Some theory and history of dollarization.

• Author: Schuler, Kurt

The unfamiliar, however simple, is often hard to understand because it requires a slightly different way of thinking. Official dollarization is a case in point. It is hard to conceive of a simpler monetary system than using somebody else's currency. There is no central bank, no independent exchange rate, and more generally no independent monetary policy. Yet precisely because most countries have central banks, people often think about dollarization by using a frame of reference derived from central banking. In this article, I make a few points about the theory and history of dollarization that I hope will provide a better understanding of the system.

All "Fixed" Exchange Rates Are Not Alike

About 10 years ago, a consensus began to develop among economists that the extremes of fixed and floating exchange rates were less prone to currency crises than intermediate exchange rates. This so-called bipolar view gained adherents after currency crises in East Asia, Russia, and Brazil in the late 1990s (Fischer 2001). The crises had severe effects on countries that had officially or unofficially linked their currencies to the U.S. dollar, neither letting them float without government intervention nor tying them so tightly to the dollar as to forgo independent monetary policy. Although the bipolar view still has adherents, Argentina's spectacular economic depression and currency crisis of 2001-02 led many observers to conclude that fixed exchange rates are more prone to crises than floating rates, so floating rates are the only consistent policy (Feldstein 2002: 14). One might call the result the unipolar view.

Before hastily agreeing with this potted summary of recent monetary history, it is worth thinking about whether all "fixed" exchange rates are really alike. Panama has no central bank, no locally issued paper money (people use U.S. dollars instead), and no exchange controls restricting trade in foreign currencies. (1) Jordan's central bank maintains an officially acknowledged exchange rate of 0.709 Jordanian dinar per U.S. dollar, and imposes no exchange controls that significantly hinder trading in foreign currency. China's central bank has maintained an unofficial but in practice rigid exchange rate of 8.28 yuan per U.S. dollar since May 1995. China has extensive exchange controls, but an active foreign exchange market exists, and most observers consider that if anything, the currency is undervalued. Cuba's central bank maintains an official exchange rate of one peso per U.S. dollar. Cuba has comprehensive exchange controls and no market where private parties can trade large amounts of foreign exchange legally; as of November 2004, the exchange rate in the black market is about 26 pesos per dollar. Are these exchange rate arrangements enough alike that we should lump them together under the same term the way textbook treatments and more advanced discussions alike often do?

I argue that the answer is no. The four exchange rate arrangements mentioned differ in two dimensions: convertibility and monetary sterilization. Convertibility is the ability to exchange domestic currency for foreign currency without restrictions. Sterilization, also known as sterilized intervention or neutralization, is action by the monetary authority to offset the effect of changes in demand for foreign currency on the supply of the domestic monetary base, or vice versa. (2)

Panama has full convertibility and no sterilization. Jordan has full convertibility and sterilization. China has limited convertibility and sterilization. Cuba has no convertibility; it also has no sterilization because as a centrally planned economy it has no officially tolerated financial markets in which its central bank could engage in sterilization.

Discussions of exchange rate arrangements frequently neglect exchange controls and their effects on the workings of the money supply. The effects can be dramatic, though, as big as the differences between Panama and Cuba. In Panama, if the public wishes to hold more "domestic" currency, it acquires more U.S. dollars without help or hindrance from the Panamanian government. In Cuba, if the public wishes to hold more domestic currency, there is no direct connection with events in the foreign exchange market, nor is there any system of rapid feedback to ensure that the central bank supplies more currency in the short term. (Over the long term, though, the central bank has oversupplied the Cuban peso, which is why it is so depreciated on the black market.) In the extreme, exchange controls totally isolate the domestic financial system from world financial markets.

Exchange controls were common from the 1930s until the early 1990s in rich countries. They remain common in poor countries, although for more than 20 years, the trend has been for controls to diminish. Having noted their prevalence, though, let us set them aside for a moment, because the matters in the next several paragraphs are simplest to think about where no exchange controls exist.

Along with economists such as Milton Friedman and Robert Mundell, I use...