This archived content originally appeared at, the predecessor site to, and does not carry the sponsorship of the Cato Institute.

Notes for an Essay on Currency Competition and Counterfeiting

I’ve long been meaning to write something about the bearing of free banking on the problem of counterfeiting—a topic I addressed only very briefly in Theory of Free Banking. Since that time, although quite a lot has been written about the general subject of counterfeiting, hardly anything has been said about whether or not having competing banks of issue would make matters worse.

Conventional wisdom has it that having many banks of issue instead of just one makes life easier for counterfeiters. This wisdom draws heavily on the United States’ antebellum experience, while overlooking the experiences of arrangements more representative of genuine free banking, including those of Scotland between 1770 or so and 1845 and of Canada during the late 19th and early 20th century. Evidence from these other episodes points to a rather different conclusion, while suggesting that extant theoretical models of counterfeiting overlook ways in which competitive note issue can actually make counterfeiting less rather than more lucrative than it is when currency is supplied monopolistically.

I hope readers of this forum will forgive my setting forth, as a framework by which to gain a better understanding of the ways in which plural note issue bears upon the problem of counterfeiting, a very simple formal representation of a prospective counterfeiter’s anticipated profit, π. Let

π = ND – CcN – FP,

where P = g(Cc, Ca, N, D, T, B), Cc = c(Ca), F = f(N,B), Ca = d(B), and T = t(B). Here N stands for the number of counterfeits placed into circulation, D for their denomination, Cc for counterfeits’ average unit cost of production, F for the penalty or a counterfeiter incurs if caught (which may depend on the number of counterfeits he is found guilty of having uttered), P for the probability of detection, Ca for the unit cost of a genuine note, T for a note’s average “turnover” time, that is, time spent in circulation before returning to its (purported) source, and B, finally, for the number of legitimate banks of issue. The presumed derivatives are Cc’ > 0, f’(N) > 0, f’(B) < 0, g’(Cc) < 0, g’(Ca) > 0, g’(N) > 0, g’(T) > 0, g’(B) < 0, d’ > 0, and t’ < 0.

Standard treatments of the counterfeiting problem, to the extent that they allow for the possibility of plural note issue at all, have tended to treat the number of banks of issue, B, as increasing the anticipated profitability of counterfeiting by directly reducing the likelihood of prosecution, an effect allowed for here by the last element in P: this effect stems from the extra information costs that must be incurred by persons seeking to distinguish genuine from counterfeit notes as the variety of genuine notes increases. The practically exclusive emphasis upon this influence reflects the tendency of researchers to look upon antebellum U.S. experience, with its huge numbers of relatively small banks, and consequent need for “Counterfeit Detectors” to keep bankers and retailers informed of the correspondingly large numbers of known counterfeits, as representing the typical consequences of plural note issue.

As anyone who is familiar with the free banking literature, or simply with more informed writings on U.S. banking history, knows, the U.S. case was a by-product of legal restrictions, chief among which were laws prohibiting most banks from having branches, while prohibiting banking altogether in some states and territories. These circumstances turned under-banked regions into dumping grounds for the worst of the authentic banknotes from other parts of the country, while in turn creating easy opportunities for counterfeiters to add their products to an already disorienting melange.

But the U.S. experience was unique. Other plural note issue systems, like those of Scotland and Canada, involved not thousands or even hundreds but no more than a few dozen banks of issue, most of which had extensive branch networks. The number of distinct banks was small enough to allow retailers to familiarized themselves with legitimate note varieties. Yet in combination with branching it was more than large enough to provide for notes’ active redemption, that is, for their regular absorption by the clearing and settlement system, by which they were sent back to their reputed sources. In the Scottish system, for instance, it took eleven days or so for a note to return to its source after having first been paid out—a speed not so different from that for modern checks. This translated into a high annual turnover (T) of about 33. In contrast, notes of a monopoly issuer might remain in circulation for several months, if redeemable in specie. If not, they might remain afloat until too badly worn to circulate, which could mean years.

What difference does turnover make? That it matters a great deal becomes apparent as soon as one considers that in the vast majority of instances counterfeits are discovered, or at least discovered and then reported, not by members of the general public or by non-bank merchants, but by bankers themselves, and especially by the banks whose notes have been counterfeited. There are two simple reasons for this. First, the tellers at these banks are the persons best equipped to distinguish their banks’ authentic notes from even clever counterfeits. Second, they have the strongest incentives both to be on the lookout for fake notes and to report fakes that they discover instead of merely refusing them or attempting to fob them off on others. The last point is true in part because, if persons other than a bank's employees report a counterfeit of a bank’s notes, they may suffer a loss equal to the value of the reported notes. (Importantly, although this last deterrent has generally been in effect in monopolistic currency arrangements, it was sometimes absent in competitive ones.)

The significance of turnover, then, is that, the higher the rate of turnover of any bank’s notes, the greater the risk a counterfeiter faces of having, first his counterfeits, and then himself, discovered. High turnover means a greater chance that counterfeits will be detected within any given span of time, and consequently a greater chance of them being detected after a small number of transfers only, which increases the odds of their original “utterers,” and therefore the counterfeiters themselves, being caught.

A second, indirect way in which currency competition deters counterfeiting, which also has something to do with turnover, is by inducing note issuers to devote more resources to making their notes counterfeit-resistant. This follows from the fact that especially convincing counterfeits can fool even a bank’s own expert tellers, causing it to suffer losses by redeeming counterfeits. The extent of such losses will be greater, for any given volume of counterfeiting, the greater the turnover (and hence the lower the float) associated with an issuer’s currency. In the extreme case of a monopoly issuer of fiat money, which is of course never called upon to redeem its notes, the risk in question is nonexistent, and the incentive to invest in counterfeit-resistant notes is correspondingly low.

Finally, the presence or absence of competing issuers can influence would-be counterfeiters’ expected profits in at least one other way. This is by having some effect on the penalties or fines to which successfully prosecuted counterfeiters are subject. Here the difference is not so much economic as political, consisting of the fact that counterfeiting of the notes of a monopoly bank of issue has often been treated as a crime equivalent to the counterfeiting of official coin, and often, therefore, as a capital offense, whereas the counterfeiting of competitively-supplied banknotes has in contrast generally been a misdemeanor, if indeed it has been considered a crime at all.

The upshot of all this is that competition in currency can serve either as a deterrent or as a stimulus to counterfeiting. Whether it serves one way or the other depends on the importance of the “turnover” and “anti-counterfeiting investment” effects on one hand and those of “information cost” and “reduced penalty” effects on the other. Theory alone—at least the sketch of a theory considered here–therefore doesn't warrant the conventional conclusion that currency competition means more counterfeiting. The matter calls for a closer look into the historical evidence, to which I will turn in a second installment of these “notes.”

  • Warren

    Is there any evidence of bankers getting ahead of the game by keeping close track of those who were good engravers and various sources of paper and ink and presses so that if they saw these things being accumulated in a manner other than towards the use in legitimate banking or other legitimate uses in various industries they would know that something untowards was afoot and then could move to interrupt the scheme?

    Because that's what I would do, and I think I could get my competitors to help out as well.

    And making the printing of legitimate notes so difficult because of the ink and paper used were rare and easily tracked and the methods used were difficult and so making it that a counterfeiter would have to lay out a hard to conceal sizable capital investment and have to recruit known competent folks at likely high wages just to be able to fit into that 11-day window would also be a disincentive.

    Plus change paper and ink recipes every so often and it gets darn near impossible to be nimble enough to counterfeit. Unless you had people on the inside who would run off uncounted notes using the actual ink, paper and presses. But that has its own complications.

    • George Selgin

      Warren, I have heard of such evidence pertaining to false coining, and suppose that alert bankers and legitimate note manufacturers, not to mention detectives, adopted the same practice. Of course wise policy dictated that the bankers themselves hire all of very best engravers.

      As for the rest, the evidence I've gathered so far suggests, consistently with my theory, that competing note issuers tended to stay on the forefront of anti-counterfeiting R&D, but that this was generally not so for privileged banks of issue like the Bank of England.

  • Rob R.

    Great article. I'm guessing that the technology already exists for making counterfeiting virtually impossible either via emoney or adding some sort of electronic verification to real notes. While an interesting topic it seems odd that people could cite this as a reason against free banking.

    • I hesitate to say "impossible", but electronic money is far less susceptible to counterfeiting, so the subject gradually becomes of academic interest only to historians, along with all the talk of "paper money" and "printing presses". For electronic banknotes backed by collateral, I need a title registry associating my notes with particular collateral and ensuring that a unit of valuable collateral has only one corresponding note. My cell phone and internet servers do this work for me. This system doesn't prevent variations in the value of collateral, but no system can. I see no need for a central monetary authority here, except to police the title registry.

    • George Selgin

      My impression, Rob, is that however good anti-counterfeiting technology gets, it eventually becomes possible for counterfeiters to tackle it, making it essential that bankers never relax either their vigilance or their innovativeness. E-money, of course, represents a whole new approach; but here I'm mainly concerned with claims that competitive issuing of paper money invites counterfeiting.

  • Richard Schulman

    It's been a delight to read such a wealth of interesting posts recently.

    I suspect that digital payments (cards, smart phones, wallet chips, etc.) are going to increasingly supplant paper and metallic currencies. If so, counterfeiting will become more of a multimedia activity. Much of the mathematics limned above by Prof. Selgin would carry over, mutatis mutandis.

    In a free banking context, digital bank notes would be a boon to merchants because competitive banks could do their best to supplant the high-cost credit card issuers with the banks' own lower-cost transaction services. The limited physical currency that customers might need weekly for their non-digital transactions, they could get at a merchant's check-out counter, much as when I use my Discover card to pay for groceries at Walmart or Safeway, I can also request cash. The bank notes that merchants would provide their customers on this occasion would be those of the bank or banks with whom they enjoy favored relationships.

    This would be a convenient way for a free bank to put more of its notes into circulation, reducing its ATM and branch costs.

  • Is there any evidence on note counterfeiting in Somalia during the 1990s and 2000s? I know that money creation was oftentimes exported abroad, and there was a large influx of "counterfeit" notes during the 1990s. I don't know, however, how Somalis adjusted to inflation (I do know that U.S. dollars began to circulate more-or-less freely).

    • George Selgin

      Hi Jonathan. I am pretty sure that Somali shillings were and still are aggressively counterfeited. Will Luther and Larry White have written about Somalia's currency; I am not as inclined as they seem to be to see that case as one exhibiting the virtues of open currency competition.

      Plural supply of identical notes is in any event not my idea of freedom of note issue.

  • will.luther

    JonCatalan: In addition to the paper George linked to, here's my paper on the Somali experience of counterfeiting:
    I discuss the poor denomination structure that emerged in Somalia here:

    George: You claim Cc’ > 0. I assume you mean that profits increase as the average unit cost of production increases. Presumably you believe that (1) increasing unit costs has a direct effect of reducing profits, (2) increasing unit costs has an indirect effect of increasing profits (by reducing the probability of detection), and (3) the net effect is an increase in profits (i.e., |(1)|<|(2)|). Have I understood you correctly?

  • George Selgin

    Will, the derivative in question just means that, the greater the cost of an authentic note, the greater will be that of making a convincing counterfeit.