Patent Experimentalism

I haven't had much time for blogging recently because I've been writing and revising a new article, Patent Experimentalism, which I'll be presenting at IPSC on August 8. This is still a work in progress, so please send me your feedback and suggestions! Here is the current abstract:

The current emphasis on national and international uniformity in patent law makes it hard to resolve the vast empirical uncertainties about patents, but unconstrained local diversity is also suboptimal due to the spillovers from local innovation policies. Instead, central planners should promote policy variation through a pluralistic, evidence-based approach. By examining different methods of policy experimentation through the lens of the patent system, this Article addresses the second-order question of how to answer these open questions: how do we promote innovation in promoting innovation?

Randomized assignment is the gold standard for evaluating the effect of a well-defined policy treatment, and more randomized patent experiments should be performed both in the field (e.g., testing pharmaceutical prizes in a random selection of therapeutic classes) and in the lab (e.g., testing how varying disclosure affects performance in implementing a software patent). Randomization should be used to evaluate well-defined policies with measurable and stable outcomes, and when spillovers between experimental groups can be minimized.

But policy randomization entails high costs and would be ineffective for nuanced, dynamic issues, such as the bounds of patent-eligibility for changing technologies in heterogeneous jurisdictions with diverse non-patent innovation policies. These uncertainties are better approached not through static, short-term experiments, but rather through adaptive governance: in “experimentalist” systems, local actors—patent examiners, judges, or even individual countries—are granted broad discretion to meet centrally defined framework goals. But unlike in “laboratories of experimentation,” local autonomy is constrained by reporting requirements and robust peer review. Experimentalism is most valuable for generating buy-in from local units; promoting local innovation with policy design; eliciting local knowledge about the applicability of different policies to heterogeneous conditions; and generating better measures of whether policies are succeeding in light of local values.

You can download the latest version of Patent Experimentalism from SSRN.