On the timing of innovation in stochastic schumpeterian growth models

"Recent work has revived the Schumpeterian hypothesis that recessions facilitate innovation and growth. But a major source of productivity growth, research and development, is actually procyclical. This paper argues that while it is optimal to concentrate growthenhancing activities in downturns, dynamic spillovers inherent to the R&D process lead private agents to concentrate too much of their R&D activity in booms, precisely when its social cost is highest. Thus, while previous literature has argued recessions promote growth and intertemporal substitution is a desirable consequence of fluctuations, in the case of R&D recessions discourage growth and intertemporal substitution proves to be a social liability"--Federal Reserve Bank of Chicago web site.