such as requiring payments after a patent expires;
- Third, resale price maintenance – for example, conditioning the license of a graphics package on a video game company’s commitment not to price the completed game below $30;
- Fourth, tying, such as requiring a company to license Patent B when it licenses Patent A; and
- Fifth, mandatory package licensing, a subset of tying requiring a company to license a bundle of packages when it wishes to license only Patent C.
These topics are especially worthy of discussion because either the Supreme Court or the Antitrust Division has recently taken new approaches to them. Further, these particular so-called No-Nos shed light not only on how antitrust thinking has evolved over the last two decades, but also reveal that more work can be done to secure healthy dynamic competition.
This is a good point to note why clarity of the laws in accordance with modern economic logic is a constant goal for the Antitrust Division.
The great Justice Antonin Scalia often joked that one of Emperor Nero’s worst practices was to post his edicts high on a column, so they would be harder to read and easier to transgress. We can all agree that is not an ideal model for law enforcement.
President Trump, as you may know, issued a Law Day proclamation on May 1 of last year. The President said, “Law Day recognizes that we govern ourselves in accordance with the rule of law rather [than] according to the whims of an elite few or the dictates of collective will. Through law, we have ensured liberty.”
As my friend, the former Deputy Attorney General Rod Rosenstein noted last year, “The rule of law is essential to commerce. It allows businesses to enter contracts, make investments, and project revenue with some assurance about the future. It establishes a mechanism to resolve disputes, and it provides protection from arbitrary government action.”
Rod also emphasized that creating a favorable environment for business to flourish and spur economic growth is one of the priorities of the Trump Administration. President Trump is mindful about the importance of promoting entrepreneurialism, avoiding unnecessary regulation, and promoting innovation. Over the last two years and certainly under the leadership of Attorney General Bill Barr, the Department of Justice is doing its part to promote an environment where businesses and particularly innovation can thrive.
Let us begin with the No-No against restrictions on the licensee’s freedom downstream to deal in products outside the scope of the patent. This No-No typically referred to exclusive dealing arrangements, which limits the ability of a licensee to use rival products. We can find an analogous argument similar to this No-No in modern cases about standard setting organizations, whose members aim to promote efficiencies by advancing interoperability among competing and complementary technologies to serve as a standard. Similar to exclusive dealing, choosing patented technology for inclusion in a standard is thought to have the possible effect of increasing the market power of a patent holder.
As many of you know, these concerns – that the standard essential patents (or “SEPs”) associated with a chosen technology have too much market power – have been growing in recent years. Many believe antitrust law ought to intervene in order to right the balance. As an enforcement agency, however, the Antitrust Division must demand sound evidentiary proof before responding to a perceived problem in a way that may unduly burden patent rights and thus innovation itself.
I find many of these concerns overstated, and have therefore made efforts to clarify the Division’s position on the issue, guided by the flexible “New Madison” approach I introduced last year.
First, I replied to growing concerns among some academics and commentators about the perceived problem of “hold-up.” For those who are unfamiliar, “hold-up” refers to the phenomenon whereby an SEP holder is alleged to seek “unfair” royalty rates after its patented technology is chosen as a standard by threatening to withhold a license from a licensee that either wishes to incorporate or has incorporated the technology into a product.
As I have said before, however, those who advocate using antitrust law to reduce the risk of an alleged “hold-up” have yet to identify any harm to the competitive process. Choosing a valuable technology as a standard – presumably with transparency and spirited debate – cannot transform an otherwise valid patent by itself into an unlawful monopoly. Further, actual evidence of hold-up remains scant even after a decade has passed since the theory was first introduced. The gulf between the theory and practice is especially troubling as many advocates ignore the real risk of hold-out by potential licensees of the chosen SEP technology. “Hold-out,” of course, refers to the countervailing problem to hold-up: when an implementer licensee refuses to negotiate in good faith with a patent holder for a license, and instead forces the patent holder either to undertake significant litigation costs or to give up IP enforcement efforts.
With a want of identifiable competitive harm, empirical evidence, or due acknowledgment of a symmetric problem, there is little, if any, basis for antitrust intervention. As a result, the Division maintains that it is inappropriate to police unilateral FRAND commitments by patent owners where contract or other common law remedies are available to complaining parties.
Our position has sparked healthy debate in the antitrust world about the proper scope of antitrust enforcement in the standard setting area. It is important to consider carefully and thoughtfully whether and to what extent antitrust law should police the exercise of IP rights, because any inaccurate assumptions can impose great burdens on innovation.
Second, the Antitrust Division withdrew our assent to the 2013 joint policy statement with the USPTO in December 2018, in large part because the statement was unduly skeptical of an SEP holder’s ability to realize the exclusionary value of its patents. Many believed the statement placed limits on an SEP holder’s ability to seek an injunction or exclusion order. Specifically, the statement suggested an SEP holder with a FRAND-encumbered patent “may harm competition” simply by seeking injunctive relief.
Yet, as I have expressed, the value of a patent stems primarily from the right to exclude, which is contemplated by the text of the U.S. Constitution itself. Questioning that right and suggesting that simply exercising it may violate antitrust law can severely threaten the incentives to innovate.
Accordingly, the joint policy statement undercut the Division’s message about the importance of innovation and no longer accurately conveyed our view on when a patent holder should be able to exclude a competitor from practicing its technologies. Instead, “[w]hen it comes to the test for obtaining injunctive relief against infringement, patent law already strikes a careful balance that optimizes the incentive to innovate.” It is the Division’s view that the four-part test the Supreme Court articulated in eBay v. MercExchange should govern whether an SEP holder gets injunctive relief – without our thumb on the scale against an injunction.
A second practice the Division considered per se illegal in the 1970s is licensing agreements involving royalties to be paid after the term of a patent expires. In 1964, the Supreme Court held in Brulotte v. Thys Co. that licensing provisions requiring royalty payments after the expiration of a patent are “unlawful per se.” The Court found that such provisions conflict with the patent law policy of encouraging patents to enter the public domain after they expire.
Brulotte has since been interpreted narrowly, although the Supreme Court reaffirmed the holding in 2015 in Kimble v. Marvel. In so doing, the Court took great pains to clarify that Brulotte was not founded on the analysis of competitive effects that antitrust practitioners and enforcers typically employ. Rather, it is based on patent law, which is more permissive about applying categorical principles such as per se rules, which antitrust law invokes only in rare circumstances. More importantly, the Court explained that although patent law bars a licensee from paying royalties after an invention “has moved into the public domain,” Brulotte nonetheless permits deferred payments and other similar financial arrangements.
Indeed, the Court expressed its general approval of these practices from the perspective of antitrust law, agreeing with the plaintiff that “such agreements more often increase than inhibit competition, both before and after the patent expires.”
In line with the Court’s observations in Kimble, the Antitrust Division does not apply rigid presumptions when considering royalties paid post-expiration. We assess whether these agreements have any demonstrable efficiencies in an effects-based analysis before condemning them as anticompetitive. For example, as the Court recognized in Kimble, deferred payments – which “typically go hand-in-hand with a lower royalty rate” – may sometimes be necessary for a procompetitive transaction to occur. Lower royalty rates during the patent term may in turn lead to lower prices or enable faster entry for more licensees to implement and to share technologies more broadly. Allowing for payments to be made past the expiration date can therefore, under certain circumstances, reduce potential deadweight loss.
The third No-No to which I’d like to turn is resale price maintenance (or “RPM”): an agreement between a manufacturer and a distributor regarding price. For decades, what was called “vertical price fixing” was considered illegal per se, regardless whether the restraint imposed a price floor or a price ceiling. Fortunately, the law has changed dramatically in response to new economic thinking and judicial experience.
The Supreme Court in 1968 endorsed per se treatment for maximum vertical price fixing in Albrecht v. Herald Co. but, in response to substantial criticism, the Court overruled it in State Oil v. Khan nearly thirty years later. In State Oil, the Court held that “vertical maximum price fixing, like the majority of commercial arrangements subject to the antitrust laws, should be evaluated under the rule of reason.”
Minimum RPM, on the other hand, was still considered illegal per se when I last spoke about the No-Nos back in 2004 – and it had been subject to that rule since at least 1911. Only in 2007 did the Supreme Court confirm in Leegin that all vertical price restraints, including minimum RPM, are to be judged according to the rule of reason. In so holding, the Court observed that economic literature about the practice is “replete with procompetitive justifications,” such as encouraging market entry and retailer attention to intra-brand promotion, which can well lead to greater inter-brand competition.
In the IP context, an analogous arrangement involves situations in which a patent holder conditions a license on the resale price of the product that incorporates the licensed technology. As of today, courts have not yet had a chance to explore how Leegin specifically affects IP licensing. As with RPM agreements applying to sales of goods, however, antitrust enforcers no longer presume licensing agreements involving resale pricing restraints are anticompetitive. Instead, the Division will apply a rule of reason analysis, weighing the competitive harms and benefits of each situation.
As in the context of goods and services, antitrust authorities will scrutinize licensing arrangements much more closely if they can be characterized as predominantly horizontal. Enforcers will treat any agreement as illegal per se when it is used to facilitate horizontal price-fixing cartels.
A fourth No-No is tying of unpatented supplies to the licensing of patented products. The Division once presumed the practice unlawful, as long as the patent owner had some degree of market power over the tying product – and, as long as the patent owner owned a patent, the Antitrust Division presumed it met the market power threshold.
In 2004, I spoke of how the agencies’ approach to analyzing tying arrangements had grown much more attuned to the potential procompetitive effects of a tie. Specifically, I explained that we no longer assume that a patent always confers market power. The Supreme Court recognized as much two years later in Illinois Tool Works, when it held that in all cases involving a tying arrangement, even when the tied product is patented, “the plaintiff must prove that the defendant has market power in the tying product.”
Nevertheless, judicial treatment of tying has varied over time. Although courts have now identified some procompetitive rationales for the practice, tying has the unique and dubious honor of being the only No-No that may still be subject to the so-called “per se” rule. Confusion remains about whether it is possible to establish per se illegality under the Sherman Act by showing simply that a company has sufficient market power in the tying product, and that the tie restrained a significant volume of commerce in the tied product.
Not every court applies such a rule broadly. When the D.C. Circuit considered the tie between the Microsoft operating system and Microsoft’s Internet Explorer in 2001, it determined that “the rule of reason, rather than per se analysis, should govern the legality of tying arrangements involving platform software products.” I discussed the case at length in remarks I presented to a conference in Israel in June.
Following more recent developments in economic thinking, it continues to be the Division’s view that tying in non-software industries also does not merit per se treatment. Put simply, tying is no longer considered a practice which, in the words of the Supreme Court in Northern Pacific Railway, categorically has a “pernicious effect on competition and lack[s] any redeeming virtue.”
Indeed, several procompetitive benefits can result from the practice, including increasing quality control, reducing risk, or avoiding double marginalization. The magnitude of these effects depends on the specific circumstances at issue – but a categorical prohibition of tying no longer seems appropriate in any industry. As we expressed in the 2017 Joint Guidelines with the FTC on the Licensing of Intellectual Property, “in the exercise of [our] prosecutorial discretion,” the Division will “consider both the anticompetitive effects and the efficiencies attributable to a tie-in.”
To be sure, antitrust enforcers will not shy away from challenging tying arrangements we find unlawful, and the Division will be much more likely to challenge a tying arrangement if the seller has market power in the tying product. As I have said before, “[w]here a company has market power, enforcers should be circumspect about conduct that does not plausibly advance a legitimate business objective and transactions that eliminate competition.”
The Division’s past prohibition on tying also extended to mandatory package licensing, including the block-booking of motion pictures. The practice of block-booking – bundling multiple films into one theatrical license – is perhaps most well-known in the context of a series of consent decrees, known as the Paramount decrees, from the 1940s.
The Division entered into the Paramount decrees after the district court and the Supreme Court held that movie distributors engaged in an industry-wide conspiracy to fix motion picture licensing practices, including fixing movie theatre admission prices, in an attempt to monopolize the first-run film distribution and movie exhibition markets during the ‘30s and ‘40s. As part of the consent decrees to address the competitive harm of the horizontal conspiracy, the defendants also agreed to stop certain vertical licensing practices, including block-booking motion pictures.
Of particular importance, these consent decrees had no termination date and continue to govern much of how the film industry conducts itself today. In the decades since we first entered into these agreements, however, the industry, the economics, and the law have changed.
The motion picture industry has undergone great transformation, including the use of different release patterns and the proliferation of Netflix, Amazon Prime, and other streaming sites as competitors. Moreover, although the original Paramount case involved a horizontal conspiracy between competitors, some of the ancillary prohibitions involving block-booking and other vertical restraints may no longer hold up in light of new economic thinking and the more recent judicial decisions I described earlier. For example, as I said in 2004, we now understand a package license in the form of block-booking can benefit a licensee by reducing the overall costs of the transaction. The treatment of block-booking and the other film licensing restraints as categorically prohibited by the Paramount decrees is therefore becoming increasingly difficult to justify.
As a result of these changes, and in the context of a broader initiative of the Antitrust Division to review all legacy antitrust decrees, the Division opened a review of the Paramount consent decrees last year to see whether specific prohibitions ought to be updated or eliminated – or whether the decrees ought to be terminated as a whole. Since then, we have received numerous public comments, and are in the process of evaluating these restrictions based on modern economics. Sometimes the decree provisions of the past may actually be restraints to innovation and competition themselves.
As we move forward, the Division will weigh evidence-based arguments to enforce the antitrust laws – and avoid improper regulation of specific industries or licensing enforcement that is based on ossified economic thinking from times past.
In conclusion, the Division’s approach to these five key forms of licensing illustrate that, as thinking has evolved over the last two decades, the room for debate as to the optimal scope of IP rights and the conditions of their licensure also has shifted. In the words of Bob Dylan,
Come gather 'round, people
Wherever you roam
And admit that the waters
Around you have grown…
For the times they are a’changin’
As enforcers, we are mindful that the ability to license intellectual property, which facilitates specialization and ensures that innovation can realize its commercial value, is a key part of a competitive, dynamic economy. Alienability of IP rights not only ensures that inventors internalize the fruits of genius, but also serves as a tool for private ordering, allowing expected returns from R&D investment and any resulting IP to grow and to be shared with consumers.
Although many of us continue to hold strong views on precisely how the two regimes – IP and antitrust law – ought to complement each other, I am grateful the conversation is now at least more nuanced than the simple “no-no” the Division gave decades ago. As Justice Louis Brandeis once remarked, “There are no shortcuts in evolution.”