Business Cycles: Real Facts or Fallacies?

Introduction Conflicting views on the causes and nature of business-cycle fluctuations usually can be attributed to different beliefs about the adjustments of prices and wages. For example, Kydland and Prescott (1990) - KP hereafter - concluded that U.S. prices had been countercyclical since 1954 and interpreted that as evidence against demand-driven models of the cycle. Those findings were based on the innovative “stylized-facts” method of KP, as further developed (Kydland and Prescott, 1991) and contrasted with the econometric “system-of-equations” approach of Koopmans (1947,1949). The main reason for KP's dismissal of econometric models seems to have been that results derived from system-of-equations models were model-dependent and represented biased evidence on the nature of business-cycle fluctuations (e.g., the “myth” that prices behave procyclically). As an alternative, KP offered their own stylized-facts method, which involved only a minimum of assumptions, thus allowing the data to speak directly, instead of through the veil of an econometric model. KP applied a filtering technique known in the economics literature as the Hodrick-Prescott (HP) filter to the raw data in order to identify and remove the trend in the individual series.