| If government patent policy includes significant renewal fees (perhaps increasing with the age of the patent), less valuable patents will be allowed to lapse, eliminating some deadweight loss from the economy. These fees, however, decrease the incentive to innovate. This thesis explores the factors determining the optimal balance of these two effects.;In a simple one-firm model, the expected value of a patent (used as a proxy for the incentive to innovate) is held fixed, and the patent fee schedule and maximum patent life are selected to minimize expected deadweight loss. Within this simple framework, it is shown that the optimal fee schedule is piecewise constant (often zero) with at most two pieces. Conditions determining the optimal schedule are derived for product and process innovations. A broad class of conditions under which no patent fees should be charged is outlined.;These results are shown to hold when the one-firm model is embedded within a series of simple models of the economy. In the first generalized model, firms produce independent innovations with revenues drawn from a single probability distribution. Government policy affects the amount of both innovation and deadweight loss. A second model involves a more explicit link between innovative effort and innovation magnitude, and a third involves a technology race between firms competing for a single innovation.;In a more complex model of the economy, an aggregation of firms create innovations with revenues drawn from a variety of probability distributions. This model demonstrates how more complex fee schedules may optimally be used to selectively encourage the innovations of the highest total societal value. |
