Lawrence H. White George Selgin

Why Private Banks and Not Central Banks Should Issue Currency, Especially in Less Developed Countries

Lawrence H. White*, George Selgin*
"The inefficiencies associated with government monopoly in currency are especially large in developing countries, where the reliability of the exchange rate (an important aspect of currency quality) is often quite low."
In all but a few areas of the world today (Northern Ireland, Scotland, and for the time being Hong Kong), currency is a nationalized industry. Treasury departments issue coins; the state-owned enterprises known as central banks issue paper notes. It was not always so. Private banks were the main issuers of paper currency in the United States and Canada a century ago, and were the sole issuers in virtually every country two centuries ago.

Nationalization of currency is largely taken for granted today, but it shouldn't be. Adam Smith praised private currency for the benefits it had brought to his native Scotland. Most economists would agree that a legally enforced government monopoly is generally an inefficient way to produce private goods and services. The post office is a prime example; other examples range from state-owned plantations to national railroads. Currency is no exception to the rule. As with other nationalized products, quality is lower than it would be under private competition. The inefficiencies associated with government monopoly in currency are especially large in developing countries, where the reliability of the exchange rate (an important aspect of currency quality) is often quite low.

The central bank's credibility problem

An important quality dimension of a currency is the reliability of the redemption pledge—namely, the issuer's promise to exchange it for another money on demand at a specified rate. In developing countries, domestic currency typically derives its value from its redeemability at a fixed rate for U.S. dollars. To the extent that the central bank actually respects this agreement, a fixed exchange rate constrains monetary expansion and thus helps avoid the high inflation to which unanchored regimes in the developing world have often succumbed. The dramatic devaluations in Southeast Asia in 1997 illustrate how unreliable are the redemption pledges made by central banks in developing countries.

Private commercial banks are more reliable. Privatizing currency means leaving it to commercial banks to issue media of exchange that are claims to the reserve asset that defines the monetary standard, and makes the enforcement of these claims a matter of commercial law rather than of public policy. It correspondingly decentralizes the responsibility for holding adequate reserves. Outside of "crony capitalism" (where nominally private banks receive special dispensation to renege on contracts), private currency issuers face incentives for quality control that do not face government monopoly issuers.

The chief weakness of central banks as currency issuers is their inability to bind themselves to their redemption promises. Their public monopoly status gives them immunity from the legal and marketplace sanctions that ordinarily prevent commercial banks from reneging on their commitments to honor their debts in full. A central bank enjoys "sovereign immunity" from claimholder lawsuits, and legal restrictions on the public's choice in currency mean that the central bank has little fear of losing customers for bad behavior. At the same time, central bankers—especially in developing countries—face political pressure to provide the short-run benefits that surprise monetary expansion can deliver (namely extra revenue to pay the government's bills, extra stimulus to the economy, or extra liquidity for the banking system).1 When devaluation is relatively costless, central banks are tempted to engage in expansionary monetary policies that ultimately force devaluation.

Private commercial bankers, by contrast, can be sued by holders of their claims when they fail to redeem those claims at par. Even if a private bank could devalue its liabilities (repay less than 100 cents on the dollar) with legal impunity, such a devaluation would deal a severe blow to its reputation, and in a competitive environment its clientele would go elsewhere. Shareholders would suffer losses. The shareholders' incentive to avoid devaluation or default compels them to limit the volume of the bank's liabilities, and makes the bank's redemption commitments credible. A larger penalty for devaluing means that a private bank will choose to run a lower devaluation risk than will a central bank. Central-bank devaluations are consequently more frequent than autonomous commercial banking defaults (i.e., defaults not associated with central bank devaluation or attacks on the central bank peg).

What would a private currency system look like? As in Scotland and Northern Ireland today, domestic banks would issue circulating notes denominated in and directly redeemable for foreign-currency assets (there, Bank of England notes).2 Each note would clearly carry the name of the issuing bank whose liability it is. Any bank that tried to issue too many notes would find them being deposited into other banks, and returning via the clearing system for redemption in reserve money.

In many developing countries, where we already observe unofficial dollarization of transactions in both financial markets and commodity markets, the U.S. dollar is the market's likely choice for the reserve currency. There the public demonstrably prefers the dollar to the money produced domestically. The domestic central bank currency, having failed the market test, survives at all only because of legal restrictions that compel domestic transactors to accept and use its currency for some purposes. Remove those restrictions, and the leading domestic and foreign commercial banks would provide dollar-denominated currency just as they now provide traveler's checks.

A common objection to dollarization is that it transfers seigniorage (the profit from currency issue) to the U.S. government. That is true only if Federal Reserve notes are used as currency, as they are in Panama today. It is not true if commercial banks are allowed to issue currency notes. Under competitive note-issue, seigniorage is transferred to the United States only to the extent (presumably very small, unless the domestic government imposes a high required reserve ratio) that the banks hold reserves of Federal Reserve notes. The bulk of potential seigniorage is kept at home, where competition among the banks distributes it to currency holders in the form of unpriced banking services. For example, as in Northern Ireland, competing banks of issue could seek customers by waiving fees for withdrawing notes from their widely distributed automatic teller machines.3


If domestic citizens want high-quality redeemable currency, they are better served by privatization of note-issue than by a central bank dollar peg. Where does this prescription apply? There is little opportunity to privatize currency in a country wedded to state-owned banking. Less of a gain in quality is to be expected where the political authorities have heavily restricted entry into banking and have severely corrupted the legal system. Our prescription therefore applies most immediately in the set of countries that have had poor monetary policy and repeated devaluation, but that also have (or could have) competitive commercial banking under a regime of effective contract enforcement and the rule of law. Casual empiricism suggests that this is not an empty set. Such countries are found in central and eastern Europe, Latin America, Africa, and Asia.

In some countries the most important reason for unreliable money has been that the fiscal authorities rely heavily on the seigniorage revenue from printing money. In several developing countries, seigniorage has funded 15 to 20 percent of state spending. In such a country dollarization and privatization of note-issue would necessitate fiscal reforms that either provide replacement revenues or cut state spending. The political economy of how to secure such reforms is beyond the scope of our discussion here. However, if replacement revenues must be found, one potential source is a tax on private note-issue. Such a tax would inefficiently suppress non-price competition in currency issue, but it would not undo the greater credibility of private currency, and it would still be better than nationalization.

But didn't private currency cause big problems in the antebellum United States?

An importance source of opposition to the idea of having commercial banks supply paper currency is the belief, based on a very partial reading of history, that such currency tends to circulate at less than its face value and thus imposes substantial exchange costs. The history of private currencies in Scotland, Sweden, Canada, and many other nations shows otherwise. As a rule, bank-issued currencies circulated at par (face value).

U.S. experience prior to the Civil War, when the U.S. currency stock consisted mainly of notes issued by some 1500 state-chartered banks, is the major exception to the rule. Instead of commanding their full face value, many state bank notes traded at discounts. The discounts reflected non-trivial costs of redeeming the notes for gold or silver, and in some cases a risk of non-redemption.

Before modern economic historians (Hugh Rockoff and others) began re-examining the period,4 typical accounts blamed the lack of par acceptance on "laissez-faire" banking policies, which supposedly fostered fraudulent "wildcat" banking. In addition, as if fraud by the bankers themselves wasn't bad enough, counterfeiters are known to have routinely imitated their notes, making it necessary for merchants and consumers to consult "banknote reporter" periodicals for lists of spurious notes. The older accounts cited such problems as the explanation for why federal authorities finally prohibited state banks from issuing notes after August 1866. Unfortunately, such myths persist.

In fact, state bank notes weren't nearly as bad as the older accounts make out, and their suppression by the Federal government wasn't really motivated by quality concerns. By the outbreak of the Civil War, sound state bank currencies were the norm. On the whole, the failure rate among antebellum banks was not much worse than the rate during other periods of U.S. banking history. A few state banking systems did produce notoriously risky currencies, especially during the 1850s, but laissez-faire policies weren't to blame (because they didn't exist), and fraud was rare. The major cause of bank failures was an interventionist regulatory regime (ironically called "free banking") that compelled banks to back their notes with high-risk state bonds.

Early on, it is true, even the notes of the better state banks sometimes circulated at a discount once they had traveled far from home. The discounts on these notes were never very large, and they fell over time with improvements in transportation and communications, particularly with the spread of railroads and telegraph lines. By late 1863, the entire stock of northern banknotes, if purchased and traded for legal tender in the New York or Chicago market, would have fetched over 99 percent of its face value.

The reason why there were any discounts at all was not a lack of government restrictions on banking, but just the opposite. State laws generally prohibited branch banking, so that most banknotes could only be redeemed at a single location. The discounts mainly reflected the cost of returning the notes to that location for redemption. Had antebellum U.S. banks been able to branch nationwide, as they were in other countries (and finally are in the United States today), their notes could have been easily redeemed at multiple points across the country, allowing them to circulate nationwide at par. The non-uniformity of U.S. currency prior to the Civil War was a byproduct of government interference with open competition in banking.

What about counterfeiting? It is true that counterfeiters imitated and altered the notes of many state banks. But they also imitated and altered later National Bank notes, and today they are no less inclined (technological advances notwithstanding) to imitate and alter Federal Reserve Notes. The counterfeiting of state bank notes was generally less profitable than the counterfeiting of today's central bank currencies, because private bank notes don't stay in circulation long. Much like travelers checks, they quickly return to issuer. Counterfeits are therefore likely to be detected by experts while the trail is still warm. Federal Reserve Notes are today frequently counterfeited, but currency users don't (yet) have the option of avoiding them in favor of safer private substitutes.

If state bank notes weren't really so bad, why were they taxed out of existence? To finance the Civil War, the Congress first empowered the Union's Treasury Department to issue over $400 million worth of U.S. Notes or "greenbacks." Congress then went on to establish a new system of federally chartered banks, which were authorized to issue another $300 million of National Bank notes provided they purchased federal bonds as backing. The Treasury realized that all this new currency threatened to cause a substantial increase in prices, but was unwilling to deny itself the fiscal advantages that the new currency would provide. Instead of issuing fewer greenbacks or further limiting the stock of National Bank notes, it made the state banks into scapegoats, forcing them to retire all $200 million of their notes, including some of the best currency the nation had ever known.5


Today's central bank currency monopolies have grown not from attempts to rectify market failures but from government's appetite for revenue. Contrary to myth, the United States' experience in the nineteenth century does not supply any grounds for restricting private banks from issuing currency. It instead points to the harmful consequences of government restrictions on bank-issued currency.

Just as it is efficient to leave the provision of checking accounts to competing private banks, rather than have a single government monopoly provider of checkable bank liabilities, it would be efficient to (re-)privatize the issue of circulating currency. By comparison to public monopoly, privatization raises the quality of currency. In developing countries, private currency can particularly improve the reliability of the issuer's pledge to redeem currency for dollars at a fixed rate.


Rockoff, Hugh (1991). "Lessons from the American Experience with Free Banking," in Forrest Capie and Geoffrey E. Wood, eds., Unregulated Banking: Chaos or Order? (London: Macmillan), 73-109.

Rolnick, Arthur J., and Warren E. Weber (1986). "Inherent Instability in Banking: The Free Banking Experience," Cato Journal 5 (Winter), 877-90.

Selgin, George (2000). "The Suppression of State Bank Notes: A Reconsideration," Economic Inquiry (forthcoming).

White, Lawrence H. (1986). "Regulatory Sources of Instability in Banking: Comment on Rolnick and Weber," Cato Journal 5 (Winter), 891-97.

White, Lawrence H. (1999). The Theory of Monetary Institutions (Oxford: Blackwell).

White, Lawrence H., and Donald J. Boudreaux (1998). "Is Nonprice Competition in Currency Inefficient?", Journal of Money, Credit, and Banking 30 (May), 252-60.

—— (2000). "Is Nonprice Competition in Currency Inefficient?: Reply", Journal of Money, Credit, and Banking 32 (February), 150-53.


For an explanation of the government revenue from monetary expansion, or "seigniorage", see White (1999, ch. 7).


The private banks that retain the right of note-issue in Scotland and Northern Ireland today face a binding marginal required reserve ratio of 100% beyond a specified "uncovered" issue. This clearly tends to weaken the competition for note-holding customers.


Stocking its ATMs with its own notes is one nonprice-competitive device a bank can use to get the public to hold them in preference to the reserve currency for which they are redeemable. Other ways are to make them physically more attractive and lower in counterfeit risk. Competition on price (interest return) appears to be economically ruled out by transactions costs: the potential interest earnings are trivial relative to the greater convenience of having a currency of a fixed face value. On the general efficiency of nonprice competition in currency see White and Boudreaux (1998) and (2000).


For a survey and extension of the "re-examination" literature see Rockoff (1991). See also Rolnick and Weber (1986) and White (1986).


On the story behind the suppression of state bank notes see Selgin (2000).

*Lawrence H. White is F. A. Hayek Professor of Economic History in the Department of Economics, University of Missouri.
*George Selgin is Associate Professor of Economics at the Terry College of Business, University of Georgia.
Return to top