Four Theories of the Innovation Commons
Four Theories of the Innovation Commons
Abstract and Keywords
This chapter proposes four theories explaining how innovation commons work, in terms of how they pool information, and what specific problems they solve in order to discover entrepreneurial opportunities. The first is the “two commons” theory in which the innovation commons is a screening mechanism by having the truly valuable commons of entrepreneurial information accessed only through the commons of technological knowledge and material innovation resources. The second is the “evolution of cooperation” theory, which draws on modern evolutionary theory (specifically multilevel selection theory and evolutionary game theory). The third is the “defense against enclosure” theory, in which the commons is a preferred institution for first movers because it raises the cost of alternative institutions and minimizes the risk of loss of control of the technology. The fourth is the “institutional uncertainty and real options” theory.
The innovation problem can be diagnosed through four different levels of knowledge problem. These were associated in chapter 3 with Deirdre McCloskey (social contract problem), Friedrich Hayek (distributed knowledge problem), Oliver Williamson (opportunism problem), and Elinor Ostrom (collective action problem). In this chapter, I want to look for solutions in the space of efficient institutions to solve the innovation problem as a knowledge problem.
My argument should by now be clear: the foundational innovation problem is not market failure but rather opportunity discovery, and the shape of this problem is largely about pooling and coordinating distributed information, knowledge, and other resources, all from behind a veil of uncertainty. Several institutional and organizational forms can be used for this, including small or large firms, relational contracts, networks, markets, clubs, commons, or government. There is diversity both among and within these different institutional forms (Hall and Soskice 2000; Williamson 2000; Ostrom 2005; Hodgson 2015). For any particular economic problem, a variety of institutional forms can solve it. The most efficient—in the sense of minimizing transaction costs—institutional form for solving the innovation problem is, generally, in the commons, namely, the “innovation commons,” referring to the rules to enable an emergent community to create and govern a common-pool innovation resource.
This particular institutional form—the innovation commons—while it has definite costs and problems associated with it, in particular that it emerges from civil society and relies mostly on informal norms and rules, will nevertheless under certain circumstances be distinctly superior to alternatives, including firms, markets, and governments, and especially so at the very beginning of the innovation process.
This is not a recent development. It has been true since the origin of market capitalism (McCloskey 2016; Mokyr 2016). Yet it has been overlooked in the innovation literature, with its strong propensity to see innovation originating exclusively from firms, and where that fails, from government policy. Yet that (p.74) issue is for later chapters. Our immediate goal is to unpack the reasons why, and the conditions under which, the commons can be an effective and efficient mechanism to solve the innovation problem. To that end, I’ve assembled four distinct theories.
The first is the “two commons” theory. The reason the commons is effective and efficient is because it works as an efficacious screening mechanism by having the truly valuable commons of entrepreneurial information accessed only through the commons of technological knowledge and material innovation resources.
The second is the “evolution of cooperation” theory. Drawing on modern evolutionary theory (specifically multilevel selection theory, evolutionary game theory, evolutionary psychology, and sociobiology), I argue that cooperative or sharing behavior is actually far more likely than the selfish rational actor model supposes. In this theory, the commons is simply an easy, low-cost solution.
The third is the “defense against enclosure” theory. In this account, the commons is a preferred institution for first movers because it raises the cost of alternative institutions and therefore minimizes the risk of loss of control of the technology to private agents or to government control. The value of this defense is that when it works well, the technology is protected from enclosure and remains open for development along all possible frontiers.
The fourth is the “institutional uncertainty and real options” theory. Early in the life cycle of a new technology there will be uncertainty about what the best institutional form to develop a technology will take, and the particular way in which property rights will develop. The innovation commons is a good way to keep options open while learning about possible uses and the property rights that will work best in those domains of application. This builds on Dixit and Pindyck’s (1990) real options approach to investment under uncertainty, and Luppi and Pasari’s (2011) asymmetric Coase theorem. A further aspect relates to reducing uncertainty of the new property rights associated with new technology.
Consider these distinct but overlapping theories of the mechanisms and evolutionary logic behind how and why an innovation commons works.
4.1 Two Commons
The first key to understanding the nature of an innovation commons is that it is not one commons, but two. And the first commons, the commons of material and technological innovation inputs and resources, is a screening mechanism for the second and more valuable commons, the commons of entrepreneurial information. The price of entrance to the second, valuable commons is cooperation in the first, material commons.
(p.75) If this theory is true, it predicts that on the surface an innovation commons will appear to be that of pooled physical and technical resources, including kit and knowledge of the sort that engineers value. A hackerspace is a prime example, in which a defined group of people pool their physical equipment (3D printers, say) or resource equivalents (club fees) along with their embodied skills and expertise for sharing within a well-defined community. Think of the technically useful resources embodied in physical and human capital as one class of resources in the commons, and with an obvious correlate with the grazing pastures, trees, water, or fish in a natural resource commons.
Economists who study innovation tend to focus on these resources. They include basic science and the translation through engineering into prototypes of technologies. These innovation resources include everything under the heading of research and most things under the heading of development. However, my claim is that these are not the most valuable resources in an innovation commons. Rather, there is another resource that also accumulates in the commons, namely Hayek’s “knowledge of time and place” about, for instance,
• how the technology works in particular circumstances
• how particular consumers use the new technology in specific instances
• price points that matter, and those that don’t
• ways to discriminate in the market
• complementary and competitive investments being made
• sources of technical expertise and talent
• qualities of different (global) locations for production and distribution from a regulatory, political, and cultural perspective
• sourcing of critical resources, and the risks and costs involved
• the prospect of potential competitive or complementary investment, and by whom
• problems that arise when attempting to scale up, and the value of particular incubators or accelerators
• specific sources of expertise in particular aspects of production and development
• regulatory and legislative risks and barriers
• possible markets that might exist and how
• the fit of different business models (and platforms) and the consequences of each
• sources of venture finance, and who might be a possible investor, and why
• how property rights work over the new technology, and so on
None of this information can be patented or protected, and all of it is context specific. Every bit of it is acquired by experience, and much of it has been accumulated without intention. Its value decays fast if not used, and information (p.76) is usually only of value when combined with other related data. Yet this is the information that entrepreneurs need.
An innovation commons is of value to an engineer, as someone seeking to put together or to access technical information about how to do something. This is usually assumed to be the prime value from the perspective of peer production of new technology (Williams and Hall 2015). But an innovation commons also has value to an entrepreneur, as someone seeking to organize resources to coordinate economic activity. What the entrepreneur needs is broader than what the engineer needs. An innovation commons is thus two commons. It is a pool of resources and information that engineers need, or those who are building the technical prototypes, but which can also be effectively assembled in alternative institutional forms. Yet this will often function as a gateway for the real reason that people are in this commons, namely to gain access to the information that entrepreneurs need, which is all of the bits of information about the costs and prospects of the idea or technology that when taken together provide sufficient information for the entrepreneur to act upon.
An innovation commons consists of two commons—one of technical resources, and one of market information. In the two-commons model the most valuable resource is the latter, but the cost of access to it is contribution to the former. An innovation commons is thus a kind of club, where the price of admission into the community is to contribute technical resources, which is in effect a screening mechanism: if you cooperate in the first you get access to the second. But once in this club, entrepreneurially relevant information circulates. That is the reason that people are there, in the innovation commons, namely to gain access to entrepreneurially relevant information.
4.2 Evolution of Cooperation
The basic argument for the efficacy of institutional solutions built with markets, hierarchies, and governments is that they economize on the need for cooperation. A market system works when each person pursues her individual self-interest and does not rely on her behaving altruistically. The mutual benefit to others, as Bernard Mandeville explained in 1714, is an emergent order—a catallaxy, as Hayek (1960) subsequently called it. Firms and governments also do not rely on altruistic cooperation, but respectively on the coercive power of the threat of contract breach, or as Thomas Hobbes put it, on the “monopoly on violence.” Obviously, things go more smoothly when agents cooperate under the shadow of these implicit threats, resulting in the “as if” cooperation better known as the social contract.
The problem is that cooperation and trust is expensive and vulnerable.1 It is not usually an evolutionarily stable strategy (Maynard Smith 1972; Axelrod (p.77) 1984). It is a well-established finding in evolutionary game theory that when an altruistic cooperator is matched with an agent who plays defect, the cooperator will lose. Evolution is in this sense “selfish,” as Dawkins (1976) famously explained. However, modern evolutionary theory has arrived at a new understanding of the evolution of cooperation in the framework of multilevel selection theory in which competition and differential selection operate not only between individuals, but also between groups (Nowak 2011). Lower-order selection operates within groups and predicts that selfish individuals will outcompete altruists. Higher-order selection operates between groups and predicts that groups of altruists (or agents who have found a way to mutually cooperate) will outcompete groups of selfish agents. We will elaborate this theory more fully in chapter 6. As technology improves, social coordination also becomes easier, and we can proceed more directly to efficient cooperative institutions. As the costs of cooperation (and coordination) fall, we can embed more of our institutions into cooperative frames. By this mechanism, innovation commons emerge and strengthen as we cooperate socially to discover individual opportunities.
4.3 Defense against Enclosure
A third explanation for the existence of the innovation commons emphasizes the role of endeavors, whether self-interested or public spirited, to ensure that a new idea of technology is not controlled or monopolized by any one group. This pursuit is driven to prevent monopoly control, regardless of the source of that control, whether from the private sector side, such as a firm controlling a technology by copyright and refusing license it or extracting significant rents from that monopoly position (e.g., pharmaceuticals), or from the public side in rendering a technology illegal or controlling it for limited purposes that meet the government’s interests with regulation or enforcement (e.g., cryptography, maps, or weapons).
The demand for the commons comes from the demand to constrain the ability of private or public actors working through private or public institutions to control a technology and its course of development. Placing an idea in the commons is a highly effective defensive gambit. The institution of creative commons licensing, for example, has been developed to achieve this.2 The innovation commons is a highly effective mechanism to ensure that an idea cannot be exclusively controlled by other private agents.3 Even the threat of the prospect of placing an idea in the commons, that is, simply the existence of the commons, can condition strategic behavior in mutual games of development of an idea toward socially beneficial equilibria.
A mutual public benefit of development in the commons is the more likely discovery of unintended consequences. The longer an idea is in the commons, (p.78) the better it gets “shaken out” and all of its potential problems and applications discovered. This has a parallel in code, and in part explains the success of open-source projects, which was expressed by Torvalds as the hope that “with enough eyeballs, all bugs are shallow.” The innovation commons can under many often-met conditions be an efficient debugging institution because it can concentrate diverse perspectives on an idea.
4.4 Institutional Uncertainty
A fourth theory to explain the existence of the innovation commons turns on the problem of institutional uncertainty with respect to the best fit between a new idea and an institution within which to develop that idea, coupled to the notion that once placed in a particular institution (e.g., private or public domain), this tends to be an irreversible investment decision (à la Dixit and Pindyck 1994). There can be substantial benefits (i.e., consumer and producer surplus) to the development of a technology or industry to a good match between idea and institution, and substantial costs (i.e., opportunity costs) to a poor match. The key insight here is that the innovation commons is a kind of “real option” for institutional choice, a way of deferring institutional choice while more information is gathered, so as to minimize the likelihood of a poor irreversible choice.
The basic model here is that there is an allocation problem of matching each new idea to a particular institution. Assume that each new idea or technology has inherent qualities that give it an institutional “true type,” which is the institutional context or environment in which the idea or technology will best flourish. However, this true type is unobservable in the beginning and only revealed through experience.
New ideas can develop the idea in public institutions, or in private institutions. Assume that different types of ideas (say, up-type and down-type) perform better under different institutional settings (private and public), and that ideas exhibit path dependency in institutions such that there is institutional stickiness and that institutional switching is costly, and in the limit impossible. With fundamental uncertainty about idea type (up or down), and therefore about optimal allocation over institutions (public or private), the innovation commons is a gateway institution. Ideas of uncertain type are first placed in the commons to induce information pooling that reduces uncertainty, eventually revealing an idea’s true type (up or down). This facilitates efficient matching of new ideas to institutions.
Innovation commons are effective institutional mechanisms for pooling information to reduce uncertainty about an idea’s true type. Innovation commons will be likely therefore to be temporary institutions. Their usefulness will eventually fade by the logic of their success in reducing entrepreneurial uncertainty. (p.79) We should expect to observe innovation commons at the nascent stages of any new idea, technology, or industry as an essential but transient component of an innovation system. An innovation commons is an emergent, temporary institution that forms in the early phases of a new idea when uncertainty is highest with respect to the institutional form that the innovation trajectory will best develop along. An innovation commons is thus a mobile institution in a national innovation system, forming at the frontiers of new technologies with high entrepreneurial and institutional uncertainty and decomposing when that uncertainty resolves. It will collapse into a public, private, or mixed institutional form once it has done its job in reducing the fundamental uncertainty about the nature of the idea and the economic opportunity it represents.
The specific form of this private uncertainty could emerge in several dimensions. It could be that the potential size of a private market for the idea is completely unknown and ex ante unknowable (e.g., genetic cloning), or development costs are unknown and ex ante unknowable (e.g., commercial space tourism), or it could be that the externalities are unknown, including whether they are positive or negative (e.g., nuclear fusion technologies). The uncertainty may be due to problems of unknown competitive or complementary investment (Richardson 1972), including knowledge of what types and scale of complementary investments might be required, and of who might undertake them, and when. The upshot is that in such circumstances, which are presumed to be common, it can be exceedingly difficult to know where to place an idea for development.
There are substantial risks and costs associated with misallocating a high-type idea in public institutions. The most obvious cost is the forgone private income that could have been associated with either ownership of the intellectual property or from owning a private company that successfully exploited the idea. The Google search engine was developed at a private university (Stanford) and quickly moved into a private company (Google, and subsequently Alphabet), currently valued at over US$400 billion. On the other hand, Tim Berners-Lee developed HTML at a public research institute (CERN) that released it as a public good in the form of the World Wide Web. Berners-Lee has received many awards and much acclaim, including a cash prize $1.3 million for inventing the web, but he is not rich (Wright 2001) compared to the economic value the web created (Goolsbee and Klenow 2006). Jonas Salk’s research and development of the polio vaccine was publicly funded; meaning that while he did receive fame, he could never have achieved the great wealth such a patented vaccine could have accrued.
And there are substantial risks and costs associated with wrongly placing a low-type idea in private institutions. One is the potential waste and duplication that comes from multiple replicated private investments where firms are not able to signal to others that the particular line of investigation is a dead end (p.80) (Akcigit and Lui 2016). Another is the social cost due to ransom and holdup problems. Bessen et al. (2011) estimate these costs on the order of hundreds of billions in relation to “patent trolls.”
More generally the risk is that the idea remains underdeveloped due to insufficient investment, inability to access complementary resources and capabilities, weakness in lobbying for regulatory exemptions or special treatment, or overly narrow application, perhaps due to exclusive licensing. For the agent developing the idea, these costs manifest in low private earnings or losses, but also in the opportunity costs of failing to access the reputational and associated benefits from developing the idea in public institutions. Examples can be seen in experimental alternative energy technologies that are flourishing under public innovation institutions (in the sense of providing income for proponents of the idea), but widely failing under private innovation institutions (examples are wave energy and wind energy and fusion reactors). In contrast, technologies for hydraulic fracturing of shale gas formations technologies were successfully developed under private innovation institutions (by Halliburton, a private US oil services company) and would likely have failed under public institutions. Recently, Lockheed-Martin (a private US defense contractor) has announced a design breakthrough for a portable nuclear fusion reactor.
An implication of this framework is that innovation institutions are sticky, or at least partially irreversible (or equivalently, that the development of an idea exhibits institutional path dependency, or that there is a hysteresis effect), such that an idea that starts under public innovation institutions can never completely transfer to private innovation institutions, or vice versa, without loss. The strong form of this claim is that ideas that begin in private institutions stay in private institutions, and those that begin in public institutions also stay in public institutions. This stickiness may be an artifact of bounded rationality and organizational problem solving (von Hippel 1994). Strong-form stickiness is likely to be observed when dominant patents are privately held. (Patent buyouts can mitigate this [Kremer 1998].) It is also likely to be observed where a group of firms collectively holds information essential for subsequent development. (Patent pools can mitigate this for these insiders [Lerner and Tirole 2004]). Stickiness in public institutions will often be due to constraints in developing new business models that do not depend upon control of intellectual property, or due to direct crowding out. Stickiness may further be due to development of specialized (research) capital such as laboratories or complementary skills (Williamson 1975; Teece 1986a) that are costly to transfer between institutional regimes.
The resolution of institutional uncertainty in the commons is a partial consequence of the Hayek-Williamson innovation problem discussed in the previous chapter. To the extent that capital has a structure, and that some capital investments are complementary (Hayek  2007; Lachmann 1956; (p.81) Richardson 1972),4 there will also be a corresponding innovation structure, so there will also arise a need for a coordinated institutional choice of each of the distributed investments in an innovation. This is a focal point coordination problem (Sugden 1995), in that the payoff is maximized when all parties choose the same institution. But it is also likely to be a path-dependent process for that same reason. Initially developing an idea in the innovation commons provides an institutional space to facilitate coordination on subsequent institutional choice.
Another way of thinking about institutional choice derives from the Coase theorem, which says that the initial allocation of property rights doesn’t matter so long as there are no transaction costs to inhibit bargaining. Luppi and Parisi (2011) point out that there is an asymmetry here, in that the transaction costs are higher for bundling and rebundling than for unbundling or fragmentation. This leads to an entropy prediction, namely that property tends to fragment. If we think of a firm as a particular or bundled structure of ideas and capital, then the asymmetric Coase theorem suggests that the emergence of an innovating firm is an unlikely event because of the transaction costs to be overcome in making the bundle.5 But the utility of an innovation commons is that it is an institutional solution (an allocation of fuzzy property rights) that minimizes transaction costs to maximize the probability of an innovating firm emerging from the commons.6
In this model, the source of the negentropy (or increased order) is from the clarification of the institutional rule through the elucidation of property rights over the idea. An idea becoming an innovation is monotonic with elucidation of institutional rules. A commons, then, is the institutional space in which the rule is still fuzzy, and so opportunities and property rights are uncertain. The logic of the innovation commons is that it is a low-transaction-cost environment to elucidate institutional rules, thus clarifying property rights, and in so doing thereby sharpening the information set necessary to reveal entrepreneurial opportunities.
Why do we expect the innovation commons to emerge at the origin of any new idea or technology? The key insight is that any fundamentally new idea or technology emerges without labeling, and therefore with uncertainty about the particular opportunities and value that it harbors. They must be discovered. And the commons can be an efficient institutional space for this discovery process to unfold by furnishing an efficient governance mechanism for pooling distributed information. The innovation commons is the missing quadrant in the theory of innovation institutions beyond private, public, and club-like institutions.
(p.82) An innovation commons is a site of peer production to reduce uncertainty to generate entrepreneurially relevant information that will yield both private and social benefit. The innovation commons exists for several distinct reasons: (1) because of uncertainty about the nature of new ideas and the entrepreneurial opportunity they represent, which requires solving a collective action problem to pool distributed information; (2) because cooperation expresses itself efficiently in a commons, such that successful groups can form that are highly technologically progressive; (3) because the commons is an efficient mechanism to distribute power in the development and adoption of an idea, making it difficult to control the path of development of a technology once it is in the commons, thus ensuring that power never attaches itself to a technology; and (4) because of uncertainty about the most effective innovation institution to develop the idea. The innovation commons can be temporary and mobile institutions that form at the origin of a new technology, and then collapse when the entrepreneurial opportunity is fully revealed, but that are nevertheless of permanent value because they shape the development of an idea and the pathways along which it develops.
(2.) Creative Commons licensing was founded by James Boyle, Laurence Lessig, and Hal Abelson in 2001.
(3.) Madison et al. (2010, 692) offer the following examples (italics added): “A commons is constructed as a defense against potential privatization of commonly useful resources. Examples of such arrangements might include constructed commons for basic biological building blocks such as the Single Nucleotide Polymorphism (SNP) consortium or the publicly available databases of genomic sequences that are part of the Human Genome Project. . . . Formal licenses and related agreements assure that participants in the commons become part of what amounts to a mutual nonaggression pact that is necessary precisely because of the possibility that intellectual resources may be propertized. So long as the resource is in the commons, it can be shared among commons members, and neither commons members nor outsiders are able to appropriate that resource, patent it, and then assert a patent claim against a commons member.”
(4.) In contrast to Keynesian and neoclassical models in which all capital goods are substitutes.
(5.) Thanks to Trent MacDonald for suggesting this model.