Public policy needs to be more supportive, and firms need to be willing to support more blue-sky projects. As a nation, we are harvesting the fruits of old enabling technologies without investing sufficiently in new ones.
Certain innovations—known as enabling technologies—provide the foundation for progress across a range of industries. Enabling technologies include mobile wireless, the laser, CT scanners, the microprocessor, artificial intelligence, and freight containerization. Such technologies drive wealth creation throughout the economy. However, the difficulties associated with monetizing this type of IP, which I explore in this article, mean that private enterprise tends to underinvest in new enabling technologies. Public policy needs to be more supportive, and firms need to be willing to support more blue-sky projects.
As a nation, we are harvesting the fruits of old enabling technologies without investing sufficiently in new ones. We are eating our seed corn.
A technology can be considered “enabling” if it exhibits at least two of the following three characteristics: (1) it impacts a wide range of activities and industries, (2) it gets better over time, and (3) it triggers complementary innovations. Consider artificial intelligence (AI), which has recently transitioned to being an enabling technology. AI encompasses a range of software techniques employed to teach computers to sense, reason, interpret, communicate, and make decisions in a human-like manner. AI-based technologies can already, with various limitations, recognize faces, understand speech, and drive vehicles. Authorities use AI to sort Internet chatter and zero in on fraud, illegal activity, and terrorist plots. The cost of AI continues to fall as it becomes available as part of the capabilities offered for hire by the providers of cloud computing. Its future applications are limited only by economics and imagination.
Academic research on enabling technologies (including a particularly valuable subset known as general-purpose technologies) have shown that their long-term impact can be massive. One study found that the adoption of container shipping over the period 1967-1982 by OECD countries raised the value of trade by 1,240%, swamping the effects of tariff reductions over the same period. Yet, despite the value created, only a small fraction of that value is converted into profits for the pioneers and the owners of the underlying intellectual property. Other studies have found similar results for the steam engine, electricity, and advanced materials.
One reason for the chasm between the social and private returns is that much of the new knowledge and opportunities fostered by an enabling technology is never paid for. Economists call these “spillover” benefits. In theory, the innovator could profit from those spillovers by taking ownership of all the complements whose values are increased by the innovation. In reality, this is seldom likely to be practical because an enabling technology affects such a wide range of asset classes. Start-up innovators of enabling technology will almost certainly need to rely on partners, and even most established firms will lack some of the relevant capabilities to derive a double-digit share of the value. Put differently, capturing value is hard because there isn’t a robust business model available to capture more than a tiny fraction of the value generated. Strong IP protection and a well-functioning market for technology are generally needed in order for inventors of enabling technology to be adequately rewarded.
A corollary is that the economy as a whole suffers from the negative demonstration effect when inventors of enabling technology encounter major difficulties in trying to capture value. Because potential inventors see the problems that current innovators face when they attempt to garner profits, they will be less likely to see enabling technologies as a worthwhile investment. Market forces will therefore under-provide new enabling technologies in the future, turning the innovator’s profit gap into an economy-wide investment gap.
Recent data suggest that the problem is getting worse, not better. While US business R&D has increased over the past 20 years, the growth was primarily in later-stage development work. Spending on basic and applied research has been essentially flat, according to National Science Foundation data. While not all basic and applied research is devoted to enabling technology, enabling technologies depend on the early, more fundamental stages of R&D.
In the rest of this article, I will discuss a number of reasons for the profit gap. Whether the technology is licensed or incorporated into products and services, innovators will be challenged to earn profits commensurate with the investment and risk involved in developing the enabling technology. The knock-on effects from this create a huge loss of potential for the wider economy and for society.
The nature of an enabling technology is that it opens opportunities in multiple downstream industries and markets. As a practical matter, this leaves licensing as the most logical way to capture value because simultaneous forward integration into multiple, disparate industries would be almost impossible. When intellectual property protection is strong, licensing will, in theory, allow an innovation’s value to be extracted for all the potential downstream uses. In practice, this pure-licensing business model faces issues of appropriability and regulation.
IP Appropriability: Many enabling technologies, like 4G wireless, are composed of multiple inventions, possibly even hundreds, which can be owned by various companies. When an enabling technology is commercialized through a licensing business model, users or implementers may simply infringe patents. They may calculate that infringing is worth the risk. A patent holder, on the other hand, will likely have to incur litigation costs associated with challenging patent infringers and trade secret misappropriators before even launching a licensing program. Given the uphill battle most patent owners face, it’s often pragmatic to give a discount to licensees to help launch a licensing program. Once a credible implementer takes a license, it’s easier to get others to follow. The discounts will, of course, reduce the innovator’s share of the collective profits and lower the expectations of potential future innovators deciding whether creating a new enabling technology is worth the investment.
More generally, patents—even if ironclad and potentially valuable—are virtually never a guarantee of profits because patents are not self-enforcing. The legal and financial requirements for upholding patent validity and for proving infringement are high. Patents can also, in some cases, be “invented around” at relatively modest cost.
Regulation: The capture of an enabling technology’s value from its users may also be limited by regulatory barriers. The profit that the highly regulated AT&T was able to earn from its 1947 invention of the transistor—the Nobel-winning basis for the post-war microelectronics revolution—was arguably limited under a 1956 antitrust consent decree. In xerography, Xerox, which had pursued a strategy of in-house manufacturing and international joint ventures, was accused of having monopoly power stemming from its patent portfolio, and in 1975 US antitrust authorities required it to divest its foreign subsidiaries and offer low-cost licensing of its key patents to foreign and domestic competitors. This both reduced royalties and enabled new rivals.
Today, successful innovators are likely to find their royalty rates closely scrutinized by foreign regulators as well as by domestic policymakers. Qualcomm, a major contributor to wireless standards, discovered this in recent years with antitrust fines levied in China, South Korea, and Taiwan. These fines have no connection to the underlying value generated. Untested and empirically unsupported antitrust theories were invoked in order to obtain access to patented technology at bargain prices.
Integrated Business Model: Challenges
When the business environment for a licensing model is unfavorable (for example, due to weak intellectual property protection), the innovator can attempt to capture value by acquiring or controlling the complementary assets that are required for commercialization. All technologies require complements—some of which may not even exist— to reach market. The science behind AI, for example, was initially developed in the 1950s, but needed cheaper and faster computing power, especially developments in the complementary hardware technologies of processors and memory, to become commercially interesting.
Complements can include a wide variety of products, from inputs to factories, and services, from administration to distribution. Some of these, the innovator can (or may have to) build rather than buy, which is typically slow and expensive, but offers strong control of the technology’s development and of the future profit stream. In some cases, partnerships will be an option but will drain value away from the innovator and may lead to over-reliance on an external supplier of a key input.
An integrated, or quasi-integrated approach to exploiting enabling technologies faces at least two major challenges: designing a profitable business model and limited access to capital.
Business Model Design: When the innovator adopts a product-based approach, the challenge is to design a powerful business model that links the company’s value proposition to a favorable structure of revenues and costs. A key decision is the make/buy/ally choice of how to access complementary goods and services. Complements that can be readily accessed on the open market are generally not worth internalizing. At the other extreme, a complement that is in short supply must be integrated (or at least made more plentiful) to avoid seeing it drain away profits as demand rises faster than supply. The design of a differentiated and profitable business model involves numerous other choices, which I will not go into here but have addressed elsewhere.
Capital Limitations: Another problem is that internally exploiting anything close to the full range of applications to which an enabling technology is suited would demand a level of resources that few firms can command. The innovator’s bargaining position with potential suppliers and partners is therefore undermined because it can’t make a credible commitment to exclusively develop and commercialize the technology and practice the patent(s) itself. The broader the applicability of an enabling technology across industries, the less complete the rent appropriation by the upstream innovator is likely to be.
An innovator can to some extent improve its position by starting with just one major application while it develops others for future product introductions. In fact, enabling technologies develop faster when there’s one large, demanding, and income-generating application sector. This was the case for semiconductors, where the U.S. Defense Department was initially the largest customer for integrated circuits. Similarly, new generations of cellular infrastructure technology receive adequate investment because of their assured worldwide market among telecom operators. But in other cases, like lasers, there was a broad array of potential applications, each with its own requirements, leading to a gradual introduction of laser-based components in telecommunications, electronics, and other applications.
Case Study: Pilkington plc
An example that reflects many of these issues is Alastair Pilkington’s invention and development of the float process for manufacturing plate glass. The float process, first put into production in 1959, reduced manufacturing cost compared with the then-dominant plate glass process by more than a third and continued to improve over time. It also facilitated downstream innovations such as flat panel displays. His company, Pilkington Brothers Ltd. (and, later, Pilkington plc) pursued a hybrid strategy of licensing plus building factories in Britain and selected overseas markets. This business model was viable because high transportation costs mean that plate glass is not sold across international borders as easily as many other goods.
Despite Pilkington’s strong intellectual property position (anchored by both patents and trade secrets) and the evident value of its technology, Pilkington licensed the float process on an exclusive territorial basis to all comers at 6% of revenues. This reflects the fact that the family-owned Pilkington was unprepared (and probably unable) to find the capital for the large-scale investment that would have been required to implement the technology by itself in all markets worldwide; so widespread licensing seemed the best alternative. As a result, a modest royalty for Pilkington effectively passed most of the benefits to manufacturers and consumers, leaving Pilkington only a small share (about 5%) of the total value created (see Appendix B of Managing IntellectualCapital for details). This strategy helped it survive UK antitrust scrutiny in the 1960s. But in 1994, after its main patents had expired, US antitrust authorities won a judgment against Pilkington for attempting to impose conditions of secrecy on its US licensees.
The Pilkington example demonstrates elements of most of the challenges described earlier. No matter how strong the IP protection, the owner of an enabling technology is unlikely to be able to avoid considerable spillovers of value. While spillovers are what make enabling technologies socially desirable, a balance needs to be struck between societal benefits and private returns to the inventors(s). The issue then is whether the realized return is adequate to encourage not only further development of the technology itself, but also development of entirely new enabling technologies by other innovators.
Addressing the Under-Investment Gap
Given the many issues faced by those trying to commercialize enabling technologies through licensing or other means, society is almost certainly not seeing as much investment in their invention and development as would be ideal. To some extent, the underinvestment gap can be addressed by government support for the necessary research. In most cases, however, government R&D spending is inadequate to support longer term, blue-sky research and to move technology beyond the lab. Business, with or without government support, is in the best position to perform the applied and generic research necessary to move new technology out of the lab to a state in which it is ready to be productized.
The other means of closing the gap is for the innovating firms to be able to maximize their profits with astute business models. Qualcomm, for example, has maintained a modestly successful two-track model, profiting both from direct licensing and from embedding its technology in the chips that it sells.
Policymakers and the judiciary can help by recognizing the appropriability challenges faced by developers and owners of enabling technologies and by recognizing the need to support reasonable royalties that come closer to reflecting the full social value of the technology. The Georgia-Pacific factors do not yet do so as well as they might. FRAND criteria are flexible enough to do so for standards-essential patents, and this ought to be the direction of travel when courts and regulators are involved. Antitrust theories such as patent hold-up still lack supporting evidence; and even if evidence of hold-up does appear, then this broader social-value criterion should still come into play.
Image Source: Deposit Photos.