Intellectual Property and the Fissured Economy

America has a monopoly problem. Most industries in the United States have consolidated in recent decades,¹ and markups and profits have dramatically increased since around 1980.² The American economy is likewise characterized by lagging investment,³ and this low investment is particularly notable among the largest and most dominant firms.⁴ A typical antimonopoly view might see this simply as the costs of insufficient competition: without needing to invest to stay ahead of competitors, monopolies get lazy, don’t expand or invest, and instead distribute cash to investors or sit on their reserves.

But if the lack of competition were the only issue, we would see markets where dominant firms—well-endowed with their own capital, workforces, and technologies—control their markets and compete only on a limited, oligopolistic basis. This is what many leading manufacturing industries, like cars, steel, and chemicals, looked like in the mid-twentieth century. Today, by contrast, what we see is a “fissured” economy, where those monopolies don’t even bother to directly employ many workers or own the physical capital needed to make the goods they sell or support the services they provide.

To take a well-known example: in 2019, Apple had ninety thousand direct employees, many of whom were highly compensated design, soft­ware, and marketing professionals at the core of Apple’s business. Nevertheless, the company claimed that its job footprint in the United States alone was 2.4 million jobs, employed through subcontractors and suppliers to make Apple products.⁵ This excludes the presumably millions more working for contractors like Foxconn in China and elsewhere in Apple’s supply chains. In other words, Apple has very few direct employees relative to its overall “job footprint.”

In August 2020, Apple’s market capitalization crossed $2 trillion,⁶ and it is one of the most profitable companies on earth, raking in $94.7 billion in 2021.⁷ At the same time, Apple has long been known to sit on large unused cash reserves, on the scale of $200 billion in 2021.⁸ Mean­while, Apple’s capital expenditures have lagged, sometimes not even meeting the depreciation costs of its existing capital stock.⁹ Compared to this dearth of physical capital formation, the total value of Apple’s intangible assets, including all its intellectual property (patents, software copyrights, know-how, and trademark), was estimated to be a whopping $1.87 trillion in 2021, making up a majority of the company’s value.¹⁰ Not only employment-light, Apple is a capital-light business.

In short, with few direct employees, Apple earns enormous profits, invests very little relative to its overall economic footprint, faces limited effective competition, owns vast amounts of legally protected intangible assets, and thus epitomizes many of the features of our current economy. Firms like Apple, taking the lion’s share of the profits, rely on the workforces and capital investments of companies they contract with.

Almost ten years ago, economist David Weil noticed these trends—specifically that the U.S. workplace had broken apart into a web of sub­contracting, franchising, and hierarchical supply-chain relationships—naming this phenomenon the “fissured workplace.”¹¹ Rather than work­ing directly for the “lead” firm that produces cars, runs hotels, or designs smartphones, a worker might be a subcontracted technician in a factory, an employee of the franchisee for a global hotel company, or a worker in a company contracted to assemble iPhones. In all these circumstances, through a legal reshuffling of the boundaries of the firm, a worker is separated from the profitable lead company.

This phenomenon occurs with respect to physical capital as well. Herman Mark Schwartz has shown that profits today are generally distributed between three types of firms.¹² First, at the top, the largest profits are generated by firms with very little physical capital and very few workers, but which instead have very large portfolios of intellectual property (IP), whether in technological patents, software and creative copyrights, or trademarks for global brands. Second, at the bottom, are labor-intensive segments of industry, where working conditions and pay are poor, and profits are hard to come by; these firms are often doing the work of the firms in the first group. This reality characterizes franchise businesses and workers in your local Subway or Hilton, as well as sweatshop workers making garments for global brands like Nike. Final­ly, in the middle are the remnants of the mass production economy: capital-intensive manufacturing industries, whether it be cars or semi­conductor chips, in which companies make significant capital investments and directly employ workers involved in production, though to a lesser degree than in the past.

Thus, by “fissuring,” I am referring to the separation of both labor and physical capital from profits, across the legal boundaries of different companies, rather than having the footprint of all three largely reside within the same company. To be clear, this is distinct from vertical dis­integration, whereby different companies make different components at different stages of a value chain, but still earn profits commensurate with their capital and labor inputs. By contrast, a company like Apple, which has high profits, yet contracts or licenses with others for most of the physical capital and labor requirements of production, is a fissured company.

This fissuring is, as Schwartz argues, likely a significant contributor to stagnant economic growth today.¹³ Whereas in a previous time, when the oligopolies of cars, chemicals, and other heavy industries earned enor­mous profits through their economies of scale, they both reinvested those profits and partly shared the spoils with a broad base of direct employees. Strong wages across large swaths of corporate America increased aggregate demand as a key driver of growth, and the corporate strategy of reinvesting most profits likewise drove economic expansion.

As Schwartz shows, however, profits have been sequestered away from capital-intensive industries (oil and chemicals, integrated manufacturing, telecommunications, cars) to high-tech but fissured industries (finance, tech, pharma, consumer brands) in which the most profitable firms earn massive profits but have a low propensity to invest.¹⁴ This misalignment—where one set of firms makes the profits and a different set of firms has large capital requirements but insufficient resources to make them—leads to a significant shortfall in investment. This is, likewise, among the reasons why the high valuations of these fissured companies have become so disconnected with economic growth and performance more generally.

Nonetheless, from a concrete policy perspective, there has only been limited exploration of why this fissuring and misalignment between profits and investment needs has happened: Julius Krein sees many of these trends in the shifts in finance to reward and prioritize asset valuations over economic growth and meaningful uses of capital, but he does not explore why this has occurred.¹⁵ David Weil emphasizes the shift to more flexible, just-in-time markets that pushed firms to modularize production and fissure their workforces.¹⁶ Others point to the rise of information technologies and globalization, which arguably created a winner-take-all market where “superstar firms” take most of the spoils.¹⁷ There is a widespread recognition of the role of antitrust rules in workplace fissuring, but this has mostly, though not exclusively, focused on the comparatively low-profit franchise sector.¹⁸

When it comes to solutions, Schwartz points to strengthening anti­trust enforcement and weakening IP rights, both seemingly with the understanding that greater competition is needed. But reviving antitrust policies that aim to contain monopoly and introduce competition, while important, would not in and of itself address this separation between profits and the physical resources that make the economy move. No amount of monopolization cases or blocked mergers will, on their own, un-fissure the economy.

I argue, rather, that the legal rights around technology and knowledge were significantly altered, specifically with respect to how those rights can be exploited through licensing and “vertical restraints.” Intel­lectual property rights have effectively become stronger in the last forty years than they have ever been before. But more importantly, the United States used to significantly restrict how owners of such property could use their rights to impose conditions on others. Free from those restrictions today, fissured companies can earn large profits from their intellectual property portfolios without having to directly control production or incur direct employment or investment spending. The origins of this fissuring are not based in technological changes, market pressures, or business strategies unilaterally chosen by corporate leader­ship. It is not technology itself that changed, I argue, but rather what companies were allowed to do with it.

This piece is a map of what policies changed, aiming to conceptually connect the specific—and often seemingly mundane—policy changes to their sometimes large economic effects. The array of policies highlighted here—licensing, the disposition of public patents, judicial reorganizations, shifts in the emphasis of antitrust, the consolidation and expansion of intellectual property rights—is partial and ignores many other factors that were also important. It was the accumulation of many small changes, reorganizations, reinterpretations, extensions, and carve-outs to existing rules that, I argue, brought us to where we are.

Antitrust and Vertical Restraints

One of the most significant policy changes driving these trends was a shift in the focus of antitrust. A common view of antitrust is that it is simply intended to ensure competition. Here, local construction firms and giant telecommunications firms alike must compete for the custom­er’s business, pushing them to provide better products and services, for better prices, and at larger scales. This view tends to focus on “horizon­tal” relationships, for example competition between two car manufacturers. By contrast, a car manufacturer, an autobody shop, and a company that makes automotive software do not directly compete. These vertical relationships often remain overlooked in the horizontal approach to antitrust.

But when antitrust is understood as more than a matter of horizontal competition, thorny questions arise: Should a manufacturer be allowed to require that an auto repair shop exclusively service its own cars, as a condition of being able to service any of its brands? Should a manufacturer be allowed to require that a software company provide software only for its cars, or require that they only service cars in a particular region? These issues, in turn, affect corporate decision-making: Should a company make a specific component or buy it from a supplier? Should it sell a product to distributors or build out its own distribution and retail network?¹⁹ Antitrust has usually resolved these questions in terms of whether vertical restrictions like this amount to restraints of trade or restrict competition. As Sanjukta Paul has argued, this implicit “firm exemption” shows how antitrust does not just encourage competition, but allocates “coordination rights” between different types of firms, commercial relationships, and indirect employment relationships.²⁰

The implications can be substantial. When a company builds some­thing in-house, the workers who build the product are employees, entitled to benefits, bargaining rights, wages, etc. Or, when a company simply buys something from another independent company, it enters into an arms-length business relationship in which both companies earn money, make investments, and have their own employees. But when business activities are contracted out with restrictive vertical restraints, the dominant company begins to have employer-like control over another company’s workers—through the price floors and caps they impose, the contract terms they make, the schedules they set, the speci­fications they require, and so on—but without having any of the responsibilities that are legally required of an employer.

More broadly, the separation of substantial capital and labor require­ments from profits earned—the “fissuring” described earlier—is made possible precisely by this vertical licensing. As in the previous example, Apple is only able to contract out critical operations because they are able to tell manufacturers what to build and how, on specific timetables, while intellectual property protections allow the company to capture high margins. Likewise, if tech companies were to develop the necessary software to operate self-driving cars, and were able to restrict that technology through intellectual property rights, they could contract out manufacturing and retailing to exact specifications and earn an overwhelming share of the profits. In this outcome, the fissuring of the automotive sector would be complete: a software firm, responsible for little of the physical capital and labor inputs, would take most of the profit. This scenario is not purely hypothetical: it is a stated concern of actual car manufacturers, afraid of becoming subservient to tech compa­nies’ control of the operating systems around autonomous vehicles, should self-driving cars become a reality.²¹

Such business practices would not have been permitted under the antitrust policies of previous eras. It is a well-known story that antitrust enforcement was scaled back beginning in the late 1970s, and particularly under the Reagan administration, inspired by the Chicago school of antitrust.²² Central to this shift was the weakening of antitrust prohibitions against “vertical restraint”—coordination between firms in different segments of the same supply chain or in adjacent industries. These restrictions are rarely enforced today.

In fact, the Chicago school’s intellectual origins were tightly tied to the question of vertical restraints. Aaron Director, a professor at the University of Chicago in the 1950s, did not have a master theory of antitrust, but rather a particular bone to pick with the existing law’s understanding of three specific practices—tying, resale price maintenance, and predatory pricing—two of which are vertical restraints.²³ Tying is a practice by which the sale of one product, such as an operat­ing system, is conditioned on a related purchase, such as an internet browser or mobile app store. Resale price maintenance occurs when a manufacturer sets the price at which a retailer can sell. Predatory pricing is not a vertical restraint per se, but involves a firm selling below cost to drive competitors out of the market, only to then be able to raise prices again once its competitors are gone. Think, for example, of how Uber ran at a loss subsidized by investor cash for well over a decade, driving much of the cab industry out of business, with the explicit goal of exerting pricing power after achieving market dominance.²⁴

Director thought all three practices should be allowed, or that the potential costs of them were tiny. But it was only his students—Robert Bork, Frank Easterbrook, and Richard Posner—who later attempted to formalize this thinking into a general approach to competition policy.²⁵

The Chicago school’s first major victory in the courts involved precisely these questions of vertical restraints, in the 1977 Contintental T.V. v. GTE Sylvania decision.²⁶ At issue was the legality of the fran­chise sales model, in which a manufacturer assigns retailers certain territories where they can sell the manufacturer’s products, prohibiting them from selling products—which they had bought and owned—anywhere else. Citing Bork’s antitrust theories multiple times, the Supreme Court ruled the practice legal, overturning its own position from ten years earlier.²⁷

This decision opened the door to allowing all sorts of other vertical restraints. In 1984, the court ruled in Jefferson Parish Hospital District No. 2 v. Hyde that antitrust allowed hospitals to use exclusive contracts with patient referrals, referring back to Sylvania.²⁸ Also in 1984, Mon­santo v. Spray-Rite decided that a manufacturer could fix the prices at which its distributors could resell its products, again primarily relying on Sylvania’s reasoning that vertical restraints were to be treated more favorably.²⁹ In 1990, Atlantic Richfield v. USA Petroleum expanded manufacturers’ ability to set maximum prices for others to resell their products, again relying on the increasing permissiveness toward vertical restraints.³⁰ This was solidified by 1997’s State Oil Co. v. Khan.³¹ In 2007, Leegin Creative Leather Products, Inc. v. PSKS decided to allow vertical restraints to set minimum resale prices, more or less allowing price fixing via vertical restraint.³²

Keep in mind that this is only a sampling of major decisions. During the 1980s, the Reagan administration more or less abandoned its en­forcement against vertical restraints, following the Chicago school, and instead brought cases against horizontal price-fixing.³³ Other administrations have broadly followed this shift ever since.

These choices were made because vertical restraints had come to be considered “procompetitive.” But vertical restraint, even if it does not affect horizontal competition, gives companies many tools to fissure their workforces. If a company does not wish to live up to its obligations as an employer—overtime pay, minimum wages, retirement and health benefits, or workplace safety obligations—vertical restraints now give companies the power to start a new company, move those employees into the new company, and then use vertical restraints and contracting to control those workers to nearly the same degree as before.

In the context of labor fissuring, this connection is well-established. Sanjukta Paul explains how the coordination rights provided by vertical restraints allow lead firms to fissure their workforces, while prohibitions on horizontal restraints prevent any countervailing bargaining power from the firms they control.³⁴ Brian Callaci demonstrates that this is in fact the effect of vertical restraints on labor fissuring within the franchise sector.³⁵ Kati Griffith shows how franchises use their control over the managers of franchisees to effectively act as joint employers, despite avoiding any direct employment obligations.³⁶

Technology Licensing

Over time, the presumption that vertical restraints were procompetitive was expanded and applied to intangible property like patents, trademark, copyright, and business know-how. These developments allowed companies owning such intangible property or technological knowledge to place greater conditions on other companies in order to maximize profits from their IP.

This form of technology licensing through vertical restraints was generally prohibited in the mid-twentieth century. In the late 1930s, Thurman Arnold became a household name as the assistant attorney general of antitrust. During his five years in that role, he filed almost half of all Sherman Act cases that had been filed since the law was passed in 1890. Less known, however, was Arnold’s relentless focus on technology, specifically patents, as the source of bottlenecks and restraints in industry. The international cartels of the interwar years were built on patent pools. Leading global manufacturers of aluminum, tungsten car­bide, incandescent bulbs, chemicals, and many other products would pool all their patents together, and then agree to divide world markets among themselves, using patent rights to secure a regional monopoly for each member company. In fighting these international cartels, Arnold submitted patent legislation to Congress, but the legislation went no­where.³⁷

Nonetheless, in the course of many, many antitrust cases filed against these arrangements, the courts began to decide in favor of Arnold and the Antitrust Division. The 1938 Ethyl Gasoline decision limited the rights of patent holders to set post-sale prices and eliminate competition between licensees for leaded gasoline.³⁸ The Morton Salt decision of 1942 held that, with respect to salt tablet depositors, it was an antitrust violation to tie the sale of an unpatented product to a patent license.³⁹ Univis in 1942 held that a company could not use a series of patents to control prices at every stage of the supply chain for glass lenses.⁴⁰ In Masonite, also in 1942, the court ruled that patent licenses for synthetic building materials did not provide any defense to the Sherman Act’s prohibitions on price-fixing.⁴¹ Two years later, Mercoid established that a company may not leverage its patent monopoly to gain a monopoly over an unpatented process or product for heating systems.⁴² Hartford-Empire in 1945 found a group of corporations had illegally acquired a series of patents for glassmaking machinery, and licensed them exclusively to cover an entire industry to restrict competition.⁴³ These are basic products today, but many of these were cutting-edge technologies at the time.

The well-known 1945 Alcoa decision was not just the condemnation of a prominent monopoly, but also a condemnation of the way the company had used patent licenses as a method of coordinating its international aluminum cartel.⁴⁴ In 1947, National Lead ruled against an international cartel that pooled patents to allocate territories to each member.⁴⁵ United States v. Line Material (1948) decided that multiple patentees may not enter a cross-licensing scheme that fixes the resale price of products manufactured under the licenses.⁴⁶ The decisions continued from there. This flurry of antitrust litigation against patents and patent licensing created a body of law and an Antitrust Division that saw patent licensing as a key anticompetitive bottleneck, to be challenged whenever possible.

These goals were expanded over the postwar period, to the point that, by 1970, the Nixon administration’s Antitrust Division opened a patent section, specifically to challenge patents and patent licenses on antitrust grounds. In 1970, Bruce Wilson, a deputy assistant attorney general for the Antitrust Division, outlined the Division’s policy in 1970, detailing “nine no-no’s”⁴⁷ of patent licensing that the Division would presume to be illegal and challenge: (1) tying the purchase of an unpatented product as a condition of a patent license; (2) requiring the licensee to grant back licenses for any improvements they make; (3) placing restrictions on the licensee’s resale of patented products; (4) re­stricting a licensee’s business outside of scope of the patent; (5) agreeing to not grant licenses; (6) requiring bundling of multiple patents as a condition of license; (7) imposing royalty provisions unrelated to the licensee’s sales; (8) restricting use of unpatented products made by patented process; and (9) requiring a minimum sale price of patented products.

While each of these principles was subject to technical considerations in specific cases, this general approach presumed that these restrictive conditions were suspect or illegal. And while this list of “no-no’s” was not a formal policy statement—it was, in fact, just a speech—the Antitrust Division’s patent litigation was rather successful in this era.

For precisely this reason—because companies were unable to license out the production of their technologies on restrictive terms—firms in the postwar era had strong incentives to undertake, in-house, the capital investments and hiring required to profit from their technologies. Com­panies either had to make the capital and wage expenditures themselves, or license the technology out with relatively nonrestrictive terms. In­deed, many companies were required to license key technologies,⁴⁸ but the share of profits they earned from licensing would not be so outsized as to prevent their capital-intensive licensees from making significant investments themselves.

This is in stark contrast to today, when most IP licensing is considered procompetitive, because licensing allows a technology to be more widely utilized (though such logic becomes dubious when lead firms are allowed to place significant vertical restraints on licensees). This ap­proach to IP was established bit by bit as the courts whittled away the prohibitions on vertical restraints and enforcers opted to interpret these practices more charitably. For example, in 1980, Dawson Chem. Co. v. Rohm & Haas Co. allowed tying as a condition of a patent license,⁴⁹ and the Patent Act was amended by Congress in 1988 to allow for tying.⁵⁰ Today, corporate legal guidance routinely notes that most of these rules are effectively dead.⁵¹

In the 1970s, the DOJ began putting out antitrust guidelines for “international operations,” which over several decades began to shape intellectual property licensing rules as well. The 1977 “Antitrust Guide for International Operations” was the first iteration, intended to clarify the rules of the time, and thus followed the federal government’s skepticism of patent licenses.⁵²

The 1988 “Antitrust Enforcement Guidelines for International Operations,” on the other hand, added a series of safe harbors for mergers and IP licensing, and took a favorable view of most vertical restraints and territorial restrictions in licensing.⁵³ Whereas the 1977 guidelines were skeptical of companies allocating territory based on know-how licensing, the 1988 guidelines were far more permissive.⁵⁴ As a later report noted, “A fundamental principle of the 1988 guidelines was that the owner of intellectual property rights is entitled to maximize the market value of its intellectual property.”⁵⁵

The 1995 “Guidelines for the Licensing of Intellectual Property” brought this logic to its conclusion, applying the new antitrust approaches from other areas to vertical restraints like tying and exclusive dealing, as well as to intellectual property. It states that “the Agencies regard intellectual property as being essentially comparable to any other form of property; (b) the Agencies do not presume that intellectual property creates market power in the antitrust context; and (c) the Agencies recognize that intellectual property licensing allows firms to combine complementary factors of production and is generally pro­competitive.”⁵⁶

While my focus here has been on technology, trademark and copy­right also play important roles. Hiba Hafiz has observed how trademark licensing rights under antitrust have contributed greatly to the fissuring of franchised workplaces since the 1970s, concentrating profits within the corporate franchisor with limited employer obligations.⁵⁷

Expanding Intangible Property

While the changing rules around licensing are central, policy and legal changes further expanded the sorts of things that could be owned as intangible property. Software code gained copyright protections, recom­bined genetic sequences became patentable, and certain categories of manufacturing processes and semiconductor chips gained specific IP protections. Technologies developed under government contract were progressively shifted from government ownership to the private sector. Courts were reformed to more consistently decide that patents were valid. Trade policies moved in tandem, such that all these categories of intellectual property rights are now protected as a condition of access to world markets.

While this article is not an attack on the notion of intellectual proper­ty, it is worth noting that the relationship between IP protections and innovation or technological development is not nearly as strong or as clear as often portrayed. Some of the best research on the topic, from economist Petra Moser, indicates that patent and other IP protections primarily determine the sort of technologies that are developed, not the amount of innovation or technological progress.⁵⁸ Comparing the new technologies between countries that did and did not have patent systems in the nineteenth century, she shows that patent protections primarily changed the direction of innovation to favor technologies and know-how that could not be protected via alternative means like secrecy. Countries entirely without patent regimes were not less innovative per se.

Software. Information and digital technologies had been developing since World War II, but really began to take off in the 1970s. In 1974, the Commission on New Technological Uses of Copyrighted Works was established and reported that software and computer code should be covered by copyright protection as a creative work. But it was not until 1980 that the U.S. Congress updated the Copyright Act of 1976 to in­clude computer programs under copyright protection.⁵⁹ In short order, and in conjunction with the changes to antitrust vertical restraints, software began to gain commercial dominance over hardware.

The timing is not entirely coincidental with one of the better known stories of the victory of intangible software companies over hardware manufacturers. In 1980, IBM realized it was late to the personal comput­er market, despite being one of the most successful computer manufacturers in the country. In an attempt to enter personal computers and to compete with Apple, IBM needed an operating system for its own computers, and it ended up contracting with Microsoft, at the time a small start-up.

Because IBM had been facing decades of antitrust scrutiny over its dominant position, the company was worried about using vertical re­straints like exclusive contracts. As a result, it gave Microsoft a nonexclusive contract, such that Microsoft could then license out the operating system it developed to other manufacturers.⁶⁰

Copyright protections for software, like Microsoft’s operating sys­tem, were established by Congress on December 12, 1980, a month after the November 6 agreement with IBM. These expanded rights were clarified through litigation throughout the 1980s, as at first courts did not have a clear sense of what counted as copyright, what counted as a patent, and what exactly other companies would be able to copy or build on. In 1983, Apple v. Franklin held that all computer programs are copyrightable,⁶¹ and this legal ruling laid the foundations for the norm of proprietary software, on which the modern digital economy runs.

These decisions and policy changes ensured that Microsoft’s business model would follow the trajectory of other fissured companies. Micro­soft’s software business requires little investment in physical capital and very little direct employment beyond the core software development, while its software is heavily protected by copyright, and the company earns an outsized share of the profits within its commercial ecosystem.

Biotechnology. In 1980, the Supreme Court also decided the Dia­mond v. Chakrabarty case, ruling that genetically altered bacteria could be patented and owned as a technology.⁶² This was expanded in the late 1980s by other decisions of the Patent and Trademark Office to allow for the patenting of genetically altered plants, animals, and foods.⁶³

Predictably, as anyone familiar with agricultural markets today is aware, seed companies like Monsanto now bind farmers within exclusive and proprietary crop systems. They rely on their large patent portfolios over crops and seeds, and their ability to license their IP out on restrictive terms that are extremely favorable to Monsanto, in order to maximize profits.⁶⁴

Public patents. Another issue is patents for technologies that are developed under government contract. Basically, if the government pays someone to do research or develop a technology, who owns the patent for that technology?

This became an issue right after World War II, when the U.S. government adopted a “title” policy in some federal statutes. A title policy meant that the government retained ownership of the patents but would license them openly for a relatively low royalty. This was the case for the 1954 Atomic Energy Act and the 1958 National Aeronautics and Space Act, and several agencies followed this approach. An executive order in 1950 attempted to establish a uniform title policy.⁶⁵ This was restated in 1963,⁶⁶ and the 1974 Federal Nonnuclear Energy Research and Development Act mandated a very clear title policy.

During the 1970s, however, a new consensus arose, holding that the private sector would be better able to successfully commercialize the technologies protected by public patents, and exploit the benefits of these innovations. This corresponded with a drift toward “license” policies, the longstanding practice of the Defense Department. With a license policy, the contractor gets the patent for the technology devel­oped with public resources, but must license it for free to the government, and allow the government to “march in” if it sees fit to do so. This shift progressively moved more and more ownership of technology into private hands.

This trend culminated in the 1980 Bayh-Dole Act, which mandated a license policy for small businesses and nonprofits, including nearly all major universities. An updated executive order in 1983 mandated the maximum expansion of the license policy.⁶⁷ Then, in 1985, the Department of Energy began giving “class waivers” of the patent titles when a title policy was required by statute, such as to large businesses, essentially privatizing patent rights even when the law required that they remain public. These class waivers are still routinely issued in bundles today.⁶⁸

Forty years later, even the Bayh-Dole Act’s explicit attempts to ensure public use of relevant technology rights have failed. One require­ment is that licensees of government-funded technologies manufacture those products in the United States, but nearly all, if not all, applications for waivers of that requirement appear to be granted at least in the pharmaceutical space,⁶⁹ allowing a category of drug patent holders to rake in massive profits from government-funded research while out­sourcing the drug and active pharmaceutical ingredient manufacturing to India and China.

Court reforms. Several reforms to the court system in the early 1980s likewise strengthened patent rights. In 1982, the Federal Courts Im­provement Act, among other things, created the United States Court of Appeals for the Federal Circuit, which consolidated appeals for patent cases in a single court. This followed a series of recommendations from late in the Carter administration, but was motivated by complaints in the patent bar that the rate at which courts upheld patents as valid was highly unpredictable and varied drastically between federal circuits.

The Federal District Court Organization Act of 1984⁷⁰ was labeled as a bill of miscellaneous reforms, meant “to make various statutory amendments with respect to: trademark clarification; Federal assistance to the States for judicial administration; copyright protection for semiconductor chips; improvement of Federal courts administration; and patent policy, as it related to Federal research and development activities.” In its details, however, the law further consolidated patent cases into the U.S. Court of Appeals for the Federal Circuit, expanded the Bayh-Dole Act beyond just small firms and nonprofits by allowing government patents to be handed off to large firms as well, and added IP protections for “mask work” rights for semiconductor chips.

Intellectual property internationalized. The 1994 trips Agreement (Trade-Related Aspects of Intellectual Property Rights) solidified these rights internationally, setting minimum standards for patent, copyright, and trademark protections as a condition of membership in the World Trade Organization. This included, among other things, copyright pro­tection for software.

Trips was the culmination of a multiyear project beginning in the early 1980s to include intellectual property as a part of trade negotiations, even though the two topics are broadly unrelated and had been considered distinct as recently as the late 1970s. But in the 1980s, the U.S. government began treating the absence of IP protections as a barrier to entry for American multinationals, not because they were barred from entering, but because they would not want to enter if their IP protections were not enforced in other countries.

What to Do?

This policy history argues that the intersection of rules around IP, vertical restraints, and licensing have allowed for the fissuring of both capital and labor from the lead firms that produce most profits today. Simply put, because technology licensing on restrictive terms allows the IP holder to maximally profit without significant investments in labor or physical capital, a subset of firms well-endowed with intangible capital can rake in enormous profits that are denied to labor- and capital-intensive industries, which do need the investment.

In a previous era, there were antitrust rules that prohibited exactly this sort of arrangement at the intersection of vertical restraints and intellectual property. Some of these issues could be resolved simply by reinstating many of the old antitrust rules, such as per se prohibitions on restrictive technology licensing. Those rules, however, were based on antitrust concerns about restraints of trade or ensuring competition, rather than the question of who captures the profits relative to who needs to make capital expenditures. Even putting aside how antitrust is today considered under the consumer welfare standard, competition is only an indirect way to get at the question of fissuring.

Particularly in light of this disjuncture, an important question in all of this history is how much American policymakers “meant” to make these changes that likely facilitated such a large redistribution of profits. The answer, unfortunately, is somewhere between mixed and unclear. Many of the antitrust changes were originally driven by the Chicago school’s beliefs about vertical restraints, but Chicago school thinkers themselves insisted that their motivation was for antitrust to prioritize efficiency, even if the effects were predictably to allow more powerful firms to dominate others.⁷¹ Meanwhile, the campaign to allow labor fissuring through franchising relationships was very much an intentional lobbying campaign by franchisors to retain profits without employer obligations.⁷² The shifts in many areas of intellectual property were driven not by a motivation to redistribute profits, but rather by a desire to promote innovation in the shadow of the America’s waning technological lead relative to the rest of the world. The Bayh-Dole Act, for example, came out of the Carter administration’s 1979 domestic policy review of industrial innovation.⁷³

Some qualifications are in order as well. Additional research could verify or disprove whether the intersection of antitrust policy and intellectual property protections is a core driver of the problem. For example, if the distribution of vertical restraints and intellectual proper­ty do not correspond with the most fissured industries and companies, then this explanation could be found lacking.

Yet the typical view of these policy areas—thought to govern entirely different relationships—is clearly inadequate. We are told that patents are needed to protect innovations and incentivize the creation of new technologies or products. Antitrust rules are supposed to ensure competition between companies, or to prevent one company from dominating the market. Licensing—allowing someone else to use your patented or protected technology in return for compensation—is sup­posed to encourage the technology to be used more broadly. In a manner that has little to do with competition, innovation incentives, or technological progress, however, the intersection of these policies has allowed profits and corporate valuations to become increasingly dis­connected from the labor and physical capital that make up the “real” economy. A better understanding of how these seemingly separate poli­cy domains interact to influence investment decisions and corporate profit distributions is needed to address the challenges facing an in­creasingly concentrated yet simultaneously fissured economy.

This article originally appeared in American Affairs Volume VII, Number 2 (Summer 2023): 3–22.

Notes