The Sraffian Revolution: Critique and Reconstruction
The cultural project Sraffa pursued, gradually defined as he worked on its various aspects, was—as mentioned above—a revolutionary one: to substitute the dominant approach to economics, the marginalist/neoclassical one based on the pillars of scarcity and preferences and on the equilibrium between supply and demand, with a return to the classical (Ricardian) approach in such a way as to make it compatible with Keynes’s new findings on the active role played by money in modern capitalist economies and the possibility of persistent unemployment. To this project, Sraffa contributed a number of basic constitutive elements: first came a critique of the then dominant Marshallian stream of the neoclassical approach (born in Cambridge, but subsequently colonizing the rest of England and nearly the whole world); second, there was the edition of Ricardo’s writings discussed in the previous section, and with it a reconstruction of the classical conceptual framework; third, we had a critique of the traditional marginalist theory of value and distribution; and fourth, a solution was proposed to the problem of value which, as we saw in the previous section, Ricardo and the Classical econo- mists—Marx included—had left open.
Before considering these two latter elements, we present a brief overview of Sraffa’s initial work on the critique of the Marshallian theory.
Sraffa’s main contribution in this respect was a long article published in 1925 in Italian when he was 27-years-old. As we saw, he had been appointed to a Lecturership at Perugia University in 1923, and had chosen Marshall’s Principles (1890) as his textbook, preferring it to Italian texts such as Pantaleoni’s Principii di Economia Pura (1889), although the latter was an author he greatly respected. Sraffa’s sharp, critical mind was thus brought to focus on that stream of marginalist economics—preferred as less abstract to the Walrasian stream of general equilibrium theory, well known in Italy through Pareto’s writings, and considered useless for interpreting the real world.
Sraffa’s ‘destructive’ criticisms—as Keynes (1930b: 79) considered them— concerned the two main aspects of Marshall’s theory: the method of partial equilibria, by which supply and demand equilibria were analysed for individual industries and firms, and the notion of laws of returns to scale, which allowed for the construction of U-shaped supply curves.
The assumption of perfect competition implies small but not infinitesimal individual firm size, and hence increasing returns (decreasing costs) at first, when the quantity produced grows starting from zero, followed by decreasing returns (increasing costs) for still higher levels of production, with an aggregation of supply curves for individual firms providing the supply curve for the industry, while competition ensures that the number of firms is such that each of them produces in equilibrium at the point of its minimum costs. The ceteris paribus assumption is also necessary to be able to focus on an individual industry or firm. However, Sraffa remarks, the reasons adduced for decreasing returns cannot in general be specific to individual firms. Nor, in fact, are they specific to individual industries: when there is a change in the total amount produced by the industry being considered, the reasons adduced to explain changes in its unit costs (such as the scarcity of a factor of production) in general also affect costs in other industries, and in general the effects will be of the same order of magnitude for various industries. As a result, the demand curve of the industry under consideration will shift, and the ceteris paribus condition will not hold, given changing prices in industries producing competing commodities.
Sraffas 1925 article concludes with a suggestion framed in the context of the analytical method of supply and demand equilibrium, namely that constant returns be assumed. However, this suggestion is soon abandoned (although many mainstream analyses of Sraffas 1960 book fail to recognize this point), reading it as half—the production side—of a general equilibrium model. Instead, in a subsequent article that Sraffa wrote at Keynes’s request for publication in the Economic Journal (Sraffa 1926), a summary of his argument is accompanied by the suggestion to move in the direction of a theory of imperfect competition, where the individual firm may be confronted by a downward-sloping demand curve, so that (slightly) increasing returns may be admitted.
Following on from this, in a famous 1930 Economic Journal Symposium on Increasing Returns and the Representative Firm, Dennis Robertson (1930) tried to answer Sraffas critique by distinguishing between the ‘simplified’ Marshall to whom Sraffa’s critiques applied, and the ‘real’ Marshall—an evolutionary theorist relying on the idea of a representative firm caught in the middle of its development, hence characterized by increasing returns to scale, while the growth and decay process of the individual firm ensures the equilibrium of the industry—who was immune to Sraffa’s criticism. Sraffa’s (1930) answer to Robertson consists in pointing out the inconsistency of such a position, based on the awkward assumption of a life cycle of firms. By then, Sraffa had developed such total opposition to the marginalist structure of Marshall’s thought that no mending was considered possible: both logical consistency and realism are required, and missing these we have to look for other avenues: ‘Marshall’s theory.can- not be interpreted in a way which makes it logically self-consistent and, at the same time, reconciles it with the facts it sets out to explain ... I think. that [it] should be discarded’ (ibid.: 93).
The fact is that Sraffa had by then already taken the revolutionary road of total abandonment of the marginalist approach to return to the classical tradition. This is testified by the Sraffa Papers and is also hinted at by Sraffa himself in the Preface of his 1960 book, where he recalls having already shown a first draft of its main theoretical propositions to Keynes in 1928. We may also stress, in this context, that Keynes is the only economist whom Sraffa thanks in the Preface, attributing to him a crucial suggestion: ‘[H]e recommended that, if constant returns were not to be assumed, an emphatic warning to that effect should be given’ (Sraffa 1960: vi; italics in original).
Ironically, in a text celebrated for its conciseness, the point about constant returns is stressed three times, at the very beginning. The insistence is due to the fact that this point is decisive for interpretation of the book. It is in this context that Sraffa distinguishes between two kinds of readers, the marginalist- and the classical-leaning, and two aims simultaneously pursued in the book, a critique of the traditional marginalist approach and reconstruction of the classical one: ‘Anyone accustomed to think in terms of the equilibrium of demand and supply’, says Sraffa (ibid.: v), alluding to the marginalist readers—‘may be inclined, on reading these pages, to suppose that the argument rests on a tacit assumption of constant returns in all industries’. Such readers, embedded with the marginalist conceptualization of the economy and hence incapable of reasoning outside of the framework of the equilibrium between supply and demand, need first to be persuaded that there are crucial theoretical inconsistencies in their theory. Thus, let them assume constant returns so that by reading Sraffas price equations as part of a general equilibrium model they can discover the logical inconsistencies embedded in the traditional marginalist theory of value and distribution. Once such a theory—and the marginalist approach based on supply and demand equilibrium—is abandoned, the reader can usefully reread the book as a positive contribution, solving a problem which the classical approach had left open: in this reading, the assumption of constant returns can, indeed, must, be abandoned, since the framework of equilibrium between supply and demand is extraneous to the classical conceptual framework (more precisely, the classical economists refer to a role played by demand and supply in the determination of market prices, which are not considered as theoretical variables).
Let us briefly summarize Sraffas criticism of the traditional marginalist theory of value and distribution. We can focus on two chapters of Sraffas book:
Chapter 6, providing a critique of the Austrian notion of the average period of production, and Chapter 12, on the choice of techniques. The analysis conducted in these two chapters shows that one of the main pillars of the marginalist construct, namely the direct relationship between the real wage rate and the capitalist intensity of production processes, does not hold.
As for the first aspect, let us recall that the notion of the average period of production had been used by the Austrian School, from Bohm-Bawerk to Hayek, as a measure of the capitalist intensity of production, with the rate of profits determined by the demand and supply of capital. Sraffa shows that the average period of production depends on the rate of profits and that there is no precise law for their relationship. Thus it cannot be utilized for measuring the quantity of the factor of production ‘capital’ in the framework of the determination of the rate of profits interpreted as the price of such a factor.
As for the second aspect, namely the choice among alternative techniques when income distribution changes, Sraffa shows the possibility of the ‘reswitching of techniques’, namely the possibility that a given technique, after turning out to be the most profitable one, is superseded by another technique when the rate of profits increases, but then turns out again to be the most profitable one at still higher profit rates. The implication of this fact is the following. However the capitalist intensity (i.e. the ratio between the quantities utilized of the two factors of production, capital and labour) of the different techniques is measured, it is not true that the quantity of capital utilized in the economy decreases (and the quantity of labour increases) when the rate of profits—that is, the ‘price’ of the factor of production capital— increases (and, conversely, the quantity of labour employed increases when the wage rate, the price of the factor of production labour, decreases). In other words, the traditional marginalist theory of value and distribution falls apart as soon as it is recognized that there is more than one single commodity in the economy. Traditional macroeconomics, based on the idea that when there is unemployment a fall in the wage rate suffices to boost employment up to the full employment level, also falls apart, together with its pillar, the inverse relationship between the real wage rate and employment.
Let us now turn to the positive contribution made by the classical approach. The classical problem is that when commodities are at the same time products and means of production, within the conceptualization of the economy as a revolving circuit of production and consumption the price of any commodity cannot be determined independently of the others nor the complex of relative prices independently of the distribution of income between wages and profits. It is therefore necessary to consider simultaneously the interrelations running on the production side between the various sectors of activity while also tackling income distribution and the determination of relative prices.
As a first step, Sraffa (1960: 3) shows that in a system of production with no surplus product, and where ‘commodities are produced by separate industries and are exchanged for one another’, at the end of the production period ‘there is a unique set of exchange values which if adopted by the market restores the original distribution of the product and makes it possible for the process to be repeated’. The situation is more complex when the system under consideration is able to produce a surplus. In this case, ‘the distribution of the surplus must be determined through the same mechanism and at the same time as are the prices of commodities’ (ibid.: 6). Sraffa shows that, if the wage rate can exceed the subsistence level, the relative prices and one of the two distributive variables—wage rate or rate of profits—are jointly determined, once the technology and the other distributive variable are taken as given. This basic result is then extended to the case of joint production and, as a specific instance of this category, to the case in which fixed capital, and a scarce resource such as land, are utilized.
Attention is also devoted to the construction of an analytical tool, the ‘standard commodity’, and to proving the uniqueness of the underlying ‘standard system’. The standard commodity has the property that the aggregate of its means of production does not change in value relative to the product when income distribution changes. Sraffa shows that this obtains when both product and means of production are but different quantities of the same composite commodity, constructed by adding up in different proportions the different sectors of the economy. Thus in the standard system the rate of profits can be determined, as in Ricardo’s ‘corn model’, as a physical ratio.
The way in which Sraffa solves the classical problem implies its specification in various respects. First, prices of production are those prices which allow for the reproduction of the economy, given the technology, and for a given value of one of the distributive variables; as a result, prices of production change when income distribution changes, or when technology changes. Second, since constant returns are not assumed, the technology assumed as given corresponds to a given set of activity levels in the various sectors of the economy: in other words, what is considered in Sraffa’s analysis is a sort of ‘photograph’ of the economy at a given moment in time. As such, the determination of prices of production and their relationship with income distribution is kept separate from the determination of technology and technological evolution, and indeed from the determination of income distribution and its changes over time as well as the determination of activity levels and employment, and again their changes over time. Such ‘separation of issues’ is a characteristic of Sraffas method, and as we saw above it also emerged in his discussions with Wittgenstein, leading the Austrian philosopher to develop a theory of language games.[1]
- [1] For a fuller illustration of this interpretation (first proposed in Roncaglia 1975), see Roncaglia (2009).
