| Arrow's (1962) popular drastic/nondrastic dichotomy of process innovations assumes perfectly competitive pre-innovation market conditions with implicit Bertrand logic, which drives the counterintuitive result that all innovations result in monopoly. When the pre-innovation industry is instead populated by heterogeneous cost firms engaged in Cournot competition, it is possible for a new technology to coexist with older technologies, so there is a wide range of post-innovation market structures. An index is introduced which generalizes Arrow's drastic/nondrastic classification for any pre-innovation market structure and any degree of licensing exclusivity. This index measures the degree of induced exit when an innovation is exclusively licensed, and identifies each innovation's drasticity, which identifies the quantity of licenses sufficient to drive out all older technologies.; This dissertation also explores the effects of asymmetric information on licensing contracts by extending Gallini and Wright's (1990) binary signaling model to three or more types. When patent licensors offer contracts that include fixed fees and performance-based royalties, the number of licenses granted may signal innovation quality. When innovation qualities are similar, the magnitude of output royalties signals both quality and the cost of inventing around the patent. Commonly observed payment structures naturally arise in Riley (1979) equilibrium contracts. |
