Marginalist Theory
In marginalist (neoclassical) theory, the tendency to full employment rests on the two mirror constructs of demand functions in labour and capital markets. These factor demand functions are constructed on the basis of factor substitution in production and, indirectly, in consumption: a fall in the wage level was regarded as leading, in the long period, to a change of techniques in production which would entail, with the same quantity of capital incorporated in different capital goods, to a higher proportion of labour per unit of capital and output (see, for example, Hicks 1932 [1971]: 18-21).
Such changes in production techniques would normally be accompanied by changes in consumption patterns also leading to a higher proportion of labour demanded with respect to capital: a fall in wages would cause, other things equal, a fall in the relative price of the commodities produced with higher labour intensity, and this in turn would cause an increased proportion of labour-intensive goods demanded by utility maximizing consumers. Both mechanisms underlie the construction of aggregate decreasing demand curves for factors of production and support the notion that, in the long period, these curves are relatively elastic (see, for example, Pigou 1933 [1968]: 40, 96-7). A reasonably high elasticity (that is, not far below one) of labour demand curves is necessary to make the explanation of distribution based on demand and supply curves plausible. With inelastic curves, large falls in wages would be needed to increase employment in order, for example, to match an increase in population, and this might cause social disruption and economic instability before full employment could be reached. In addition, it has recently been argued that, if demand curves are not very elastic (as suggested by several econometric studies), this would favour the emergence among the workers of social norms tending to prevent competition over wages, since the latter would prove damaging for the social group as a whole (Solow 1980).In marginalist theory the decreasing relationship between the real wage and employment ensures that if wages are flexible, competition tends to bring the real wage rate to the full employment equilibrium level. Wages will be equal to the full employment marginal product of labour (or, even with fixed production coefficients, to the marginal contribution to consumers’ utility). Where different types of labour exist, wage differentials will reflect different marginal contributions as determined by the interplay of labour endowments, consumer tastes and the costs of acquiring different skills, given technology and the amount of capital. Within this framework, human capital theory as an explanation of wage differentials simply adds that differences between types of labour may be brought about by education and training, which entail a cost that may be regarded as a form of investment. Wage differentials would thus reflect the return on such investments, which competition (arbitrage) ought to render equal to what can be obtained on other forms of investments, such as those in physical capital (a conclusion, however, that has not been confirmed by empirical evidence). In neoclassical models however decreasing returns on investment in education and training are assumed, so that the return on human capital does not depend only on the costs of acquiring the skill, but on its cost at the margin, and hence also on demand factors (that is, consumer’s tastes) determining the relative scarcity of different types of labour (Becker 1975).
Decreasing labour demand curves and wage flexibility, however, are not enough to ensure that the economy will remain at full employment. The further condition is that “Say’s law” applies, that is that, on aggregate, the volume of full employment output will be entirely sold, that is, the income derived from production will be entirely spent. This means that saving decisions at full employment output must be matched by an equal amount of aggregate investments.
In the marginalist approach this actually rests on different premises than those found in the classical economists. It rests, that is, on a decreasing demand function for capital with respect to the interest rate. From this curve is derived a decreasing function of investments with respect to the same variable (equilibrium between demand and supply curves of capital as a stock requires that the changes in those stocks, that is the flows of investments and savings, are also brought to equilibrium). Thus in the marginalist theory, unlike in the classical approach, the so-called Say’s law is based on an economic mechanism whereby the decisions to save and to invest, taken by different subjects, are brought to equilibrium by the interest rate owing to the fact that aggregate investments are a decreasing function of the latter. This was in fact the version of Say’s law against which Keynes levelled his attack.Before Keynes’s General Theory, economic theory was characterized by the view that the economy would tend to be at, or close to, full employment in the long period, but this went along with the acknowledgement of economic cycles. Indeed, several interpretations of the business cycle and of temporary states of less than full employment were advanced. Of particular interest, owing to their similarities with recently advanced models and ideas, are the contributions by Pigou and Wicksell.
Writing in 1933, in the middle of the Great Depression and with unemployment peaking, Pigou still perceived the roots of unemployment as due to a lack of real wage flexibility, for which he provided various explanations. Wicksell saw the roots of economic fluctuations and episodes of price inflation and deflation in the existence, in a money economy, of a developed banking system capable of creating credit money. The endogeneity of credit money could cause (temporary) divergences between the actual and the equilibrium (“natural”) interest rate, particularly following alterations of the latter such as could be caused by technical change. The divergence between actual and equilibrium interest rate could accordingly cause phases of over or under investment with respect to full employment savings, which however would soon be corrected by a return to the natural or equilibrium rate of interest thanks to the direct response of the banking system to such situations or, failing this, owing to the changes in bank reserves that would eventually determine the required adjustment (Wicksell 1934, II: 206-7).