Fluctuations in R&D Investment and Long-run Growth: The Role of the Size Distribution of Innovating Firms

R&D investment drives productivity growth. Therefore, its uctuations over the business cycle a ect long-run dynamics. I show that taking into account the size distribution of innovating rms generates new insights on this link. I write an endogenous growth model with heterogeneous rms assuming, in line with empirical evidence, that small rms have a relatively higher innovation capacity than large rms. In addition, the model predicts that small rms reduce R&D more than large rms after negative aggregate shocks, especially nancial ones. These di erences between rms generate two novel predictions at the aggregate level. First, a nancial shock leads to a rightward shift in the size distribution of innovating rms. This ampli es the shock's e ect on productivity growth and makes it persistent over time. Second, there is a positive correlation between the share of an industry's innovations done by small rms, the industry's average productivity growth rate and the volatility of that growth rate. Firm-level empirical evidence supports the greater reaction of small rms' R&D to negative aggregate shocks, and the model's aggregate predictions are consistent with a line of stylized facts.