Optimal Taxation and R&D Policies

We study the optimal design of corporate taxation and R&D policies as a dynamic mechanism problem with spillovers. Firms have heterogeneous research productivity, that productivity is private information. There are non‐internalized technological spillovers across firms, but asymmetric information prevents government from correcting them in first best way. highlight key parameters for (i) relative complementarities between observable investments, unobservable inputs, firm (ii) dispersion persistence firms' productivities, (iii) magnitude firms. estimate our model using firm‐level data matched to patent quantify policies. In data, high firms get disproportionately higher returns investments than lower Very simple innovation policies, such linear taxes combined nonlinear subsidy—which provides marginal subsidies at levels—can do almost well unrestricted Our formulas theoretical numerical methods more broadly applicable provision incentives settings spillovers, generally.