The free banking debate seems to be a perennially reoccurring event with no resolution in sight. On Twitter, George Selgin recently had a series of tweets and threads again criticizing the “Rothbardian” position on free banking. Although partly due to the limitations of the medium, the description of the Rothbardians’ objections to free banking was less than accurate. Since we are never going to get anywhere in the debate if the two sides keep misunderstanding each other, allow me to here present a short guide to the main critiques of free banking from the Rothbardian side.
Given this emphasis, this guide will necessarily present a lopsided picture of the debate, and I therefore suggest that the neophyte also read the able expositions of the free banking side, including online overviews such as this one and this one, Larry White’s and George Selgin’s foundational books, and this brief(ish) academic article by Selgin. Fundamentally, the point at issue is this: Is free banking, i.e., a system of banks freely and competitively supplying money backed by fractional reserves, economically stable and efficient, or an unstable source of inflation and distortions in the economy?
“Free Banking Is Fraudulent”
One point often critiqued by the free bankers is the claim made by Murray N. Rothbard and his followers that free banking is necessarily fraudulent. In recent years Rothbardians have downplayed this point, understandably so, perhaps, since they have wanted to emphasize the purely economic critiques of free banking. However, the point contains a substantial critique and deserves more than to be passed over in silence.
There are really two strands to this critique:
First, that money substitutes—banknotes and deposits—are really warehouse receipts or property titles, and therefore it is tantamount to fraud to issue notes and deposits in excess of the reserves of money proper. Or in the Roman law framework favored by Jesús Huerta de Soto, loan contracts that confuse the deposit and loan function of banks and grant several persons an equal claim to the same underlying property are self-contradictory and therefore void.
Second, that the historical emergence of fractional reserve banking was due to a fraud perpetrated on the public, as the first bankers—whether London goldsmiths or Italian merchants—lent out the gold or silver deposited with them or issued more notes when they discovered that their notes circulated as money substitutes. Ever since, fractional reserve banking, whether free or supported by a central bank, has systematically relied on deception and government privilege.
Ad 1: It’s Not, Except It Is
By no means all Rothbardians insist on the first point: Guido Hülsmann in 2003 (see also Joseph T. Salerno’s brief summary) suggested the theoretical possibility of fiduciary media (i.e., money substitutes in excess of reserves) that take the legal form of an IOU with a redemption promise. If the banker can convince his customer to accept such a claim on par, no fraud occurs, but there will be a tendency on the free market for such callable loans to trade at a discount to money proper and fully backed titles to money due to the default risk inherent in such mere promises. I myself tried to expand on this point more recently, arguing that the difference between fiduciary media and fully backed titles—one carries the risk of default while the other doesn’t—means that an individual who accepts callable loans at par with money and considers them part of his cash holding commits an entrepreneurial error. In the free market there will be a tendency for all errors to be corrected, including this one, leading to the virtual suppression of fiduciary media.
This does not mean that the older fraud arguments must be disbanded, however. They point to a basic point of contention between free bankers and their critics: when a person holds money, what he wants is immediate access to a fund of purchasing power. This is true also in the case of fiduciary media: the person accepts banknotes because he thinks they give him immediate and total control over a sum of money; that is, he thinks he is the owner of the sum in question in the economic sense if not in the legal sense.
Free bankers disagree. Yes, we all agree when it comes to money in the narrow sense. But it is different when it comes to bank money: Steven Horwitz argues that the banking system supplying fiduciary media to money demanders is really a subset of the loanable funds market. Selgin says that “[t]o hold inside money is to engage in voluntary saving” and that holders of money lend out the sum in question through the banking system. He repeated the same point in his critique of Bob Murphy’s take on market monetarism in 2020.
Yet as we just saw, the fraud argument is a result of the claim that two persons cannot have equal, mutually exclusive property rights in the same good—but this is exactly what the free bankers say happens in a free banking system: claims to the same sum of money multiply within the system and everyone thinks he has immediate and total control over it. Huerta de Soto in his 1998 article points out this confusion between the concepts of saving and of demand for money. An increase in the supply of fiduciary media, even when it corresponds to an increase in the demand for fiduciary media, does not constitute an increase in real savings.
There is a basic disagreement here over the nature of savings in the form of money. The free bankers consider cash holding as a form of capitalistic saving that releases resources for investment. While Hülsmann in his 1996 critique agrees that holding money is a form of saving—holding any kind of durable good is saving—it is not capitalistic saving. It would be more correct to consider it akin to what Ludwig von Mises termed plain saving, the piling up of consumer goods for later use. Since the effects of changes in money saving are different from changes in other kinds of saving, it would perhaps be best to put it in its own category, monetary saving.
Whether money is held in the form of money proper, money titles, or fiduciary media, what the holder wants is immediate access to purchasing power. Just because a money substitute takes the legal form of a loan, this does not mean the holder of the claim has given up present goods. He has not, or rather, he did not intend to. This is why Huerta de Soto considers fractional reserve banking illegitimate. The intentions behind the contracts are contradictory. This is also why Rothbard suggested antibank vigilante leagues would be beneficial in a free banking setting: the public would be misguided in thinking their money safe in the bank, and the vigilantes would attempt to spread awareness of the practice and basic unsoundness of fractional reserve banking.
There is, then, more substance to the fraud critique—it is not simply a misunderstanding about the nature of callable loans.
Ad 2: Define “Fraud,” Your Honor
The second variant of the fraud argument is mainly historical, as it concerns the interpretation of the various historical episodes claimed as instances of successful free banking. One classic example is the London goldsmiths. Rothbard argues in his Mystery of Banking that goldsmiths’ note issues in excess of the gold deposited with them was fraudulent—an argument that puts Rothbard square in the mainstream of opinion on this issue—while Selgin has quite ably shown that the story is not so simple.
Beyond fraud, the Rothbardian charge is more generally that some kind of legal privilege was always necessary for fractional reserve banking to get off the ground. The locus classicus of his disagreement is the case of Scottish free banking. Rothbard at first agreed with Larry White’s positive interpretation of the Scottish experience but by 1988 had changed his mind. Several features of the Scottish institutional setup suggested to Rothbard that the system was not so much free banking as privileged banking, since the bankers were to some extent protected from prosecution for failure to pay debts if they didn’t honor their notes. That is of course one kind of freedom. The most recent round in the Scottish debate, to my knowledge, was between Joakim Book and myself. Bob Murphy also deals with Scotland (and the Canadian case, described by Selgin here) in a 2019 paper.
There are dozens of other examples of historical free banking, but I will not go into further details here. This is not to disrespect the scholarly work done, much of it of great interest (it’s for instance pretty clear that American “free banking” was nothing of the kind), but rather to point to a more fundamental issue. If by free banking we mean the complete absence of government interference, then such a thing has never existed. I think everyone agrees on this. But it’s the same with free markets: if what is meant is the complete absence of government regulation, then they never existed, and we wouldn’t therefore claim that arguments for the superiority of economic freedom to interventionism are vacuous.
In other words, free bankers can claim that free banking was free enough so long as government intervention was minimal. This is a too cavalier attitude, I think, both when it comes to free banking and economic history in general. What we must do is look at what the specific government interventions were and what their effects were. Only after having done so can we declare whether or not they were irrelevant to the evolution of the market. This is exactly what is at issue in the Scottish case: one side claims the interventions were irrelevant, the other argues that they had important distorting effects. The historical record does not tell us which is true: this is a matter of interpretation.
In other words, history cannot take the place of theory: we must first settle the issue of free banking by economic reasoning, and then apply our results to history. Is Rothbard wrong to say that the key decisions in the British legal system that led deposits to be considered loans rather than bailments were wrong? Is Huerta de Soto wrong in his interpretation and critique of the Roman law treatment of deposit contracts as it evolved and was applied in the early modern period? Does Selgin prove the immaculate conception of fractional reserve banking when he shows that London goldsmiths did not in fact break the law? Answering these questions without regard to the results of economic theory would simply amount to smuggling in our own assumptions about the matter. If Rothbard’s economic analysis of deposit banking is correct, then he is also correct to deplore the British legal evolution that legitimized fractional reserves; if not, then he is not.
Notwithstanding disputes over historical episodes, the main debate is in the field of theory. I have already described one fundamental disagreement over the nature of money: the Rothbardians claim that demanding money always means demanding a present good, while free bankers argue that demanding bank money is the same as lending money through the banking system. I will not repeat the arguments, but only point to an interesting paper by Michael Bauwens from 2017 suggesting that the dispute boils down to a basic difference in philosophy. Critics of free banking espouse (at least implicitly) a realist social philosophy while free bankers have a more nominalist orientation.
Free banking theory argues that a free banking system will be stable and resilient. An equilibrium will emerge where banks supply a stable stock of money that only varies in accordance with the demand for money. Rather than causing instability and shocks, free banking alleviates monetary shocks and prevents them from causing economic damage. A rise in the demand for money, for instance, is compensated by an equal rise in the supply of bank money, and this means that the flow of spending is not disrupted and the economy does not have to adjust to a painful exogenous shock.
Philipp Bagus and David Howden in their critiques of free banking (here, here, and here) argued that this is not so. The supply of bank money is not simply passive in response to demand for money, and the constraint on money expansion from the need for precautionary reserves is overstated. Lending of reserves between banks can accommodate an expanding bank, even when it holds lower reserves than other banks. Nikolay Gertchev has explained how an interbank reserve loan market effectively takes the place of a central bank by permitting credit expansion.
The supply of money, critics argue, is therefore not constrained and tightly regulated in a free banking system. One reason for this is that free bankers are wrong to claim that the demand for bank money is exogenous to the banking system. On the contrary, demand for bank balances depends on the terms on which they can be had. The interest rate at which banks make loans is thus an important determinant of the demand for bank money. Following Mises, Rothbardians claim that banks can lower the rate of interest (at least temporarily) below the natural rate and that this incentivizes bank borrowing and the holding of bank money.
If banks issue more money, this has consequences beyond simply allowing greater nominal demand for money. There is an implicit assumption in free banking theory that the issue of bank money, when kept within the bounds set by the demand for money, is neutral—it does not affect relative prices or the distribution of real incomes and wealth. Rothbardians, on the contrary, argue that since an increase in the supply of fiduciary media is by definition inflationary, it leads to the outcomes described by the Cantillon effect and (possibly) also to a business cycle. The first receivers of the new money gain while later receivers lose, and if the first receivers are businessmen who use the new money for investment projects, the result is the process of malinvestment described by business cycle theory.
Selgin claims that fractional reserve banking is only inflationary if the reserve ratio is persistently declining. This is not so, the Rothbardians say. If banks can expand credit, an increase in reserves leads to bank credit inflation and a business cycle, even if the reserve ratio is increasing. Rothbard’s own work on the Great Depression shows how this happened historically, with the inflow of gold from abroad contributing to the inflation. It is also a thoroughly Misesian point: Mises explained how when banks merely maintain the stock of media in existence without adding to it there is no inflation and no malinvestment but that any issue of fiduciary media that increases the stock does cause disruptions. Nowhere does Mises claim that the impact of credit expansion depends on a change in the reserve ratio. What matters is the money supply and changes to the money supply, not the ratio of money substitutes to reserves.
While free bankers maintain that fractional reserve banking can maintain monetary equilibrium, their critics attack the very idea of monetary equilibrium. All agree that when analyzing money, we need to take account of more than simply exchange demand. Setting out from the brilliant British economist Edwin Cannan, reservation demand is included as a key cause influencing the value of money.
Yet here free bankers and critics part: to the free bankers, monetary equilibrium simply means adjusting the stock of money to satisfy the reservation demand for money, while the exchange demand for money—the supply of goods and services—is fundamentally unrelated: so long as there are no disturbances from the monetary side, the real economy can go on functioning smoothly. In terms of the equation of exchange (MV = PY), the money side is MV and the goods side is PY.
Not surprisingly, perhaps, the Rothbardians reject this way of conceiving of money and the economy. As Salerno shows in a 2006 article, while it is correct to follow Cannan (and Rothbard) in looking at reservation demand as well as exchange demand for money, this does not mean that we can ignore exchange demand. In order to understand the formation of prices, the demand for money, and the value of money, we need to always keep the exchange demand for money in mind; i.e., the supply of all the goods and services against money (exchange demand) is a key factor not only in the formation of specific prices but in the total demand for money.
Hülsmann made much the same point in his review of Lawrence White’s Theory of Monetary Institutions. Any change in monetary conditions—be it from the demand or supply side—leads to a change in the purchasing power of money, the very quality for which it is demanded: “Thus the supply of money does not have to be adjusted to the demand of money. Unlike all other commodities, money itself constantly adjusts to the conditions of the market.”
In her article against monetary disequilibrium theory, Laura Davidson expands on this critique. We cannot simply cancel out exchange demand for money because there necessarily is an equal exchange supply of money. We must look at the total demand schedule of money, and that includes its exchange demand. Each good on the market has its own supply-demand schedule in terms of money, and money therefore has a partial supply-demand schedule in terms of each good. Davidson puts it eloquently:
Thus, when the social reservation demand for money changes, it is not a single supply curve that shifts, rather it is the partial supply curves of money with respect to goods and services individually (and hence those goods’ demand curves) that shift, all to varying extents. And they do so precisely because a change in the social reservation demand for money is nothing more than a change in its marginal utility as it moves up or down each market participant’s value scale, a value scale that encompasses all goods including money.
Descending from the heights of disequilibrium, price stickiness and menu costs have been put forward as reasons for why an uncompensated increase in the demand for money will cause economic dislocation. Since prices cannot adjust immediately, a decrease in the flow of spending will lead to painful dislocations as factors of production become unemployed and output declines. Bagus and Howden have criticized this line of argument in an academic exchange with Anthony J. Evans and Horwitz and Selgin. The claim is not that prices will instantly change in response to every change in market data. There is some price stickiness built in, but only because entrepreneurs and consumers want it this way (e.g., long-term contracts with fixed prices help businessmen engage in economic planning, since they are insulated from unforeseen price changes). In fact, if prices were in constant flux, planning would become very difficult, if not impossible. That said, prices are sticky only so long as and to the extent that this is desired by market actors. If they change their minds, prices will change too.
Of course, it is a different case if government interventions cause stickiness. Price controls can cause permanent unemployment, for instance in the case of unions violently enforcing minimum wages for various kinds of work. Then, an increase in the money supply to lower real wages can be a help, if only very temporarily—for if the union had the power to enforce one minimum wage, what is to prevent them from getting wise to the deceit and enforcing a higher or index-linked minimum wage? F.A. Hayek showed long ago that inflation to combat union power is a futile endeavor.
Conclusion: Misesian by Any Other Name Is Just as Sweet
There is thus plenty of meat on the Rothbardian critique of free banking. While I count myself among these critics, I’ve here tried to simply present the main points and the principal authors, although I’ve doubtlessly neglected and forgotten some.
One final note on the use of terms, however: I’ve throughout humored the free bankers by accepting the term “Rothbardian” as a descriptor of the critics. Now, this is in part because it is correct: Rothbard was a main critic and all later critics have drawn from his work. However, it is also misleading: it suggests that hostility to free banking is something that flows from ideas unique to Rothbard. In reality, the critique of free banking is Misesian. The critics cleave to a Misesian understanding of money and its impact on the economy and on this basis criticize free banking theory. The only uniquely Rothbardian argument not found in Mises is that fractional reserves should be outlawed as fraudulent. But that is down to a difference in legal and ethical philosophy between the two: Mises the utilitarian versus Rothbard the natural law theorist. Yet as we saw, the fraud argument rests on economic theory and on a desire to make legal categories match economic reality. If Mises refused to go the route of 100 percent reserves, it is only because he thought free banking the better policy for the suppression of fractional reserve banking.