Intellectual property scholars have vigorously debated the merits of patents versus prizes for encouraging innovation, with occasional consideration of government grants. But these are not the only options. Perhaps most significantly, the patents-versus-prizes (or patents-versus-prizes-versus-grants) debate has largely neglected the role of tax incentives in innovation policy, despite the tens of billions of dollars spent globally on tax breaks for R&D activities each year. How should R&D-related tax incentives figure into this debate, and what criteria are relevant for policymakers selecting among the various tools?In this Article, we develop a new taxonomy of innovation policies that allows direct comparisons among patents, prizes, grants, and tax incentives. This taxonomy highlights the overlooked efficiency benefits of tax credits: like patents, they elicit privately held information about the expected value of R&D projects; like grants, they reduce the social-welfare costs of frictions in imperfect capital markets. Our taxonomy also sheds new light on nonefficiency dimensions of R&D policy. Grants, tax credits, and prizes generally require all taxpayers to subsidize R&D regardless of whether they use the resulting products, whereas the patent system imposes R&D costs primarily upon the consumers who purchase patented products. In some contexts (e.g., life-saving drugs), the user-pays aspect of the patent system is difficult to defend on distributive justice grounds. In other contexts (e.g., luxury goods), the user-pays aspect of the patent system may make patents normatively preferable in comparison to alternative incentive mechanisms.
Ultimately, optimal innovation policy will depend on a range of factors that are likely to vary across contexts. For example, grants may be optimal where the government has a comparative advantage in evaluating potential projects, while tax credits may be optimal where potential innovators have private information about project prospects and limited access to outside capital. We argue for a pluralistic approach to innovation policy that incorporates each of the four main incentive mechanisms, and we provide examples of this pluralistic approach in practice.