Paolo Mauro

The IMF’s mandate to conduct surveillance over international exchange rates has led IMF researchers to consider the desirability of and the requisite supporting policies for, many currency unions in past decades. Given its broad country membership, the IMF analyzes all currency unions, including unions that consist of small developing countries, a subject often underexplored by researchers in developed countries. These small-country currency unions are of interest intellectually and in terms of policy; they also provide lessons of broader relevance. This article surveys the work on currency unions done at the IMF.

Work related to exchange rate regimes and currency unions is at the core of the IMF’s mandate. Member countries of the IMF choose their own exchange rate regimes, but the IMF plays an important advisory role on whether a country’s exchange rate regime is consistent with other policies. IMF researchers’ interest in currency unions may also have been spurred by Robert Mundell’s stint in the IMF Research Department in the early 1960s: his first seminal contributions to the theory of optimum currency areas (OCA) were published about the time he joined the department.¹

IMF researchers have made significant theoretical contributions in this area of study. Bayoumi (1994) developed possibly the first formal OCA model, obtaining some of the key insights (which had been expressed only discursively in earlier papers) on the role of openness, diversification, labor mobility, and correlation of economic shocks in a general equilibrium context with microeconomic foundations.² Ricci (1997) not only examined the real aspects of OCA theory, but also its monetary aspects and the interaction with real aspects, and derived the conditions under which openness makes a currency union more attractive.³

Research has also been conducted on fiscal policies within currency unions and on the institutional features of a common central bank. Debrun (2000) argues that fiscal rules are necessary to provide credibility to a monetary union even with an independent central bank.⁴ Bayoumi and Masson (1998) provide evidence that fiscal stabilization policies conducted by individual countries within a currency union tend to be less effective than supranational stabilization policies that provide insurance across countries.⁵ Debrun (forthcoming) shows that bargaining among founding member countries of a currency union tends to result in a supranational central bank resembling a national central bank with the strongest inflation-fighting credentials.⁶

Empirical work has shown that observed exchange rate and monetary policies can be related to variables identifed as important by OCA theory. Bayoumi and Eichengreen (1998) find that pressures on the exchange rate mainly reflect asymmetric shocks, whereas intervention reflects a country’s small size and large trade links—the variables that cause countries to value stable exchange rates according to OCA theory.⁷ Ricci and Isard (1998) show that the relative variability (against external currencies) of the euro and a basket of predecessor currencies depends on the relative sizes of countries, their sectoral patterns of trade, and the relative importance of different shocks.⁸

Much of the recent work on currency unions at the IMF has focused on whether particular groups of countries have the requisite economic and institutional characteristics for the successful operation of a currency union. As there is no definitive empirical criterion for determining whether a group of countries constitutes an OCA, most empirical analyses compare key OCA variables between existing and proposed unions. Pioneering work in this vein was done by Bayoumi and Eichengreen (1993), who showed that economic disturbances were more correlated among the states of the United States than among the countries of Europe.⁹

With a focus on the European Economic and Monetary Union (EMU), Bayoumi and Prasad (1997) examine sectoral data on output and employment for eight U.S. regions and eight European states, and find that labor market adjustment is slower in Europe than in the United States.¹⁰ Bayoumi and Eichengreen (1997) estimate an OCA index for the European countries, and find that Europe looked more like an OCA in 1995 than it did in 1987, suggesting that economic integration increased the desirability of monetary integration.¹¹ Kopits (1999) and Masson (1999) argue that participation in the EMU would be desirable for the Central and Eastern European countries that aspire to European Union membership, and outline possible strategies for their exchange rate policies in the interim.¹²

Bayoumi and Eichengreen (1994) find that the underlying economic disturbances are more diverse across the countries of the North American Free Trade Area (NAFTA) than across the countries of EMU, implying that the costs of abandoning the exchange rate instrument would be higher in the case of NAFTA. Similarly, they find the correlation of disturbances across Mercosur countries to be small and insignificant.¹³

Bayoumi and Mauro (1999) and Bayoumi, Eichengreen, and Mauro (2000) find that on economic criteria (including direction of trade, correlation of aggregate supply shocks across countries, size of shocks and speed of adjustment to them, and degree of similarity of the monetary transmission mechanism across countries) the Association of South East Asian Nations seems to be less suitable for a regional currency arrangement than the European Union was before the Maastricht treaty, although the difference is not large; political commitment would be a key factor in improving the chances of success of the proposed union.¹⁴

Jadresic (forthcoming) argues that, for the countries in the Gulf Cooperation Council, moving from individual currencies currently de facto pegged to the U.S. dollar, to a single common currency, which would also be strongly linked to the dollar, would yield little added benefit or cost. Nevertheless, a properly implemented currency union might enhance economic efficiency, deepen regional integration, and foster development of the non-oil economy.¹⁵

Van Beek and others (2000) provide a thorough and comprehensive review of the institutional characteristics of the Eastern Caribbean Union, and of recent economic developments in the area.¹⁶ They argue that the common currency arrangement has promoted fiscal discipline, despite the absence of fiscal harmonization targets.

Hoffmaister, Roldós, and Wickham (1998) show that external shocks seem to have a greater impact on CFA franc zone countries than on non-CFA countries in sub-Saharan Africa, and argue that this finding may reflect the absence of the exchange rate as a shock absorber for the CFA countries.¹⁷ Masson and Pattillo (2001) argue that a proposed monetary union in West Africa (ECOWAS) could create the temptation for fiscal profligacy through prospects of a bailout, or costs diluted through the membership; such a union could promote fiscal discipline only if the fiscal authorities are also tied by a strong set of fiscal restraints.¹⁸

Guillaume and Stasavage (2000) consider several past, present, and potential currency unions in sub-Saharan Africa, including the Rand Common Monetary Area. They argue that currency unions help improve policy credibility, but only under stringent conditions.¹⁹ On the basis of trade patterns and correlations of economic disturbances, Bayoumi and Ostry (1997) find little evidence that sub-Saharan countries would benefit in the future from larger currency unions.²⁰