Right Idea, Wrong Place? Knowledge Diffusion and Spatial Misallocation in R&D

A few cities perform a high share of research and development (R&D) in the United States. If R&D generates local knowledge spillovers, then the social returns to R&D may vary across cities and the geographic distribution of researchers may be inefficient. Equally important, but less well understood, is whether the private returns to R&D vary systematically across space. In this direction, I document a new fact from the market for technology: patent sales from inventor to firm decline steeply with distance, other things equal. My interpretation is that it is hard for inventors to commercialize their ideas in distant markets. Through the lens of a spatial growth model, I then infer that the private returns to R&D are low in remote regions. By contrast, spillovers are relatively flat across space because patent citations decline slowly with distance. Place-based R&D policy subsidizes research not in dense cities, but in remote locations where private returns are low. The optimal policy increases patenting by 3.6% and aggregate consumption by 1% in the long run, with minimal effects on inequality across regions or workers