Abstract
Barnett (2022) criticises the assumption that intellectual property (IP) holding companies can be seen as monopolists which “tax” downstream innovation by charging excessive royalties for patent licences. This study examines the “License as Tax” fallacy, a concept that presumes that patent owners exercise unchecked market power. The author points out that this argument is not backed by empirical evidence. Instead, SEP licensing often facilitates value-creating transactions that drive technological innovation. Examples of this can be found in plenty of markets, such as in wireless communications, audiovisual technologies, or biopharmaceutical research. The paper illustrates how SEP licensing supports market entry, reduces transaction costs, and promotes competition. In addition, IP holding companies enhance consumer welfare through mechanisms, such as vertical licensing, patent pools, and IP giveaway strategies. The author advocates for a “rule-of-reason” approach to assess market power in SEP licensing. This stands in contrast to more conservative and outdated antitrust policies that impose broad restrictions on SEP holders. Ultimately, IP licensing should be seen as a bottom-line pro-competitive and welfare-enhancing mechanism when evaluated within a dynamic market context.
Summary
1. INTRODUCTION
‘Words matter!’ This is the argument with which Barnett introduces his study. He shows that there is a historical pattern in academic and judicial discourse, where intellectual property (IP) rights are framed as monopolies. IP licences, in turn, are seen as taxes increasing end-product prices. The “license-as-tax” analogy lays the focus on the risk of IP owners using contracts to unfairly expand their IP rights. As an example, the author names a Supreme Court case in 1944. Here, the court characterised “tying” in patent licence agreements as “a graphic illustration of the evils of an expansion of the patent monopoly”. Other examples where the “IP = monopoly” fallacy was enforced are a Supreme Court decision in 1962 or a motion in front of the Federal Trade Commission (FTC) in 2017. In both cases, the institutions used terms such as “tax” or “monopoly”. Barnett argues that this is not just about terminology. As noted in an 1890 patent textbook, whether a patent is a monopoly is more than a matter of wording.
In the context of the “Internet of Things” the question of ‘to what extent patent and antitrust law should impose limitations on IP licenses’ remains open. Property-rights arrangements behind technology in wireless computing and communication devices drive a multi-billion-dollar market. These arrangements depend heavily on whether courts and regulators view IP licences as value-depleting taxes or value-enhancing tools. IP licences facilitate the transfer and pricing of IP assets among innovators and participants in the innovation ecosystem. However, the current policy climate is focused on IP licensing scepticism. These perspectives can influence competition regulators significantly.
For example, from the late 1970s and onwards, courts relaxed most rules of per se liability of IP licensors. The idea that licensors are per se liable is sceptical of IP licensing and understands licensors as monopolists with market power. However, this view met with plenty of scholarly criticism in the 1950s and 1960s. The DOJ’s and FTC’s 1995 Antitrust Guidelines for the Licensing of Intellectual Property[1] formalised this shift by rejecting any presumptive market power of IP owners and recognised that IP licences foster the commercialisation of IP assets. Nowadays, this view is being challenged again by the US courts and worldwide. Since 2017, several Supreme Court decisions have reflected a return to a more rigid view, by which IP licensors impose a risk for market competition. This approach assumes that licensing inflates the costs of end products “excessively”, with the need for institutions to “protect” the consumer. The return to the “IP = monopoly” equation suffers from two errors. First, IP licensors rarely exercise market power. Second, in many cases, IP licences enable value-enhancing transactions in technology markets. The mentioned conclusion overlooks these social benefits and systematically underestimates welfare gains generated by IP. This can lead to misguided antitrust interventions. Such actions impose restrictions on legitimate business practices and favour large firms with vertically integrated research and development (R&D) while disadvantaging smaller firms reliant on licensing. Ultimately this may harm competition, discourage market entry, and protect incumbents rather than foster a competitive environment.
Barnett aims to correct the licence-as-tax analogy. He focuses on the large pool of negotiated IP licences rather than IP litigation court cases. These court cases are less common but receive more media attention. Focusing on real-world transactions in the licensing market allows the author to illustrate the value-creating mechanism. This mechanism facilitates the assembly of complementary IP and non-IP inputs and the production of goods for end-users. IP licensing allows for two types of efficiencies in the distribution of patented inputs. First, licensing allows for vertical relationships between upstream innovation-specialised firms and a downstream network of companies that execute the production and distribution. Second, licensing contributes to the exchange of informational assets in joint ventures and other horizontal relationships. Firms might otherwise avoid these relationships due to the risk of knowledge leakage to competitors.
Barnett underlines that IP licensing allows for net welfare gains by enabling value-enhancing arrangements that would otherwise not be viable due to expropriation risks. However, the author acknowledges one important exception. Horizontal agreements involving IP assets that are substitutes may lead to collusion concerns. Apart from that case, in general, IP licensing may facilitate the entry of specialised R&D firms. This is due to the lack of capital or technical capacity of those firms to produce goods for the target intermediate or end-user market.
2. LEGAL ENCROACHMENTS ON INTELLECTUAL PROPERTY LICENSING
The shift in recent years from a more open view of IP licensing to the strict assumption of the “IP = monopoly” equation arose through a combination of U.S. courts and institutions. The U.S. Supreme Court decisions and statements by the U.S. competition agencies both started around the mid-2000s. It is widely acknowledged that the law can take two polar approaches to patent licensing. The laissez-faire approach treats IP licences no differently than any other agreement under the common law of contract. In the interventionist approach, on the other hand, IP licences are subject to multiple mandatory provisions that prohibit or require the use of certain terms. Intervention may even go as far as requiring IP licensors to license all interested parties at a judicially determined “reasonable” royalty.
Historically, up until the early twentieth century, the laissez-faire approach was common. An example is the 1902 Supreme Court decision in E. Bennett & Sons v. National Harrow Co. The Court established the “general rule” of “absolute freedom in the use or sale of rights under the patent laws.” Following the 1930s, Roosevelt’s cabinet took a more restrictive approach towards IP licensing by adopting the New Deal. In the following years, Congress undertook actions with a focus on cross-licensing that purportedly resulted in the formation of cartels within various industries. At the same time, the DOJ led a campaign of antitrust enforcement against IP licensors. Following these actions, the U.S. Supreme Court ruled on key decisions, increasing the risk of being held subject to antitrust violations for IP licensing. The Court ruled that resale price maintenance in patent licences was equivalent to a horizontal price-fixing cartel. This included that tying and resale price maintenance clauses in IP licences were deemed per se antitrust violations. An example is the Morton Salt decision. Here, the Court found that tying clauses constitute a misuse of patents even without violating antitrust laws. Thus, the enforcement of patents or licences was further complicated. This IP-sceptical approach was retained in the following decades. Another example is United States v. Loew’s from 1962. The court ruled that the existence of a patent without more leads to tying practices being per se anticompetitive. When it comes to antitrust agencies it is observed that they undertook over 100 enforcement actions that resulted in compulsory licensing orders in the 1950-60s. In addition, a White House task force found patents to be one of the principal sources of monopoly power in 1968. The suggestion was that any patentee who chooses to license a patent must do so on a non-exclusive basis. In the early 1980s, antitrust liability for IP licensing eased as the Supreme Court reformed antitrust rules on vertical restraints. This movement was influenced by scholarly critiques of strict per se rules and broad patent misuse applications.
The Fortner II decision marked a changing point. Tying claims were rejected due to a lack of proof of market power. The Court enabled, in multiple instances, the removal of the per se treatment of all non-price vertical restraints. By 1984, the shift to a more nuanced evaluation can be seen in the Jefferson Parish Hospital District No. 2 v. Hyde decision. A “modified” per se rule was used. It required that plaintiffs demonstrate both market power in the tying product market and an unreasonable restraint on competition in the tied product market. Between 2001 and 2006 the U.S. Supreme Court confirmed this approach.
Furthermore, Courts and Antitrust Agencies adopted a more liberal policy towards patents. In 1986, the Federal Circuit tightened patent misuse standards. It emphasised economic views that IP licensing is typically pro-competitive and rejected per se liability for most licensing practices. In 1988, Congress required proof of market power for patent misuse claims involving tying, effectively rejecting the Morton Salt decision. The 1995 Guidelines issued by antitrust agencies endorsed three key principles. First, in contrast to the “IP = monopoly” equation, IP rights were not considered any more to imply market power. Second, the Guidelines acknowledged that IP rights may lead to efficiency gains. Hence, competitive concerns should be assessed under a rule of reason standard. Third, the Guidelines underlined that IP licensing might lead to net anticompetitive effects in certain instances. Nonetheless, the Guidelines cautioned against the reflexive application of per se liability rules.
Contrary to recent movements, in 2017 the Supreme Court reverted to a more IP-hostile ruling in Impression Products, Inc. v. Lexmark International, Inc. This was no surprise, as multiple court decisions since the mid-2000s took a more IP-sceptical view. In this specific case, Lexmark sold two types of printer cartridges: refillable higher-priced ones and restricted-use lower-priced ones. Despite knowing the restrictions, Impression Products bought, refilled, and resold the lower-priced cartridges, undermining Lexmark’s price discrimination strategy. The Court concluded that the initial sale of the cartridges had triggered the exhaustion of the underlying patents. Therefore, as a matter of patent law, those restrictions had no legal force against subsequent users such as Impression Products. In line with a per se analysis, the Court avoided any inquiry into whether the patent owner exercised market power. It also did not examine whether the use restrictions caused net negative welfare effects. Ultimately, this overruled a long-standing Federal Circuit interpretation of patent exhaustion. This initial interpretation provided patent licensors and licensees with the freedom to negotiate a wide range of transactional structures in technology markets without the risk of triggering exhaustion. In the end, the Supreme Court’s broad interpretation of exhaustion forces manufacturers like Lexmark to use uniform pricing unless they find contractual or technological alternatives. Eliminating Lexmark’s price tiers likely raised costs for low-intensity users, and reduced efficiency by harming innovation incentives. In addition, it shifted wealth from lower- to higher-income consumers under a uniform pricing model. The Supreme Court’s decision represented a departure from the established legal consensus, overturning the 1992 Mallinckrodt ruling that upheld use restrictions on patented products through express conditions.
The Supreme Court was criticised for the rigid interpretation of the exhaustion doctrine that lacks empirical support. There was no evidence that markets were harmed under the previous legal framework. On the contrary, the Mallinckrodt era coincided with market growth, cost efficiencies, and consumer benefits, particularly in technology sectors dependent on licensing and complex supply chains. According to Impression Products, explicit use restrictions are unenforceable under patent law and often unenforceable under contract law. This results in uncertainty for patent holders, which affects their monetisation strategies, often forcing them to rely on less efficient methods. These constraints on licensing freedom may distort transactions, undermine competition, and increase costs for end consumers.
Competition law enforcers from around the world began to rewrite the rules that govern the licensing infrastructure in the global smartphone market. Regulators are concerned that two theories of harm might become reality. The first one suggests a patent hold-up, where standard essential patent (SEP) owners demand excessive royalties from producers who have made investments tailored to the relevant technology standard. The second one claims that royalty stacking might occur, where uncoordinated rate-setting by monopolist patentees leads to excessive aggregate royalties, driving device prices to high levels. In view of these theoretical concerns, regulatory and judicial actions were adopted that significantly limited the enforcement and licensing capacities of SEP owners. These owners are the principal sources of technology inputs for the smartphone market. It should be noted that no empirical study yet has been able to confirm these concerns.
Following this policy trajectory, antitrust agencies in major jurisdictions have scrutinised and acted against leading SEP holders’ licensing practices in the smartphone industry. They have promoted three key principles, limiting SEP holders’ enforcement and licensing powers.
Initially, there has been a trend to not entitle SEP owners to injunctive relief against infringers unless the licensees tried to negotiate on “fair, reasonable and non-discriminatory” (FRAND) terms. The Courts have awarded attorney’s fees to the infringer because the SEP owner was found to have pursued injunctive relief in circumstances inconsistent with its FRAND commitment in at least two cases. Furthermore, the DOJ encouraged standard-setting organisations to bar contributors from pursuing injunctions as part of the FRAND commitment. Similar actions were taken in the UK, the European Union and Germany. Here, courts adopted weakened forms of the “almost-no-injunction” rule. This emphasises the infringing user’s obligation to negotiate in good faith to enjoy effective immunity from the threat of injunctive relief. In practice, this means that a well-advised infringing user can easily comply with this good-faith standard even in a drawn-out negotiation process. In China, SEP holders rarely succeed in obtaining injunctions and often face antitrust counterclaims from infringers.
Secondly, US courts set prices for SEP owner’s licences based on component price. This goes against standard market practice. In the wireless communication industry, it has been a long-standing practice to assess the value added using the sale price of end devices in the consumer market. Here, the value created is larger. In 2011, the FTC recommended that courts base their royalty calculations on the value of the smallest priceable component containing the patented invention. This approach was said to reflect what the parties would have hypothetically agreed upon during FRAND negotiations. In 2019, the District Court in the FTC v. Qualcomm litigation adopted this view. The court argued that device-level licensing is inconsistent with Federal Circuit law on the smallest saleable patent practising unit (SSPPU). However, this decision was overruled by the Ninth Circuit. The smallest saleable patent practising unit is not a per se rule for reasonable royalties, according to the court. Furthermore, there is nothing wrong with using market values when calculating infringement damages. Although courts have not mandated this principle, standard-setting organisations like the IEEE have voluntarily encouraged component-level licensing. An example is the 2016 FRAND definition update for Wi-Fi standards, which was supported by the DOJ’s letter endorsing the SSPPU as a FRAND-compliant royalty base. The DOJ’s stance influenced competition regulators in other countries, including South Korea and China. While Chinese competition regulators initially sought to enforce SSPPU-based royalties, they ultimately allowed device-level royalties. A reason for that was reduced rates for local manufacturers. However, this success was short-lived as major technology contributors resisted, especially in the U.S. In 2020, the DOJ updated its earlier letter, clarifying it was not an approval of the IEEE’s policies. By 2021, the DOJ further distanced itself from the initial letter.
The third principle follows the logic that “excessive” SEP royalty rates can constitute an independent competition law violation. This principle is not compatible with U.S. competition law but exists in European and Chinese competition law. While it is rarely applied in Europe, it is a substantial part of China’s Anti-Monopoly Law. The idea is that “excessive pricing” can be an independent basis for a competition law violation in the case of firms that hold a “dominant” market position. In 2013, a Chinese court ruled against a SEP owner, InterDigital Corporation. The company had filed an infringement case against Huawei, a local device producer, and was found liable for assessing “excessive” royalties. In 2015, China’s competition authority applied this principle when imposing a $975 million fine against Qualcomm. In both cases, the result was a settlement which significantly reduced royalty rates. While the District Court in FTC v. Qualcomm also thought royalties were set unreasonably high, the Ninth Circuit overruled this. Ultimately, this principle was deemed incompatible with U.S. antitrust regulation, because the ability to charge monopoly prices is a key aspect of the US free-market system.
Barnett proceeds to present multiple antitrust actions taken against SEP owners in recent years, which are based on these three mentioned principles.
Competition regulators’ ambitious effort to reshape SEP licensing markets is based on theoretical models of patent hold-up and royalty stacking. The author argues that these predictions of market failure can be empirically tested. Hence, one can compare the predictions with the performance of real-world markets. In the case of patent hold-up and royalty stacking we should observe increasing consumer prices, decreasing output and, over time, less entry into SEP-intensive markets. The wireless device market is particularly vulnerable to this outcome due to the fragmented ownership of SEPs required for the technology standard. However, in the last three decades wireless communications do not appear to suffer from market failure. We observe frequent implementer market entry and continuous innovation in upstream R&D, all while quality-adjusted prices are falling. In addition, aggregate royalties are moderate, in the range of three to five percent of global handset revenues and constant over time. Ultimately, this is evidence that SEP markets do not show any signs of market failure or abuse of dominant positions by patent owners. The reason why the “license as tax” fallacy does not stand the test of reality is because it lacks dynamic economics: IP patent owners are not one-period profit maximising monopolies.[2] In fact, IP holders repeatedly maximise revenue over multiple generations in wireless technology markets.
For example, an IP owner maximises revenue over 2G, 3G, 4G and 5G, rather than a single iteration of the relevant standard, for example 3G only. This repeat-play nature stems from ongoing R&D cycles with overlapping technology generations. Should a 4G patent holder set excessive royalties in one generation, they risk losing market position with the launch of 5G. This is because licensees will choose from competing technologies, influenced by the licensor’s reputation. Even if a patent owner were to charge excessive royalties in one generation, this would not necessarily maximise revenue in that period. The increased costs of downstream implementers would hinder the adoption of the technology and limit sale numbers (and ultimately revenue) of the downstream product. Consequently, the licensor increases the risk that his patent will not enter the standard, resulting in massive sunk costs in R&D.
Despite the overwhelming evidence from wireless markets, competition regulators decline an integration of these empirical studies. They resist revising the existing policy approach toward SEPs and the FRAND requirement in wireless markets. This risks a reversion to the “IP = monopoly” equation and per se liability rules. Crucially, IP licensing has not had the market growth deterrent effect of taxes that was predicted. Instead, it has promoted rapid and widespread adoption among intermediate and end-users. There is strong confidence that the current system leads to a net-positive welfare outcome, supporting innovation for R&D firms and allowing access for end-users. Therefore, there is no clear evidence of market failure for extensive antitrust interventions to change established licensing practices.
3. THE ENABLING VIEW OF INTELLECTUAL PROPERTY LICENSING
The author advocates for a more open view of intellectual property licensing: in many aspects, SEP owners do not impose extractive “taxes” but rather facilitate value-creating transactions with other entities and enable market pricing of intellectual assets. IP scepticism views licensing as a tool for patentees to block competition and extend monopolies through litigation. The “Chicago” school critiqued this stance from the 1940s to the 1970s. It challenged the economic logic of per se liability rules applied to clauses like tying, exclusivity, and resale price maintenance. Scholars sought to challenge then-prevailing assumptions in economics and law. It was shown that several agreements previously seen as anti-competitive were not showing signs of market failure. However, there exists no follow-on research that developed a more precise understanding of economic functioning specifically attributable to IP licensing. This study considers the economic characteristics of intangible goods and the mechanisms that are used to convert an intangible good into a commercially feasible product. This perspective suggests that IP-intensive content and technology markets typically engineer transactional structures that facilitate efficient arrangements.
Contrary to the “license-as-tax” fallacy, this approach does not understand the IP licensors to be unrestricted monopolists. Instead, IP owners typically do not impose an extractive “tax” but rather, as is widely recognised in the case of property rights generally, enable market pricing of intellectual assets. This facilitates value-creating transactions with other entities that hold complementary IP or non-IP assets. In addition, this study argues that IP licences lower transaction costs by enabling value-creating exchanges of informational assets among business parties. It is obvious that IP licences impose access costs on users through royalties and negotiation expenses. However, these costs must be weighed against the innovation and economic benefits they generate. Hence, the net welfare effects should be analysed under a tailored rule-of-reason standard of antitrust law, as competitive harm is not always immediately evident.
One example of IP increasing welfare is vertical licensing in markets with Hub-and-Spoke structures. In these markets, there exists one large downstream incumbent which faces economies of scale for the commercialisation of innovation assets. Hence, the commercialisation of innovation comes with large fixed costs, such as in the case of pharmaceutical products. Upstream of this incumbent exists multiple smaller firms owning SEPs, facing diseconomies of scale in R&D. This means, breakthrough types of innovation tend to arise in smaller-firm environments where entrepreneurs have a large degree of control over the company. For small downstream companies, it would be difficult to achieve net positive returns due to the large fixed costs needed to develop and bring IP-intensive products to market. However, large incumbents can absorb this risk by funding diversified portfolios of technological projects, where the collective returns across projects are more likely to yield net positive returns. The result is the market following a “hub-and-spoke” structure. This model consists of two parts. The hub consists of a few large companies that can handle the expensive and high-risk tasks of commercialisation. They participate in capital-intensive production, marketing, and distribution activities. The spokes are populated by a larger number of smaller firms that focus on lower-cost but higher-risk innovation activities. These companies lack the scale to bring their innovation to the end-consumer market. Examples of these market arrangements can be found in content markets such as motion pictures, in which studios finance and source content from smaller production companies. Another example is technology markets such as biopharmaceuticals, in which “Big Pharma” sources R&D inputs from small biotech firms.
In these markets, patents protect the profits of smaller upstream companies while at the same time leading to overall more R&D than in a market where players would be vertically integrated. Secure IP rights and licensing agreements are a crucial part of the hub-and-spoke structure. For example, film-studios (hubs) and production companies (spokes) collaborate through contracts, sharing roles in the supply chain. IP licensing allows production companies to negotiate terms ensuring financing and revenue sharing with the studios. Without enforceable IP agreements, firms would internalise production to avoid imitators, limiting opportunities for independent creators. Reliable IP rights reduce entry barriers and distribute economic benefits across the ecosystem.
A second example where IP increases welfare is markets where a single IP asset can create plenty of downstream “derivative products”. Here, usually one major upstream IP holder lacks the downstream commercialisation capacities to convert that asset into an end-product. Instead, they license the patent to a downstream pool of intermediate users, who then embed the technology in a wide range of applications for the end-user market. The cases of Dolby Laboratories for audio technology and Qualcomm for wireless communication technology are examples. These companies have used licensing mechanisms to disseminate their technology and enable broad applications in the end-user market. Crucially, it is the licensing structure that creates royalties, allowing the upstream innovator to recoup R&D costs. Furthermore, it helps fund ongoing technology development for downstream production and distribution. While in theory the patent might constitute a legal monopoly, it would be difficult for the innovator to exert market power. This is because patents could be substituted by other processes or products. This can be, except for pharmaceutical markets, observed empirically in most licensing markets. Even when licensors have market power, they may choose not to exercise it. For example, in the case of fundamental recombinant DNA techniques, Stanford University held a landmark patent. Nonetheless, the patent owner chose to license on a non-exclusive basis below market value. There is an economic rationale for this. The patent owner maximises revenues by licensing all interested firms at a relatively modest rate that would then elicit widespread adoption. This leads to higher returns than setting a higher rate, which discourages smaller firms from taking a licence.
Another example where IP licences present welfare increases is patent pools. The academic theory of patent thickets hypothesises that the intensive issuance of IP rights and the dispersion of those rights among multiple holders may slow down innovation. However, this is not observed empirically. Both content and technology markets are unusually adept at engineering licensing solutions to potential patent or other IP rights thickets. For example, in the music industry collective licensing arrangements were introduced as a solution to public performance rights relating to musical compositions being dispersed among multiple owners. Similar structures can be found in the automotive or aircraft market. If IP rights generate a transactional barrier to the profitable exploitation of IP rights, companies have an incentive to find solutions around these barriers. Contrary to the patent thicket theory, real-world IT markets show companies using licensing agreements and pooling. This allows simplifying the complex licensing, and gaining efficiency by replacing many separate deals with one. Patent pools are a structure that explains in part why IT markets have largely avoided the high prices, low output, and slow growth as anticipated by the IP thicket thesis. Their structure as a one-stop shop presents opportunities to avoid any potential patent thickets by reducing transaction costs. Since patent pools raise the risk of collusion, they were de facto prohibited until the late 1990s. Current patent pools balance promoting pro-competitive solutions to IP roadblocks by mitigating the collusion risks of joint IP licensing, even under third-party administration. Four conditions maximise efficiency gains and minimise collusion risks. First the pool is open to all interested licensees. Second, the pool is restricted to complementary patents that are deemed “essential” for the relevant standard. Third, licensors are free to license independently of the pool. Fourth, the pool does not specify prices in the relevant product market.
Lastly, we observe increased welfare by IP giveaway strategies in some markets. Here, IP holders fully or partially give away their patented technology assets to maximise long-term revenues. This has proven welfare enhancing in several cases, such as Ethernet by Xerox, Intel and DEC, the USB interface by Intel, Java by Sun Microsystems, and Bluetooth technologies by a consortium of firms. Firms set zero or minimal royalties to encourage widespread adoption of standards. They rely on the fact that once their technology becomes widely adopted, it generates long-term revenue through follow up innovation or complementary products. This is enabled by minimising access costs of the crucial technology in the first place. This strategy contrasts with the “license-as-tax” model and illustrates that licensing can be used to lower access costs, fostering innovation and market growth.
4. REVISITING THE “LICENSE AS TAX” ANALOGY
Coming back to the “license as tax” analogy, the author concludes that viewing SEP owners as dominant players with significant market power is outdated. The “IP = monopoly” equation assumes that licensing royalties exceed, at least to some extent, the required amount to incentivise the investment that generated the innovation. This argument is, however, contestable on two grounds. For one it is the exception, not the rule, that IP rights give patent owners the ability to extract excessive rents from implementers. For another, viewing licences as taxes overlooks the efficiency gains they create even if licensors were able to extract excessive rents. Hence, a nuanced case-by-case assessment is necessary to estimate the welfare effect of IP licensing. Antitrust actions that deviate from fact-based analysis risk discouraging efficient licensing. Such deviations can also hinder the entry of innovation-driven firms that rely on licensing to monetise their R&D investments.
The study argues that the “license-as-tax” fallacy is a special case of the “IP = monopoly” equation, whereby the licensor can dictate the terms of use of its technology. However, drawing the conclusion that intellectual property leads to monopolistic behaviour of R&D companies is false. While patents represent legal monopolies, economically there might be a different input good for the implementer that might work as a substitute for that IP technology. Thus, the licensor still has no market power. We do not observe empirically any widespread extraction of monopolistic rents by patent owners, where royalties usually lie in the single-digit percentage of revenue generated.
In the current literature, there is wide consensus that the assumption that IP is equivalent to monopoly has only limited application in the real world. In fact, there are only two cases where these concerns are true. One is biopharmaceutical markets with above average royalty rates. The other is markets with smaller venture capital-backed firms, mostly in medical device and information technology hardware industries. In the latter case, smaller firms with IP-based advantages need pricing power as an incentive to challenge incumbents with cheaper, non-IP market advantages. However, in the broader IP licensing sphere these examples remain the exception. Consequently, competition authorities cannot assume that patent owners are per se welfare deterring monopolies. Instead, one should establish an appropriately calibrated rule-of-reason approach for determining the market power of patent owners.
One can see in several antitrust and patent laws and guidelines, that the mere existence of IP rights is a poor indicator for market power. An example is the 1995 Guidelines, or the Patent Misuse Reform Act of 1988. In these laws and guidelines it was required to prove that market power existed, rather than simply assuming it. Regulators and courts should use a balanced rule-of-reason approach. They should consider if an IP owner’s licensing terms result in a net anticompetitive effect only after market power is proven. The return to IP scepticism in recent years is reason to worry. If it is presumptively assumed that an SEP owner enjoys a secure legal and economic monopoly position, patent hold-up could seem the consequence. However, as previously shown, this does not match real world observations.
Competition regulators in SEP licensing should consider market- and firm-specific factors. Analysing real wireless markets, rather than theoretical models like the licence-as-tax analogy, could reveal competitive pressures on SEP licensing practices. The reason for that is the fact that SEP owners typically operate under a multi-period payoff structure. This creates incentives for relatively modest royalty rates. These rates promote the adoption of a new technology standard and build reputational goodwill, supporting adoption of future technology generations. Second, the conventional model overlooks key factors in SEP licensing. Non-vertically integrated SEP owners cannot credibly forego licensing revenue from major device manufacturers like Apple, Oppo, and Samsung, which control a large share of the market. These manufacturers have an incentive to delay negotiations or engage in “hold-out” behaviour to pressure SEP owners into lowering royalty rates. Meanwhile, they continue to profit from the technology. This creates an imbalance in negotiating power, as seen in Apple’s withholding of billions in royalties from Qualcomm during their legal disputes in 2017-2019. Crucially, it is not credible that SEP owners are typically able to raise royalty rates above socially efficient levels in wireless device markets. More likely, it is SEP users who may have the ability to depress royalty rates below socially efficient levels. This could be due to the practical difficulties of injunctive relief and the R&D investments made by innovators before standard selection and adoption. Additionally, it may result from the challenging-to-replicate suite of production and distribution assets held by leading implementers.
The “licenses-as-tax” fallacy contradicts the previous 1995 Guidelines and their 2017 revision, which suggest that IP does not automatically equate to a monopoly. The fallacy advocates for a lower evidentiary standard for judicial practice in IP licensing. Such a theoretical approach brings regulators and courts broad discretion when considering licensing agreements. Building on that fallacy, they might use that discretion to protect the market from high licensing fees imposed by IP owners. In a weak IP environment, innovators face three options for monetising their R&D: a) exit the market, b) build their own integrated production and distribution infrastructure, or c) seek acquisitions or employment with firms that already have such infrastructure. If building infrastructure is not feasible, the lack of reliable IP enforcement may limit options for monetisation. Ultimately, this protects incumbents, encourages premature exits or acquisitions, and discourages new entrants to the market.
The argument that antitrust policies may harm competition was anticipated by the Supreme Court’s decision in Continental T.V., Inc. v. GTE Sylvania Inc. in 1977. This decision in 1977 foresaw the potential harm of such antitrust policies to competition. Specifically, in United States v. Arnold, Schwinn & Co. (1967), the Supreme Court addressed an antitrust challenge to Schwinn’s distribution system, where exclusive territories were assigned to retailers. In that case, the Court ruled that antitrust violations depended on whether a retailer bought and took title to the products, applying a per se rule to sales transactions and a rule-of-reason standard to non-sales transactions. This decision was overruled in 1977 in Sylvania, recognising that the distinction between sale (per se treatment) and non-sale transactions (rule-of-reason treatment) lacked economic substance. Instead, a more economically reasonable approach was established. However, 2017 marked the return to the IP sceptical approach, reviving a formalist distinction between “sales” (which trigger exhaustion) and “licences” (which do not) with the Impression Products decision. According to the author, the Sylvania court correctly rejected arbitrary distinctions between sales and licences in determining the legality of use restrictions. Applying these distinctions between sale and non-sale makes little economic sense. Similarly, current arguments to limit IP licensing to protect consumers often lack factual evidence of competitive harm. While concerns about patent overreach may be plausible in theoretical models, there is little evidence that patent owners typically have the incentives to impose excessive licensing fees. Such strategies are self-defeating and do not maximise revenue in the long run.
The “license-as-tax” analogy and “IP = monopoly” assumption are poor principles for guiding IP licensing policies. Antitrust law should clearly prefer a rule-of-reason approach over per se rules to avoid wrongly suppressing efficient practices. Though costlier in assessing, the rule-of-reason approach allows courts to address genuinely anticompetitive licensing without harming beneficial arrangements. IP licensing enables inter-firm relationships that allow asset owners to profit without bearing the full costs of commercialisation. This structure supports innovation by allowing R&D specialists and production specialists to focus on their strengths without acquiring each other’s capacities. The increased scepticism toward IP licensing puts the legal foundations at risk, pushing firms toward internal commercialisation and reducing the variety of market-based transactions. Interventions in IP licensing markets favour large, integrated firms and disadvantage smaller innovation-driven firms reliant on licensing revenue. This could give rise to market structures that are less competitive and more resistant to new entrants. Without evidence of market failure, such interventions risk inefficiencies that contradict the goals of competition law.
5. CONCLUSION
In summary, this study contends that SEP licensing should be seen not as a restrictive “tax” on innovation but as a dynamic, efficiency-enhancing mechanism that fosters technological dissemination. SEP licensing supports R&D, and enables a competitive downstream market, challenging the notion that patent ownership inherently implies monopolistic control.
[1] United States. Department of Justice, & United States. Federal Trade Commission. (1995). Antitrust guidelines for the licensing of intellectual property. US Department of Justice, Federal Trade Commission. In 2017, the FTC and DOJ largely reaffirmed the substance of the 1995 Guidelines. United States. Department of Justice, & United States. Federal Trade Commission. (2017). Antitrust guidelines for the licensing of intellectual property. US Department of Justice, Federal Trade Commission.
[2] The fallacy does not consider that economic variables can change over time, the long-term effects and interactions of the market with outside factors. One-period profit maximising monopolies are firms that aim to maximise their profits in a single period. They do not consider future market conditions or potential long-term strategies such as reinvestment or market entry.
