Fluctuations in productivity growth rates and input utilization in U.S. manufacturing

Abstract

Recent work suggests that the utilization rate of inputs (that is, the intensity with which inputs are used) varies over time and partly explains procyclical productivity. To confirm this, the panel data of the “NBER Manufacturing Productivity Database” (450 4-digit industries from 1958 to 1991) [Bartelsman and Gray, 1994] were analyzed to determine the year-by-year fluctuations in total factor productivity growth rates and the utilization rates of nonmaterial inputs. The utilization rate was found to be procyclical. When the gross domestic product (GDP) growth rate is 1 percentage point above normal, the utilization rate increases 2.5 percent. After this correction, productivity growth rates are still found to be procyclical, though its correlation with GDP growth is half that of the uncorrected Solow residual. That is, demand shocks can explain half of the procyclicality of productivity growth rates, but real business cycles still cannot be ruled out. Lastly, the results are compared with the Federal Reserve's capacity utilization statistics. It is shown that the latter actually measures a combination of capacity utilization and productivity shifts and, hence, overstates changes in capacity utilization.