EXTERNAL EFFECTS IN UNITED-STATES PROCYCLICAL PRODUCTIVITY

In this paper we highlight a new dimension of the aggregate procyclical productivity phenomenon. We show that estimates of the degree of returns to scale are larger for manufacturing as a whole than for two-digit industries. Since this difference must be due to factors that are only internalized at the most aggregate level, we term it an external effect. This result rules out explanations based on own-input variation - such as true increasing returns and unmeasured factor utilization tied to own-activity - as the sole explanations for aggregate procyclical productivity. We explore several potential explanations of this external effect.