You can find the paper here.

Title: Misallocation in the Market for Inventors

Abstract: How much misallocation is there in the market for inventors? This paper introduces a novel growth model where firms search for inventors in a frictional labor market, and where the efficiency of innovation depends on the quality of the match between the inventor and firm. The model features both knowledge spillovers and congestion in the labor market, hence the resources invested in search can be either too low or too high. I use the model to quantify this tradeoff and to study what policies can rectify the inefficiencies in this market. I document that inventors are 40% more productive when moving across firms, a finding not driven by differences in firm innovativeness. I find that search frictions and the resulting misallocation of inventors across firms are of first-order importance for growth: shutting down search frictions and moving to a world where inventors are perfectly allocated across firms increases growth from 2.0 to 2.6 percent. Moreover, I find that firms invest too little in searching for inventors, and that conventional R&D subsidies do not address the inefficiencies in the market for inventors. Finally, I show that partial-equilibrium effects of R&D subsidies dramatically overstate their general equilibrium effects due to a novel channel in my model: when firms innovate and creatively destroy their rivals’ product lines, they also break their matches with inventors.