Using the recent EC Commission report `One Market, One Money' as a point of reference, we consider the merits of a single currency in Europe. The main benefit is the reduction in transaction costs, which the report estimates at 0.4% of European Community (EC) GDP (but much less in countries with sophisticated financial systems). A weak case is made for other benefits. The principal potential cost, and the main concern of this paper, arises from macroeconomic instability, the stochastic equivalent of the problems stressed in the traditionally Keynesian literature on `optimal currency areas'. Since the EC has a low level of labour mobility, a negligible fiscal offset to national shocks, and a fair degree of short-run nominal rigidity in wages and prices, one would expect that under European Monetary Union (EMU) the loss of the exchange rate as a stabilizing mechanism would be damaging. This duly emerges from our stochastic simulations of both the world and the UK Liverpool models. Our multilateral world simulations reveal that, under both fixed money supply rules and strategically responsive monetary policy, floating is superior to EMU for all countries, and that even if the rest of the Community proceeds with EMU, the UK is better off outside it. This latter conclusion is reinforced in the exercise on the more refined UK quarterly model. We examine critically the EC's stochastic simulation exercise using the IMF Multimod world model. The simulations yield a positive result for EMU by inserting shocks in risk premia into the Uncovered Interest Parity relationships and removing these between EC currencies under EMU. This approach is, we argue, internally inconsistent and econometrically flawed. Additionally, the EC simulations are based on asymmetric assumptions about monetary policy under floating and EMU. As a result, the comparison between floating and EMU is seriously biased.
