The General Theory
The General Theory of Employment, Interest and Money appeared in February 1936. It immediately found a wide readership, although not repeating the success of The Economic Consequences of the Peace. However, it did have a more solid influence, concentrated in the field of professional economists: many young scholars soon adopted it as a basic reference point in their own research work and teaching.
The General Theory is not an easy book, and many ‘Keynesian’ economists did not read it. Only this fact could have rendered possible philologically untenable interpretations, like the idea that Keynesian theory was based on the downward rigidity of wages and prices (when chapter 19 of the book explicitly rejected this idea).
First of all, let us recall Keynes’s aims and his views on the economy. Defence of a liberal political system based, among other things, on freedom ofindividual initiative in the economic arena required, according to him, that the limits ofthe pure laissez-faire system be recognised; hence the need for active intervention of the state in the economy, in the interests not only of fairness but also of overall efficiency. Economic agents take their decisions under conditions of uncertainty, as defined in the Treatise on Probability. At the methodological level, this led to preference for open models, specifically designed for the problem under consideration, to be built with caution, and pondering the conditions under which individual causal relations hold. Various other aspects of Keynes’s theory also derived from uncertainty, such as the role of financial markets, which not only acted as intermediary between agents with active and passive financial positions but also, and above all, provided flexibility for inter-temporal choices and allowed entrepreneurs to take decisions on production levels and investments concerning the future. It is in this context that we find the distinction between short- and long-run expectations, the former concerning choices on current production, such that they can be promptly adjusted to results, the latter concerning decisions on investment in fixed capital; for these the impact of uncertainty is indeed strong.
The analytical structure of the General Theory rested on three pillars: the notion of effective demand, the multiplier mechanism and the theory of interest. All these are well-known aspects, but they occasionally suffer some distortion - the first and third in particular - when illustrated in university textbooks, so let us take a brief look at them here.
Chapter 3 of the General Theory is devoted to the principle of effective demand. The point of effective demand is defined as the point of encounter of two curves: an aggregate supply function and an aggregate demand function, conceptually different from traditional supply and demand curves, since they relate the entrepreneurs’ evaluations regarding costs on the one hand and receipts on the other to the number of employed workers. More precisely, the aggregate supply function Z indicates the minimum expected proceeds necessary to persuade entrepreneurs to employ N workers, while the aggregate demand function D indicates how much entrepreneurs expect to earn by selling on the market the product they hope to obtain through the employment of N workers. Both curves thus express the point of view - the evaluations - of the same category of economic agents, the entrepreneurs, not of two distinct and opposed groups of buyers and sellers (consumers and producers).
Both expected costs and expected proceeds increase with the number of employed workers. Thus both functions are increasing ones. However, Z increases ever more rapidly (its second derivative is positive), while D increases ever more slowly (its second derivative is negative). D is in fact made up of two components, consumption and investment; according to Keynes, because of a psychological law consumption increases less than income, and hence than employment, while investments depend on the entrepreneurs’ long-run expectations, so that they may be considered as given in the context of determination of the point of effective demand.
As far as Z is concerned, in the Marshallian context of Keynes’s theory it was natural to assume that when the number of employed workers increased (while, in the short-period context, it was assumed that the productive equipment remained unchanged), the marginal cost turned out to be increasing.
The point of effective demand is the one at which D = Z. It thus tells us the expected level of employment, and hence of production, given the entrepreneurs’ short-run expectations regarding costs and proceeds.
Decisions concerning consumption and investment are taken by different categories of economic agents (respectively, families and firms) and thus follow two completely different logics. Consumption (and savings, defined as their complement to income) essentially depend on income. Investments depend on the entrepreneurs’ decisions (hence on their expectations) and are independent from income. As a consequence, it is investment decisions that determine the equilibrium level of income, while savings adapt. More precisely, equilibrium income has to be such as to generate an amount of savings corresponding (in the simplified system without taxes and public expenditure and with no relations with foreign countries) to the amount of investments generated by entrepreneurs’ decisions. It thus depends both on the level of investments I and on the propensity to save s (s = S/Y, where S are savings and Y is income); more precisely, on the equilibrium condition I = S (equality between inflows and outflows in the circular income flow) and on the definition of the propensity to save we get Y = I/s: the multiplier, namely that multiplicative coefficient which, when applied to the level of investment, gives equilibrium income, is equal to the inverse of the propensity to save. When identifying the multiplier as the second of the three pillars of the General Theory, we are referring not simply to this equation but to the active role attributed to investments and the passive role attributed to consumption and savings in determination of income.
For the theory of investment, as for that of effective demand, Keynes adopted the entrepreneur’s point of view. The latter decides whether to invest by evaluating expected returns on investment and comparing them with the monetary rate of interest indicating return on financial investments, which constitute an alternative employment of available funds. As pointed out previously, expectations relevant to investment decisions concern the ‘long period’, since they cover the whole foreseen life of the productive equipment acquisition of which is under consideration, and decisions taken on their basis may be revised within such a time interval only at high costs, while the expectations relevant for decisions on current production levels and employment concern the ‘short period’, open to ready revision with relatively low if not zero costs. Long-period expectations are not stable; on the contrary, precisely because they concern so long an interval of time as to elude sufficiently precise and reliable evaluation, they are far less stable than short-period expectations.
The third pillar of Keynes’s General Theory was represented by the theory of monetary and financial markets, with the rate of interest conceived as premium for foregoing liquidity. This theory has two main aspects, often misinterpreted in macroeconomic textbooks. Firstly, behind the mass of large and small savers deciding what form to keep their financial assets in loom the financial institutions, the true protagonists of the decision-making process described by Keynes. Secondly, the decision-making process itself does not concern flows, on which traditional theory focussed, but the allocation of stocks; thus the speculative demand for money dominates the transaction demand for money. On well-developed financial markets transaction costs are very low, and it is possible to revise daily, or even from one hour to the next, decisions on the allocation of financial holdings between the various possible assets.
Simplifying the issue, Keynes considered two kinds of assets: money, extremely liquid since commonly accepted for all kinds of transactions but not yielding income, and bonds yielding a predetermined yearly coupon. As we know, the market price of pre-existing bonds increases when the rate of interest decreases, and vice versa. As a consequence, those who expect a fall in interest rates by the same token also expect an increase in bond prices and will be buyers on the bond market, while those expecting an increase in the rate of interest operate in the opposite direction, offering bonds in exchange for money. In the presence of different opinions on the prospects facing the monetary and financial markets, the rate of interest is set at each instant at that level which corresponds to equilibrium between the two opposite ranks, the ‘bulls’ and the ‘bears’.
Thus, everything depends on the expectations of the financial operators. If for a moment we assume that these remain fixed, it is clear that when the rate of interest decreases, the number of operators who expect a subsequent increase (and thus offer bonds in exchange for money) rises: the demand for money thus turns out to be an inverse function of the interest rate. However, this relationship has very thin foundations, since expectations regarding financial events are extremely volatile. It is quite possible, for instance, for a reduction in the interest rate to induce many operators to revise their expectations and foresee further interest rate reductions, preferring bonds to money even more than before: a direct, rather than inverse, relationship would then hold between changes in the rate of interest and changes in the demand for money.
Within Keynes’s analytical framework, the theory of speculative demand for money distanced interest rate determination from the traditional mechanism of comparison between savings and investments, respectively understood as supply of and demand for loanable funds. According to Keynes, decisions to save are logically distinct from those concerning the kind of financial asset (money or bonds) in which to invest the savings. Contrary to the interpretation advanced by many commentators, the main point was not that the amount of savings depended more on income than on the rate of interest - a point also acknowledged by a theoretician like Pigou, chosen by Keynes as paradigm of the traditional theory he was attacking. The point was the separation between the two kinds of decisions concerning, respectively, the amount of savings and the financial asset to invest the savings in; it was this latter decision that concurred together with the monetary policy followed by monetary authorities in determining the current level of the interest rate.
Also Hicks’ idea, embodied in his famous IS-LL model (Hicks 1937), to set transaction demand and the speculative demand for money side by side, as if they were on the same plane, lost sight of the difference in nature between the two kinds of decisions. Speculative choices concern the allocation of the stocks of savings cumulated over time and thus dominate over the transaction demand for money, namely the liquidity requirements to finance the flow of current exchanges. This is all the more evident when the stocks of savings to be allocated between bonds and money are confronted not with yearly income and exchanges but, as is in the nature of continually revised financial choices, with daily flows. We thus have a hierarchy of influences: financial expectations dominate the allocation of the stock of savings, and hence the determination of interest rates, relegating to a secondary level all other factors, including the transaction demand for money. It is, then, the interest rates thus brought about, together with long-run expectations, that determine the level of investments, while the latter in turn, through the multiplier mechanism, determines income and employment.
This scheme of hierarchical relations was in sharp contrast to general economic equilibrium schemes, in which each variable depends on all other variables and on all the parameters of the system. It is precisely in this aspect that Keynes’s theory, following through with the short causal chains methodology, fully revealed its Marshallian foundations, emphasised by the pragmatism characterising all Keynes’s work. And, indeed, it is this aspect that has been submerged in the interpretations of Keynes’s thought dominating successive generations of macroeconomics textbooks.
-  The assumption of downward rigidity of wages and prices maybe utilised to get ‘Keynesian’results in a different analytical framework, as was the case with the so-called neoclassicalsynthesis, but not to interpret ‘what Keynes really meant’.
-  This implied an inverse relation between real wage and employment analogous to the onepostulated by Marshall and all versions of marginalist theory in support of the thesis of anautomatic tendency to full employment. In Keynes’s theory, which rejected this adjustment mechanism, the inverse relation was not essential; in fact, Keynes was ready toabandon it when confronted with Dunlop’s 1938 and Tarshis’s 1939 empirical criticismsand a sizeable mass of empirical evidence on the pro-cyclical nature of real wage movements. Indeed, as is obvious, abandonment of that assumption reinforces the Keynesiancritique of the thesis of an automatic tendency towards full employment equilibrium.
-  Therefore, it should not be interpreted as a point of equilibrium between two oppositeforces of demand and supply, let alone as a stable equilibrium, as macroeconomicsmanuals have long done. Keynes’s viewpoint and the textbook one only coincide underthe assumption that the entrepreneurs’ short-period expectations are always fulfilled, sothat expectations and uncertainty exit the scene, while Keynes’s thesis that supply (production) adapts to demand remains.