Do shocks to animal spirits cause output fluctuations?

摘要

We evaluate whether shocks to "animal spirits" affect real outcomes at business cycle frequencies, using data on consumer confidence and region- and state-level coincident activity indexes in a vector autoregression (VAR) with long run restrictions. We assume that animal spirits shocks do not affect the long run level of activity. Innovations in fundamentals explain most variation in the level of activity, but animal spirits have economically significant effects at business cycle frequencies. A positive innovation in animal spirits causes real activity to rise by between 0.2 and 0.6% in the medium term, and animal spirits may explain up to about half of real fluctuations in the medium run.