Monetary Unions
Optimum sizes

Recent events in the European Community and the former Soviet bloc have brought the theory of optimum currency areas (OCAs) to the forefront of discussion. In Discussion Paper No. 847, Research Fellow Jacques Mélitz integrates this theory with modern approaches to international trade and macroeconomics. He likens the benefits and costs of widening a currency area to a reduction in transport costs and a slowing of the terms-of-trade adjustment respectively. With many trading nations of different sizes, each engages in inter-industry trade based on comparative advantage and intra-industry trade based on economies of scale. Foreign trade imposes particular costs, which are in part monetary, relating to multiple currencies and units of account; its non- monetary costs include not only transport but also information, distribution, litigation and insurance arising from differences in language, custom and law.

As the union expands in size, the monetary sales costs of its members' trade fall while their welfare rises but at a diminishing rate. With sticky prices in goods and factor markets, the costs associated with a monetary union of any size will depend on its composition. High levels of internal intra-industry trade imply that its members' industrial structures are similar so the terms of trade among themselves will change little, while low levels of non-monetary sales costs indicate that its members are geographically, culturally and juridically close, which also facilitates trade adjustment. Finally, the union's costs also rise as the union and national currencies' equilibrium real exchange rates diverge. As the union's size increases, the quality of the best potential new members progressively worsens, so enlargement entails a rising marginal cost. Taken together with the diminishing marginal benefits of enlarging a monetary union, this implies that there does exist an optimum size for a currency area.

The Theory of Optimum Currency Areas, Trade Adjustment and Trade
Jacques Mélitz


Discussion Paper No. 847, October 1993 (IM)