May 19, 2000
« More Economic Letters
Dollarization as a Technology Import
Alan M. Taylor
The debate over dollarization has arisen in several countries, but it is often at its most fervent in Argentina. In that country, a decade of currency board experience with dollar-peso convertibility has brought the economy as close to being dollarized as one can be without going all the way. But not close enough, doubters say. Argentina remains susceptible to the shocks of global financial crises because the Argentine peso is not fully credible. These shocks result in widening bond spreads, which leave the economy with inefficiently high credit costs, slow investment, and slow economic growth. Thus, hard-core dollarizers advocate nothing less than the abolition of the domestic currency and the substitution of dollars for pesos.
This Economic Letter argues that it is informative to see the debate over the costs and benefits of dollarizing in light of both historical experience and an understanding of money as a technology that provides services.
Governments can finance their expenditures by levying taxes, by borrowing from the public, and by printing money. In nineteenth-century post-independence Argentina, the first two of these avenues were essentially shut off: the tax base was exhausted by crises and wars, and domestic and foreign credit often was unavailable. So printing money to finance expenditures was the norm from the 1820s to the 1880s. During this period of monetary anarchy, the state and provincial banks acted as money printers to their respective governments.
The revenue generated by printing fiat token money—fiat because it is money by state decree, and token because it is not based on a commodity with an intrinsic value—is called seigniorage. It amounts to the difference between the actual cost of printing the money and the face value of the money. When a government tries to extract too much seigniorage to cover its expenditures, the result is inevitably inflation and a debasement of the currency. And that is exactly what happened in Argentina.
To try to stabilize the price level, a convertibility commitment was made in 1881. This commitment essentially tied the value of the peso to gold by backing the peso with gold reserves. But it proved short-lived. The banks continued to overissue money, and gold reserve losses soon mounted. After 1884, a temporary “dirty float” of the paper-gold exchange rate was announced, but the gold drain continued, aided by a continued policy of sterilization whereby the government note issue was not contracted despite an ongoing gold drain. Attempts to fix the banking laws were ineffective, even counterproductive. The fiasco ended in a classic speculative attack with no gold left in the vaults whatsoever.
By 1890 a major macroeconomic and financial meltdown was underway, the so-called Baring Crisis. Observers of recent turmoil in global financial markets will find that it involved some very familiar ingredients: persistent budget deficits, an unsustainable commitment to a fixed exchange rate peg, and problems with currency and maturity mismatches—banks that borrowed short abroad in gold and lent long domestically in pesos. It was, in essence, the first modern emerging market crisis.
During the 1880s, Argentines resisted the increasingly burdensome inflation tax and shifted dramatically from paper money to gold. In the beginning of the decade, currency in the hands of the Argentine public was almost 100% paper pesos; as the value of the peso depreciated rapidly, people shifted to gold, holding as much as 80% of their currency in the metal during the thick of the crisis. Such a phenomenon of currency substitution is now very familiar, given the frequent postwar dollarization of Latin American economies during hyperinflations—as we recently saw in Ecuador, a country now embarking on de jure dollarization. The Baring Crisis only stands out as one of the earliest such episodes.
One clear lesson emerges from this history. Despite the legal tender status of fiat token money, the public can, through uncoordinated private action, spontaneously substitute away from it and towards an alternative money if the costs associated with fiat money get too high—that is, the people can effect a de facto currency substitution if the government attempts to extract too much seigniorage.
As we have noted, the temptation to abuse seigniorage power has gotten out of hand in not a few countries, including Argentina. Thus, in 1991, in an act of political will, Argentina adopted the so-called Convertibility Plan. Now, with small exceptions, the only securities allowed on the central bank’s balance sheet are safe U.S. dollar-denominated foreign assets. In other words, instead of tying the peso to gold, the government is tying the peso to the U.S. dollar.
An important consequence of such a plan is that the central bank cannot use open market operations to absorb domestic government debt paid for with newly printed money, the true seigniorage. Thus, the Argentine central bank cannot use open market operations to expand the balance sheet and control the money supply.
The accounting implications are clear. The government has lost its seigniorage revenue because it has been limited from putting its own interest-bearing liabilities onto the bank’s balance sheet. However, the central bank can engage in some maturity transformation by holding interest-bearing dollar assets. From this activity, a modest profit arises.
This is akin to the profits in a private bank. But the analogy is weak. If a private bank held 100% reserves in low-risk foreign securities, only a small profit would obtain. Such a low-risk bank (termed a narrow bank) could not survive competition. The Argentine central bank can survive as a public narrow bank because it has a unique product protected by a legal, politically determined monopoly—the issue of base money.
The Argentine public can threaten those profits—and the existence of the central bank itself—by dollarizing, either de facto or de jure. De facto dollarization could come in a crisis if the Argentine public, by individual choice, swapped pesos for dollars in an act of decentralized currency substitution. If there is no crisis large enough to wipe out base money from circulation, such individual action is unlikely, and the peso will circulate and some central bank profits will result. De jure dollarization would mean that the Argentine public works through the political process to effect a smooth and coordinated currency substitution toward the dollar. A collective choice to dollarize might be seen as a preemptive move to rule out the possibility of a future chaotic decentralized switch. Given the network externality properties of money, which we discuss next, dollarization via collective political action might be the only way for such a change to occur in normal times.
The best—perhaps the only—argument for the public regulation of money rests on the coordination question: Is it possible for a suboptimal money to establish itself as a standard when “better” monies exist? Yes, if the use of a certain money as a medium of exchange exhibits path dependence—that is, if its current market dominance derives not from its intrinsic superiority but from historically determined circumstances.
How do these circumstances arise? Think of money as software, which—in a very real sense—it is. This is more than just a reference to the recent rise of quasi-monies, and even e-money, as a substitute for cold hard cash. Money is a type of technology. In this sense, it has many properties in common with software. Excluding commodity money, most money has a trivial intrinsic value, but has value for the transaction services it facilitates. Likewise, one can “print” software as easily as money onto very cheap disks that cost a fraction of the value of the services rendered by the software itself.
But more important, money as technology exhibits network externalities. Specifically, if my benefit from using a technology depends on the number of other users who have adopted it, then one firm can soon dominate the market. Once everyone else has chosen a certain technology as a de facto standard, then I have little incentive to choose a less popular, albeit better, technology. This idea of a network externality, or technological lock-in, has been invoked to explain, say, the triumph of VHS over Betamax in video, or the victory of Microsoft over everybody in software. Thus, just as certain software protocols have become established standards, our custom of using a certain money as a medium of exchange makes that money a standard—a protocol for exchange transactions in the economy.
Can the market ever coordinate a switch from a “worse” to a “better” money technology? Yes, but only under conditions of economic chaos, as in recent Ecuadorean experience or in historical panics like the Baring Crisis, when the costs of using the incumbent money rise too high. A currency switch is then a de facto outcome, where custom overrides the law and establishes a new monetary standard.
The Argentines are shopping for a new money. At present, they are merely window shopping, and the choice is limited. In the “marketplace” of monies, the choices are just the national monies of other countries. While there may be sufficient choice to offer a “better” technology, would it be worth the price? And what is the price?
If the Argentine government dollarizes—that is, converts its dollar bonds to dollar cash, thereby losing its revenue stream, and then exchanges the dollars for the pesos in circulation—this looks, on the surface, like a pure fiscal loss to Argentina and a transfer to the U.S. Dollarization will not happen, some say, unless the U.S. agrees to rebate some or all of this lost revenue when countries make the switch to a dollarized money base.
We can look at it a different way, however. What if the public takes the view that the government’s balance sheet is their own balance sheet, too, since households are ultimately responsible for the government’s fiscal position? In that case, the interest stream from the bonds is not really the government’s to lose, because the people “own” the government. Suppose the public then demands all central bank assets with their future interest stream, the full opportunity cost of their pesos. On the consolidated national balance sheet, if we transfer the assets this way there is no loss of national welfare; it is a purely internal transfer.
What now? Households would have interest-yielding illiquid dollar securities, but no cash. For transactions purposes, a cash would need to be agreed upon and obtained. To obtain cash they then have to shop for money, as it were, at the global “money store” of currencies. They must pay for the cash with the securities.
In contrast to most consumer scenarios, the Argentines in this thought experiment would be going shopping not with money, but for money. They forgo interest to have a liquid cash instrument. But this is so for all consumers everywhere. The difference here is that they have imported this monetary technology and are buying it from a foreign source. Thus the interest forgone—the price of the money—has been paid to the exporter—the U.S., if they select dollars—instead of to the Argentine government. The only novel element here is that we must now augment our concept of international financial accounts to embrace the notion that “money services” may be traded—and that only some central banks (maybe only one or two?) have a “comparative advantage” in producing them.
If Argentina dollarizes, their money would be denationalized. This is akin to privatizing a state industry. Frequently, the rationale for privatizing a state industry is that it is a drain on the national coffers, both public and private. However, with dollarization we encounter the seemingly odd spectacle of a government being asked to shut the doors on a profit-making business!
Just as consumers “own” the government bonds at the central bank that back the money issue, they also “own” the money monopoly profits—the fiscal revenues—that derive from the assets. Such is the downside of dollarization. Still, given the credibility problem, the public might reasonably feel that the Argentine government ought to get out of the money monopoly business. The government may not agree, because fiat money is an easy way to raise revenue—perhaps too easy, for, when in the wrong hands, even fiat money can have its costs, and better alternatives may exist in the “world market.”
Different national monies offer competing services, provided that national legal barriers are lowered and policy coordination is available to counter lock-in and make a switch. In such a market for “money technology” only monies of the highest quality are likely to survive. Consumers will be willing to switch to those technologies, even for a price.
To coin a phrase: when it comes to money, you get what you pay for.
Alan M. Taylor
More Economic Letters
Opinions expressed in FRBSF Economic Letter do not necessarily reflect the views of the management of the Federal Reserve Bank of San Francisco or of the Board of Governors of the Federal Reserve System. This publication is edited by Sam Zuckerman and Anita Todd. Permission to reprint must be obtained in writing.
Please send editorial comments and requests for reprint permission to