In his paper on optimal currency areas, Robert Mundell (1961) proposed several conditions under which a monetary union between two or more regions would define a feasible regime of exhange rates. First, factors of production, capital and labour, should be highly mobile across regions. With one region facing an adverse demand shock, say, moving factors from that region would bring markets back into equilibrium. Second, macroeconomic shocks should be synchronized, i.e. a shock hitting one region should hit other regions in same direction. Third, nominal factor prices, including wages, should be flexible, in order to restore equilibrium in the factor markets after a shock. For a currency area to function smoothly, at least one of these conditons should be met.
This general proposition has been applied at the national level, bringing up the issue whether a particular economy is suitable for joining a larger currency area, of which the European Monetary Union (EMU) is a recent example. Several advantages have been listed from joining such a union. First, the reduction of transaction costs that are incurred in exchanging from one currency to another. Second, the elimination of exchange rate risk, reducing real interest rate uncertainty, and in turn, the accompanying premium. Third, more transparity in prices across countries. These effects are likely to improve market efficiency in general, enhancing economic activity and growth. Among the disadvantages from entering a monetary union are: first, a loss of a degree of freedom, the nominal exchange rate, as a means of reacting to macroeconomic shocks. Second, the surrender of national sovereignty, as the monetary authority no longer conducts an independent monetary policy.
Since the emergence of EMU, there has been an ongoing debate in Iceland whether or not to join, mostly in the general media. However, some scholarly contributions have also been made. Buiter (2000) compares the late 1990's regime of price stability (coupled with a fairly flexible exchange rate) adopted in Iceland and monetary union membership. He concludes that, on balance, neither regime overwhelmingly dominates the other. The lack of real factor mobility and nominal price rigidities would not be major obstacles to EMU membership, but the current arrangement might provide better macroeconomic stabilization in face of temporary real shocks. Gudmundsson et. al. (2000) conclude that Iceland does not meet the optimal currency area criteria for EMU-membership. Economic fluctuations are, according to them, rooted mainly in real supply shocks, largely uncorrelated with shocks in other economies. However, they discuss the limitations of these criteria, and conclude that the feasibility of joining the EMU remains an open issue. Finally, due to the nature of the fluctuations and the inherent difficulty of maintaining a unilateral stable exchange rate policy, they conclude that Iceland should adopt a floating exchange rate, so long as it keeps its own currency. Agnarsson et.al. (2000) conclude that adopting the euro currency would limit the Icelandic economy's ability to absorb adverse supply shocks, at least on the scale seen in the past. However, they point out the potential for more monetary discipline from EMU membership, which might contribute to more flexibility in the labour market.
This paper focuses on the suitability of alternative exchange rate regimes in Iceland, from the viewpoint of macroeconomic stability. To that end we develop a simple small open economy model that builds on the real business cycle (RBC) literature. Early examples of such modelling include Cardia (1991) and Mendoza (1991) (single consumer good), and Einarsson (1992) (two consumer goods). These studies focus on certain statistics pertaining to open economies, such as the current account and the correlation between savings and investment, under some given exchange rate regime, typically fixed rates. Among more recent examples is Fernandez and Kehoe (2000), who develope a model with one traded and one nontraded good to examine the effects of Spain's abolition of capital controls in the late 1980's and early 1990's. Cooley and Quadranini (2001) develop a model with one home produced and one imported intermediate good and compare the welfare implications of adopting the U.S. dollar in the Mexican economy to monetary independence.
They conclude that a 'dollarization' may not be Pareto superior to an independent monetary policy. Mendoza (2001) formulates a model with a traded and a nontraded good, with households facing a potential borrowing constraint in an international bond market. He concludes that 'dollarization' may entail significant welfare benefits for emerging economies such as Mexico's: first, by eliminating price and wealth distortions induced by the lack of credible stabilization policies; second, by improving the efficiency of financial markets by reducing frictions originating in information structure or institutions. Schmitt-Grohé and Uribe (2001) consider a small open economy model with households who have access to an international bond market. Two types of goods are produced, exportables and nontradeables, and three goods are absorbed, exportables, importables, and nontradeables. Two sources of nominal rigidities are assumed: 'sticky' prices of nontradeables, and transaction costs that are decreasing in households' money balances. Calibrating the model to Mexican quarterly data, it is found that 'dollarization' is slightly Pareto inferior to the alternative regimes examined, including constant money growth and inflation targeting.
