Critique of the Marginalist Approach
While advancing a theory of production prices within the framework of the classical approach, Sraffa’s book also offered the tools for a critique of the foundations of the marginalist theory of value and distribution. Preliminarily, however, we need to clear the path of a misunderstanding, namely interpretation of Sraffa’s contribution as a general equilibrium analysis conducted under the assumption of constant returns to scale, in which it would have been possible to explain prices by focusing attention on production costs - the supply side - and dropping the demand side. Sraffa explicitly rejected- three times, in the preface to his book - the idea that his analysis would require the assumption of constant returns. ‘No question arises as to the variation or constancy of returns. The investigation is concerned exclusively with such properties of an economic system as do not depend on changes in the scale of production or in the proportions of “factors”.’ ‘This standpoint, which is that of the old classical economists ..., has been submerged and forgotten since the advent of the “marginal” method’ (Sraffa 1960, p. v). We can, however, utilize the analytical results reached with regard to prices of production for an internal criticism of logical inconsistency of the marginalist theory of value and distribution by transposing them into the marginalist conceptual framework.
The results in Sraffa’s book that can be directly used as the foundation for a critique of the marginalist theories of value and distribution concern the average period of production and the choice of techniques. The average period of production had been propounded by Bohm- Bawerk (1889) as a measure of the capital intensity of production. Sraffa showed that, depending as it does on the rate of profits, it cannot be used to measure capital in the ambit of an explanation of the rate of profits taken as the price of this factor of production. The difficulty had already been sensed by Wicksell (1901-6), but modern exponents of the Austrian school, including Hayek (1931), were later to return to the notion of the average period of production.
With regard to the choice between alternative techniques of production when the rate of profits changes, Sraffa (1960, pp. 81-7) pointed out the possibility of a reswitching of techniques; in other words, a given technique that proves the most advantageous for a given rate of profits may be superseded by another technique when we raise the rate of profits but may once again be preferable when the rate of profits rises still higher. The implication is that, however the capital intensity of the two techniques (or in other words the ratio between the quantities utilised of capital and labour) is measured, the general rule that the marginalist theory of value rests on remains contradicted. This rule takes the distributive variables, wage rate and rate of profits, as prices of the corresponding factors of production determined by the law of demand and supply, so that the quantity of capital employed in production should diminish (and the quantity of labour increase) as the rate of profits rises (and the wage consequently falls). With the reswitching of techniques, violation of the marginalist rule is unavoidable: if the rule holds when one technique gives way to another with a rising rate of profits, the contrary occurs when from the second technology the economy turns again to the first as the rate of profits rises yet higher.
This criticism gave rise to widespread debate, while the crucial question of its relevance received relatively scant attention. It applies not only to the aggregate production function but also to all those cases in which, while acknowledging the fact that capital is a collection ofheterogeneous means of production, the rate of profits is still interpreted as the price of a factor of production capital, however it be defined (aggregate of value, waiting, average period of production). In particular, Sraffa’s critique undermines the very foundations of the idea - crucial to marginalist macroeconomic theory - of an inverse relationship between real wage rate and employment, such that a competitive labour market in a closed economy would automatically tend towards full employment equilibrium, since the decline in real wages brought about by unemployment would prompt an increase in the labour-capital ratio and hence, given the endowment of capital, an increase in the quantity of labour employed. Sraffa’s critique not only rejects the idea of the existence of equilibrium (optimal) values for the distributive variables, wage rate and rate of profits but also sides with Keynes’s critique in denying the existence of an automatic tendency of competitive labour markets towards full employment.
-  Harrod (1961), in a review of Sraffa’s book, persisted in defending the Austrian theory ofvalue by recalling that for any level of the rate of profits we may univocally define theaverage period of production, though in the presence of compounded interest. Sraffa(1962) replied that this fact is not sufficient, since here we fall into a logical vicious circle:the rate ofprofits must be known in order to determine the average period ofproduction tobe utilised, as a measure of the capitalistic intensity of production, in determining the rateof profits.
-  More or less simultaneously with the publication of Sraffa’s book, Garegnani (1960)developed a direct critique ofsome ofthe main theoretical contributions in the marginalisttradition. Publication of Sraffa’s book was then followed by lively debate. An initialskirmish (Harrod 1961; Sraffa 1962) has already been recalled in the previous footnote.A second clash began with Samuelson’s 1962 attempt to depict the aggregate productionfunction as a ‘parable’ not betraying the essential characteristics of a market economy andby Levhari’s (1965) attempt to show that the problems raised by Sraffa (such as thepossibility of the reswitching of techniques) referred only to the single industry and not tothe economic system as a whole. These propositions were immediately refuted byGaregnani (1970), Spaventa (1968) and Pasinetti (1966); Samuelson 1966 and Levhari(with Samuelson, 1966) were themselves to recognise the erroneousness of their theses.Pasinetti (1969) then criticised the recourse on the part of Solow (1963, 1967) to theFisherian notion of the rate of return, considered as index of the quantity of capitaldefinable independently of the profit rate and thus utilisable for explaining the latter.
-  Incidentally, this means that the so-called structural reforms insisted upon in the past fewyears by the so-called Washington consensus, aiming at reducing workers’ bargainingpower, are certain to affect income distribution, while the effects on employment may turnout to be positive or negative depending on circumstances (and Keynesian theoryindicates that negative effects are more likely). We may add that Sraffa’s revival ofthe classical approach appears to be compatible withthe Keynesian approach. Abandoning the approach to the determination of equilibriumprices and quantities based on the demand-supply mechanism leaves us free to considerthe determination of prices and the determination of quantities as separate analyticalissues, which opens the way to a Keynesian determination of the levels of production and