Working Papers

---

"Authorizing Bespoke Therapies," with Daniel Chen and Timothy Yu.   

Novel medical treatments that skip, mask, or edit genetic mutations are capable of solving previously incurable ailments. Such ultra-individualized treatments pose challenges for the existing system of 1) premarket regulation, 2) pharmaceutical incentives, and 3) tort compensation. On top of these challenges, the structure of N-of-1 precision medicine creates a further complication. Each N-of-1 precision treatment uses shared modalities to deliver individualized treatments; this means that information created in one treatment’s development can benefit the development of another treatment. The Article reframes the interconnected nature of N-of-1 precision therapies as a positive network externality and incorporates the platform economics literature to show that such externalities can be well-managed in a multi-sided platform system. Onerous ex-ante premarket approval would be replaced by standards-based good practice review of pre-registration designs, similar to the regulatory structure currently governing laboratories. Rather than relying on the patent system to provide incentives to create, laboratories would be paid for sharing data from pre-registered studies. This data sharing not only reduces costs of development but also helps insurance markets price the risk of covering such treatments. Finally, the gaps left by products liability and medical malpractice claims would be filled by monitoring the pre-registered designs.  

“Behavioral   Impediments to Beneficial Experimentation,” with W. Kip Viscusi.  Prior evidence suggests that people do not   experiment enough, and that when they do experiment, they do so poorly. Without sufficient instruction about when   experimentation would be beneficial, consumers may be too reluctant to   experiment with uncertain treatments. This project explores consumers’   sensitivity to estimate precision and data quality in their willingness to   experiment by allowing participants to choose between a traditional drug and   a “new" drug, whose efficacy is less certain. The survey explicitly   discloses several key data: the success rate and sample size associated with   the traditional drug and any existing evidence for the new drug, as well as   the quality and sample size of additional data about the new drug. It also   provides visual representations of the mechanics and weaknesses of different   sampling techniques (random vs. voluntary samples). Using an   incentive-compatible experimental design, we measure participants’   willingness to pay for additional information about new treatments and   ability to update their prior beliefs based on this additional information.   Results demonstrate that participants' willingness to pay for additional   information is sensitive to quality of data offered, the existing data   available, and the potential selection observed in the data. Once purchased, participants incorporate   data by discounting heavily selected samples in their updated beliefs.