At a lunchtime meeting on 23 March, Patrick Minford presented results of recent research on European Monetary Union. Minford is Edward Gonner Professor of Applied Economics at the University of Liverpool, a Research Fellow in CEPR's International Macroeconomics programme and member of the panel of seven independent advisors to HM Treasury. His remarks were based in part on his CEPR Discussion Papers Nos. 767, `Time-inconsistency, Democracy and Optimal Contingent Rules', and 774, `The Political Economy of the Exchange Rate Mechanism'. The views expressed by Professor Minford were his own, however, not those of CEPR, which takes no institutional policy positions.

Minford first noted that the rational expectations revolution has given rise to prescriptions for `tying the hands' of politicians that are strongly recommended by many policy-makers especially central bankers who would in many cases disown the rational expectations hypothesis. In particular, the Exchange Rate Mechanism of the EMS supposedly allows the non-German members countries' `dependent currencies' to borrow credibility from the Bundesbank. These currencies tend to become systematically overvalued over time, a process that is accompanied by large-scale subsidies to manufacturing.

Minford accounted for this pattern using a model of a small, open economy with traded and non-traded sectors under rational expectations with a fixed exchange rate. The government seeks re-election by satisfying vested interests that require it to support the traded goods sector, while at the same time seeking to control inflation and limit distortions to the domestic economy. Bad states give rise to inflation through devaluation, but in good times there is no such pressure for devaluation. In normal times, however, expected inflation is a weighted average of these two states and therefore reflects a constant risk of devaluation. When this actually happens, inflation rises above its expected value, but usually expected inflation is above its actual value. Constant upward pressure on wages in anticipation of a non-existent devaluation therefore leads to overvaluation of the currency and reduced competitiveness in the traded goods sector, which in turn induces lobbying for subsidies.

Minford maintained that the rule of tying the government's hands can in fact be supplemented by a degree of discretion to react to excessive shocks. In a standard BarroGordon model, `surprise' inflation will not raise output as the government intends, and there will be stagflation. The government must therefore either tie its hands and forgo monetary policy altogether or use pure discretion to stabilize shocks and live with the high inflation that results. Minford argued that this set-up must be highly artificial, since it requires voters in a democracy who desire both high output and low inflation systematically to reward behaviour that produces the precise opposite. If the electorate is sufficiently sophisticated to understand the Phillips curve, it will be equally able to set up and monitor a two-part incentive scheme. This would reward a government either for maintaining low inflation when there are no shocks between the time of its promise and the election or for its intelligent use of stabilization policy when there are such shocks, but not for seeking to raise output above its natural rate. Provided the electorate has full information, this `modified discretion' will dominate the existing trade-off; with limited information, it will still dominate `no discretion', but whether it will dominate pure discretion remains unclear.

Minford concluded that there is no need for EC member countries to stick with the ERM to gain credibility for anti-inflationary commitment; they should consider alternative policies that permit stabilization of shocks as well as monitoring of rule-based behaviour.