Research

We study the design of online platforms that aggregate information and facilitate transactions. Leading players in the industry (e.g. the Booking Group) hold two types of platforms in their portfolio: revealing platforms that disclose the identity of transaction partners (like Booking.com) and anonymous platforms that do not (like Hotwire.com). Anonymous platforms offer discounts but lead to inefficient matching between consumers and firms. We develop a model in which horizontally differentiated firms sell to heterogeneous consumers both directly and via a platform that enlarges the pool of consumers they can attract. The platform charges firms for transactions it intermediates and can choose to offer an anonymous sales channel in addition to a revealing one. We show that offering both sales channels is profitable not only because it allows the platform to implement price discrimination, as suggested by the literature on opaque selling, but also because it improves rent extraction. The anonymous channel breaks the link between the price on the revealing channel and the firms' outside option; moreover, it can reduce double marginalization. The welfare impact of the anonymous channel is ambiguous: while it sometimes leads to market expansion, it also causes inefficiently high transport costs.

There is widespread concern about the lack of transparency regarding standard essential patents (SEPs). This paper examines the proposal to introduce essentiality checks, a certification scheme for declared SEPs. We develop a framework that allows us to evaluate how essentiality checks would impact licensing, litigation, and incentives to innovate. In our model, an upstream innovator invests in R&D and privately learns about the likely essentiality of its patents for a standard. The innovator then licenses the patents to a downstream implementer who can contest the essentiality of the patent in court. We identify a tradeoff whereby essentiality checks can reduce litigation but also provide excessive incentives for R&D investment. Their overall welfare effect depends on the level of the "fair, reasonable, and non-discriminatory" (FRAND) royalty rate.

This paper studies how to reward research when a key issue is separating obvious and nonobvious ideas. Because obvious ideas might still value rewards like publications or patents, screening with fees or delay is not possible. Instead the planner must use costly scrutiny. We show that when scrutiny is imperfect, there is a tension between screening for obviousness and the full development of ideas: nonobvious ideas are less developed than they would be in the full-information optimal allocation. This is true even if development does not impact the signal about obviousness. The constrained efficient allocations have features that look like the publication process, including tiers of journals and a top tier that is very exclusive. The model is further applied to registered reports: under what conditions should scrutiny be applied before full development of the research?

Self-imposed price caps have recently been proposed as a solution to competition problems arising, in particular, in the context of mergers, patent pools, or standardization. A common feature of many of the relevant settings is that they involve vertical relationships. Using a standard vertical-contracting framework, we show that price caps can be used by upstream firms to solve the opportunism problem of Hart and Tirole (1990) and thus restrict the quantity supplied to the downstream market. Our results imply that price caps can be anticompetitive in vertically related industries.

Scholars and policy makers have tried for years to solve the tenacious and harmful crisis of low quality, erroneously granted patents. Far from resolving the problem, these determined efforts have resulted in hundreds of conflicting policy proposals, failed Congressional bills, and no way to evaluate the policies’ value or impact or to decide between the overwhelming multiplicity of policies.

This Article provides not only new solutions, but a new approach for designing and assessing policies both in patent law and legal systems more generally. We introduce a formal economic model of the patent system that differs from existing scholarship because it permits us to (1) determine how a policy change to one part of the patent system affects the system as a whole; and (2) quantify the impact of policy changes. Existing scholarship typically analyses a policy by assessing its effect on just the targeted element of the patent system, but legal systems are complex with interrelated components and players react along multiple margins, so these analyses are incomplete and sometimes incorrect. Our approach fixes this problem, providing a comprehensive understanding of how a policy change affects the patent system from beginning-to-end. It also permits us to conduct complex analyses such as varying multiple policies at once. Further, much existing scholarship fails to quantify the magnitude of a policy’s effect, and even empirical scholarship can only measure the effect of an already-implemented policy, not predict the effect of a proposed change. Quantification is critical because policies generally have multiple effects, often in countervailing directions. Quantification—as shown using our model—permits scholars to determine the overall direction and size of a theoretically ambiguous effect. Quantification also allows us to compare the social welfare effects of different reforms so that policy-makers know where to focus their efforts.

We apply our model to several of the most prominent policy debates in patent law. We conclude that certain reforms such as regulation of settlement licenses and increased examination intensity yield large gains in social welfare and should be prioritized. Other reforms that are popular with scholars, including decreasing the availability of injunctions and reducing litigation costs produce surprisingly small gains in social welfare. Often existing scholarship operates too much on intuition, which, we show, can be wrong. Our new approach to patent reform provides an approach that offers deeper understanding and a more effective evaluation framework.

Many observers have voiced concerns that standards create essentiality and thus monopoly power for the holders of standard essential patents (SEPs). To address these concerns, the academic literature advocates structured price commitments, whereby SEP holders commit to the maximum royalty they would charge were their technology included in the standard. We consider a setting in which a technology implementer holds private information about profitability. In this setting, price commitments increase efficiency not only by curbing SEP holders' market power, but also by alleviating distortions in the design of the royalty scheme. In the absence of price commitments, the SEP holder distorts the implementer's output downward to reduce information rents. Price commitments reduce this distortion. We derive conditions under which price commitments can be implemented using a simple royalty cap as used in practice.

Critics claim that patent screening is ineffective, granting low-quality patents that impose unnecessary social costs. We develop an integrated framework, involving patent office examination, fees, and endogenous validity challenges in the courts, to study patent screening both theoretically and quantitatively. In our model, some inventions require the patent incentive while others do not, and asymmetric information creates a need for screening. We show that the endogeneity of challenges implies that courts, even if perfect, cannot solve the screening problem. Simulations of the model, calibrated on U.S. data, indicate that screening is highly imperfect, with almost half of all patents issued on inventions that do not require the patent incentive. While we find that the current patent system generates positive social value, intensifying examination would yield large welfare gains. The social value of the patent system would also be larger if complemented by antitrust limits on licensing.