The Delors Report has prompted debate on whether to move rapidly towards monetary union or to adopt a more evolutionary approach by tightening the limits on the frequency and size of permissible realignments within the current EMS. The Bundesbank now sets German monetary policy in accordance with domestic anti- inflationary priorities, enabling other member countries easily to acquire anti-inflationary reputations by fixing their exchange rates vis-à-vis the Deutschmark. Such best-practice inflation policy will only follow from the Delors process if the new Eurofed has the same credibility that the Bundesbank already enjoys.

In Discussion Paper No. 472, Research Fellow David Currie, Paul Levine and Joseph Pearlman compare the costs of EMU if the Eurofed has no credibility, with those of the `hard-EMS' if member countries can credibly fix nominal exchange rates to the Deutschmark. In their model governments minimize welfare-loss functions based on deviations of output, inflation and interest rates from targets, and Germany sets greater store by low inflation than do the other members of the system. Their results overstate the true costs of EMU to the extent that the Eurofed builds up a reputation over time. They also apply the same model to two bench-mark cases with floating exchange rates: `non-EMS', in which only the Bundesbank enjoys credibility, which allows German leadership; and the fully-cooperative optimal policy, in which all national central banks have credibility.

In their natural rate model, in which wage and price inflation are always equal, Germany has zero inflation under non-EMS, while other members experience an inflationary bias. Hard-EMS and the fully-cooperative policy each impose zero inflation on both blocs, while inflation under EMU is higher than for any other regime, so cooperation without reputation does not pay. Extending the model to allow for short-run persistence in wages and prices the authors simulate the various regimes. They find that for member governments with ambitious output targets facing permanent stochastic demand and supply shocks, hard-EMS is unambiguously superior to EMU; and if the weighting on German preferences is high then hard-EMS improves markedly on both non- EMS and EMU for the non-German members. For hard-EMS to outrank EMU in the face of transient shocks requires at least some compromise from the German authorities.

Currie, Levine and Pearlman conclude that hard-EMS has superior anti-inflationary and stabilization properties to EMU if the Eurofed has no reputation and if the Bundesbank considers other members' interests. Once the Eurofed acquires a reputation, however, EMU will improve slightly on hard-EMS with no German compromise. But compromise under such a `reputational' EMU would be more likely than hard-EMS to lead to the adoption of welfare criteria that would make stabilization policy favourable to the non-German members. Since building this credibility will require time and effort, these results suggest the need for caution in the transition to EMU.

European Monetary Union or Hard-EMS?
David Currie, Paul Levine and Joseph Pearlman

Discussion Paper No. 472, November 1990 (IM)