Mises, the Regression Theorem, and Free Banking

     

My thoughts about The Theory of Money and Credit by Ludwig von Mises are inevitably colored by the fact that my first introduction to the discipline of economics was at Grove City College in 1967 in a class taught by Hans Sennholz, a devoted student of Mises. All of us who studied under Sennholz were encouraged to read Mises, and I soon tackled The Theory of Money and Credit. The only other book on monetary theory I had read before that was Murray Rothbard’s What Has Government Done to Our Money? (1964). Consequently, Mises’s delving into advanced and somewhat abstruse monetary controversies and his critiquing of other authors with whom I was entirely unfamiliar left me a bit lost, particularly since I was a history major. It is only with subsequent study over the intervening years that I came to understand fully Mises’s contributions. I therefore find myself in general agreement with Larry White’s appreciation of The Theory of Money and Credit. So I will focus my comments on two of the issues he discusses, offering a few elaborations, reservations, and unresolved questions.

Mises employed his Regression Theorem to explain the acceptance of fiat moneys as media of exchange. According to Larry, “[A]ll governments that have successfully launched new fiat monies have first given them a fixed redemption value in terms of a commodity or an established fiat money.” Yet if we interpret the term “redemption” strictly, this is not quite historically accurate. Among the earliest fiat moneys in the West were those issued by the British colonies in North America. Rarely if ever were these “bills of credit,” as they were called at the time, directly and immediately redeemable for gold or silver at a fixed rate. Instead, they usually were indirectly linked to commodity money at a fixed rate through taxes (and often also through legal tender laws that applied to debts enforced in the courts). In other words, unlike Federal Reserve notes, which initially could be exchanged for gold before they became fiat money, bills of credit were denominated in the colony’s unit of account and could be used to pay taxes and other government levies in lieu of the monetary commodity. But bills of credit were rarely immediately redeemable, although they were often accompanied by a promise that at some future date they would become so (Michener 2011).

This reliance on taxation and future redeemability applies also to the Continental currency issued by the U.S. during the American Revolution, to the Greenbacks issued by the Union during the Civil War, and to the currency issued by the Confederacy during the same conflict. Indeed, Confederate currency, unlike Union Greenbacks, was not made legal tender in private transactions but only publicly receivable for taxes. George Selgin’s extensive treatment of fiat money’s emergence (1994, 816–21), which invokes Mises analysis, is somewhat difficult to interpret. On the one hand, he seems to require only a fixed exchange rate rather than strict redeemability for successful launching of fiat money, thereby encompassing the historical instances I have mentioned. And when George explicitly rejects taxes as a sufficient mechanism, he appears to be referring to a new fiat money using an entirely new unit of account, completely unrelated to and floating against the existing commodity money. However, he also claims that “[p]ublic receivability laws can confer value on a new money only” if “there be real resources at the disposal of government to be exchanged for the new money at administered prices,” which was hardly the case with bills of credit, Greenbacks, and Confederate currency. Whatever George’s position, Mises was clearly wrong when he stated that fiat money emerges exclusively “when the coins or notes in question have already been in circulation as money-substitutes,” by which he meant “perfectly secure and immediately convertible claims” (emphasis mine; 1953, 78).

These historical counterexamples are particularly interesting because Mises, in other passages, inadvertently describes them. He divided what he called “money in the narrower sense” (or what has come to be called outside money in modern monetary theory) into not just two categories: commodity and fiat money. He introduced a third category, which in the English translation was designated “credit money.” The Theory of Money and Credit defined credit money as “a claim against any physical or legal person,” but one that “must not be both payable on demand and absolutely secure.… Credit money … is a claim falling due in the future that is used as a general medium of exchange” (1953, 61–62). Mises continued to posit these three categories in the final edition of Human Action (1966, 429). He considered the notes of banks that had temporarily suspended redemption in specie (gold or silver coin) as one example of credit money, but irredeemable moneys launched through taxation fit the category equally well. In fact, Mises (1953, 153) referred to Confederate currency as a specific example of credit money. The Theory of Money and Credit went so far as to speculate that most forms of so-called fiat money might in reality be credit money: “Whether fiat money has ever actually existed is, of course, another question, and one that cannot offhand be answered affirmatively. It can hardly be doubted that most of those kinds of money that are not commodity money must be classified as credit money. But only detailed historical investigation could clear this matter up” (1953, 61).

Outside of Sennholz in his lectures, the only Austrian economist (to my knowledge) who has pursued Mises’s distinction between fiat and credit money, albeit cursorily, is Joe Salerno (2010, 68–70, 586). Yet this distinction may be more than a terminological quibble. One recent development in monetary theory is the Fiscal Theory of the Price Level (FTPL), most ably advocated by John Cochrane (2005) of the University of Chicago. It is a highly mathematical attempt to formally integrate the impacts of both monetary policy and fiscal policy, and one of its crucial underlying assumptions is that the current value of fiat money depends on people’s expectations of the government’s future taxes. In other words, it explicitly treats fiat money as what Mises called credit money, and moreover contends that future expected taxes have a significant effect on money’s velocity (that is, the reciprocal of the portfolio demand for money) and therefore on the price level. In this respect, the FTPL is building on an extensive literature arguing that expected taxes played the major role in determining the price level in the early fiat (credit) money issues in America. (Smith 1984,1985a, 1985b, 1988; Wicker 1985; Calomiris 1988a, 1988b; critiques of this literature include Michener 1988; McCallum 1992; and Sumner 1993; for surveys of the debate see Michener and Wright 2006 and Michener 2011.) It is also consistent with Michael Sproul’s (2003) much cruder “backing theory of money,” which denies the existence of fiat money altogether and about which Sproul (2013) is currently debating Kurt Schuler at the Free Banking blog.

The FTPL does not necessarily undermine the validity Mises’s Regression Theorem as an explanation for the launching of fiat (or credit) money. Nor is it entirely inconsistent with a sophisticated interpretation of the quantity theory of money. Yet it has some far more intriguing implications. Tyler Cowen once emailed me that the FTPL “is formally correct but not so useful; just a way of re-explaining the traditional boxes,” while Greg Mankiw in personal conversation similarly described it as merely “a new vocabulary.” But I believe they are both mistaken. If correct, the FTPL implies that neither fiat nor credit money are true outside money in the sense of being assets only, with no offsetting liability. Instead they are really what current monetary theorists refer to as inside money, with future taxes representing the offsetting liability-side, making them much like shares of stock, whose value depends on an anticipated future income stream. Not only does this conclusion eliminate any real-balance effect that can result from fiat or credit money constituting net wealth (unlike commodity money), but it impinges on the long-standing debate over whether a pure inside-money economy would be feasible. Initiated by Knut Wicksell (1936) and addressed by Don Patinkin (1965), this debate was part of the wide-ranging development during the late 1980s and early 1990s of the New Monetary Economics, with Leland Yeager (1997), Tyler Cowen and Randy Kroszner (1994), and George Selgin and Larry White (1994) major participants. I am not myself convinced that the FTPL is correct, but it deserves more attention and discussion than those influenced by Mises’s Theory of Money and Credit have given it.

Larry credits The Theory of Money and Credit with “putting the analysis and conclusions of the 19th-century Free Banking School on firmer footing.” Here he is reiterating his interpretation in an earlier article that coincidentally appeared in a Sennholz festschrift (White 1992). Larry was somewhat more cautious in that version, admitting that the first edition of The Theory of Money and Credit “endorsed free banking mainly by implication” before drawing upon Mises’s subsequent writings to buttress his conclusion. Undoubtedly Mises, like Sennholz, favored unregulated banking and, by the last edition of Human Action (1964, 443), was skeptical of any legal imposition of 100 percent reserves, as advocated by Rothbard. But we must carefully distinguish between favoring free banking as a legal regime and predicting how it would operate in practice. I think Larry goes too far when he seems to imply that Mises had in mind the kind of free banking that he (1999) and George (1988) predict would emerge without regulation: that is, a system in which reserve ratios are extremely low and banks adjust the money supply to demand in a way that stabilizes velocity. As much as I may agree with their prediction, I can assure them that Sennholz repeatedly affirmed his belief that unregulated competition among banks would drive reserve ratios up very high and possibly close to 100 percent, and he left the impression that such was Mises’s opinion as well.

Mises himself was never entirely clear whether or not he advocated free banking as a means of approximating Rothbardian ends, neither in The Theory of Money and Credit nor in Human Action. Yet revealing is his division of “money substitutes” (i.e., inside money) into “money certificates,” that portion 100 percent redeemable for outside money, and “fiduciary media,” the remaining portion exceeding the amount of outside money. The passage that Larry cites where Mises (1953, 312 [note my pagination is different from Larry’s because I am using a different printing of the same edition]) admits that free banking might “help stabilize the objective exchange-value of money” appears to be no more than a minor concession to the Banking School’s penchant for currency elasticity rather than an expression of something Mises found economically desirable. And in Human Action (1966, 443) he evinced an unambiguous desire to keep fiduciary media tightly constrained: “Free banking is the only method available for the prevention of the dangers inherent in credit expansion. It would, it is true, not hinder a slow credit expansion, kept within very narrow limits, on the part of cautious banks… . But under free banking, it would have been impossible for credit expansion to have developed into a regular—one is tempted to say normal—feature of the economic system. Only free banking would have rendered the market economy secure against crises and depressions.”

The key phrase, “credit expansion,” did not appear in The Theory of Money and Credit until Mises added Part Four to the 1952 edition. But it relates to Austrian business cycle theory, which Mises had first sketched out in the original edition of that work. I have never been entirely comfortable with this theory (Hummel 1979), despite believing that it contains some penetrating insights. Mises defined credit expansion as any net increase in fiduciary media in Human Action (1966, 434). Although he earlier had conceded that such increases through fractional-reserve banking had, as world output grew, prevented the “undesirable consequence” of a “tremendous increase in the exchange-value of money” (i.e., significant secular deflation), with its “additional harm” of increasing commodity money’s resource cost (1953, 298–99); in the final edition of Human Action he is insisting that “the trade cycle” can arise in a “pure market economy” from any credit expansion (1964, 573–74). This tension simply highlights one major difficulty with Austrian business cycle theory, no matter which variant from its assorted advocates we examine. It is a theory that hinges on specifying two firm dividing lines: (a) between those financial instruments that constitute inside money and those comprising what Mises considered genuine manifestations of people’s savings, and (b) between those increases in the money stock, however defined, that generate a self-reversing boom and those that do not. Alas, after myriad attempts, no consensus has emerged on either question among Austrian economists.