On the cyclicality of R&D activities

Abstract

This paper develops a real business cycle model with endogenous innovation, in order to explore cyclical fluctuations in R&D activities. A calibrated version of the model is used to discriminate between different R&D production technologies on their ability to account for the observation that R&D in the U.S. is procyclical. The analysis finds that the knowledge-driven specification, commonly used in the growth literature, fails to account for important features of the business cycle, most notably the procycical movement of both R&D investment and R&D employment. An alternative specification that allows for multiple inputs including scientists, staff, and final goods can account for the business cycle phenomena of the U.S. economy, including the procyclicality of R&D activities. An implication of this particular specification of the R&D process is that productivity shocks amplify the business cycle by stimulating more ideas in good times.