by C. Fred Bergsten, Peterson Institute for International Economics
Testimony before the Joint Hearing of the Subcommittee on Economic Policy and the Subcommittee on International Trade and Finance
Committee on Banking, Housing, and Urban Affairs
United States Senate
April 22, 1999
Dollarization of foreign economies is a fairly widespread phenomenon. A recent IMF compilation indicated that foreign currency (mainly dollar1) deposits in 1998:
In fact, ratios of 30-60 percent prevailed in most transition economies in Eastern Europe and the former Soviet Union during 1990-95 and were prevalent in Latin America as well. Most of the countries are very small; the largest are Turkey (with a foreign currency share of 46%), Argentina (44%), Russia, Greece, Poland and the Philippines (all about 20%). Mexico is only at 7%. Bolivia had the highest ratio at 82%. The only fully dollarized countries remain Panama and Liberia.
From a United States standpoint, it is noteworthy that about two-thirds of all dollar currency is held outside the United States. About three quarters of recent increases in such cash holdings is accumulated beyond our borders. US exports of dollar currency totaled $44 billion to Russia and $35 billion to Argentina alone during 1989-96.3
It is important at the outset to distinguish between de facto dollarization and policy dollarization. The former occurs as residents of foreign countries seek refuge in the dollar (or other strong currencies) to hedge against inflation, default or confiscation by their own governments. One new phenomenon in the 1990s is that de facto dollarization has occurred beyond those risks, at least in their most obvious manifestation, in at least some parts of the world.
This has led in turn to a view, in some quarters, that the benefits of maintaining a national currency are outweighed by the benefits of adopting someone else's. Hence Argentina, and a few other countries to lesser extents, have begun to consider policy dollarization. Europe's creation of the euro has also sparked interest in the possible payoff from explicit abandonment of monetary sovereignty (albeit in that case for a common currency rather than someone else's).
The Policy Choice
Countries have three basic choices in determining the monetary linkage between their economy and the rest of the world, assuming that they maintain a currency of their own as most do:
There is increasing intellectual and policy consensus that "fixed but adjustable" pegs, the traditional means of "fixing" the exchange rate, do not work well for emerging market economies, the focus of this hearing.4 Hence they must either float to some extensive degree or fix convincingly. Both courses have clear costs as well as benefits.
Floating permits a country to maintain a degree of national control over its monetary policy (and other national macroeconomic policies) because it does not have to use them to defend the exchange rate. However, markets can substantially overshoot the economic fundamentals; they can thus push a currency far below its underlying economic value, generating inflation and large debt servicing costs, or far above that level, hurting the country's competitiveness and throwing its trade balance into large deficit.
Fixed exchange rates can avoid those costs if the authorities can successfully set the rate at a sustainable level and convince the markets of both their ability and will to keep it there. Moreover, fixed rates reduce the transactions costs of international trade and investment. In addition, a fixed rate can provide a useful anchor for price stability by linking a small country to the world economy (and especially to a large country with relatively stable prices, like the United States or Germany).
However, governments may set or try to sustain a rate at unsustainable levels and private capital flows will eventually then force devaluations that can be extremely costly. Successful defense of a fixed rate can often be costly too, requiring a country to raise interest rates and/or otherwise slow its economy to avoid speculative attack. There is thus a clear trend away from fixed rates in emerging market economies, especially in the wake of the Asian/global crisis where a number of relatively fixed-rate countries were forced to abandon their pegs and countries with various types of floating rates generally fared better.
In the present world economy of huge private capital flows, there is in fact a growing consensus that countries can enjoy the benefits of fixed exchange rates only in two ways: by maintaining extensive capital controls, to directly limit their vulnerability to private speculation, or by adopting such a credible commitment to fixity that the markets will believe them through thick and thin and desist from speculative attack. Capital controls obviously reduce a country's access to some types of foreign investment but there is increasing interest in that approach and it has been adopted by a few countries (Chile, Colombia, Malaysia recently and, in different forms, China and India).
If a country wants to avoid the costs of both currency flexibility and capital controls, and therefore seeks credibility for its fixed exchange rate, one approach is to install a currency board as in Argentina since 1991 and Hong Kong since 1983. Under that approach, a country only issues local currency that is fully backed by a foreign currency (the dollar in both these cases) at a fixed rate. Such an arrangement proscribes the national authorities from changing the rate or conducting an autonomous monetary policy. The currency board thus substitutes for a central bank (which inter alia means that the country will probably not have a lender of last resort to respond to internal financial crises, a point to which I return below).
Even a currency board may not achieve full credibility, however. The country could change the laws by which the board was established, or at least the ironclad rules under which it is supposed to operate. Even Argentina, whose currency board has achieved considerable credibility, still experiences interest rate increases of about 3 percent whenever US interest rates rise by 1 percent—whereas dollarized Panama experiences a rise of less than 0.5 percent.5 Hence there has developed considerable interest in countries such as Argentina, Mexico and even Canada for adopting the dollar as their national currency.
Benefits and Costs
The case for dollarization is essentially the same as the case for currency boards, the gold standard in bygone days, or any system of irrevocably fixed exchange rates as outlined above. Hence it makes sense for two types of countries6:
Relatively few countries meet these criteria. Certain countries for which currency boards or dollarization have recently been proposed—Brazil, Indonesia, Mexico, Russia—clearly do not. Hence I believe there is a very limited universe of countries where this approach would make sense.
However, one additional criterion has recently been suggested for dollarization: close economic links between a country and the United States. The argument is that such a country, in addition to enhancing its monetary stability, could achieve significant savings in transaction costs because of those extensive trade and investment links. For example, studies show that transactions between Canadian cities are 20 times those between similar Canadian and American cities despite ten years of a free trade area and the geographic propinquity, common language and culture of the two countries; the main barrier may be the different currencies and the fluctuating exchange rate between them. This is one of the main economic arguments for creating the euro as a common currency for countries that have already developed extensive trade and other economic ties in Europe.
Hence dollarization could lead to a further increase in trade and investment between countries that are already integrated to a substantial degree. The associated costs would be reduced if there are already relatively free flows of capital and labor between the country and the United States, since such flows represent the alternative mechanisms of adjusting to economic conditions that affect the two economies differentially.
On this criterion, Mexico and Canada—the NAFTA trading partners of the United States—might seriously consider dollarization. About one-third of total Canadian output, and about one-fifth of all Mexican production, is exported to the United States. (By contrast, the ratio is less than 1 percent in Argentina or Brazil. Dollarization would not shield Argentina from the substantial impact of devaluation by Brazil, by far its largest trading partner.) Hence full financial integration with the United States, as the European Union has chosen via the euro in part because of the members' extensive trade ties, has some appeal—along with the presumed advantages of greater price stability and lower interest rates, as noted above and as achieved already by some of the less industrialized members of Euroland (especially Italy, Portugal and Spain).
The Bottom Line: A "Currency Board Plus"
Dollarization can thus be conceptualized as a "currency board plus,"7 in three senses. First, it imparts even greater credibility to a country's commitment to forever renounce the devaluation option. Second, it assures that the country will import stable prices and presumably lower interest rates from the United States. Third, it minimizes transaction costs and promotes further long-term integration with the US economy.
All this of course comes with a price. The country gives up two major policy instruments, which have traditionally been viewed as integral elements of national sovereignty: monetary policy and the exchange rate. In addition, by abolishing its central bank, it abolishes its lender of last resort so has no agent to respond to a domestic financial crisis (which is still possible, despite dollarization, if the domestic banking system is unsound). The absence of a central bank also eliminates the normal supervisor and regulator of the financial system, although a separate agency can be created for that purpose.
This last consideration raises the possibility that dollarization can become a "currency board plus plus." The second "plus" can result from adopting the Federal Reserve as supervisor/regulator of the banking system, either by negotiation with the Fed itself (which is unlikely, although Argentina has raised the issue) or by turning all or most of the financial system over to American banks. At the cost of a further reduction of sovereignty, this latter approach takes advantage of the Fed's authority over the parents of the local banks to ensure their safety and stability—thus, at a stroke, overcoming all the problems of banking weakness that have been so central to the Asian/global crisis and so endemic throughout the developing (and developed) world for the past generation. Argentina has already gone fairly far in this respect (as have a few other countries).
Given the tradeoffs involved, it bears repeating that, in my judgement, dollarization makes sense only for countries in one (or some combination) of three positions: a very small and open economy that has no real autonomy over its exchange rate anyway, a country desperate to overcome a legacy of hyperinflation, or a country that is already deeply integrated with the United States. All others should continue to seek alternative exchange-rate systems more suited to their positions as relatively closed economies that are relatively noninflationary and not tightly linked to the United States.
Implications for the United States
The implications of dollarization for the United States are very modest if the phenomenon is limited, as at present, to a set of relatively small and relatively troubled economies whose citizens are using its currency as a refuge. Foreign currency holdings of perhaps $300 billion represent only about 5 percent of gross US external liabilities of more than $6 trillion (and only about 20 percent of our net foreign debt of about $1.5 trillion). Being "currency in circulation," this portion of the "dollar overhang" is even less likely than other parts of our debt to be converted precipitously in the case of some sharp deterioration in US economic or policy performance.
One modest cost for the United States could be the co-optation of Fed supervisory authority by a country that turned its financial system over to US banks as well as dollarized. However, the Fed already regulates those banks and their expanded foreign role would add only marginally to its responsibilities.
The situation would of course become different if a large part of the world, or even a large part of the neighborhood (say, Canada and Mexico), made a policy decision to dollarize. Economic developments in those areas would then have a considerably greater impact on overall monetary conditions in the United States itself, requiring greater consideration thereof in the conduct of US monetary policy. At some point, we would have to contemplate accepting those countries as new Federal Reserve districts and giving them seats on the Fed's Open Market Committee, and even devising some form of fiscal federalism to provide an alternative adjustment device to compensate for the abolition of the exchange rate instrument between them and us.
There would be both costs and benefits to the United States from such a far-reaching systemic change. The United States could gain substantial seigniorage from much wider use of its currency by nonresidents. The implied increase in financial stability in the region would be positive for the United States.8
On the other hand, the United States would then have to participate more directly in the adjustment of the dollarized foreign economies to economic dislocation. In essence, we would either have to send more capital to them or accept more workers from them in case they ran into a recession. This would not be totally different from the current situation, where the United States already takes some of the impact of adjustment vis-à-vis neighboring countries via the exchange rate and the trade balance. However, the distribution of the effects would be different and the transparency (and hence political awareness) of the impact on the United States could become much greater. (This is the same problem that Germany has faced vis-à-vis the rest of the European Union, incidentally, as it agreed to formalize its de facto "DM zone" via first the European Monetary System and now the euro.)
The only plausible scenario for widespread policy dollarization is widespread economic integration in part, or all, of the Western Hemisphere. Europe created the euro at the culmination of fifty years of economic integration of its continent. The existing NAFTA, or its evolution into a Free Trade Area of the Americas as pledged by the Miami summit of 1994, could lead in a similar direction over a similarly long period of time. Dollarization would be much more likely to emerge at the end of the process, as in Europe, rather than at its outset.9
The correct policy for the United States at this time is thus (1)benign neglect, or even a cautious helping hand, for smaller individual countries that might want to dollarize in the near term to promote their domestic monetary stability but (2) insistence on thorough discussions, covering both the substantive and institutional implications of the move, if any large number of larger countries were to develop serious intentions of moving in that direction. The latter scenario is clearly some (perhaps an infinite) distance away, given the cost-benefit calculation for most countries described above, even if Western Hemispheric economic integration comes to appear more likely than at present. In either case, there will be plenty of time to consider the idea carefully before deciding on a course of action.
3. Data provided by Financial Crimes Enforcement Network, US Department of the Treasury, as recorded in the Customs Service Currency and Monetary Instruments Reports (CMIR) forms and reported in IMF, op. cit., p. 10.
7. A country that already has a currency board could move to dollarization very quickly. The details for Argentina are laid out in "Monetary Union for the Americas," J.P. Morgan Economic Research Note, February 12, 1999, p. 10. That note also describes the prospects for dollarization in Canada, Mexico, Brazil and the Andean Pact countries.
8. This factor leads to a proposal for active US promotion of dollarization throughout the Hemisphere, including through a one-time allotment of dollar bills and lines of credit to members, "particularly less reliable countries such as Brazil and Mexico", …"to compete with the recently established euro area" in Robert J. Barro, "Let the Dollar Reign from Seattle to Santiago," Wall Street Journal, March 8, 1999.
9. Moody's has indicated that, in the advent of monetary union with the United States, "Argentina's country ceiling rating might rise to Aaa (from its current Ba3) as was the case with members of the European Monetary Union." See "Moody's Views on Argentina's Dollarization," March 1999.