Inflation targeting for the United States?

• Author: McCallum, Bennett T.

Is inflation targeting suitable for the United States? I am inclined to say "Yes," but there are some issues and qualifications that need to be discussed. First, what is inflation targeting? Second, are there political problems that are likely to arise? Third, what about the arguments suggesting that inflation targeting has not provided benefits to those nations that have adopted it? Fourth, what about the analytical components of models typically used to analyze inflation targeting? These topics will be taken up in the sections that follow.

What Is Inflation Targeting?

When the term "inflation targeting" first began to be used frequently, around 1990, it was understood to mean a monetary policy strategy that made the achievement of a designated low rate of inflation the sole objective of monetary policy. Low inflation had, after all, been one of the main stated objectives of the Federal Reserve (the Fed) and other central banks for a long time (although it was not an explicit objective under the gold standard and the date at which maintenance of the gold standard ended is certainly open to dispute (1)). The concept of inflation targeting (henceforth, IT) gained prominence, I believe, from the policy regimes adopted by the Bank of Canada and the Reserve Bank of New Zealand--especially the latter, because of its spectacular provision concerning possible discharge of the governor if inflation was not contained each year within a target range of 0 to 2 percent. A major objective of these developments in Canada and New Zealand was, as I understood it, to get away from the then-standard view that monetary policy needed to strike an appropriate balance between goals pertaining to inflation and real output (or employment) stabilization. Thus early writings on IT emphasized the idea that, from a long-run perspective, central banks could have virtually no influence on output while being almost totally responsible for the behavior of inflation. (2) In order to put this idea into practice, it was deemed desirable to stress the primacy of inflation control to the point that a formal representation of a central bank objective function would, logically, include inflation (measured relative to target) as its only argument. During the later 1990s, however, this concept was replaced in academic--and probably central-bank--work with one that included both inflation and aggregate output (or employment) goals in the central bank's objective function. Writings by Svensson (1997, 1999) were particularly influential in this respect.

What, then, was the justification for a terminology in which policy regimes of this type would be called inflation targeting? The evolution in terminology went, I believe, something like the following. In the early 1990s, both the United Kingdom and Sweden were forced by market and political pressures to give up their fixed exchange rate regimes. They needed some nominal anchor, and wished to avoid any suggestion of "monetary targeting" as the latter would seem too much like a "monetarist" policy, which had been discredited in the eyes of the public and was never embraced by central banks other than the Swiss National Bank and (to some extent) the Bundesbank. Also, they wished to give more weight to inflation prevention than had been the case earlier in the 1980s and believed that the New Zealand-Canada idea of an explicit numerical inflation target was promising. So, the Bank of England and the Sveriges Riksbank chose "inflation targeting" as the label for their new (and not yet fully developed) monetary policy regimes. Writers on the subject who wanted to promote policies that stressed inflation control combined with output stabilization and more explicitness (or "transparency") therefore adopted the IT label. And since these writers were more numerous than ones who favored a sole operational objective--believing that this approach would result in excellent output/employment performance--they were able to capture the IT designation and to use it to refer to both types of regimes (i.e., with only inflation or with both inflation and output in the objective function).

Gradually, any inclination to make inflation the only variable in a central bank's objective function moved ever farther into the background. This occurred partly because of accumulating experience in Canada and New Zealand and also because of academic or partly academic writings. Particularly influential examples would include Clarida, Gall, and Gertler (1999) and Bernanke et. al. (1999). The definition given in the latter is that an IT regime is "characterized by the public announcement of official quantitative targets (or target ranges) for the inflation rate over one or more time horizons, and by explicit acknowledgment that low, stable inflation is monetary policy's primary long-run goal" (Bernanke et. al. 1999: 4). From this statement it seems to be the announcements, rather than the behavior, of the central bank that are crucial in determining whether a regime is an IT regime.

Meanwhile, most formal analytical studies of IT conducted today by academics, or by central bank economists, use formal frameworks (models) in which both inflation and output objectives appear in the central bank's objective function--which is often taken to mimic the objective function of private households--with no explicit requirements concerning specification of the relative weights attached to these variables. Furthermore, in many cases the model is...