Monetary Policy and Inflation Targeting

Abstract

In the 1990s, several countries shifted to a new monetary policy regime: an announced quantitative target. The reason for this shift was the unsatisfactory performance under previous regimes. New Zealand, Canada, Australia, and Spain all introduced targets under persistently high inflation; the United Kingdom, Sweden, and Finland did so after having abandoned fixed exchange rates, which had failed to achieve low and stable and had been subject to dramatic speculative attacks. Inflation targeting has received much recent attention, both among policymakers and academics. In the United States and in Europe it is debated as a possible monetary policy strategy for the Federal Reserve System and the future European Central Bank, respectively. Academic research on targeting, both theoretical and empirical, has grown quickly.(1) My own research in the last few years has largely dealt with understanding targeting in relation to other monetary policy regimes and investigating how practical monetary policy can best be conducted under targeting. Practical targeting has several common characteristics: 1) an announced quantitative target, varying across countries between 1.5 and 2.5 percent per year, in most countries with a tolerance band of plus/minus 1 percentage point around the target; 2) no explicit rule on how the central bank shall set its instrument; 3) a floating exchange rate (except for Finland and Spain, which are members of the Exchange Rate Mechanism, although the wide exchange rate bands there so far have not created any conflict between the target and the exchange rate target); and 4) a high degree of transparency and accountability. Commentators also often describe targeting as a regime without an intermediate target for monetary policy (instead, targeting directly). I have argued in some of my research that this is misleading and that targeting actually implies a particular intermediate target, namely the central bank's forecast. Inflation Targeting as a Remedy Against High Inflation Inflation targeting can be seen as a potential remedy for persistent high inflation. Other remedies discussed and suggested in the literature include: 1) accepting that the long-run Phillips curve is vertical and implicitly, or explicitly, setting any output or employment target equal to (rather than above) the natural level; 2) creating an independent and central bank; and 3) setting up a performance contract (an inflation contract) for the central bank governor or governing board. In one of my papers, I examine the relation between targeting and these remedies. Inflation targeting indeed can involve elements of all three remedies. By announcing a rather low target and creating some degree of commitment to it, targeting can help to reduce inflation, even if an inflationary bias remains, and if more often exceeds than falls short of the target. This creates a conservative central bank in the sense of having a lower target rather than, as is common in the literature since Rogoff's classic 1985 article, identifying with a larger weight on a given target. Incidentally, this interpretation of conservatism solves an empirical puzzle about independent central banks, inflation, and output variability. If independent central banks are more in that they give more weight to a specific target, then lower should be correlated with higher variability of output. A large literature instead has stated that more independent central banks in industrialized countries are associated with lower rates, but not with higher variability of output. This finding is instead consistent with independent central banks simply having lower targets.(2) Price-Level Targeting versus Inflation Targeting Inflation targeting implies base drift of the price level, even if the target is set at zero: if overshoots its target, then the target for the next period is related to the new price level. …

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 67 citations exist upstream.