This week’s Law and Economics Workshop featured Professor Jonathan Barnett and his paper, entitled, “Intellectual Property as a Law of Organization,” available at http://www.law.uchicago.edu/files/files/barnettdraftfeb1.pdf.
What It’s About
If innovation happens even in industries with weak IP rights, what is the point of having IP rights? In the prior literature, some academics had raised concerns that having a costly system of patent rights could be redundant and unnecessary when private measures, such as contracts, barriers to reverse-engineering, and limiting the amount of outsourcing, could protect a firm’s intellectual capital just as well. Professor Barnett’s paper counters this viewpoint by arguing that patent rights can incentivize innovation indirectly in other ways, by changing the way firms bring innovations to market (e.g., patent rights can give businesses more flexibility in deciding how to roll out their products, from testing to structuring the supply chain).
No matter how innovative an idea is, there’s got to be some way to bring the idea to market in order for the public to capture some benefit. According to Professor Barnett, the problem with a world without patent rights is not so much the lack of innovation, but rather the way that firms will react to the possibility of expropriation. In this world, firms will still come up with new ideas, but they will be more reluctant to partner too closely with other firms because it will be harder to prevent their partners from stealing their ideas without IP rights. For example, if an idea-generating firm wanted to make a contract with an idea-buying firm, it would have to disclose the idea before the firms can come to an agreement. Once the seller discloses the idea, though, the buyer already has the idea; the buyer has no reason to agree to give something back to the seller, if the seller didn’t hold something back (Arrow’s Paradox).
Professor Barnett predicts that idea-generating firms will react to this paradox by integrating down the supply chain. Since the firm can’t rely on outside help without endangering the product itself, each step of the commercialization process that requires knowledge of the idea would have to be done in-house. Professor Barnett sees this as a potentially inefficient outcome because some firms that would have been better at specializing in just making ideas would no longer be able to outsource later steps in product rollout, such as testing, manufacturing, marketing, and distribution. The inability to outsource would also raise barriers to entry in an industry, because firms couldn’t just specialize in one part of the product development process—firms would have to enter ready to invest in multiple stages of the product rollout, if they are to enter at all.
What Was Discussed
Workshop participants honed in right away on the possibility of using mechanisms other than patents or supply chain integration to protect intellectual capital. If firms wanted to specialize as “idea-generating firms” rather than integrating, perhaps they could try to protect their ideas using covenants not to compete, trade secret protections, or more precise contracting. However, as Professor Barnett noted, these are tools better suited for guarding against internal threats of expropriation by employees, rather than external expropriation by competitors. (Intuitively, that seemed like a persuasive distinction to me. Employees are easier for firms to control, since they may have long-term loyalties and human capital tied up with the firm and may be less incentivized to be opportunistic with company secrets. Covenants not to compete between firms, on the other hand, might not even be legal, let alone effective.)
The discussion on alternatives to integration led one participant to suggest that, to get around Arrow’s Paradox, contracts could mimic patent priority rules: Suppose firms are concerned about disclosing their ideas because they fear the other party will claim that the idea is unoriginal, even if they actually stole it. The patent system seems to be proof that these factual controversies could be sorted out with evidentiary hearings and detailed investigations into who came up with the idea first. Why not create contracts that agree to arbitration if a dispute arises about whether the buyer knew about the idea before the seller disclosed it? Professor Barnett responded that third-party “expert agencies” would be one way to protect intellectual capital in the absence of integration or patent rights, but noted that third-party experts could still pose an expropriation threat to idea sellers. He then explained that the patent system, by contrast, is shielded from third-party opportunism because government agencies are not profit-seeking entities and would have no interest or capacity to expropriate.
From there, the workshop participants moved toward discussing the evidence supporting Professor Barnett’s theory that weak IP rights leads to greater vertical integration. A few participants noted that a firm’s decision to integrate may depend on the relative importance of human capital to that particular industry, rather than the strength of patent rights in the industry. For example, the success of a law firm depends heavily on human intellectual capital, and the institutional knowhow of an experienced firm is hard to transfer or steal, by its very nature. Even though none of the intellectual output of a law firm is protected by patents and the legal theories themselves cannot be protected by copyright, a law firm would feel no need to be particularly protective about its nonprivileged information or to “integrate” by doing all of its legal work in-house. And indeed, law firms outsource doc review, refer clients’ matters to other firms, and share litigation strategy with corporate counsel, without any real fear of forfeiting their core intellectual capital.
Professor Barnett conceded that certain industries seem to be able to self-regulate without integration or patent rights, and noted that the case studies in his paper (e.g., machine tools, petrochemical, biotechnology, semiconductor industries) show correlation rather than causation. It is not clear whether disintegration happens because strong IP rights in that area of technology make disintegration possible, or because of other peculiarities specific to the nature of the industry. For example, one participant pointed out that the auto industry is a counterexample to Barnett’s theory, where there is weak IP protection, yet largely disintegrated supply chains. Another participant noted that the trend toward disintegration might not be related to a contemporaneous trend in stronger IP protections after all, but instead might be correlated with the increased availability of financing from the rise of venture capitalism and private equity that happened at about the same time as stronger IP protection. Professor Barnett noted in response, however, that the two developments are related. He explained that the existence of external capital markets for financing R&D relies on the ability of smaller entrants to use patent rights to enter concentrated markets. The fact that the rise of venture capitalism coincides with the strengthening of patent rights does not have to mean that one has to settle for one explanation and reject the other.
Another line of criticism focused on the question of whether a system that increases innovation is even desirable and, if so, what kinds of innovation to favor. As one participant framed the issue, the ideal level of IP should be based on the goal of reaching an efficient level of innovation, rather than just looking to maximize innovation. There is a cost to protecting the fruits of innovation, just like anything else. Professor Barnett took up the idea and elaborated on it, suggesting that strong IP rights could actually be imposing a hidden tax on certain industries that would work well without IP protections (e.g., the financial services industry, which, as Professor Barnett noted, relies on reputational capital and other private instruments). He observed further that even private mechanisms, such as using reputation to prevent opportunistic exploitation of others’ innovations, can be societally very expensive to maintain. He linked this point back to his paper’s position that integration, as one form of private protection, also could come at significant social cost by raising barriers to entry for firms.
If there is a social cost to protection, there could be such a thing as too much protection. There would need to be some principled way of determining at what point firms are likely to “overintegrate” so that IP rights can be strengthened there, while at the same time cabining the principle so that the government does not reward IP rights every time IP rights could have an effect on disintegration, as one participant feared. For example, it seems unlikely that the government would strengthen IP rights in areas where it is hard to establish definitions that distinguish between innovations (e.g., business method patents, perhaps). Awarding IP rights in such areas could allow firms to specialize even further, but would we necessarily even want them to? (Does it make sense to have a consulting firm that specializes in patenting and selling business processes that they have devised divorced of any real implementation? Shouldn’t we be especially wary of patent trolls in certain industries where innovations are hard to define and happening at a rapid pace?) The point here seemed to be that not all organizational changes are worth promoting.
Professor Barnett seemed to acknowledge this point, although he noted that nothing in a regime with strong IP rights would force businesses to disintegrate if it were not in their interest. As his reasoning goes, at least in a world with strong IP rights, a firm has the choice to integrate or specialize based on actual considerations of what is most efficient for the firm, rather than being forced to integrate to protect their intellectual investments (or not make intellectual investments at all, if the cost of integration is too great), as in the world with weak IP rights.
Professor Barnett explained at the workshop that he hadn’t reached a stage in his research where he had compared the effectiveness of patent IP rights with the effectiveness of alternate mechanisms for protecting intellectual capital. But that seems like the next natural step to take for the theory to progress. If firms who really want to specialize have viable alternatives to integration, then it may become more plausible to think that firms that integrate in the face of weak IP rights do so because they actually find it efficient, not because they are forced to do so due to a lack of other options.
The discussion on third-party mediators seemed especially interesting as a possible alternative to integration. If Arrow’s Paradox is one main obstacle to specialization, perhaps it would be possible for third-party agencies to exist that bind themselves especially heavily in contract not to have any interactions with the industry outside of resolving expropriation disputes (to reduce the risk that the third parties will also expropriate). As one participant noted, there is nothing especially impossible about resolving the factual question of who came up with an idea first. The very fact that this sort of arbitration option exists as a last resort could reduce opportunistic behavior, and companies will adjust by keeping good documentation of their ideas, so that such factual disputes become easier to resolve in the future. Another possibility would be to employ some sort of “idea escrow” during negotiations. Would it be possible for parties to signal what they already have and what they are looking for to a third party and have this third party sort out whether there is a match? This would assume that there are strong constraints on the third party to make sure that the third party doesn’t steal the idea either, but that doesn’t seem so unrealistic. Transactional lawyers, for example, have strong professional constraints, conflicts of interest limitations, and regulations (e.g., rules against insider trading) that prevent them from opportunistically disclosing what they may discover at the negotiating table. Could there be some similar non-industry-participant third party who chooses to forego certain market opportunities in exchange for being ideally situated to work as an intermediary?
Because we can’t tell what a modern market would look like without IP rights, it is completely speculative whether firms would actually organize this way instead of integrating. This ordering system, like integration or the use of reputational effects, imposes social costs, and it may turn out that integrating is simply the least expensive way to protect intellectual resources. One way to try to get empirical evidence could be to look at where different stages of product rollout happen. For example, do firms choose to do all of their product development in countries with strong IP rights, while outsourcing less risky stages of rollout, like manufacturing, to places like China? In some industries, ideas may be vulnerable to expropriation for more stages in the rollout process than in other industries. Do those industries tend to partner with firms in different countries than industries where the nature of the product makes expropriation inherently difficult?
On an ending note, it may be good to revisit the concern of one participant, who expressed concern that there seems to be no way to cabin the move toward stronger IP rights if the government starts recognizing rights whenever it could give firms more choices on how to organize. There needs to be some comparison between the costs of various alternatives—whether patents, integration, contracts—so that there could be some principled way of evaluating which alternative gives the best balance between incentivizing innovation and maintaining the social costs of such a system. In his paper, Professor Barnett already acknowledges the need to weigh the benefits of the patent system against the social costs of maintaining the system. His point, for now, is that current attempts to strike a balance ignore a whole different category of benefits (i.e., the benefit of structural flexibility) that the patent system provides. The paper sparked a great deal of interest at the workshop; as the paper develops further, perhaps a future version will give further insight into how the patent system measures up to its alternatives.