RDP 8605: On Some Recent Developments in Monetary Economics 2. The Neoclassical Inheritance

Neoclassical economics provides a vision of the world in which resource allocation depends on endowments and preferences of consumers across the spectrum of commodities. Markets always clear, so that economic equilibrium can be described by relative prices alone.

Many of the classic questions of monetary economics concern the linkages between financial markets and markets for labour and commodities. With the possibility of significant imperfections in all markets, particularly for capital commodities and most forms of labour, description and analysis become complicated.

When markets clear, as in the well-developed Arrow-Debreu model, prices convey all information necessary for the optimal functioning of the economy. Models of non-clearing markets are less well-developed. However, it is known that when markets fail to clear, not only prices but also traded quantities contain useful information about the state of the market. This information can affect not only behaviour in a single market, but also behaviour in all markets in which affected parties are engaged. Disequilibria can therefore be spread between markets.

In the neoclassical market-clearing model, money is a “veil” which has no role in the real economic process (except possibly during phases of adjustment).[3] In providing a unit of account, its only function is to determine the absolute level of accounting prices in the economy. Monetary theory consists therefore of a demand or quantity equation for money which, given the stock of money and the level of real transactions, yields the aggregate price level.

The writing of Patinkin elaborated the process of absolute price determination through the real balance effect, described by Patinkin as the “sine qua non” of monetary theory.[4] Still, however, the demand for money remains the central issue: for a determinate price level in Patinkin's model, the money demand equation must be homogeneous in prices and income together.[5]

Monetary economics in the 1960s and 1970s was largely centred on the study of the demand function for money.[6] While empirical studies of the demand for money proliferated (almost invariably running simple regressions of a money measure on a succession of interest rate and income or wealth measures), analytical studies explored the effects of the properties of particular money demand functions on the behaviour of the relevant models (usually simple macroeconomic models of the Hicksian tradition).

The models of the 1960s and 1970s were not always clear (or even explicit) about the way in which changes in the supply of money (or more generally in “monetary policy”) influenced the economy. With controlled financial markets, changes in monetary policy were sometimes difficult to define. Models handled this in a variety of ad hoc ways. Most models also adopted an approach in which both real incomes and prices were assumed sticky in the short run.

If interest rates were free to move, the monetary “transmission mechanism” was often defined in terms of an initial impact on interest rates, which introduced changes in economic activity and then (with a long lag) in prices.

There were various problems with this approach. Interest rates and/or lending controls were sometimes relevant, which introduced so-called “credit rationing” effects. Even when interest rates varied, effects on real activity were difficult to pin down. Links from varying activity to inflation were not always well specified, although in the 1970s the so-called “expectations augmented Phillips curve” was a useful development. This raised the question of possible direct “money supply” effects on inflation or inflationary expectations.[7]

Throughout this period, the stability of the demand function for money was virtually an article of faith (despite evidence of other economic functions being variable and/or hard to measure). This was true even of models in which “monetary disequilibrium” was an integral part of the transmission mechanism.[8] In these models, stocks of money balances are held passively in the very short run but agents respond to the gap between the actual and the longer-term desired levels of money stocks. The stability of the longer-term demand function remains central, even though observed money need not correspond to this desired level.


Johnson (1978) and Niehans (1978) give a comprehensive account at the neoclassical theory and some extensions. [3]

Patinkin (1956). That the real balance effect was a process was first emphasised not by Patinkin, but by Archibald & Lipsey (1958). Implications for macroeconomic models of adjustment are discussed in Jonson (1976). [4]

Gale (1982) provides a useful discussion. [5]

The money demand literature is surveyed in Laidler (1985); Davis and Lewis (1978) and Veale et. al. (1985) cover the Australian literature. [6]

In the Australian literature two relevant studies were by Jonson and Mahoney (1973), and Valentine (1977). [7]

See Laidler (1984) for a survey of this field. [8]