Introduction

Alessandro Roncaglia has been analyzing Piero Sraffa’s contributions and how they fit into the context of the developments of economic theory since the time of the classical political economists on.

While I greatly respect his arguments and provision of evidence in the public domain, I want to argue here that the Marxian stream is the most appropriate one, both for interpreting Sraffa’s own views and inclinations and for providing relevant theoretical developments to aid our understanding of the structure of, and processes at work in, the modern capitalist world. In arguing this I am comforted by the fact that, for example, the late Krishna Bharadwaj, whose understanding of Sraffa and his works was second to none, Giancarlo de Vivo, the late Pier Luigi Porta and Luigi Pasinetti are in overall agreement with such a take on these important matters.

Many years ago I published an exploratory and speculative paper, “Marshall, Sraffa and Keynes: Incompatible Bedfellows?” (Harcourt, 1981). Its principal objective was to examine the role of the concept of a center of gravitation in the contributions of these three great political economists.

Marshall wished to have a theory of long-period normal prices that included as ingre­dients normal profits and normal wages. His time periods—market period, short period, long period—were used to establish his theory. He argued that they could merge imper­ceptibly one into another and could be short or long according to the immediate purpose at hand, and actual or potential according to the realism of the factors locked up in the ceteris paribus pound in any particular case. He vacillated between whether they were actual or potential, sometimes naming periods of calendar time as illustrative of what he had in mind, wishing to have it both ways. That is to say, he wished his normal prices to be real centers of gravitation, making sense of observations in historical time, yet he knew that if he simultaneously had in mind an economy moving forward through time with technical progress and accumulation occurring, there were puzzles associated with arguing that they could be revealed from observations on actual prices. He in fact says that it is only in the stationary state that the actual and normal coincide, can be “convert­ible terms” (Marshall, 1890, 372).

He never solved his conundrum—hence his use of time-period analysis became less and less satisfactory and illuminating, the longer the period (or run) he had in his sights. Moreover, he was not able to draw on a consensus of views from the evolutionary biologists of his time because there was none. After his death, as Neil Hart has shown (2012, 2013), evolutionary theory itself moved more toward a consensus among evolu­tionary biologists and allowed great strides forward to be made by modern evolutionary economists.

While John Maynard Keynes was very much a Marshallian in method, even in The General Theory, he was not as handicapped as was his mentor because for most of the time he was analyzing short-period systemic problems. There, I argued, the application of the concept of a center of gravitation provided a useful shortcut for establishing illuminat­ing theory and providing sensible and realistic policy proposals. The concept had a part to play in the Keynesian context of short-period rest states (and, possibly, even longer sustained rest states associated with deep and sustained depressions), a view I believed could be supported from Keynes’s own writings and by Jan Kregel’s (1976) classic inter­pretation of them.

When I wrote in 1981, there was an ongoing debate concerning whether the concept should or should not be maintained in a Marxian and/ or post-Keynesian analysis of a modern capitalist economy moving forward through time. I argued that their natural place was in the theory of pricing that characterized important sections of such econo­mies, in the normal cost-pricing hypothesis and the connection of price making with the investment decision. In particular, I thought then, and I do so now, that there may be a role for the concept of the natural rates of profits and of wages—that they are macro­economic concepts associated with the working of the system as a whole, imposing them­selves as norms on group behavior within the system.

Similarly, if we accept the snapshot view of what Sraffa was doing—as Roncaglia has long argued and which the emerging evidence from the unlocking of the Sraffa archives more and more establishes—Sraffa too is not be deviled by Marshall’s conundrums.¹ Recently, Ajit Sinha (2013) has argued that the uniform rate of profits in Sraffa’s system is not necessarily based on Smithian (natural prices), Marxian (prices of production) and Marshallian (normal long-period equilibrium prices) centers of gravitation, but is instead a mathematical property of Sraffa’s system. I think there is merit in this argument; how­ever, it is certainly not one that has been accepted by other Sraffa watchers.

Be that as it may, I want to argue that there is a similarity between the role that the dominance of the sphere of production over the sphere of distribution and exchange in Karl Marx’s analysis, and the role that Sraffa prices play in modern price theory, espe­cially if allied with the promising methods and advances associated with the development of the cyclical growth models of Richard Goodwin (1967) and (late) Michal Kalecki (1968). Put together, I shall argue that we potentially have a structure that overcomes the inability of the mainstream structure to link the short period with the long period because the theory of the medium period in between is incoherent, to say the least.^[9]

Moreover, it is unclear whether the long period is an actual state but a theoretical concept, as Marshall sometimes had it, or a stretch of historical time in which changes in the capital stock, methods of production and the supplies of different types of skilled labor are occurring. Suppose, though, that the role of the long period is interpreted as the gathering together of expected long-term forces and factors that have an impact on deci­sions made in the short period. Then, if this is accompanied by the impacts of expected short-term factors, we may analyze how activity and so on is established short period by short period, run by run, over actual historical time.^[10]

The proponents of the long-period method—Pierangelo Garegnani, Heinz Kurz, Neri Salvadori and others—while they have always stressed that they want an apparatus with which to handle the interrelationships of persistent forces, have never been able to incorporate the two most characteristic sustained and persistent forces—an inescapable environment of fundamental uncertainty and continuous technical changes—into their formal analytical structures. In contrast, the approach outlined here does, I argue, allow us to start to tackle these problems, and Sraffa’s contributions are a vital, central part of what may evolve.

for accumulation and other uses is the outcome, period by period, of the relationship between the current state of the class war between the capitalists and the wage earners and the existing methods of production in the current capital stock, itself the outcome of past accumulation associated with the use of previous realized surpluses. The potential surplus is the ultimate source of the potential size of profits and the rate of profits real­ized in the sphere of distribution and exchange—that is one of the core meanings of Marx’s labor theory of value (Harcourt and Kerr, 1996). On the one hand, realization of the surplus depends on the factors determining aggregate investment expenditure (a relationship between expected profitability and planned accumulation), and, on the other hand, the relationship between actual accumulation and actual profitability, itself associated with the impact of the distribution of income between profits and wages on aggregate saving.

For the present purposes, the moral to be drawn is that the relationships in the sphere of production dominate and determine what may happen in the sphere of distribution and exchange without there being any need for a complete one-to-one mapping from the actual size of the potential surplus onto the corresponding sizes of profits and the rate of profits. I argue that exactly the same features are present in Sraffa’s system of prices in relation to the determination of the sizes of markups on costs and their corresponding prices in the multinational oligopolistic market structures of the modern world.

In Harcourt (1981, 261), I noted the puzzles that hound the concept of prices of pro­duction when trying to incorporate them as operational concepts in the analysis referred to above. The dynamic nature of capitalist development, with the embodiment of tech­nical changes through investment expenditure, may be so rapid in most periods (runs) as to not allow sufficient historical time for centers of gravitation of a lasting nature to be formed. There is not the time, asJoan Robinson put it, for traders to become familiar with what is the norm through actual experience, so that when their bearings are cut loose, they are all at sea, rudderless, as are also the overlooking economists. The factors needed to be held constant theoretically so as to allow the centers of gravitation they imply to be ultimately struck are in fact changing so fast that there is not the time needed for their ultimate values to be established. I still do not see how John Eatwell’s (1979, 2) statement that the forces that determine the centers of gravitation are the most domi­nant, systematic and persistent, and that “[whether] this centre of [gravitation] is a tem­poral constant, or takes different values through time, does not affect the essence of the method,” overcomes this puzzle.

3.