The Normative Implications of Heterogeneity in the Frequency of Price Adjustment

What inflation rate should central banks target? Following the work of Bils and Klenow (2004), who were the first to document the large amount of heterogeneity in the frequency of price changes across different categories of goods and services in the United States, a growing literature has investigated the positive implications of this heterogeneity, especially with respect to the neutrality of money. On the other hand, the normative implications of the observed degree of heterogeneity in price stickiness remain largely unexplored. From a theoretical point of view, a well-known principle is that stickier prices should carry a disproportionate weight in the inflation stabilization objective of a central bank, since these prices contribute more significantly to the overall distortion in the economy than more flexible ones (Benigno, 2004.) In the two-sector limit, in which one sector has perfectly flexible prices, this principle implies that the monetary authority should focus exclusively on stabilizing inflation in the sticky price sector (Aoki, 2001.) In practice, this is often interpreted as implying that central banks around the world should target "core" inflation—in the United States, inflation ex food and energy. In this paper, we investigate quantitatively the accuracy of this prescription as an approximation of the optimal (Ramsey) policy in a microfounded model with heterogeneity in the frequency of price adjustment, calibrated to match the data in Bils and Klenow (2004) and Nakamura and Steinsson (2006.) The result of this analysis is what we call CONRI, a Cost-of-Nominal-Rigidities-Index, which weighs different prices as a function of their relative contribution to the economic distortion associated with nominal rigidity, rather than of their expenditure share alone, as in cost-of-living inspired price indexes such as the CPI. We argue that CONRI is the price index that central banks should focus on stabilizing, at least if they believe that the role of monetary policy is to minimize the distortions associated with the inefficient dispersion in relative prices caused by the presence of nominal rigidities. Finally, we compare this index to more traditional versions of "core" inflation, such as the CPI and PCE ex food and energy, and compute the welfare losses associated with targeting these indexes rather than CONRI.