Separating the Business Cycle from Other Economic Fluctuations, NBER Working

Macroeconomists—especially those studying monetary policy—often view the business cycle as a transitory departure from the smooth evolution of a neoclassical growth model. Important ideas contributed by Friedman, Lucas, and the developers of the sticky-price macro model generate this type of aggregate behavior. But the real-business cycle model shows that the neoclassical model implies anything but smooth growth. A purely neoclassical model, devoid of anything resembling a business cycle in the sense of transitory departures from neoclassical equilibrium, nevertheless ex-plains most of the volatility of GDP growth at all frequencies. Monetary policymak-ers looking to a neoclassical model to provide the neutral levels of key variables— potential GDP, the natural rate of unemployment, and the equilibrium real interest rate—need to solve a complicated and controversial model to find these constructs. They cannot take average or smoothed values of actual data to find them. Further, low-frequency movements of unemployment suggest a failure of the basic idea that departures from the neoclassical equilibrium are transitory. I discuss new theories of the labor market capable of explaining the low-frequency movements of unemploy-ment. I conclude that monetary policymakers should not try to discern neutral values of real variables. Some branches of modern theory do not support the concepts of po-tential GDP, the natural rate of unemployment, and the equilibrium real interest rate. Even the theories that do support the concepts suggest that measurement in real time is impractical