Power Relations and Economic Paradigms,or Why Post-Keynesian Theory Is Not Dominant

1.1 The Diagnosis

The book is divided into three parts (called “books”): parts one and three are analytical; and part two describes the intellectual environment of postwar Cambridge (UK), providing the biographies of the five most prominent members of the Keynesian school (Kahn, Joan Robinson, Kaldor, Sraffa, and Goodwin).

The introductory part 1 reproduces two lectures Pasinetti delivered in Rome in October 1994 (the “Federico Caffé lectures,” in memory of this eminent Italian Keynesian). Pasinetti starts by explaining why Keynes’s General Theory was a “scientific revolution” in the Kuhnian sense of the sharp break from the scientific paradigm of “exchange” and subjective utility prevailing at that time, to the classical paradigm of “production” and labor as source of value. The principle of effective demand marks this paradigmatic shift. ¹

However—Pasinetti argues—this scientific revolution was unaccomplished and this explains why it did not end up with the primacy of the classical/Keynesian paradigm over the marginalist-neoclassical one. Very early, Hicks attempted to reconduct the disruptive flow of Keynesian ideas into the riverbed of the mainstream through the IS-LM model, a “rescue” operation that typically occurs when the dominant paradigm is called into question by some disturbing arguments or evidence. More fundamentally, Keynes neglected some essential building blocks of his theoretical construction, and the Cambridge Keynesians did not fill the gap completely. They suffered from a lack of “vision,” engaging them in producing a multitude of models without the objective to work as a team in order to strengthen the theoretical foundations of their school. “By paying too little attention to the positive aspects of each other’s contributions, [they] have not encouraged the search of [such] a comprehensive theoretical framework” (37–38). The results of their scientific efforts were very valuable (see, for instance, Harcourt 2006), but the coherence between all these models is not always established, and one is left with the impression of a too wide heterogeneity. Other, less essential, elements conspired against the success of the Keynesian revolution. Pasinetti cites rivalries and prima donna behaviors of some of the Keynesian group, the impalpable opposition of academia resulting from the fact that, in the context of the cold war, they were perceived as leftist, etc. The author sums up complaining that the numerous schools grouped under the post-Keynesian label have been “insensitive . . . to the necessity . . . of building up an overall, . . . solid theoretical framework as an alternative to the prevailing orthodox stream” (46). Of course, the post-Keynesian scholars share the same general analytical approach that, for instance, can be found in the “Theoretical Perspectives” of the EAEPE (European Association for Evolutionary Political Economy: www.eaepe.org). However, Pasinetti deems that “all this [is] perfectly correct, yet widely insufficient” (47). For presenting a new paradigm, “something much deeper is necessary: something more specific . . . capable of originating those characterisations as consequences” (id.). The third part of the book is devoted to precisely such a task.

The propositive part of the book is preceded by the beautiful part 2, where Pasinetti presents the life, the scientific achievements, and the failures of the “pupils of the first hour”: Kahn, “co-author” of the General Theory; Joan Robinson, the woman who missed the Nobel Prize; Kaldor, “always bubbling with new ideas”; and Sraffa, the “master of all critics.” Special attention is devoted to this latter scholar and to his unrealized aim to reconstruct the whole economic theory on a classical basis. An essay on Goodwin—the missed Keynes-Schumpeter connection—completes this bio-bibliographical part.

Pasinetti concludes by summarizing the following nine points, the characteristic features and building blocks of the Keynesian school, which could serve as an appealing starting point for further developments:

1.2 Towards a Production Paradigm

Part 3—devoted to a proposal to complete the classical-Keynesian paradigm—begins with a chapter showing the inadequacy of mainstream theory (ch. 8: “Beyond Neoclassical Economics”). Recalling the historical background of economic analysis, it appears that the exchange paradigm is indeed typical of the phase of trade that preceded present industrial society. Then, to support their claim for generality, the neoclassicals fell into a “methodological reductionism”: rather than attempting to explain reality, they constructed by assumption a hypothetical world to fit into the pre-conceived general equilibrium scheme (264). At the macroeconomic level, the unavoidable consequence was a one-commodity world, where production is formalized with the Cobb-Douglas function and, to cope with the heterogeneity of individuals, the “representative agent” appeared, something that amounts to positing that society is formed by a series of absolutely identical individuals. In this way the neoclassicals committed the mistake of the “fallacy of composition,” resulting from the assumption that “the whole system always and exclusively is generated by the sum of the independent actions of the single individuals” (267).

In the following chapter 9—the most important of the book—Pasinetti traces the way to complete the classical-Keynesian paradigm. His suggestion turns on a methodological device, which he calls the “separation theorem.” Economic investigation—he argues—should be split into two distinct phases: first the stage of “pure” theory and, subsequently, the stage of institutional analysis. The first level of analysis consists in disentangling the foundational basis of economic relations, what classical economists called “natural.” In his model of structural change Pasinetti (1981) works out this methodological approach in detail with his concept of “natural system,” studying “the ‘primary and natural’ determinants of the variables characterising an economic system,” which are prior to, and independent of, any institutional set-up (Pasinetti 1981: 149). “The [theoretical] problems . . . that emerge at this stage are either in terms of necessary relations, if certain goals are to be achieved (e.g. full employment, price stability, etc.), or in terms of logically consistent relations, or in terms of normative rules, or in terms of those problems which are generated by the basic forces at work in a dynamic context” (Pasinetti 1994: 41). All these relations can be developed without referring to specific behavioral and organizational assumptions; they reflect the basic characteristics of any modern industrialized economy. In addition, the “natural system” has a normative dimension because it singles out optimal (efficient) positions, which are also socially fair. Thus it should set the target for economic policy.

In Pasinetti’s view, the members of the Cambridge school, “while implicitly taking advantage of this methodological . . . separation approach, have not explicitly acknowledged it, which . . . go[es] a long way to explain why they found it so difficult to face . . . the task of sketching out . . . [a common] comprehensive theoretical scheme” (277). Then, to illustrate the foundational framework that should inspire all post-Keynesian scholars, Pasinetti presents his simplest version of the “natural system,” a pure labor economy where Sraffa’s scheme is completed and extended to the long run by the introduction of technical change and structural dynamics (279–302). The choice to illustrate his proposals by such an unrealistic model where commodities are produced by labor alone is motivated by the aim of contrasting the production approach to the exchange approach at the most abstract level; however, his more elaborate models fully incorporate capital goods (Pasinetti 1981, 1988).

Chapter 10 is devoted to the second phase of investigation—the institutional analysis—suggesting how to transform the “natural” positions into actual outcomes. Considering that the former positions are ideal ones, a comparison between the “actual outcomes with their corresponding natural configurations . . . provide[s] the criterion for the justification or non-justification . . . of the institutional mechanism through which they are . . . pursued” (307). The concluding chapter 11 suggests that, within the production paradigm and the separation theorem, it is possible to reabsorb a lot of research that has been carried out within the rival paradigm of exchange. This holds, for instance, for the non-competitive market structures and mechanisms, strategic behavior and non-cooperative games, the economics of imperfect information, etc. In Pasinetti’s view, all these contributions find an appropriate place at the institutional level of analysis.

1.3 Assessment

Reading Pasinetti’s book was an enjoyable and fruitful experience, an experience that I recommend to all colleagues. They will discover an extraordinary richness of analysis and will receive numerous stimuli for their research. Pasinetti’s lucid discussion of the Keynesian school shows the stumbling blocks to be avoided if we want to assert the validity of a paradigm. On a substantive level, the methodological device he proposes to complete the classical-Keynesian paradigm (the “separation theorem”) is very useful and requires careful attention. Within this context, institutional enquiry has a very broad scope. It is interdisciplinary by its nature, as it involves, among other things, the study of the social and power relations that favor or oppose progress. Since social change is ultimately the result of social struggles, institutional analysis should be at the service of social progressive movements by providing a coherent and feasible project of society. We rejoin here the purpose Marx assigned to the economic and social enquiry: to know for change.

Two points deserve further analysis: the notion of “natural system” and the relation with Marxian political economy.

2.1 Preliminary: Dissipating a Possible Misunderstanding

In Pasinetti’s analysis the concept of “natural” must be conceived in the sense of fundamental, which is at odds with the meaning classical economists attributed to the adjective “natural.” In fact, with the exception of Marx, for other classical economists, the “natural system” was in fact the capitalism prevailing at their time, the “natural” character being extended to institutional forms of such capitalism and to its exploitative nature. The “natural wage,” for instance, is “that price which is necessary to enable the labourers . . . to subsist and to perpetuate their race, without either increase or diminution” (Ricardo 1817: 93). Later on, the neoclassicals spoke of a “natural” rate of unemployment—something that has nothing “fundamental” in Pasinetti’s sense. Moreover, the classical concept of “natural price”—as determined by the cost of production plus a uniform rate of profit resulting from competition—is indissolubly linked with what Pasinetti would call an “institutional aspect.” In fact, for classical economists competition is the result of inter-sectoral mobility of capitals, which means no barriers to entry and to exit, a condition that can typically be fulfilled at the institutional level.

2.2 “Natural” Profit and Exploitation

Pasinetti’s contributions strengthen Marxian political economy by providing analytical tools that confirm some of Marx’s fundamental insights. I am referring to the notion of “natural” profit, which gives further analytical content to the notion of exploitation, and to the definition of dynamic values, that offers a novel solution to the problem of “transformation.” This will confirm the complementarity between the post-Keynesian and Marxian approaches to political economy (Reati 2000). Consider now the “natural” rate of profit.

In the “natural” system, income distribution is governed by a “pure labor principle”: entrepreneurship is taken as a sort of skill and it is not remunerated with a private appropriation of surplus value. Thus the “natural prices” cover, as costs, the depreciation of fixed capital as well as the salary of managers and other supervisory personnel, plus the “natural” rate of profit, which is the only thing accruing to capital. Within this context, the rate of profit corresponds to what is required by the accumulation of capital in a growing system. In each final sector i of the economy (i=1.2, . . .,n), the natural rate of profit π i * is thus the sum of:

π i * = g + r i

(1)

This can be linked to the question of exploitation in the following way. Let us distinguish two components of exploitation, which are quantitatively encompassed in surplus value: I would call them exploitation from “pure alienation” and exploitation from “pure appropriation.” The first type of exploitation comes from the fact that, in a capitalist society, the technical relation of investing a share of social product to expand productive capacity becomes a social relation, which is performed by a class of people (the capitalists) who enjoy the privileges associated with that social role, in terms of the ownership of the means of production as well as in terms of income, power, and prestige (Medio 1972: 329). The second component of exploitation (“pure appropriation”) is the difference between total capitalist profit (higher than the “natural” one) and what is required for the expansion of the economic system. ² Thus, what accrues to capitalists as the difference between the two notions of profitability is subtracted from workers’ consumption. This appropriation aspect is further exacerbated by the already noted fact that capitalists become the owners of the capital goods acquired with the “natural” profit. In conclusion, the “natural profit” provides the way to quantify what, in a capitalist society, corresponds to the two types of exploitation and, by showing the exact magnitude of what is technically required for accumulation in a growing system, it strengthens the awareness that capitalists usurp such a function.

2.3 “Dynamic” Labor Values and the “Transformation” Problem

The “natural” rate of profit also provides the basis for solving the “transformation” problem. As we know, in a static context labor values at any time t include the total labor required to replace the worn out capacity during the same period (the flow of depreciation). However, for the reproduction of the system in a dynamic context—where the demand of each sector grows at its specific rate—this is no longer sufficient, because all sectors now have to anticipate the expansion of their productive capacity. This implies a redefinition of labor values (Pasinetti 1988). In a dynamic context, the labor value of each commodity cannot just embody the worn out capacity, but should also encompass the labor required for the reproduction of this commodity at time t+1. The “dynamic labor value” that results should thus develop according to the percentage increase of demand of each commodity, given by (g+r _i ). Supposing, to simplify, that there is only circulating capital and denoting by

l the row vector of direct labor requirements

A the input matrix

G the diagonal matrix of the (g+r _i ) at time t

the row vector of the “dynamic” labor values is

l ( t ) = l [ I − A ( I + G ) ] − 1

(2)³

To show how the “dynamic” labor values solve “transformation,” let us write the vector of the prices of production ⁴

p ( t ) = l w + ( 1 + π ) p A

(3)

where π is the uniform (capitalist) rate of profit.

As already noted, this capitalist rate of profit is higher than the “natural” rate of profit because of the “appropriation” component (δ_i )

π = π * + δ i

In matrix terms, taking into consideration formula (1):

∏ = G + Δ

(4)

where

Π is the diagonal matrix of the uniform rates of profit

Δ is the diagonal matrix of the δ_i, i.e. π − π i * , for each i,

both referring to time t.

Rewriting formula (3) taking into consideration (4) we have

p ( t ) = l w + p ( t ) A + p ( t ) A G + p ( t ) A Δ p ( t ) = l [ I − A ( I + G ) ] − 1 w + p ( t ) A Δ [ I − A ( I + G ) ] − 1

Since in the “natural system” Δ = 0, Pasinetti obtains the remarkable result that the dynamic “natural prices” are directly proportional to the dynamic labor values

p ( t ) = l ( t ) w

(5)

Thus, at the fundamental (“natural”) level, the labor theory of value holds perfectly, and the principle of the conservation of value during “transformation” is respected. Pasinetti’s result confirms Marx’s insight that, fundamentally, prices of production are nothing other than transformed values, but capitalist structures distort the original ratios to conform to the requirement of a uniform rate of profit.

3.1 Labor Value—The Logical Implication of the Production Paradigm

The paradigm of production is shared by a host of contemporary heterodox schools. Among the scholars who explicitly refer to such a paradigm, one may quote the post-Keynesians narrowly defined (as the direct heirs of Keynes in Cambridge), the Sraffians, the radical/Marxians, the Kaleckians, and part of the Schumpeterians. ⁵ Other scholars do not make such a reference explicit, but their works are fundamentally in line with it. I think, for instance, of the evolutionary and institutionalist scholars belonging to EAEPE.

The most logical implication of the paradigm of production is the theory of labor value. Keynes was fully aware of this. In fact, in chapter 16 (on “Sundry Observations on the Nature of Capital”) of the General Theory, Keynes (1936: 213–14) writes:

I sympathise . . . with the . . . doctrine that everything is produced by labour, aided by [the] . . . technique, by natural resources . . . and by the results of past labour, embodied in assets. . . . It is preferable to regard labour . . . as the sole factor of production. . . . This partly explains why we have been able to take the unit of labour as the sole physical unit we require in our economic system, apart from units of money and of time. (emphasis in the original)

This quotation shed lights on what we find in chapter 4, on “The Choice of Units.” Keynes (1936: 40–4) argues here that macroeconomic aggregates such as net output or total demand can be rigorously quantified only on the basis of two units of measurement, i.e. their money value and the volume of employment they embody. This latter unit of measurement can be expressed in terms of hours of ordinary labor. This is not a simple technical device to transform ex post monetary variables into their real equivalents, but implies that Keynes was in accordance with the classical approach to value.

However, very few post-Keynesians share this theory of labor value. Pasinetti is the most important exception, but usually his colleagues of the Keynesian school do not follow him on this ground. Sraffians, in particular, strongly oppose the labor theory of value charging it with redundancy and inconsistency (Steedman 1977). They do not recognize the decisive support of this theory that comes from Pasinetti’s work, and admit not having a specific theory of value. Alternatively, they take as theory of value the Sraffian prices of production (Screpanti 2003; see Reati 2005 for a discussion). Some of them, more Marxian sympathetic, try to justify capitalist exploitation without referring to surplus-value (e.g. Screpanti 2001). I can hardly understand this situation.

A paradigm, to be accepted, should be logically consistent (“waterproof”) and, consequently, any internal failure should be put into evidence. From this standpoint what Steedman (1977) has done is very positive; among other things he has clarified many issues connected with the Marxian theory of labor value. However, it seems to me that the long discussion that followed Steedman’s book has not at all destroyed such a theory, which also benefits from strong empirical evidence. ⁶ Thus, the hostility towards labor values shown by so many Sraffians appears to me unjustified and I plead for avoiding further sterile quarrels and follow Pasinetti’s exhortation to work in a co-ordinated way to strengthen the production paradigm.

3.2 What Makes the Success of a Paradigm?

In economics, the validity of a paradigm is measured not only by the fact that the majority of scholars share the same “research program” but also by the fact that the theories that spring from that paradigm aptly guide economic policy. In my view, this second component of success is as important as the first one: if the economic policy resulting from a given theory gives good results for society, this does not prove by itself the soundness of the underlying theoretical apparatus but nevertheless gives a clear indication in this sense (a tree is judged according to its fruits!). From this perspective Keynesian theory was very successful because it inspired the economic policy of most countries for about 40 years—from the depression of the 1930s till the end of the “golden 1960s”—with very positive results. Thanks to the support of demand, the world economy was drawn out of the depression and, subsequently, “fine tuning” demand management avoided pronounced recessions in the Western world. In Europe this was supplemented by a set of structural policies with the aim of reducing the unbalances and inequalities which result from an unfettered capitalist dynamics. The “Fordist consensus” of this second period prevented major social unrest by linking wages to productivity, a typical post-Keynesian recipe.

On theoretical grounds, however, the Keynesian school did not obtain the same success. Pasinetti’s analysis reported above provides an excellent explanation of this partial failure. Is this the end of the story? In my view it is not. To assess the possible future of the classical/Keynesian paradigm, we should consider it within the general context of the political and power relations that characterized the period where Keynesian policies were dominant and when, on the contrary, the rival theory gained momentum.

Historical experience shows that the success or failure of an economic theory does not depend exclusively on its theoretical strength, but also on the fact that the theory says the right things at the right time. Obviously, theoretical validity is a strictly necessary condition, but it does not suffice by itself to guarantee the prevalence of a “good” theory. Pasinetti (1981: 11–14) provides an interesting example referring to the 19th century, when in the 1870s the marginalist school took over the then dominant classical school. The decisive reason was not an alleged intellectual superiority but, rather, the fact that marginalist theory was very functional to the interests of the European bourgeoisie of that period. In fact, the industrial bourgeoisie—that, when the classical paradigm was framing its way, was the mounting class, struggling with the old class of landowners for political power—was firmly settled in power in the 1870s. The theory of labor value was functional for this ascent, since it emphasized the importance of productive activity as opposed to the passive role of the rentier, but its interest vanished when the industrial bourgeoisie did not have to justify its position any more. Moreover, the classical theory of labor value had taken a dangerous turn with Marx who, relying on the same classical postulates, demonstrated that at the origin of profit there is exploitation. Finally, the social situation was upsetting: Europe was threatened by a revolutionary wave (e.g. the Commune de Paris, 1871) and socialism was gaining momentum. In such circumstances, “if only one could find an economic theory that made no reference to labour, no reference to means of production, possibly even to production itself . . . that would surely be the sort of thing that a frightened Establishment could not but most warmly welcome. Marginal utility provided precisely that” (Pasinetti 1981: 13). ⁷

Let us now consider what happened to the two rival paradigms in the post-World War II period in Europe and the United States. We shall see that, also in this case, the decisive factors putting to the fore neoclassical theory were far from depending exclusively on pure theory.

The economic situation was characterized by a long upswing until the mid-seventies, followed by a depression and long stagnation. It was the fourth long wave in the history of capitalism; the upswing was driven by radical technical and institutional change and, when the technological revolution exhausted its effects, the system fell into long stagnation.

On the political side, during the long expansion the workers’ movement progressively gained power, reaching its maximum strength towards the end of the 1960s. This situation ended with huge unemployment resulting from the downswing, which inflicted a severe defeat on the working class. The balance of power was completely upset in favor of capitalists, and this created the ground for a resurgence of neoclassical theory.

On the economic policy side, depression and stagnation showed the inadequacy of the demand-management policies that, being essentially short-term policies, worked very well when the underlying long-term trend was growing, but were inappropriate to solve the problems of structural change raised by the long stagnation.

At the level of economic thought, the Keynesian school was strong during the whole economic upswing; however, when growth and employment declined, there was a sort of theoretical vacuum. In fact, to interpret and govern the new reality, two things would have been necessary: (i) in the field of “pure theory,” an extension to the long term of the classical paradigm, and (ii) new conceptual tools at the level of institutional analysis. These theoretical needs were eventually fulfilled, but with such long delays that the possibility for the post-Keynesian school to keep the lead in both theory and policy was compromised.

Pasinetti achieved the fundamental task of extending the classical paradigm to the long run quite early. The first version of his model of structural change was presented at a Vatican conference in 1963 (Pasinetti 1965), and at that time Pasinetti was on the brink of publishing a book with the full version of the model, but he was detained by further theoretical problems. As a result, the book appeared only in 1981, when the neoclassicals had already prevailed in academia and on economic policy. At an applied level, the revival of long-wave theory (Mandel 1976) stimulated a mass of research that enriched very substantially the production paradigm. I am referring to the Schumpeterian school of Christopher Freeman, Dosi, Perez, et al., to the French régulation, and to its American counterpart of the Social Structures of Accumulation. Unfortunately these contributions also arrived too late to counteract the neoclassical offensive and, in addition, they were relatively poor in terms of economic policy prescriptions.

All this provided fertile ground for a reflourishing of neoclassical theory. As a matter of fact, neoclassicals considered that they need do nothing substantial to change their exchange paradigm in order to cope with the new facts. By applying their methodological reductionism, the structural crisis of the 1970s was interpreted as a short-term cyclical downturn produced by the first oil crisis, an error of perspective that was also committed by several post-Keynesians. Unemployment—instead being analyzed as the result of structural change—was attributed to the (mal) functioning of the labor market. The insufficiency of Keynesian demand policies to cope with the long stagnation was interpreted as a failure of Keynesian theory, etc. The new conventional wisdom was a faithful reliance on general equilibrium theory and on neoclassical growth models, indicating that the system could grow indefinitely provided governments do not interfere with free markets. The political weakness of the working class favored the formation of conservative governments in most countries, and neoclassical policies became dominant.

In conclusion, we see that the present defeat of the classical/post-Keynesian paradigm is not due to theoretical weaknesses but is the result of the unfavorable social and political conditions of the last decades. As this will not continue forever, our task is thus to deepen our work along the lines indicated in Pasinetti’s book, in order to support the social movements with a solid analytical and policy background to be used when, thanks to social struggles, the balance of powers will shift again toward the working classes.