The 100% money proposal and its implications for banking: the Currie–Fisher approach versus the Chicago Plan approach

ABSTRACT

The literature on the 100% money proposal often reveals some confusion when it comes to its implications for the banking sphere. We argue that this can be partly explained by a failure to have distinguished between two divergent approaches to the proposal: the “Currie–Fisher” (or “transaction”) approach, on the one hand, which would preserve banking; and the “Chicago Plan” (or “liquidity”) approach, on the other hand, which would abolish banking. This division among 100% money proponents stemmed, in particular, from different definitions of money, and different explanations of monetary instability. The present paper attempts to clarify this divergence of views.

KEYWORDS:

• 100% money
• banking
• Chicago Plan
• Lauchlin Currie
• Irving Fisher

JEL CODES:

• B22
• B31
• E42
• E50
• E51
• G20
• G21

Acknowledgments

I would like to thank Juan Carlos Acosta, Robert W. Dimand, Rebeca Gomez Betancourt, Laurent Le Maux, Ronnie J. Phillips, Roger Sandilands, and Adrien Vila for helpful comments or discussions in relation with this paper. I am also grateful to two anonymous referees for their valuable remarks and recommendations.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

¹ The idea that private banks should be prevented from creating money can be found in the writings of David Ricardo already, as well as in the Currency School arguments which led to the adoption of the English Bank Charter Act of 1844. All those writers, however, focused on the issuance of bank notes, and failed to consider transferable bank deposits as money. The proposal of a 100% reserve requirement behind checking deposits appeared in the mid-nineteenth century, in the United States, with authors such as Charles H. Carroll (see Mints 1945, 154–156). Léon Walras, in 1898, also suggested that all checking accounts be kept in a central “Transfer Bank” [“Banque de virements”], with 100% reserves behind them, so that “monetary circulation would not be disturbed by the arrangements of credit” [“que la circulation monétaire ne soit pas troublée par les combinaisons du crédit”] (Walras 1898, 376, 395, my translation). Early twentieth century advocates of 100% reserves included Frederick Soddy ([1926] 1933, 229–231), whose proposal, based on “fiat money” reserves, might have influenced the Chicago Plan directly (Phillips 1995, 46).

² This group included Garfield Cox, Aaron Director, Paul Douglas, Frank Knight, Albert G. Hart, Lloyd Mints, Henry Schultz, and Henry Simons. They circulated three memoranda in 1933 (see Phillips 1995, 47–68): the first in March (Knight et al. [1933] 1995), the second – a revised version of the first by Simons – in April (unpublished), and the third – mainly the work of Simons – in November (Simons et al. [1933] 1994). As noted by Laidler (1999, 231), the term “Chicago Plan” was coined by Hart (1935), whereas “100% money” was Fisher's phrasing.

³ See, for example, Huber and Robertson (2000), Kotlikoff (2010), Benes and Kumhof (2012), Jackson and Dyson (2013), Sigurjónsson (2015), Levitin (2016), or Huber (2017). Patrizio Lainà (2015) has recently provided a historical overview of what he terms “full-reserve banking proposals”.

⁴ One can also refer to Phillips (1988) and Whalen (1994) for discussions about the Chicago Plan; to Sandilands (2004) about Currie's plan; and to Allen (1993), Dimand (1993) and Fisher ([1935] 1997b, editorial content) about Fisher's plan.

⁵ “Banking” could be more properly defined as consisting of two sets of activities, one related to the administration of the payment system (deposit-keeping and transferring), the other to the intermediation between savers–depositors and investors–borrowers (deposit-lending). In this paper, however, we tend to use the term to refer to this latter activity specifically.

⁶ One can refer to Table 5, in the concluding section, for a summary of these divergences.

⁷ These divergences between the “transaction” and “liquidity” approaches can also be found in the proposals for “narrow banking” which appeared in the 1980s (see, for example, Litan 1987).

⁸ For discussions regarding the implementation of the 100% money proposal under its various versions, one can refer, for example, to Hart (1935), Watkins (1938), or G. R. Barber (1973).

⁹ For Fisher (as for Simons), the stakes were even higher: “[t]he best available safeguard against the overthrow of capitalism is the 100% system, combined with money management, to give us a stable dollar. Of all people, bankers should, therefore, favour this proposal if only in self-defense. Otherwise, by the irony of fate, they may someday be the ones to upset capitalism” (Fisher, [1935] 1997b, 219).

¹⁰ Those should be distinguished, however, from the 100% gold reserve proposals, which have been developed by economists of the Austrian school in particular – see, for example, Mises ([1912] 1953, 408) or Rothbard (1962). One can refer to Huerta de Soto ([1998] 2012, 715–735) for a historical overview of the 100% gold reserve theory – which, although its analysis of economic instability is most interesting, falls outside the scope of this paper.

¹¹ The 100% money idea should not, either, be confused with the “modern monetary theory” (MMT) expounded by authors such as Wray ([2012] 2015). Indeed, while stressing the monetary role played by the state, MMT insists that the banks should remain involved in money creation too.

¹² Because the state, via its monetary authority, would be buying its own bonds, it was usually argued that the public debt would not be aggravated – and could even be reduced – by money injections.

¹³ This definition was implicitly contained, but not explicitly formulated, in his book 100% Money.

¹⁴ For the sake of simplification, we limit our analysis here to those two kinds of money. Historically, however, other instruments have been used as means of payment, such as various kinds of commodities, commercial paper, shares of money market mutual funds, or, more recently, cryptocurrencies.

¹⁵ This quantity of reserves used by the banks to cover their checking deposits (equal to MoMʹ), however, does not represent the totality of bank reserves, insofar as part of those are kept for other purposes – for example, as reserves behind savings deposits, or for the banks’ own transaction needs.

¹⁶ Allais thus defined the money supply as: M(t) = M₁(t)+*(t, θ)p*(t, θ)dθ + _i(t)q_i (t), “where M₁(t) is the volume at time t of the basic money in circulation outside the banking system plus demand deposits held by private individuals; p*(t, θ)dθ is the volume of time deposits at time t whose term lies between θ and θ + dθ; *(t, θ) is the average of the corresponding substitutability ratio, and the q_i(t) are assets other than deposits at time t with rates of substitutability against cash of σ_i(t)” (Allais 1987, 508–509, original italics; see also 1975, 126).

¹⁷ With the exception, however, of Milton Friedman, whose arguments for 100% reserves fell under neither of the two approaches discussed in this paper. Friedman assigned the “inherent instability” of the fractional-reserve monetary system only to the “decisions by holders of money about the form in which they want[ed] to hold money and by banks about the structure of their assets” ([1960] 1992, 66). He further criticised that system for “involv[ing] extensive governmental intervention into lending and investing activities” (Friedman [1960] 1992, 66). But he was not specifically opposed to the creation of money out of bank loans. Indeed, alternatively to 100% reserves, he was ready to consider completely opposite solutions: either to allow the banks “to issue currency as well as deposits” (Friedman [1960] 1992, 68–69), or “to permit ‘free’ deposit banking, without any requirements about reserves” (Friedman [1960] 1992, 108).

¹⁸ Fisher referred to the fractional-reserve money system as the “10% system”, in contrast with the “100% system” he was calling for.

¹⁹ With P here representing all prices (including, for example, asset or house prices), although, in practice, Fisher usually recommended the use of “a fixed index of the cost of living” as a criterion of stability (Fisher [1935] 1997b, 97).

²⁰ The constant instability of this ratio had been noted by Fisher in 1911 already ([1911] 1997a, 55–73).

²¹ On this point, Fisher's position would find support in empirical works. Clark Warburton, studying velocity changes over the period 1919–1947, concluded: “Factual information for the period since 1919 does not support the assumption that variations in monetary velocity are an initial factor in business depression. The data do, however, indicate that in some cases a declining velocity of money has accompanied and in other cases has followed downward deviations from trend in the quantity of money. In fact, after a business recession has run for a time and the quantity of money has been reduced, there is almost uniformly a slowing down in velocity, relative to trend, which is reversed only when the shrinkage in the money supply is known, or believed, to have been stopped. There is no evidence that disturbances to economic equilibrium originate in an erratic rate of use of money, but there is much evidence that such disturbances result in, and are in turn intensified by, variations from trend in the rate of use of money” (Warburton 1949, 91). Friedman and Schwartz (1963, 682), studying the period 1867–1960, reached convergent conclusions.

²² Fisher insisted that some discretionary powers of action should be left to the monetary authority, allowing it to successfully stabilise the general price level (Fisher [1935] 1997b, 24). He thus advocated what we would call today a “constrained discretion” for the Currency Commission.

²³ However, there did not seem to be a complete consensus among the authors on this point, as the following passage indicates: “some of us are inclined to feel that the disturbances occasioned merely by changes of velocity are unlikely to be of serious magnitude” (Simons et al. [1933] 1994, 42). Friedman (1967, 12) would also disagree with his former teacher on this issue: “The movements in velocity—which Simons took as an independent source of instability—come later than the movements in the quantity of money and are mild when the movements in the quantity of money are mild. They have been sharp only when there have been sharp movements in the quantity of money”.

²⁴ According to Whalen (1988, 541), “the cycle theories of Minsky and Simons share a number of essential features”, even though they are “not identical”. One difference is that “[u]nlike Simons, Minsky expresses his analysis without reference to the equation of exchange” (Whalen 1988, 536).

²⁵ This evolution of Simons's analysis seems to have escaped Friedman, who regarded “[w]idespread borrowing on short-term in order to finance long-term obligations” as the “key to instability” in Simons's view (Friedman 1967, 5).

²⁶ Of course, there were certainly very good reasons why, historically, promises to pay (issued either by banks or businesses) came to be used as means of payment – if only, to bring elasticity to the volume of money, which a metallic currency was ill-suited to provide. It follows that a major challenge for the monetary authority, under a 100% money system, would be to adjust the money supply flexibly enough to the volume of transactions. Otherwise, economic agents might be pressured to break the law and devise alternative means of payment. Discussing this essential issue, however, is beyond the scope of our present study.

²⁷ The distinction between Mo and Mʹ could even completely vanish, if all checking accounts were held on the books of the central bank directly – leaving place to a simple identity: M ≡ Mo. This was suggested by Tolley (1962, 299–300): “Let the deposit liabilities of the commercial banks be transferred to the Federal Reserve banks. If ‘reserves’ are defined in the usual way, Federal Reserve liabilities connected with deposit money, there would be 100 per cent reserves in the sense that deposit money and reserves would be identical. The physical arrangements in the use of money could be continued as at present, located in the commercial banks with servicing expenses paid for on a contract basis by the Federal Reserve banks.” Many of the most recent 100% money proposals have adopted this suggestion, although with differing practical arrangements.

²⁸ As one could have expected, the proposal was understood quite differently by the bankers in general, although several of them endorsed the plan. As Dimand (1993, 70) reported: “The banking community remained, however, generally hostile. Such writers as Robinson (1937) and Hackett (1945) viewed deposits from the standpoint of the banks whose liabilities they were, instead of concentrating on which deposits could serve as means of payments. Chequing and savings deposits both enabled fractional-reserve banks to channel savings to borrowers, so Robinson and Hackett saw no case for treating the two types of deposits so differently.”

²⁹ It seems, moreover, that had Fisher considered the variations of V as a leading factor of instability, his proposed solution would still not have been the abolition of banking. Instead, he would have considered a system of taxing the currency, on the stamp scrip model, so as to increase or decrease its velocity of circulation (see Fisher [1935] 1997b, 102). When asked if the “V” in the equation of exchange had not been neglected in his 100% plan, Fisher replied: “This is quite true. I, at one time, tried to introduce into the plan a tax method to control the influence of ‘V’. But I left it out as soon as I found that ‘V’ is really nearly constant under conditions which would prevail if the 100% plan were in operation” (Fisher, letter to Theodore Morgan, 25 September 1945, in Fisher 1997c, 242).

³⁰ For this reason, it is obviously misleading to refer to the 100% money concept as “full-reserve banking”. Fractional-reserve banking, under the Currie–Fisher approach, would still exist; only, there would be full-reserve money, separated from banking.

³¹ Fisher would, personally, have these requirements strengthened somewhat ([1935] 1997b, 13), while Currie would have them reduced to zero (Currie [1934] 1968b, 199; [1938] 2004, 361).

³² Fisher et al. (1939, 32), for example, recommended that savings deposits “should be withdrawable only upon adequate notice”. See also Currie ([1934] 1968b, 200) and Fisher ([1935] 1997b, 165–166) for other suggested safeguards.

³³ Of course, those deposits held by the loan department should be included in the money supply calculation, if one defined M as MoMʹ. But they would be excluded if, instead, one restricted M to currency in circulation plus checking deposits held by the non-bank public. Such limited definition led Robinson (1937, 442–445) and Watkins (1938, 440) to consider that M would still vary endogenously, under a 100% system, whenever money would be transferred to or from savings accounts.

³⁴ The old nineteenth century proposals seem to have followed this approach as well. Carroll, in 1860, had called for 100% reserves behind demand deposits, “but he would not have interfered with the operation of savings departments” (Mints 1945, 156). Under Walras's plan for a Transfer Bank, too, private banks would still finance loans and investments out of savings deposits: “Discount banks would receive interest-bearing deposits of a fixed term of one month, three months, one year, and have in their portfolios all the securities of the circulating capital; their cash balances as short-term credit entrepreneurs would be kept at the Transfer Bank” [“Les banques d'escompte recevraient des dépôts à intérêt à échéance fixe de un mois, trois mois, un an, et elles auraient en portefeuille tous les titres du capital circulant ; leur encaisse d'entrepreneurs de crédit à courte échéance serait à la Banque de virements.”] (Walras 1898, 396, my translation).

³⁵ However, Tobin's proposals for a “deposited currency”, which he developed in other writings (Tobin, 1985, 1987a), followed more specifically the concept of narrow banking, rather than that of 100% money. Under his plans, indeed, the liabilities attached to segregated funds, invested in eligible safe assets (other than cash), would still be allowed to circulate as means of payment (Tobin 1985, 27, 1987a, 173). The main objective of Tobin's proposals – like those of narrow banking – was to secure the payment system. In contrast, the main objective of the 100% money proposal was to end the “perverse elasticity” of the money supply; “safeguarding depositors” was seen by Fisher (1937, 296) as a major benefit of the reform, but of “secondary importance”.

³⁶ If one followed this approach, then a case could also be made for eliminating all “government facilitation of ‘safe asset’ creation by the shadow-banking sector”, which assets tend to be regarded as “free of credit risk and hence deposit-like”, as Levitin (2016, 417–418) recently argued.

³⁷ The paradoxical fact that Friedman, despite his analytical divergences with Simons, came to support an essentially similar banking scheme was noted by Phillips (1995, 208).

³⁸ In a later text, however, Minsky (1995, 8) would – like Tobin – tend to assimilate the concept of 100% money with the idea of backing checking deposits with safe assets (that is, narrow banking).

³⁹ Allais (1987, 498), designating by A*(t, θ) “the total amount at time t of the asset items maturing at or before time θ”, and by P*(t, θ) “liability items at time t falling due on or before θ”, formalised his proposed rule as the following: P*(t, θ) ≤ A*(t, θ), for any t and θ (Allais 1987, 525).

⁴⁰ With the exception, once again, of Friedman (1967, 3), who, despite his advocacy of the Chicago Plan reform scheme, seemed to regard this transformation activity favourably.

⁴¹ Here, Simons wrote “demand deposits”, but this was obviously a typing error.

⁴² See Neuman (1937, 62), Robinson (1937, 42), and Reeve (1943, 324).

⁴³ See Lehmann (1936, 44), Neuman (1937, 61), Robinson (1937, 440), Watkins (1938, 442), Brown (1936, 312), Thomas (1940, 315), and Reeve (1943, 324).

⁴⁴ See Robinson (1937, 442), Brown (1936, 312–313), and Higgins (1941, 94).

⁴⁵ See Lehmann (1936, 43), Neuman (1937, 62), Robinson (1937, 439), Watkins (1938, 445), and Thomas (1940, 317).

⁴⁶ See Neuman (1937, 61), Robinson (1937, 440), Thomas (1940, 315, 323), and, more recently, Goodhart and Jensen (2015, 23).

⁴⁷ See, for example, Wolf (2014, 210), Turner (2016, 10), King (2016, 262), or Glasner (2017, 32). An opposite confusion – that the Chicago Plan would, like Fisher's Plan, still allow banks to lend from savings deposits – can be found in Allais (1987, 523–524) and Levitin (2016, 419).

⁴⁸ The 100% money idea has been criticised, for example, by Turner (2016, 188–190) mainly on the ground that it would prevent the banks from performing maturity transformation, and by King (2016, 262–264) mainly on the ground that it would prevent them from performing risk transformation.

⁴⁹ Of course, this distinction should not be interpreted too rigorously, and any classification of 100% money proponents according to the two approaches should be made carefully. Indeed, some authors, while following one or the other approach about banking reform, would at the same time rather follow the opposite approach when it comes to defining money or explaining monetary instability.

⁵⁰ With M representing here the volume of the means of payment, and V their velocity of circulation.