When I agreed to review this book, advertised as presenting a discussion between Paul Samuelson and Pierangelo Garegnani over the theory of value and distribution, I made the mistake of assuming it would contain the transcript of a conversation. Such conversations can be very valuable in that in a face-to-face exchange, people typically engage each other more directly than in publications that follow the more formal conventions of academic writing. It was, therefore, very disappointing to find that the book did no more than reprint, with a short editorial introduction, four previously published pieces, all of which I had previously encountered.

The debate concerns the interpretation of Piero Sraffa’s Production of Commodities by Means of Commodities (1960). This book used a model of production in which commodities are produced using commodities and labor. Capital is the value of commodities used in production, which means that price equations can be written in the form,

$${p_i} = ({a_{i1}}{p_1} + {a_{i2}}{p_2} + ... + {a_{in}}{p_n})(1 + r) + w{l_i},$$

where p_i is the price of good i, a_ij is the amount of good j used to produce a unit of good i, l _i is the amount of labor to produce a unit of good i, r is the rate of interest, and w is the wage rate. Sraffa, inspired by David Ricardo, whose Principles of Political Economy and Taxation he had recently finished editing, implied that such a system could form the basis for a new theory of value and distribution in which the technical conditions of production determined the surplus produced by society; the distribution of that surplus between profits and wages was determined outside the system. Using this model, Sraffa derived a number of propositions that he considered fatal to the neoclassical theory of value and distribution. The book was concise to the point of being enigmatic, leaving ample room for different interpretations of what he was doing.

Samuelson’s view was that Sraffa, who claimed to have developed these ideas in the 1920s, was using a model that should be familiar to anyone who understood the work of Wassily Leontief and John von Neumann, and the ensuing literature on linear modeling. He claimed that, because Sraffa did not know this mathematical literature, he had failed to understand some points, such as the Hawkins–Simon condition for such a system to be productive (his misses). However, Sraffa had nonetheless scored some hits, for he had proved some results that were not known prior to his book, notably that the relationships among consumption, the capital stock, and the rate of interest might be different from those implied by a “neoclassical” production function. To make this point, Samuelson attributed to Sraffa assumptions, such as constant returns to scale, that were, he believed, necessary to make sense of his system. Against this, Garegnani argued that Samuelson had missed Sraffa’s point. Sraffa had not made any assumption about returns to scale because his concern was to construct a new theory of value and distribution, not a special case of the neoclassical model. There was, he claimed, an alternative classical paradigm that was represented by Sraffa’s system, and Samuelson’s unwillingness to see this had led him astray: Sraffa was not doing the same as Leontief.

Though these issues are no longer widely discussed, outside a small group committed to the “classical” approach as interpreted by Sraffa and an even smaller number of people who engage with them, they should be of interest to historians of economics, for they were once actively debated in the leading journals. It is convenient to have these pieces in the same volume. However, consider an alternative purchase. For only $90, it is possible to buy The Collected Scientific Papers of Paul A. Samuelson, volume 6 (Samuelson Reference Samuelson and Murray2011), which contains Samuelson’s side of the exchange, including “Sraffa’s hits and misses” (Samuelson Reference Samuelson and Kurz2000), the article to which Garegnani is responding. Garegnani’s two articles are from the European Journal of the History of Economic Thought (2007) and can easily be downloaded, free of charge to subscribers or those whose library has a subscription. In place of Kurz’s ten-page introduction, the reader can download his informative obituaries of Samuelson (Kurz 2010) and Garegnani (Kurz 2012), from the same journal. Aside from saving $40, this alternative purchase has the advantage of including no fewer than seventy other articles, including a reprint of an eleven-page comment by Kurz and Neri Salvadori (2000) on the article to which Garegnani is responding, Samuelson’s reply to their comment and a dozen or so other articles by Samuelson on Sraffian economics, and several more articles on Ricardo and Marx. Given this, it is difficult to recommend the book under review, except to those who face no budget constraint, those who cannot access online journals, or those whose bookshelves have insufficient space for the larger volume.

Leaving aside the issue of value for money, which reflects the publisher’s decision to produce such a slim volume of easily accessible material at a high price, how should the collection be evaluated on more conventional grounds? A full evaluation would involve placing the exchange with other literature such as the Samueson–Kurz–Salvadori exchange mentioned in the previous paragraph (Kurz would obviously have been very well-equipped to do this in the introduction), something that is beyond the scope of a short review, so I confine myself to a few points.

Samuelson considered Sraffa’s book to be important—sufficiently important that it would have justified awarding him the Nobel Memorial Prize. Samuelson and Garegnani both dissent from the majority view among economists that the issues involved are irrelevant. Where they disagree is that Garegnani sees Sraffa as having laid the foundation for a new paradigm in value theory; whereas Samuelson sees nothing that cannot be accommodated within a single tradition running from Smith to the present. My conjecture is that the exchange will change few minds; in particular, neither Samuelson nor Garegnani will persuade skeptics that the issues matter.

As one would expect, each tries to shift the debate on to their favored ground, rather than tackling the issues as formulated by the other party. The result is that, to a certain extent, they speak past each other. Samuelson justified his approach of not countering Garegnani’s points but instead focusing on what he thought important by saying that he had “an antipathy to adversary procedures.” Garegnani did the same, though recognizing that this results in one party’s criticisms not being answered by the other. This is why a face-to-face discussion would have been so valuable.