The purpose of this chapter* is two-fold, namely: not merely to criticize the ‘real balance’ approach to employment but also to demonstrate a theoretical possibility of underemployment equil­ ibrium in terms of expectation functions. Such a demonstration may throw further light on the fundamental question of whether the price mechanism in general and the interest mechanism in particular are or are not capable of automatically restoring a full-employment; posi­ tion. That question has been revived recently by the ‘liquidity trap’ v. ‘real balance effect’ debate between Professor J. R. Hicks and Pro­ fessor D. Patinkin in the pages of the Economic Journal.1 The basic difficulty with the Hicks-Patinkin debate is that both of

them have, wittingly or unwittingly, forgotten Keynes’ first lesson in monetary dynamics: that ‘changing views about the future are capable of influencing the present situation’.2 As a consequence, Hicks has found it necessary to invoke the Keynesian assumption of an infinitely interest-elastic liquidity-preference function in order to prove his case of the ‘liquidity trap’, while Patinkin has found it expedient to invoke the Pigouvian assumption of a real-balanceelastic saving function in order to provide a neo-classical escape from the Keynesian-Hicksian ‘liquidity trap’. It will be shown that the former assumption is as unnecessary as the latter assumption is in­ sufficient in proving the case against and for automatic full-employ­ ment equilibrium. The essence of the Hicks-Patinkin debate seems to be as follows:

Hicks argues that the falling wages and prices needed for the ‘real balance effect’, while they reduce the interest rate (via their primary effect of decreasing the demand for transaction-money and hence of increasing the supply of speculative or asset money, given the fixed quantity of money), nevertheless provoke an infinitely interestelastic liquidity function (LM curve) at a low interest rate thus

brought about so as to make it intersect the investment-saving function (IS curve) at a point corresponding to an underemployment equilibrium income. Accordingly, the economy is trapped in that underemployment equilibrium position, inasmuch as the infinitely interest-elastic liquidity function prevents any increase in the total quantity of money from lowering the interest rate further to a level for which investment can equal full-employment savings. This is where Patinkin comes to the rescue. Patinkin counter-argues that the falling wages and prices also have the effect of increasing real balances and hence of reducing savings (via the ‘Pigou effect’) into equality with full-employment investment. Analytically this ‘real balance effect’ is shown as shifting the IS curve upward and rightward to intersect the LM curve at a point corresponding to the full-employ­ ment equilibrium values of interest and income (see Fig. 2 in his ‘Rejoinder’, ibid., p. 585). Thus the economy has presumably extri­ cated itself from the Keynesian liquidity trap. This is as far as the Hicks-Patinkin debate takes us. It is necessary

to go further in order to see whether a regime of complete wageprice flexibility, such as envisaged by Pigou and Patinkin, may not affect the behaviour of investment, saving and liquidity-preference functions in a destabilizing manner. For this purpose we would have to introduce expectations as explicit variables in the saving, investment and liquidity-preference functions based on dynamic assumptions. Let us build an alternative model of an economy with wage-price expectations as well as wage-price flexibility. Such a model can be described by means of the following system of

difference equations and behavioral assumptions:

( U )

< 0> (M ) It = St; Yt = Y V t , S t), (1.5)

MJpt = M/pt, (1.6)

Lt = L ( r t, Yt, ~ ^ Wt}, (1.7)

d(log w) < ° ’ (1,8) MJpt = L t; rt = r \ L t, M t\pt). (1.9)

Here the subscripts t and t + \ denote the variables in terms of the present period and the next period; I and S stand for real investment and real savings; Mjp and L are the amounts of real balances sup­ plied and demanded, M being the fixed quantity of nominal balances held by the public; Y° and r° are, respectively, the equilibrium values of real income ( Y) and the interest rate (r); log w and log p represent percentage changes in the money wage-rate and the price index. Here we suppress the labour market equation on the simplifying assump­ tion that employment is a linear function of income, so that the ‘full-equilibrium’ position of our general system consists in the simul­ taneous solution of equations (1.5) and (1.9) for r and Y, with the other independent variables serving as shift parameters. The key vari­ ables of this system are the expected percentage changes in the money wage-rate and the price index, (wm - wt)lwt9 (pt+1 - p t)/pt, in equations (1.1), (1.3) and (1.7) corresponding to the crucial assump­ tions expressed by (1.2), (1.4) and (1.8). Equation (1.1) gives current investment as a function of an ex­

pected percentage change in the money wage-rate as well as of current interest and income, while (1.2) specifies investment as an increasing function of a percentage change in the money wage-rate. This latter functional shape is based on the assumption that private investment would be profitably withheld in anticipation of lower and lower labour cost, and vice versa. In the present context of unlimited wage falls this means that investment will decrease with wages over time, the amount of that decrease depending on the prevailing elasticity of wage expectations. Equation (1.3) gives current savings as a function of an expected percentage change in the price index as well as of current interest, income and real balances (a decreasing function of real balances via the ‘Pigou effect’ assumed), while (1.4) specifies savings as a decreasing function of a percentage change in the price index. This last relation, in the present context of complete down­ ward price flexibility, means that the consuming public would with­ hold current consumption expenditure in anticipation of lower and lower prices and so increase current savings, the amount of that withholding depending on the prevailing elasticity of price expecta­ tions. Equation (1.5) states the condition for equilibrium in the com­

modity market and the implied dynamic determination of equil­ ibrium income. Equation (1.6) gives the supply of real balances as a function of a constant stock of money and current prices, while equation (1.7) gives the demand for real balances as a function of an expected percentage change in the money wage-rate as well as of interest and income. (1.8) specifies the demand for money in real terms as a decreasing function of a percentage change in the money wage-rate, implying that in the present context of complete downward

wage flexibility the public’s demand for real balances would rise in anticipation of further wage falls and greater uncertainties that accompany those wage falls. Equation (1.9) states the condition for equilibrium in the money market and the implied dynamic determina­ tion of the equilibrium rate of interest. The equilibrium solutions of equations (1.5) and (1.9) for Y and r would represent an under­ employment position characterized by

where Yf is full-employment real income. I (4.10)

MJpt = L t; r°(Lt, M j p t) > r„ j where rf is full-employment interest. This position of underemployment equilibrium expressed by (4.10) can be further clarified by reference to Fig. 1.