This paper provides a framework for Institutionalist analysis of the determinants of inflation and inflation targeting policies based upon principles in F. Gregory Hayden's book, Policymaking for a Good Society (2006).
Milton Friedman famously argued, "inflation is everywhere and always a monetary phenomenon." In debates with orthodox Keynesians regarding the appropriate method of modeling the transmission of monetary policy, Friedman preferred analysis via the quantity theory of money and reduced-form models rather than detailing within structural models the links between monetary policy and inflation. Of course, in addition to the money supply, the natural rates of unemployment and inflation expectations also were key in Friedman's understanding of the transmission of monetary policy to inflation.
Central banks in the real world target interest rates, not monetary aggregates, while neither economists nor central bankers have uncovered a particular rate of unemployment--or a methodology for doing so in real time--that was reliably linked to the inflation rate. As the only part of Friedman's transmission mechanism that remains, inflation expectations are front and center within the current "New Consensus" framework for analyzing monetary policy. (1) In the New Keynesian Phillips curve--standard in the "New Consensus" literature--inflation is determined by expected inflation and the current output gap. In a core "New Consensus" paper, Clarida, Gali, and Gertler (1999) argue that inflation expectations are set by current and expected future paths of the output gap, while the output gap depends on current and expected future paths of monetary policy. As Michael Woodford put it, "not only do expectations about policy matter, at least under current conditions, little else matters" (2004, 16; emphasis in original). In order to "anchor" these expectations and keep inflation low, "New Consensus" adherents propose monetary authorities follow credible, systematic monetary policy strategies, such as Taylor's (1993) famous interest-rate feedback rule, in which a predictable, aggressive policy reaction occurs when inflation exceeds its target. In short, while the "New Consensus" emphasis on expectations "may bypass money ... it has retained the key [Friedmanian] conclusion that central banks ultimately determine the inflation rate" (Meyer 2001, 3).
The "New Consensus" research program retains Friedman's lack of interest in empirically investigating and explicitly modeling the avenues of monetary transmission through the economy to the inflation rate. Consider, for example, the growing literature seeking the best "short run" measure of inflation. In this literature, the primary criterion for an appropriate "short run" measure--be it "core" inflation, "trimmed mean" inflation, "median" inflation, or some other adjustment to a price index--is that it be the best predictor of "longer run" trends in a complete price index (e.g., Webb 2004; Dolmas 2005; Rich and Steindel 2005; Khettry and Mester 2006). While "experience indicates that food and energy prices are subject to large short-run disturbances that are beyond the ability of monetary policy to control" (Poole 2006, 161), this is now insufficient grounds for excluding food and energy prices from the "short-term" measure of inflation being targeted by the Fed as "core" measures of inflation have done. Instead, inclusion of prices in the energy, food, or any other sector into a "short run" inflation measure depends upon whether "longer-run" inflation predictions improve. As with both Friedman and the "New Consensus," it is taken as given that "long-term" inflation is set by monetary policy.
Toward an Institutionalist Approach
An Institutionalist approach to analyzing inflation targeting would differ markedly. Specifically, the Personal Consumption Expenditures Price Index (PCEPI) and the Consumer Price Index (CPI) are broadly accepted empirical measures of aggregate prices in the United States. Whether these measures are appropriate monetary policy targets depends upon whether there is a reliable link between the actions of the monetary authorities and these indexes. In other words, whether central banks can target inflation is an empirical question, and this is particularly the case where generally accepted measures of the aggregate price level are already in use.
This point is illustrated in Figure 1, taken from Hayden (2006, 17 and Chapter 5), which presents an Institutionalist approach to policy analysis. As an accepted policy goal, low inflation fits into Figure 1 as a primary criterion of policy and largely a qualitative goal according to the qualitative/quantitative spectrum at the bottom of the figure. At the far right of the figure would be the CPI and PCEPI, which are the quantitative measurements that serve as indicators for assessing whether the policy goal...