Dr. Glasner, an economist with the Federal Trade Commission, is the author of Free Banking and Monetary Reform (Cambridge University Press, 1989) and the editor of The Encyclopedia of Business Cycles, Panics, Crises, and Depressions (Garland, forthcoming 1995). The views expressed in this article do not necessarily reflect those of the Federal Trade Commission or of individual Commissioners.
About five years ago I published a book, Free Banking and Monetary Reform, that proposed a radical reform of our monetary system. Competing banks, I argued, should be free to supply any monetary instrument, including currency or banknotes, while the government would perform the limited but vital function of establishing a currency unit (e.g., the dollar) in terms of which privately supplied monetary instruments could be defined. And to ensure optimal stability of the purchasing power of the currency unit, I proposed a mechanism of indirect convertibility tied to a price (or, preferably, a wage) index.
I have been disappointed but not surprised to detect no groundswell of popular support either for free banking or for any of my specific proposals. I do not believe that this lack of enthusiasm betrays any shortcomings with free banking or my proposals. What the indifference to free banking reflects is rather a salutary, “if it ain’t broke, don’t fix it” sort of conservatism. As long as inflation remains low and the banking system is not collapsing, practical people will not undertake the effort required to effect a reform of this magnitude. The potential benefit from such a reform is not big enough to outweigh the perceived risk in trading the monetary system we know for one we don’t. I have therefore concluded that my proposals for free banking are less relevant for the United States and other developed countries with stable monetary systems than for less developed and former Eastern Bloc countries now lacking the monetary stability necessary for economic development. Without secure monetary institutions, these countries have far less to lose than do advanced countries by experimenting with free banking. Nor, for reasons that will become apparent, can less developed countries simply copy the monetary systems of the advanced ones. Free banking is, therefore, ideally suited for overcoming the systemic problems that now frustrate the attempts of less developed countries to achieve monetary stability.
To understand why free banking is so well suited to the circumstances of less developed and former Eastern Bloc countries, we must first consider how money and banking can contribute to economic development. The role of money is familiar and obvious—it is a medium of exchange. Money facilitates exchange by allowing us to trade without having to identify, as we do in barter transactions, a double coincidence of wants. Reducing barriers to trade promotes economic progress by allowing resources to be shifted from less to more valued uses. Such shifts create new opportunities and new demands for resources, triggering an upward spiral of output and wealth.
It is worth observing that the capacity of money to perform this extraordinarily valuable social function poses something of a puzzle. The existence of an instrument that serves only as a medium of exchange, providing no real services, seems to contradict the usual assumption of economists that self-interest motivates economic decisions. Why do people accept money, which (despite its social utility) has no direct use for them individually, in exchange for real commodities or services that do have direct use for them? The acceptability of money is sometimes attributed to an implicit understanding among people to act in the common good rather than pursue selfish goals or to a command by the sovereign imposed through legal-tender laws. But neither recognition of the common interest in having a medium of exchange nor laws commanding that an instrument be accepted as legal tender could make people use as money an instrument that they would not have otherwise, in their narrow self-interest, chosen so to use.
Self-Interest and Exchange
How then does self-interest cause anyone to accept a money that has no use except to be exchanged for something else? Whether it is in my self-interest to accept money in exchange for real goods and services depends critically on whether I expect other people to accept money in exchange for real goods and services. If I expect other people to refuse money that I offer in exchange for their goods and services, then my self-interest is to refuse money in exchange, too. But if I expect other people to accept money that I offer in exchange for their goods and services, then my self-interest may dictate accepting money in exchange for the goods and services that I supply, because doing so may allow me more easily to sell what I want to sell and more easily to buy what I want to buy than if I try to barter. The less confident I am that it will retain its value, the less willing I shall be to accept it in exchange. So whether money is acceptable is a matter of degree, not a simple yes or no question.
It is, at any rate, clear that money cannot function well as a medium of exchange unless people are confident that it will be acceptable at roughly its current value in the future. Whatever undermines people’s confidence or trust in the future value of money threatens its capacity to serve as a medium of exchange. The delicate web of mutually supporting expectations that allows a medium of exchange to function can easily unravel or collapse if the trust underlying those expectations is eroded or betrayed.
In primitive conditions, the medium-of-exchange role of money can be performed without the aid of banks. Money could circulate hand-to-hand, either in the form of precious metals, coins, or currency (convertible or fiat) issued by the state. However, the transfer of deposits within or between banks through checks (and now electronically) is an exceptionally efficient way to convey money in trade. To engage in monetary exchange through banks, people must hold deposits with them. By holding bank deposits instead of some other form of money or wealth, people lend banks capital which the banks then lend to borrowers (who typically borrow to finance investment, not consumption). Thus, by providing a convenient way for the public to hold money and execute transactions, banks channel the savings represented by the public’s deposits to investors. As intermediaries between ultimate savers and ultimate borrowers, banks increase the return to savers from savings and reduce the cost to borrowers of borrowing, promoting economic development within the areas they serve.
Creating and Maintaining Confidence
Having considered how money and banking promote economic development, we can now ask which institutions will support a stable system of money and banking. Since money cannot function well as a medium of exchange unless people have confidence in its future value, the fundamental task of monetary institutions is to create and maintain that confidence. How can such confidence be created and maintained? The answer for a private supplier of money, i.e., a competitive bank, is very different from the answer for a state that supplies money. And it is on that difference that I am going to rest the case for free banking as the solution for chronic monetary instability in less developed and former Eastern Bloc countries.
Why does it matter whether money is supplied privately or by the state? When a private bank creates money, it does so by issuing a special type of IOU against itself. The IOU allows the owner of the IOU or anyone he assigns to demand its instant redemption in terms of a fixed amount of a specified asset. For example, when Citibank creates a demand deposit, it is promising to redeem that deposit in terms of an equivalent amount of U.S. currency to the depositor or to anyone to whom the depositor writes a check up to the amount of the deposit.
A bank’s contractual obligation to redeem its IOUs on demand does not automatically create the confidence in their future value required for them to function as money. If the bank is widely expected to default on its IOUs, those IOUs, regardless of the bank’s net worth or financial soundness, will not function as money, because IOUs that people do not expect to be honored will be unacceptable in exchange. For a bank to create confidence that it will continue redeeming its IOUs, it must convince people that it would lose more by defaulting than it would gain. Whether people will trust banks with a substantial net worth to honor their contractual obligations depends in large part on the legal consequences of default for the bank. If the legal system under which banks operate strictly enforces contractual obligations and penalizes default, default will appear unlikely.
One might question whether, if they were not legally required to make their moneys convertible into an asset whose supply or value they could not control (e.g., currency or gold), private banks would voluntarily obligate themselves to redeem their IOUs on demand in terms of such an asset. But no private bank has ever issued irredeemable money without a state edict declaring the money legal tender and acceptable for discharging tax liabilities. Moreover, the theoretical argument denying that a private bank can issue irredeemable money is compelling. The magic a bank performs by creating money is to impart a value to something whose only use consists in being valuable. The bank performs this conjurer’s trick by legally committing itself to redeem instantly its IOUs in terms of another asset whose value it cannot control. That credible promise allows the bank’s IOU to take on a value identical to that of the redemption asset. But without the promise, people, recognizing the potential profit from issuing valuable IOUs and redeeming them for nothing, would never place a value on such IOUs any higher than their expected final redemption value, namely, zero. Inconvertible bank IOUs must be worthless.
Inside and Outside Money
It may be helpful to distinguish here between inside and outside money. The money private banks supply is called inside money because it represents a debt the bank creates against itself. Outside money can be a physical commodity (e.g., gold) that has become acceptable as a medium of exchange, or a fiat currency issued by the state that has become acceptable as a medium of exchange. Like gold, outside money is an asset without being anyone’s liability. Private banks cannot create outside money; they can only create inside money which, to make acceptable, they promise to convert on demand into some outside money.
A sovereign may choose to issue inside money by committing itself, like a private bank, to convert its money on demand into some asset whose supply and value are beyond its control, or to issue outside money in the form of fiat currency. Should it do the former, its IOUs are apt to be less acceptable than those of a private issuer for one very powerful reason: while the default by a private bank on its obligation to redeem its IOUs triggers immediate insolvency, allowing creditors to seize its assets, a sovereign is immune from such sanctions when it defaults. Indeed, a sovereign’s default, isn’t even called a default, but a devaluation. People would therefore have much better reason to expect a sovereign to default on its obligations than to expect a private bank to do so.
Should the sovereign seek to issue an inconvertible fiat currency, it faces credibility problems of a different sort. Sovereigns have, indeed, successfully issued fiat currencies in numerous instances, so fiat currencies can maintain their value for long periods of time. However, there have been more unsuccessful than successful fiat currencies. So it certainly is not the case that a sovereign, just by declaring a fiat currency legal tender, can ensure its acceptability in exchange.
A sovereign wishing to issue a fiat currency must overcome two problems. First, how can it create a demand to hold the currency it is issuing? Why should anyone sacrifice any real goods or services just to hold pieces of paper that have no use other than to be exchanged for something else? For the pieces of paper to be used as money, people must want them enough to sacrifice something else of value for them. To assume that such pieces of paper have value because they are money begs the question why people accept them as money in the first place.
Second, a positive demand to hold an asset clearly does not ensure its acceptance as money. As I explained above, for an asset to be accepted as money, people must share expectations about the stability of its future value. So even if the sovereign can create a demand for fiat currency, how does it create sufficient consensus among the public about the currency’s future value for people to use it as money?
Creating Demand for Fiat Money
Let us consider first how the State can make its currency more valuable than the paper it’s written on. A legal-tender law that requires flat currency to be accepted in the discharge of debts doesn’t, by itself, impart any value to the fiat currency. It simply provides a way for some people to discharge debts that they previously incurred but does not compel anyone to accept legal tender in exchange for real goods and services. So why would anyone prospectively supply something of value in exchange for fiat currency? If what the State declares legal tender is not acceptable, people can avoid accepting debt instruments that could be discharged by the proffer of legal tender.
A more powerful way to create a demand to hold fiat money is for the sovereign to require its currency to be used in discharging tax liabilities. If the public owes the sovereign enough at certain times of the year (e.g., April 15), then the demand to hold currency may be sufficient (even during periods of zero or negative net tax liability) to give the currency positive value throughout the year.
But though feasible, such a strategy may still be impractical, which raises our second problem. Even if government tax certificates have a positive value, what would create a sufficient consensus about the expected future value of these tax certificates for them to serve as money? Without such consensus, a positive value will not enable them to serve as money. Indeed, there are few if any historical instances in which a fiat currency was successfully introduced without its first having been made convertible into another money. Thus, although requiring taxes to be paid using a fiat currency seems to be necessary to prop up its value after convertibility has been suspended, acceptability as payment for taxes may not suffice to enable a sovereign to create a new fiat currency.
To sum up: A private issuer of inside money has more credibility than a sovereign issuer of inside money, because people generally understand that a sovereign has less to lose than a private bank by reneging on a convertibility commitment. A defaulting bank forfeits its assets to its creditors while a devaluing sovereign forfeits only its reputation. On the other hand, a private bank cannot issue outside money, while a sovereign can. But a sovereign’s capacity to issue a fiat currency without first making it convertible into some other money or asset may be quite limited. Even if requiring its currency to be used in paying taxes gives the currency a positive value, the currency may still not be an acceptable money. Unless the currency, upon introduction, is made convertible into an accepted medium of exchange, the consensus about its future value required for a fiat currency to serve as money may be lacking.
My argument thus far has two basic implications for less developed and former Eastern Bloc countries. First, such countries cannot create stable monetary systems based on inconvertible fiat currencies, because their political regimes lack the credibility to impose stable monetary institutions by fiat. Lacking a widely accepted money about whose future value expectations are secure, such regimes cannot create an acceptable money out of thin air, because they cannot impose a consensus about the future value of a fiat currency.
Creating a new fiat currency inevitably creates a hyperinflationary environment, because with no public confidence in the future value of the currency, the public will be willing to hold very little of it. This does not mean that the demand for money of any kind is small. The demand for a stable money in which people had confidence would be much greater than the demand for a new fiat money. An attempt to force more than this small amount into circulation, say, to finance the government deficit, causes rapid inflation. And the inevitable attempt to overreach the limits of that revenue source by printing even more money triggers a vicious inflationary cycle that causes a complete monetary breakdown.
All Moneys Are Not Equal
Conventional monetary models make two unwarranted assumptions that lead to disastrously misguided policies for developing countries. First, they assume that all money is alike, so that there is a given demand for money, which any instrument so designated can satisfy. Second, they assume that total output is independent of the amount of money. But not all moneys are equal and an inferior money in which people have no confidence cannot perform the services that a superior money could. Moreover, money serves as working capital for households and businesses adding to their productivity, while monetary stability provides a kind of intangible infrastructural capital that adds to the productivity of all economic agents independent of the amount of money they hold individually. Policies aimed at achieving monetary stability in developing countries by restricting the quantity of the available fiat money treat a minor symptom but ignore the fundamental problem, which is that distrust of the available money makes it useless and deprives the economy of desperately needed monetary services.
Thus, the second implication is that in such circumstances the only feasible way to create a consensus about the future value of a currency is to make it convertible into another money, e.g., the dollar, about whose future value expectations are secure. But governmental commitments to establish and maintain convertibility, as the recent Mexican fiasco has shown yet again, are obviously not credible, because a sovereign that defaults on such a commitment faces no effective sanction. Devaluations are a dime a dozen.
Nevertheless, given sufficient reserves, and given some institutional constraints on money creation and on government borrowing, governments can maintain a fixed exchange rate for a period of time. With sufficient resolve, they may do so indefinitely. However, such pegs are extremely fragile. Once an economic or political shock occurs, the expectation of a future devaluation becomes almost irresistible even for a developed country.
One way a government could increase confidence in its commitment to maintain convertibility is to create a currency board whose sole function would be to issue domestic currency in exchange for an equivalent amount of some foreign currency in terms of which the domestic currency would be defined. Such a system, if maintained, converts the domestic currency into a denomination of the foreign currency in terms of which it is defined. However, even a currency board cannot prevent a government from devaluing if that is what the government decides to do.
It is now clear why free banking is so well suited for less-developed and former Eastern Bloc countries. By making the commitment to maintain convertibility, one which holders of money can enforce through legal means against private banks instead of one that can be abrogated by the sovereign at will, free banking avoids the barrier that sovereign irresponsibility places in the way of creating monetary confidence.
Under free banking, private banks would be allowed to issue currency (banknotes) and create deposits denominated in units of their own choosing. Thus, if the public wished to use dollars, free banks would be willing to create money denominated in dollars. However, since there would be legal problems in issuing banknotes denominated in dollars, free banks would instead define new denominations (say, the crown) in terms of dollars (one crown equals one dollar), so that prices could be quoted interchangeably in either dollars or crowns. Because it would allow private banks to supply the hand-to-hand currency needs of the public, free banking would be preferable to simple dollarization which would require a country to import the dollars required for hand-to-hand circulation by means of a costly export surplus.
By setting an appropriate tax rate on bank profits, governments would, in the end, derive more tax revenue from the profits of the competitive banks than they could have by issuing fiat currencies and seeking to exploit their monopolistic control over those currencies. But even this gain would be dwarfed by the general increase in tax revenue that would result from an economic boom triggered by the provision of sound and stable money by a system of free competitive banking. 
1. Yet, as I pointed out in my book, continuing financial innovation and technological progress toward a cashless economy will, like it or not, gradually lead us over the long term at least part of the way toward free banking even with no deliberate redesign of our basic monetary institutions. The same point has been emphasized by Sir Samuel Britten in two recent columns in the Financial Times, June 9, 1994 and June 16, 1994.
2. A money can survive even if there is some expected depreciation in its value, but the more depreciation is expected the less useful and the less acceptable money becomes. Because of the positive feedback effects between your willingness to accept money and my willingness to accept money (now referred to as a network externality), monetary collapse can come very suddenly in response to a seemingly small change in expected depreciation.
3. The legal restrictions requiring U.S. banks to hold reserves of currency or otherwise restricting their behavior are, for purposes of this discussion, irrelevant. What matters is that in creating deposits, banks are offering an IOU that they are contractually obligated to redeem in terms of an asset (U.S. currency) whose supply or value they cannot control. If banks were legally allowed, as they once were and as they would be under free banking, to create banknotes circulating from hand to hand as currency, the legal status of banknotes could be similarly characterized, except that the bank’s obligation to redeem would be to the bearer of the banknote not to the original holder of the deposit.
5. On the role of prior convertibility in establishing a fiat currency, see G. Selgin, “On Ensuring the Acceptability of a New Fiat Money,” Journal of Money. Credit and Banking, November 1994, pp. 808-26.
6. The government might still profit from devaluation by reducing the real value of debt denominated in terms of the domestic currency, but it is not clear that the government so circumstanced would have anything to gain by denominating debt in terms of its domestic currency rather than the foreign standard currency since denominating in terms of the foreign standard currency would enable it to borrow at a reduced interest rate.