Working Papers
Screening Property Rights for Innovation (SSRN), with Mark Schankerman (Online Appendices)
Revise and Resubmit, Econometrica
Coverage: C-IP2
Recording of intro: Spotify Acast Apple Amazon Music Audible
We develop a dynamic structural model of patent screening incorporating incentives, intrinsic motivation, and multi-round negotiation. We estimate the model using detailed data on examiner decisions and employ natural language processing to create a new measure of patent distance that enables us to study strategic decisions by applicants and examiners. The estimated parameters and counterfactual analysis imply three main findings. First, patent screening is moderately effective, given the existing standards for patentability. Second, examiners exhibit substantial intrinsic motivation that significantly improves the effectiveness, and reduces the net social costs, of screening. Third, limiting the number of negotiation rounds strongly increases the speed and quality of screening. We quantify the net social costs of patent screening and find that the annual social cost of the existing system is $25.5bn, equivalent to 6.5% of U.S. R&D performed by the private sector.
Risk-Based Borrowing Limits in Credit Card Markets (SSRN) (Online Appendices) [Extra Note]
Coverage: FCA
Recording of intro: Spotify Acast Apple Amazon Music Audible
I use novel statement-level data on the 2010–2015 UK credit card market to show that lenders primarily individualize contracts through risk-based credit limits. Though shared with other European credit markets, this feature contrasts with the US counterpart, where interest rates are also individualized. To quantify the welfare implications of this distinction, I estimate a structural model that explains credit limit distributions with lender-specific credit scores. I evaluate a counterfactual where lenders can freely individualize prices and credit limits, which the existing environment precludes. Lenders control default risk with credit limits and use prices to extract surplus from inelastic borrowers.
Multivariate Ordered Discrete Response Models (SSRN), with Tatiana Komarova
We introduce multivariate ordered discrete response models with general rectangular structures. From the perspective of behavioral economics, these non-lattice models correspond to broad bracketing in decision making, whereas lattice models, which researchers typically estimate in practice, correspond to narrow bracketing. In these models, we specify latent processes as a sum of an index of covariates and an unobserved error, with unobservables for different latent processes potentially correlated. We provide conditions that are sufficient for identification under the independence of errors and covariates and outline an estimation approach. We present simulations and empirical examples, with a particular focus on probit specifications.
Policy Reports
Patent Quality in the United States: Findings and Suggestions for Policymakers, with Matthew Chervenak, Ani Harutyunyan, Mark Schankerman, and Nishant Shrestha
Coverage: IP Watchdog Law360 MarketWatch ClaimWise
Council For Innovation Promotion (C4IP) Panel Discussion (2nd October 2024):
Other Work in Progress
Linear Simultaneous Sample Selection models