We investigate the effects of trade secret (TS) protection and relationship-specificity on supplier exclusivity in vertical industrial relations. We highlight how manufacturers restrict their suppliers' buyer network when facing weak TS protection to safeguard sensitive information while ensuring market access to encourage suppliers' investment. Using a unique dataset of automotive parts' contracts, we confirm this intuition by identifying a positive relation between robust TS protection and the number of Original Equipment Manufacturers (OEMs) a supplier serves, particularly for parts of intermediate relation-specificity, aligning with theoretical insights. Consistently, this association amplifies with greater OEMs' bargaining power. We validate these findings by examining supplier attributes, including size, patent portfolio, cartel involvement, and ownership structure.

In many jurisdictions, the existence and contents of patent applications are unknown to third parties until the application is published by the patent office at least 18 months after the initial filing. In contrast to other features of the patent system, this publication lag has received little analytical attention. The patent applicant can expedite public awareness of the existing application and the respective technology by announcing the patent application before its automatic publication. In our model, the applicant balances a negative effect of disclosure on its informational advantage in the short run (value of secrecy) with a positive long-run effect stemming from potential deterrence of a rival’s R&D (value of deterring innovation). We give conditions under which announcing the pending patent deters a rival’s innovation. We show that, in equilibrium, the applicant’s decision to announce and the rival’s decision to innovate are non-monotonic in the strength of the application and the strength of the patent. We present evidence supporting core predictions of our model by identifying press releases, one channel for disclosing business information, that announce nothing but the recent filing of patent applications. Using a technique suggested in the corporate finance literature, we estimate a measure of the nature of competition for all major NAICS codes. In doing so, we are the first to provide broad evidence supporting the prediction dating back to Gal-or (1986) that cost disclosure depends on the nature of competition in an industry.

This paper explores the impact of the COVID-19 pandemic on the interactions between firms and universities. Although it might be expected that restrictions during COVID-19 made interactions between business firms and universities more challenging, there has been little evidence on this relationship. Drawing on a large-scale survey carried out in the UK during the pandemic which focused on pre-pandemic business interactions with universities, we explore what factors shape whether firms’ interactions with universities were negatively affected by the crisis. We find that firms with broad interactions and with different university partners were negative impacted, whereas those firms with pre-existing formal collaborations and high absorptive capacity were less affected. We explore the implications of these findings for policies to support business-university interactions in the post-pandemic era. 

Large-scale survey project studying business-university interaction, funded by Research England. Findings published as "The Changing State of Business-University Interactions in the UK" by the National Centre for Universities and Business, 2022. Summary of the key findings available here. The anonymised dataset is available via the UK Data Archive.

It is well known that not every patentable invention will be patented. I demonstrate that there is also an intensive margin to the patenting decision: patent applicants can choose how much of an invention to include in any individual patent. I use the staggered enactment of the Uniform Trade Secrets Act across the United States to estimate the effect of strengthened protection of trade secrets on the number of patents and the number of independent claims in a patent. More effective secrecy protection impacts both variables negatively, but patents and claims do not necessarily react simultaneously. About a third of the total effect of the UTSA manifests itself in less disclosure per patent.

By computing the text-similarity of claims in a patent, I produce a measure of each claim's marginal disclosure. I then obtain evidence in support of a simple model of allocation of inventive components into patents. I suggest that the design of patents and the average marginal disclosure per patent claim affect how patentees will react to the weakened incentive to patent, i.e., whether they reduce patents or claims per patent. 

Finally, I illustrate how the endogeneity of patent claims can distort the measurement of innovation with patent data. I offer an explanation for the conflicting results of previous studies determining the effect of competition on innovation. Measuring innovation by the number of (sufficiently different) patent claims, I find an inverted-U-shaped relationship for US patentees, in line with the finding of Aghion et al. (2005) for UK patentees.

A recurring theme in the analysis of patent systems for at least the past 20 years is that their usefulness increases in the size of the patent holder. More recent discussion about the state of the US patent system has focussed on the perceived problem of meritless patent litigation by non-practising entities, or ``patent trolls''. Partly motivated by the lower extent of the problem in European countries, a widely suggested remedy for this latter problem is the shifting of legal fees to the losing party, which is expected to make defending more profitable for manufacturing firms who find themselves accused of patent infringement by a patent troll. This paper studies the complementary question how fee shifting changes the patenting incentive for smaller manufacturing firms by affecting their ability to enforce their patents. In a model incorporating endogenous court judgements, fee shifting, patent quality, and a variable capturing the importance of legal spending, fee-shifting does indeed decrease the profitability of patenting, and small firms are generally the first to forego patenting. The extent of fee-shifting necessary to affect the patenting rate however depends negatively on the importance of spending and on patent quality. To the extent that the United States are characterised by a higher importance of spending in litigation and lower patent quality compared to many European countries, a comparably small amount of fee shifting might be sufficient to reach the effect that requires a much higher amount of fee-shifting in Europe. Introducing the ability to settle litigation, I additionally show that very small patent owners may have an incentive to "be infringed" when their bargaining position is strong and that this incentive is relatively independent of the extent of fee shifting. 

It is commonly assumed that a patent race ends when the first firm applies for a patent. Even if each application was granted protection with certainty, this does not have to be true since patent offices around the world have almost universally adopted the policy of publishing patent applications with a delay of 18 months. Compared to some estimates of the average length of patent races this can be a sizeable amount of additional time in which each firm other than the winner potentially continues to spend R\&D resources that have an effective return of zero. The empirical academic literature has focused on the effect of publication of patent applications at all on the diffusion of knowledge. The theoretical literature on competition in R\&D has neglected this feature of the patent system. This paper contributes a first analysis of the effect of a statutory delay in patent publication in races for a single innovation, and then studies cases in which firms may have an incentive to announce their pending patents before they will be published by the patent office. Such incentives may exist when the patentee of a first innovation benefits from her competitor working on a second innovation, which can be the case when, e.g., the first innovation allows generation of licensing revenues (cumulative innovation) or there is some degree of complementarity between the two innovations. Finally, I show that if policy makers were to set the statutory delay to zero there can be an incentive to delay the patenting of some complementary innovations.