The author would like to thank Francesco Caramazza, Mark Griffiths, Graham Hacche, Timothy Lane, Flemming Larsen, Paul Masson, Donogh McDonald, and Massimo Russo for comments on a previous version of this paper. Comments received from seminar participants at the Central Bank of Sweden are also gratefully acknowledged. The opinions expressed in this paper should, however, be solely attributed to the author.
The “core” is defined as the group of EU countries which have managed to maintain relatively stable exchange rates against the deutsche mark since at least 1987, and consequently, have synchronized their monetary policy fairly closely with the anchor country. The “core” could, thus, include Austria, Belgium, Denmark, France, Germany, Luxembourg, and the Netherlands. “Encompassing” monetary union refers in this paper to one in which all, or most of the European Union countries, become part of the single currency zone.
The discussion in this section draws on the EMI’s Annual Report for 1995, and the EMI’s document, “The Single Monetary Policy in Stage Three: Specification of the Operational Framework,” January 1997.
Reserve requirements enhance the controllability of the monetary aggregate by increasing the interest elasticity of money demand. The degree of controllability will vary according to how the reserve-requirements provisions are defined—whether lagged, contemporaneous, or averaged. See Goodfriend (1987) and Feldstein (1993) for a discussion of these issues. The EMI has also emphasized the need for reserve requirements to stabilize money market rates, and for dealing with structural liquidity shortages.
Svensson (1994) and Obstfeld and Rogoff (1995) argue that these developments have increased the probability of self-fulfilling speculative attacks on fixed exchange rates, particularly when the economy is subject to asymmetric real shocks in an environment characterized by nominal rigidities.
In the United States, for instance, targets were set for both broad and narrow money in 1975, but the Federal Reserve Board has focused mainly on the broad monetary aggregate since the late 1980s—see Friedman (1996) and Friedman and Kuttner (1996) for details. Germany started off by targeting “Central Bank Money”—a weighted average of M3 between 1974 and 1987—but then shifted to targeting M3 itself—see, in this context, von Hagen (1994) and Clarida and Gertler (1996). Other industrial countries also had their own versions of monetary targets to serve as focal points for monetary policy following the collapse of the fixed exchange regime in the early seventies.
See Baumgartner and Ramaswamy (1996) for a discussion of how the inflation targeting framework can combine flexibility with the features of a rule.
In practice, the distinctions among explicit inflation targeting, implicit inflation targeting, and monetary targeting are likely to be of a less stark form than is implied by the analytical discussion in the paper. Nevertheless, the conceptual distinction made between the different monetary frameworks serves the useful purpose of providing a sharp focus on the criteria for choosing one particular framework over another, and the implications of that particular choice.
Implicit inflation targeting is not incompatible with the central bank also having output as an argument of its objective function. In fact, if demand shocks predominate, targeting inflation should not be very different from targeting nominal income. With supply shocks, however, there could be a conflict between targeting inflation and targeting nominal income. See Baumgartner and Ramaswamy (1996) for a review of this discussion.
It should, however, be noted in this context that the Bundesbank at times provides justifications for overriding monetary targets.
More precisely, the Taylor rule is a policy rule in which the monetary authority increases its operational interest rate either when current inflation is above the target rate, or output is above potential. This feedback rule assigns equal weights to the deviation of inflation from target, and the deviation of output from potential. The Taylor rule appears to fit the actual policy performance of the Federal Reserve Board fairly well. See Taylor (1993). In this context, monetary targeting can be perceived as a special case of inflation targeting in which the monetary aggregate is assigned a weight of unity.
Measured credit premia is currently only in the range of 30 to 40 basis points even in the highly indebted countries of the European Union. However, it is quite likely that the eurobond yields issued by the different countries in an encompassing monetary union will vary to a greater extent than is indicated by current measures of credit-premia. The reasons often cited in support of this conjecture are that the removal of the currency premia makes credit risk the focal point, and that measured currency premia in practice incorporates some credit risk. Consequently, measured credit risk is likely to higher in a monetary union than is suggested by current estimates of it.
Cassard, Lane and Masson (1997) provide evidence for the existence of a stable demand for broad money for a “core” group of European countries—Germany, France, the Netherlands, Belgium, Denmark, and Luxembourg. The study by Artis, Bladen-Hovell, and Zhang (1993) also indicates a more stable demand for the monetary aggregates of a group of seven European Union countries (Germany, France, Italy, the Netherlands, Belgium, Denmark and Ireland) than is the case with the individual money demand functions. In a comprehensive analysis of EU-wide money demand functions, Monticelli and Papi (1996) present evidence that money demand for the EU area as a whole is stable and predictable for a wide variety of specifications.