Introduction

The debate over the adoption of a common currency in the European Community has focussed on savings of transaction costs as one of the leading advantages of a monetary union. As international transactions continue to expand, and as they become more complex, the small losses involved in exchanging currencies may grow to sizable fractions of total values, and the savings realized through a common monetary standard may therefore be nonnegligible. The EC has estimated direct, ‘mechanical’ savings to be of the order of 0.5 per cent of Community-wide GDP (European Economy, 1990), with unequal distribution across countries. The belief that transaction costs are one of the main motivations for a common currency is shared by policy-makers and financial journalists (see for example the editorials of The Economist in the last two years), by academics (Canzoneri and Rogers, 1990), and, I believe, by public opinion in general.

But if transaction costs are important, they must play a role in determining the composition and the size of the different markets. A common currency would then imply not only lump-sum savings, but a change in the partition of traders between domestic and international activities. This is the view studied in the paper.

Since a monetary union requires a common monetary policy and a common inflation rate, if inflation is distortionary there is a second channel through which monetary unification may affect the formation of markets.