Real balances, the price level and the unit of account: from Wicksell to Patinkin and beyond - Special Invited Issue: Money, Trust, Speculation and Social Justice - Part 2: Trust and Money - impact of Knut Wicksell's book on value of money on Don Patinkin's work on price level

American Journal of Economics and Sociology, The, Oct, 1998 by Mauro Boianovsky

I

Introduction

This paper is an inquiry into how Knut Wicksell's Interest and Prices: A Study of the Causes Regulating the Value of Money (originally in German, 1898; translated 1936 by Richard F. Kahn) influenced Don Patinkin's Money, Interest and Prices: An Integration of Monetary and Value Theory (1956; 2nd ed. 1965; 2nd ed., abridged, with a new introduction, 1989). Patinkin (1956, p. xix) wrote in the preface to the first edition that "from the text the reader will . . . see how much my thinking has been colored by Knut Wicksell's classic Interest and Prices." In the same vein, in a symposium held in Frostavallen (Sweden) in 1977 to celebrate the 50th anniversary of Wicksell's death, Patinkin (1978, p. 135) stated that "from the days that I first read Wicksell's works over twenty-five years ago, I have in a sense always felt that he was one of my teachers." Furthermore, in his 1967 "Wicksell Lecture," Patinkin (1972b, p. 143) acknowledged the "specific and significant intellectual debt which I feel I owe to Wicksell's works in monetary theory." He did not make clear the precise character of his debt to Wicksell, however. Because the "central message" of Money, Interest and Prices is the integration of monetary and value theory by means of the "real balance effect" (see Trescott, 1989; Fischer, 1993), and because, moreover, Patinkin (1965, pp. 19, n. 12, and 581-2) pointed out that Wicksell's ([1898b] 1936, pp. 39-40) description of the real balance effect was "unique in the literature," one could imagine that Patinkin's reading of Interest and Prices was the source of inspiration for the role of the "effect" in his 1956 book. However, his 1995 autobiographical piece shows that this is not the case. Patinkin (1995, pp. 380-1) recalls how he "suddenly realized," while working on his 1949 Econometrica article (see Patinkin, 1949, pp. 21-7), how to make the commodity equations depend on the absolute price level, without any mention of Wicksell.(1)

I argue that the main influence on Patinkin of Wicksell's monetary theory in general (and of Interest and Prices in particular) can be found in the pivotal role of stability analysis - defined as "the nature of the corrective market forces that would be brought into play should the absolute price level deviate from its equilibrium level" (Patinkin, 1965, p. 186) - in Money, Interest and Prices. Stability analysis was not part of the articles Patinkin wrote in the late 1940s and early 1950s about the inconsistency that resulted from the "classical" assumption of zero homogeneity (in absolute prices) of excess demand functions for commodities (see Patinkin, 1965, p. 666). This changed in the book, where the main charge against traditional monetary theory was not the "invalid dichotomy" per se but rather the absence of stability analysis (see Patinkin, 1972a, pp. 281-2, n. 11, and 1965, chap. viii, sec. 1; cf. Johnson, 1969, p. 19; and Mizen and Presley, 1996, pp. 3940). At the same time, Wicksell - who had been included in the group of writers Patinkin criticized (see, e.g., 1949, p. 12, n. 5; and 1951, p. 149, n. 30) was - in 1956 - indicated to be the sole exception to the lack of stability analysis of the absolute price level in neoclassical monetary theory (1965, p. 168). Patinkin's change of mind about Wicksell's monetary thought can be traced to his correspondence with Dennis Robertson in the summer of 1951 (see Patinkin, 1954, p. 118, n. 2; and 1965, p. 586, n. 18; relevant parts of Robertson's letter of 25 July, 1951, are reproduced in Mizen and Presley, 1996, p. 37), who convinced him that there was no inconsistency between Wicksell's treatment of money as unit of account and as store of value on pp. 23-4 and 39-40, respectively, of Interest and Prices (see also Becker and Baumol, 1952, pp. 369-70). The first evidence in print of Patinkin's emphasis on the role of stability analysis in Wicksell can be found in his study of the Wicksellian "cumulative process" (1952; see, especially, p. 844, first paragraph) and in his 1954 (p. 114, top, and n. 1) reply to Becker and Baumol, but it was fully developed only in the book, where Patinkin (1965, chap. iii, sec. 3) based his own explanation of the forces that stabilize the price level on Wicksell's previous exposition (see also Patinkin's 1955 French summary of the argument of Part I of the book).

Patinkin, like Wicksell, first discussed the stability of the price level - and, by that, its determinacy - assuming an economy without bonds, which corresponds to what Wicksell used to call "pure cash economy" (see Patinkin, 1965, chap. iii; Wicksell, [1898b] 1936, pp. 39-41; and chap. 6, sec. A). In this case, the determinacy of money prices depends on the existence of a real balance effect in the commodity markets. When the analysis is extended to an economy with bonds (issued by individuals only), the real balance effect can also be found in the bond market (Patinkin, chap. iv); this corresponds to Wicksell's (chap. 6.B) "simple credit" case. Wicksell did not discuss in detail the determination of the price level in such an economy (see [1898b] 1936, p. 62; and, especially, 1904, p. 102), but focused instead on his third case, an "organised credit economy" ([1898b] 1936, chap. 6.C), with bonds issued by banks also. The banking system is conspicuous by its absence in the first part of Patinkin's 1956 book; it is introduced in the second part ("macroeconomics") in connection with Patinkin's interpretation of "classical" monetary theory in general and Wicksell's cumulative process in particular (1965, pp. 241-4). The interest rate is now a crucial factor in the determination of the price level, especially under the assumption that bank deposits have completely replaced the individual's cash balance as a store of value and, consequently, all transactions can be effected by checks and bookkeeping transfers, as in Wicksell's "pure credit economy" ([1898b] 1936, pp. 68-74). Patinkin (1956, chap. viii, sec. 1, p. 86, excluded from the second edition; cf. 1965, p. 195) generally regarded the pure credit system as essentially a barter economy, with prices measured in terms of an abstract unit of account. In the second edition, however, as part of the new sections on the effects of a banking system (1965, chap. xii, secs. 5 and 6) introduced as a reaction to Gurley and Shaw's (1960) distinction between inside and outside money, Patinkin (p. 303) treated Wicksell's pure credit case as a monetary (pure inside-money) economy. In this case, the profit-maximizing behavior of banks and the absence of reserves cause indeterminacy of the price level. While this is closer to Wicksell's meaning, it should be noted that Patinkin (and other commentators as well) overlooked the fact that Wicksell usually dealt with a pure credit economy in which gold was still the standard of value. Under these circumstances, the price level is governed by the industrial demand for gold as a commodity, which decides the marginal cost of production of gold. As Wicksell ([1898b] 1936, p. 24) made clear, the stability of the price level is then decided by the demand for base money as a commodity, not monetary demand as store of value. As we shall see, the Swedish economist put forward a proposal to replace gold for "the unity in which the accounts of banks are kept" (1907b, p. 218) as a measure of value, assuming absence of demand for outside currency. Wicksell's concept of a "pure credit economy" has been compared recently (see, e.g., Cowen and Kroszner, 1994, pp. 143-7) to the accounting systems of exchange of the so-called "new monetary economics" of Black ([1970] 1987) and Fama (1980), but, in contrast with these authors, he did not envisage a separation between the unit of account and the medium of exchange. Wicksell's suggestions on how to render a pure credit system determinate are discussed below and compared to Patinkin's (1972b; 1965, chap. xii, sec. 6) notion of the "minimum requirements" of a monetary system. t, II