I was a little surprised to read in the Christmas edition of the Frankfurter Allgemeine Sonntagszeitung (not yet online, subscription wall, in German) that Robert Mundell seems to have changed his mind. In his seminal 1961 paper about monetary integration, he famously stated that “the optimum currency area is not the world”. Now it appears he favors a sort-of worldwide currency union, initially comprising Dollar, Euro, and Yen (apparently, he’s also made that point earlier this year in Lib?ration (subscription wall, in French)).
Mundell’s theory of optimum currency areas deals mainly with ‘asymmetric real shocks’, that is a demand shift of the assumed only product of region A to the only only good produced in region B. He also assumes that price level and wages are downward inflexible and that labor is completely immobile between the two regions. In a currency union such a demand shock has the effect of causing unemployment in region A and inflation in region B because a common monetary policy can only accommodate either the macroeconomic needs of region A (monetary expansion) or region B (monetary contraction). This leads to the first central result that “the optimum currency area is not the world“, because by changing the exchange rate between region A and B the demand shock can be absorbed without creating unemployment in region A and causing inflation in region B.
But Mundell himself noted that these assumptions are limiting the usefulness of the model for practical reasons, as money is – the second central result usually disregarded by opponents of integration in the pre-Euro discussion – also a “convenience”, and that currency unions can be a tool to greatly reduce the transaction costs of economic exchange if their creation is at all viable given the structural characteristics of the economies involved – particularly with respect to the diversity of production structures and labor market flexibility. I think, among others, the eminent US economist and NY Times columnist Paul Krugman once remarked that, with respect ot production structures, the 12 US federal districts may be less suitable for a currency union than the current Eurozone.
Thus, while it is inherently difficult to determine an optimum, any currency area is a trade-off of regional macroeconomic policy independence and gains from reduced transaction costs made possible by a single currency – the more tightly two economies are intertwined, the higher the gains from monetary integration. Consequently, even proponents of regional monetary integration usually believe, despite the rapidly increasing international division of labor, that Mundell’s first point is still valid with respect to the current global economy: the optimum currency area still is not the world. But now it seems Mundell doesn’t believe that anymore.
The interview in the Frankfurter Allgemeine is not too detailed, unfortunately. But my interpretation of Mundell’s proposal to create a sort-of currency union of Dollar, Euro, and Yen (to begin with) is that he must have come to the conclusion that the transaction costs of flexible exchange rates are so high (now?) that he believes the savings would offset the loss of regulatory independence even given a degree of economic integration far below that of, say, the Eurozone.
I’m sure there must be reactions to such a challenge from “flexible rate”-proponents that go beyond regulatory inertia and pure lobbying. Suggestions very welcome.