The Theory of Value and Distribution in Economics

Discussions between Pierangelo Garegnani and Paul Samuelson

Pierangelo Garegnani, Paul Samuelson, Heinz Kurz

• Publisher: Taylor & Francis
• Year: 2012
• Print ISBN: 9780415519595
• eBook ISBN: 9781136244803
• Edition: 1
• Topic: Economics
• Subtopic: Economic History
• Index: Economics

This new volume explores two alternative economic theories – the classical theory and the marginalist or neoclassical theory- through a discussion between two eminent economists, Pierangelo Garegnani and Paul Samuelson.

The key themes of the volume are the difference in approaches to the explanation of the distribution of income and relative prices, and therefore different approaches to all other economic problems, in particular capital accumulation and economic growth. The book discusses whether there is a 'classical' approach to the theory of value and distribution at the core of economic theory that is fundamentally different from the later marginalist or neoclassical theory. In the volume, the late Pierangelo Garegnani argues for the validity of Piero Sraffa's position on this issue, whilst the late noble laureate Paul Samuelson vehemently contests it.

At a time of economic crisis, the future of the discipline is far from certain, and so it is extremely important to bring these debates back into the light, by reproducing them together for the first time. A comprehensive introduction by Heinz Kurz sets the debate in this context, and provides crucial background to the arguments.

1 Is there a ‘classical’ alternative to marginalist theory?

Some introductory remarks

Heinz D. Kurz

When this volume was planned, Pierangelo Garegnani was still with us. He would have loved to see it in print, but was not given the opportunity. He passed away on 15 October 2011 at the age of 81.¹

Since spring 2011 I had known that he was suffering from some illness relating to his lungs, but he did not disclose to me precisely what it was. Since at the time I was also suffering badly from a lung infection he perhaps did not want to alarm me unnecessarily. However, I remember him complaining on one occasion, talking on the phone, about the ‘nuisance’ of having to die. He still had so many things to do, ideas to work out, papers to complete, etc. I did not then interpret what he said as an indication of what he expected to happen before long. I had always experienced Pierangelo as a man of the long, ultra-long run, who forbade himself to take into account the finiteness of his own life. I did not even hear the alarm bells ring when, two days before he died, he called me, asking: ‘Heinz, will you do the big thing?’ I immediately thought that he was referring to the edition of Piero Sraffa’s unpublished works and correspondence of which he, in his capacity as Sraffa’s literary executor, had invited me to be the general editor.² I answered that, being on the mend after my illness, which had put me out of action for several months, I was about to get back to the Sraffa project. But he replied with a sigh: ‘No, no. I am referring to the edition of my debates with Samuelson.’ I assured him that I would do it and that as soon as I had drafted the introduction around the turn of the year I would get back to him. He was relieved, thanked me and we said goodbye to each other. It did not occur to me that his call was meant to settle things that were important to him before it was too late. Two days later, in the morning, Neri Salvadori called me from the University of Rome III, where a conference was taking place and where Garegnani had taught until his retirement and still acted as Chair of the Fondazione Centro Piero Sraffa di Studi e Documenti. He informed me that Pierangelo was no longer with us.

1 Is there a ‘classical’ approach to economic theory?

In the general introduction in Volume I of his edition of The Works and Correspondence of David Ricardo (Ricardo, 1951), Piero Sraffa had indicated that Ricardo’s approach to the problem of value and distribution was founded on the notion of social surplus. This approach was not an early and rude version of later marginalist theory, which tried to explain prices and income distribution in terms of the opposing forces of demand and supply, as Alfred Marshall and others had contended. It was a fundamentally different theory, which, alas, had been ‘submerged and forgotten since the advent of the “marginal” method’, as Sraffa was to stress in his Production of Commodities by Means of Commodities (Sraffa, 1960, p. v). That book was designed to accomplish two closely intertwined tasks: first, to reformulate and develop the ‘standpoint . . . of the old classical economists from Adam Smith to Ricardo’ in a coherent form; and second, to lay the foundation ‘for a critique of [the marginal theory of value and distribution]’ (Sraffa, 1960, pp. v, vi). Sraffa added: ‘If the foundation holds, the critique may be attempted later, either by the writer or by someone younger and better equipped for the task’ (1960: vi).³

Garegnani was one of those younger scholars. Others included, to mention but a few, Luigi Pasinetti, also a student of Sraffa’s, the late Krishna Bharadwaj, John Eatwell, Geoff Harcourt, Bertram Schefold and Ian Steedman. Garegnani was at the forefront of those who carried out the critique in the so-called ‘Cambridge controversies in the theory of capital’ (for summary accounts, see Harcourt, 1972; Garegnani, 1990a; Kurz and Salvadori, 1995, ch. 13). It is in this context that he first got involved in debates with Paul Samuelson, the main representative of the Cambridge, Massachusetts, side in the controversy.

In 1953 the Cambridge, United Kingdom, economist Joan Robinson had published a frontal assault on the concept of the aggregate production function and the marginalist concept of capital (Robinson, 1953). She based her argument to a considerable extent on ideas she had picked up in conversations with Sraffa. Paul Samuelson from the Massachusetts Institute of Technology (MIT), in an attempt to counter the attack, contended that even in cases with heterogeneous capital goods some justification can be provided for the employment of simple neoclassical ‘parables’ which assume that there is a single homogeneous factor called ‘capital’, whose marginal product equals the rate of return on capital. His paper on ‘Parable and realism in capital theory: the surrogate production function’ was published in 1962 (Samuelson, 1962) and quickly became a focal point in the controversy in capital theory.

Garegnani spent 1961–1962 at the MIT on a fellowship of the Rockefeller Foundation. He met Paul Samuelson and was given the opportunity to read the latter’s paper prior to its publication. He immediately spotted a major shortcoming in the argument and told Samuelson accordingly: in constructing the ‘surrogate production function’ Samuelson had boldly assumed uniform capital–labour ratios as between the machine-producing and consumer-good producing processes.⁴ This was an extremely special case, in which the labour theory of value explained relative prices in a straightforward way, whereas the argument does not apply in the only interesting, while realistic, case of different capital–labour ratios. Therefore Samuelson’s construction could not be considered a defence of the ‘Clark–Ramsey Parable’.

Before I continue, it is perhaps worth noting that in a note written as early as 16 January 1946 Sraffa had anticipated ante litteram the flaw in Samuelson’s argument. Sraffa had written:

The Irony of it is, that if the ‘Labour Theory of Value’ applied exactly throughout, then, and only then, would the ‘marginal product of capital’ theory work!

It would require that all products had the same org.[anic] comp.[osition]; and that at each value of r [rate of interest or profits] each comm.[odity] had an ‘alternative method’, and that the relations within each pair should be the same (i.e. that marg.[inal] prod[uct]s. should be the same; + also the elasticities should be the same); so that, even when the System is switched, and another Org. Comp. came into being, it should be the same for all products.

Obviously this would be equivalent to having only one means-product (wheat).

Then, commodities would always be exchanged at their Values; and their relative Values would not change, even when productivity of labor [sic] increased.

(Sraffa Papers D3/12/16: 34; Sraffa’s underlining)⁵

Garegnani worked out his criticism of Samuelson’s argument and showed not only that the aggregate production function is generally untenable, but also that all long-period microeconomic versions of marginalism starting from a concept of capital whose quantity can be ascertained independently of income distribution and prior to the determination of prices are invalid. In particular, there is no reason to presume that the capital–labour ratio moves inversely to the interest–wage ratio.⁶ By assuming equal proportions of labour and capital Samuelson turned the ‘real’ economy with heterogeneous goods into an ‘imaginary’ economy with a homogeneous output. In other words, the surrogate production function was nothing more than the aggregate production function. The publication of Garegnani’s essay was delayed. Submitted to the Review of Economic Studies in April 1963, it was accepted for publication subject to revision shortly afterwards. Garegnani submitted a substantially enlarged version in October 1968. The essay eventually appeared in 1970 entitled ‘Heterogeneous capital, the production function and the theory of distribution’ (Garegnani, 1970). It soon became one of the most often cited works in the capital controversies.

Prior to this, Garegnani was involved in a debate that was also indirectly triggered by Samuelson, who had asked a student of his, David Levhari, to investigate whether the return of the same technique, its ‘reswitching’, at different levels of the rate of interest was logically possible. In his paper Levhari (1965) contended that it was not and that entire systems of production can be ordered monotonically according to ‘degrees of mechanization’. This claim was disputed by Luigi Pasinetti (1966), Garegnani (1966) and others and gave rise to a symposium published in the Quarterly Journal of Economics. Samuelson and Levhari in their joint paper to the symposium and Samuelson in his ‘Summing up’ paper (Samuelson, 1966) frankly admitted that the no-reswitching ‘theorem’ was wrong. Samuelson also provided some numerical examples which illustrated in simple terms why reswitching and capital reversing are possible.

But how important were these phenomena and did they effectively undermine marginalist theory? To Samuelson the most disturbing result was the finding in steady-state capital theory that consumption per head could be positively related to the rate of interest. But since, he contended, reswitching and capital reversing do not upset Pareto optimality their importance was limited, a view that was strongly contested by Garegnani and others.

In addition, in a number of papers, the most important of which is perhaps his essay ‘The canonical Classical model of political economy’ (Samuelson, 1978), he maintained that there are no fundamental differences between the theories of the classical authors and those of the later marginalists. To him, Smith, Ricardo, etc. were essentially marginalist theorists waiting to be born. He maintained that there is a fundamental unite de doctrine across all ‘schools’ of economic thought, once each particular doctrine has been purged of concepts that lack clarity, errors of reasoning and special assumptions. While Samuelson had influential followers in this regard, including John Hicks, he drew criticisms from numerous scholars coming from different fields and orientations in economics, and especially from historians of economic thought, who felt that he was patching over important differences between and within different schools of thought. However, the main criticisms of his point of view came from scholars working in the tradition of Piero Sraffa, especially Garegnani.⁷

Before we turn to this, two further events ought to be mentioned. First, in his keynote address at the History of Economics Society meeting in Boston in 1987, reflecting his unite de doctrine point of view, Samuelson proposed a programme for what he dubbed the ‘Whig history of economic science’ (Samuelson, 1987a). By this he meant a re-orientation of the history of economic thought ‘toward studying the past from the standpoint of the present state of economic science’ (p. 52). He motivated his proposal by asserting that there were no ‘Kuhnian breakthroughs in current economic science’ and that ‘ours is not an age of heady accomplishments and new exciting syntheses’ (p. 52). As an exemplification of the alleged cumulative character of the normal science of economics he mentioned Piero Sraffa’s edition of The Works and Correspondence of David Ricardo (Ricardo, 1951–1973) and his reformulation of the classical approach to the theory of value and distribution (Sraffa, 1960). This was, of course, a deliberate provocation of those convinced by Sraffa’s message that there is a classical alternative to marginalism.

Second, prior to this Samuelson had been invited to contribute a paper to a conference in Florence in 1985 commemorating the twenty-fifth anniversary of the publication of Sraffa’s book. While he could not himself participate in the meeting, he sent a paper entitled ‘Revisionist findings of Sraffa’ (Samuelson, 1990), which was published in a volume edited by Krishna Bharadwaj and Bertram Schefold (1990), followed by comments by Eatwell, Garegnani and Schefold. Samuelson’s paper consists essentially of a follow-up to his entry ‘Sraffian economics’ in The New Palgrave (Samuelson, 1987b), reiterating the view that some of Sraffa’s propositions, especially those concerning the problems of constant returns to scale, joint production and the Standard Commodity, cannot be sustained and that Sraffa’s analysis does not constitute an alternative to marginalism (see, in particular, Samuelson 1990, section 14). An implicit reply to Samuelson was contained in Garegnani’s entry ‘Surplus approach to value and distribution’ in The New Palgrave (Garegnani, 1987),⁸ and an explicit one in his comment in the Bharadwaj–Schefold volume (Garegnani, 1990b). In the latter, Garegnani tried to clarify the relationship between the analyses of Ricardo, Marx and Sraffa on the one hand, and their differences from the marginalist theories on the other. He concluded that

Sraffa is a very difficult author. The difficulty is made even greater than it needs be, because Sraffa’s work has in effect been so little discussed on its own terms. Professor Samuelson is therefore to be thanked for what we must hope will be the beginning of a fuller discussion – a beginning that, like all beginnings, is bound to suffer from the fact that the necessary clearing of the ground has yet to be effected.

(Garegnani, 1990b, p. 297)

Apparently, there was a wide consensus among the debatants, especially Samuelson and Garegnani, that many of the problems raised had not yet been satisfa...

Table of contents

• Front Cover
• The Theory of Value and Distribution in Economics
• Routledge studies in the history of economics
• Title Page
• Copyright
• Contents
• Acknowledgements
• Notes on contributors
• 1 Is there a ‘classical’ alternative to marginalist theory? Some introductory remarks
• 2 Sraffa’s hits and misses
• 3 Professor Samuelson on Sraffa and the Classical economists
• 4 Classical and Neoclassical harmonies and dissonances
• 5 Samuelson’s misses: a rejoinder
• Subject index
• Name index