RDP 1978-01: Two Essays in Monetary Economics 2. Arguments that “Money” has a Special Role in Economic Dynamics, and Some Implications

Much of the literature on economic fluctuations is concerned with the implications of monetary disequilibrium, and the empirical work stimulated by this literature has established a strong a priori case for the importance of “money”.^[1] The word money is emphasised because the literature has not been precise enough to define the most appropriate measure, and as a result many definitions have been used in empirical work. Similarly, it has not been convincingly argued that a particular class of liquid assets, rather than all assets (wealth) is the appropriate empirical counterpart to various theoretical constructs.

At the most general level, there would appear to be a broad measure of agreement about the transmission mechanism, at least in qualitative terms. What I regard as the consensus view focuses on the interrelationship of stocks of assets and flows of expenditure, and envisages a complex chain of responses following a monetary or other disturbance. The view that money is just one of a vector of assets, and that disequilibrium in all asset markets is in principle relevant for the transmission mechanism, is put by Friedman and Schwartz:

“It is this interconnection of stocks and flows that stretches the effect of shocks out in time, produces a diffusion over different economic categories, and gives rise to cyclical reaction mechanisms. The stocks serve as buffers or shock-absorbers of initial changes in rates of flow, by expanding or contracting from their ‘normal’ or ‘natural’ or ‘desired’ state, and then slowly alter other flows as holders try to regain that state.

In this stock-flow view, money is a stock in a portfolio of assets, like the stocks of financial assets, or houses, or buildings, or inventories, or people, or skills. It yields a flow of services as these other assets do; it is also subject to increase or decrease through inflows or outflows, as the other assets are.”^[1]

Somewhat ironically, in reviewing the Monetary History of the United States, Tobin puts the case for money having a special role when he says:

“Some day perhaps Friedman will tell us what transient income is used for. We know from his work on the consumption function that none of it is consumed. We now know that none of it is saved in monetary from. Does it all go into the stock market? A priori, I should have thought that money balances were a likely repository of windfalls. After all, Friedman and Schwartz define money as a ‘temporary abode of purchasing power …’ ”^[1]

In their well-known paper on “Pitfalls in financial model building”

Brainard and Tobin are even more explicit, saying that one of their pre-conceptions is that:

“… New saving is initially accumulated as demand deposits, later to be distributed among other assets if holdings of demand deposits are too large.”^[2]

This puts the case that “money”, defined in this case as demand deposits, is a buffer stock which soaks up discrepancies between income and expenditure in the short run. Disequilibrium in money balances will be distributed across portfolios, and, if the relevant portfolio includes consumer durables, excess money balances would be expected to induce additional consumer expenditure, with parallel effects in equations describing adjustment in other asset markets. The direct effects of monetary disequilibrium on the balance of payments, emphasised in the so called “monetary approach to the balance of payments”,^[3] could be interpreted as reflecting the net effect of monetary disequilibrium in all domestic markets.

The discussion in terms of a whole set of asset disequilibrium effects in Friedman and Schwartz (1963a) could be made consistent with the view that money has a unique role if the money supply process was specified in a way that made money the residual of all other decisions in the short run. The emphasis in Friedman's writings on the role of disturbances from the supply side of the money market is obviously consistent with this view, and considerations of risk and uncertainty suggest that it may be rational for economic agents to use an asset or group of assets as a buffer stock. If a temporary drop in income occurs, for example, a consumer is able to use his assets to maintain consumption; conversely, if there is an unexpected temporary increase in income the consumer may find it sensible to accumulate assets against the possibility of later shortfalls in income. If the increase or decrease of income turns out to be permanent, a persistent rundown or build up of the buffer asset will occur, signalling the desirability of altering consumption or other behaviour. This notion can readily be formalised in simple models with one good and one asset as in the seminal contribution of Archibald and Lipsey (1958), and the contributions of Friedman and Tobin cited above apply the same basic idea in a more general framework.

The question as to the appropriate definition of the buffer asset or assets arises naturally at this point. From a broad point of view, the existence of assets of any sort provides a buffer against disturbances, or more generally enable a preferred consumption path to be attained for any given income stream.^[1] This view would have an asset disequilibrium effect upon current expenditure defined in terms of all assets, or net wealth, rather than the particular set of liquid assets usually defined as “money”.

Set against this, however, is the role of money as a means of exchange; “Money buys goods and goods buy money, but goods do not buy goods.”^[2] Essentially, this argument for the uniqueness of money is based on the existence of information and transactions costs in an uncertain world.^[1] Economic agents use nonmonetary assets, which are generally more costly than money to convert into purchasing power, to attain a desired consumption path given foreseeable fluctuations in purchasing power. A stock of money balances will however provide a buffer against random or unforeseeable disturbances, and, as a corrolary, economic agents will allow money balances to be determined by the net outcome of all other decisions in the short run. This view would not assign any unique role to a particular measure of money; one could imagine a hierarchy of buffer stocks, with currency the first line buffer for daily transactions, current deposits the next, then savings deposits, bonds, etc. Depending on the size and duration of any given disturbance, the shock will be absorbed first by fluctuations in highly liquid assets, and less liquid assets will be adjusted if necessary. A signal that longer term decisions need to be adjusted is provided by deviations of holdings of the buffer asset from ‘normal’ or ‘natural’ or ‘desired’ values.

This view would leave the most appropriate definition of “money” to be determined by empirical criteria such as the most stable relation to nominal income and interest rates in the long run, and to fluctuations in expenditure, prices and the balance of payments in the short run. If the current arguments about the role of money in economic dynamics are correct, the relative stability in the short run of demand/functions for money estimated in the conventional way^[2] is not in general a useful criterion for determining the appropriate definition of money for economic analysis.

There have been few explicit tests of what the residual asset or assets are in macroeconomic models; indeed, systematic examination of a variety of alternative dynamic specifications would require an extensive series of whole model tests. One important piece of indirect evidence is provided by the tendency for demand functions for money estimated in the conventional way to break down when there are sizeable disturbances from the supply side of the money market. This result suggests that, although the traditional approach with economic agents adjusting to a desired demand for real balances with a convergent first order adjustment process, as in equation 1

may be more or less appropriate in relatively stable periods, it is likely to be seriously defective when there are sizeable changes in the money supply.

Disturbances to the supply of money are likely to be correlated with real disturbances, if only because of policy reaction to the real disturbances, and examples of periods when sizeable real and monetary disturbances have occurred include both world wars, the great depression^[1] and the early 1970's. Significantly, authors in various countries report that money demand equations estimated in the conventional way have broken down in recent times,^[2] and studies of the demand for money with longer runs of data need to include special variables for highly disturbed periods such as the world wars.

The approach which views money as a buffer stock has been implemented in some studies by replacing equation 1 by one in which the quantity of money is determined in the short run by changes in international reserves and in domestic credit, as in equation 2:

The particular arrangement of this identity depends on the hypothesis that governments find it convenient to use the printing press as a source of residual finance and that economic agents' in the private sector use decision rules in which money is a “temporary abode of purchasing power”.

Some authors, notably Darby (1972), have implemented the view that money is a buffer stock by specification of demand for money functions in which transitory income has a high weight. The answer to Tobin's question (cited on p.3 above) given by this approach is that economic agents choose to put a high proportion of any unexpected windfalls, or any divergence between income and consumption (with consumption dependent on permanent income) into money balances in the short run. This answer is not necessarily inconsistent with the notion, encapsulated by equation 2, that any level of money balances generated by decisions concerning the balance of payments and about domestic credit expansion will be held in the short run, and that divergences between actual and desired money balances will be adjusted through the sort of price, output and balance of payments effects to be discussed in more detail in sections 3, 4 and 5. A formal reconciliation of Darby's approach with that implied by equation 2 has not been carried out, however, and would be quite difficult.

In the foreseeable future, the appropriate dynamic specification for macroeconomic models may be decided by the simplicity of the alternative approaches. It is clear that substitution of the determinants of the balance of payments and of domestic credit expansion into equation 2 would produce an equation for the quantity of money with current and permanent income, lagged real money balances, domestic and world prices and interest rates, which could be interpreted as a (highly complicated) demand function for money. It is noteworthy that a tendency towards demand functions for money which are more complicated is observable in the conventional studies mentioned in footnote 2, page 6. The alternative approach, which views money as being determined residually in the short run, with relatively simple theories for the determinants of balance of payments and domestic credit expansion, is simpler and also more consistent with views about the adjustment mechanism in terms of a disequilibrium real balance effect which have been discussed above.

To summarize, the view that money is used, as a buffer stock implies, firstly, that studies of the demand for money function which seek to identify money demand in equations (such as (1) above) with the level or changes in the quantity of money on the left hand side are likely to be seriously deficient or extremely complicated when there are major disturbances from the supply side of the money market. A second implication is that the gap between actual and desired money balances is likely systematically to influece expenditure decisions and perhaps other decisions as well. This is not to deny that disturbances in other asset markets might not be important in particular episodes, and that the most general dynamic model might, not have a whole vector of asset disequilibrium effects influencing all economic decisions.¹ Nevertheless, for most of this essay, the focus will be on the role of monetary disequilibrium, as the argument that this has a unique role in economic dynamics has relatively simple and apparently relevant implications for the adjustment of output, prices and the balance of payments.

Footnotes

For broad surveys of empirical work on the effects of monetary variables, see Fisher and Sheppard (1972), Brainard and Cooper (1975) and Laidler (1977). [1]

Friedman and Schwartz (1963a), page 234. [1]

Tobin (1971), page 487. [1]

Brainard and Tobin (1968), page 352. [2]

Discussed, for example, in Johnson (1972) and Frenkel and Johnson (1976); mention should also be made of the early work of Polak (1957), Prais (1961), and the various contributions of Mundell (e.g. 1968) and his students. [3]

As in the path-breaking analysis of optimal savings behaviour carried out by Ramsay (1926). [1]

Clower (1967), p.5. [2]

See Brunner and Meltzer (1971) and Laidler (1974) for relevant discussions. [1]

As in equation (1) below. [2]

The money supply disturbance at this time was especially important for the U.S. economy. [1]

See Adams and Porter (1976) for Australia, Artis and Lewis (1976) for the U.K., and Goldfeld (1976) and Enzler, Johnson and Paulus (1976) for the U.S. [2]