Market Process
vol. 1, no. 1 (January), 1983, pp. 3-5

Classical and Neoclassical Theories of General Equilibrium: Historical and Mathematical Structure
by Vivian Walsh and Harvey Gram bottled equilibirum
New York: Oxford University Press, 1980, pp. xvi, 426

This intriguing volume on alternative theories of equilibrium should be of interest to those working in the Austrian tradition--even though the book deliberately downplays the contribution of the Austrian school. When a book is well written and its arguments are laid out in their essential form, the reader can learn much, but sometimes what he learns is not exactly what the authors are striving to teach. So it is with Walsh and Gram.
        The economic theories dealt with range from pre-Smith to post-Hahn. The ideas are organized in a way suggested by the book's subtitle. The first half of the book presents ideas chronologically by author. Tracing the development of the thought from Sir William Petty to Leon Walras involves identifying the important developers, pointing out the relevant biographical considerations, and analyzing their contributions to general-equilibrium theory. In the second half the individual contributors recede into the background and the material is presented logically as opposed to chronologically. The chapters alternate between classical and neoclassical analysis as the corresponding models increase in complexity. The equilibrium conditions are expressed verbally, graphically and mathematically, and the basic structures of the classical and neoclassical models are compared and contrasted.
        The fundamental and recurring message in the book is that classical theories and neoclassical theories are to be distinguished in terms of their respective themes. Concerns about subjectivism, methodological individualism and even marginalism take a back seat to the underlying thematic difference between the two kinds of theories. The classical theme is the accumulation and allocation of surplus output; the neoclassical theme is the allocation of given resources among alternative uses. As Walsh and Gram sift through the history of economic thought, it is this thematic distinction that serves as their sieve.
        Clearly, the authors have an affinity for theories that allow for an economic surplus. Progress in the development of economic thought is to be associated with theorists who increased our understanding of the surplus. The major figures include those who have been involved in the current "classical revival" as well as the classical theorists themselves. More specifically, the development of the notion of a surplus is traced through the writings of William Petty, Richard Cantillon, Francois Quesnay, Adam Smith, David Ricardo and Karl Marx. Then, after a long dry spell of over a half century, further development occurred with the efforts of Pierro Sraffa, John von Neumann, Joan Robinson and a number of lesser contributors. The long dry spell, of course, is coterminous with the birth and growth of the Austrian school and of neoclassical economics in general. The economic theory that dominated in this period (from 1871 to the 1930s) is described by Walsh and Gram as a "highly special new version of allocation theory." It was highly special because it did not dwell on the allocation of the economic surplus.
        Readers who have a special appreciation for the contributions of the Austrian theorists can gain many insights from this particular perspective on classical and neoclassical theories, but the ultimate assessment of the theories is bound to be far different from those of Walsh and Gram. The importance of an adequate theory of capital, the meaning of an economic surplus, and the role of economic classes and institutions in theoretical constructs are three issues that are likely to be on the mind of the Austrian-oriented reader throughout most of the book. A brief consideration of each of these issues can convey much of the substance and flavor of this volume. It can also serve to contrast the Austrian resurgence with the classical revival.
        From Carl Menger onward the Austrian authors have been aware of the lack of an adequate theory of capital in both classical and neoclassical writings. This is inadvertently dramatized as Walsh and Gram construct the bare-bones models of these two alternative theories. The classical notion of a surplus requires that both inputs and outputs be measured in physical terms. But if surplus is to be reckoned by subtracting the (physical) quantity of inputs from the (physical) quantity of outputs, the input good and the output good must be the same good. This accounts for the dominance in classical theory of one-good models. In the Ricardian corn model, corn constitutes both the input (seed and subsistence for the workers) and the output. The amount of corn produced minus the seed corn and the corn consumed by workers is the surplus. Walsh and Gram begin their presentation of the classical vision by showing (mathematically) that for an economy to be "viable," the output of corn has to be at least as great as the input, and for a surplus to exist, it has to be greater. The Austrian-oriented reader will have no trouble following this logic, but those who have studied Böhm-Bawerk's Capital and Interest and Hayek's Pure Theory of Capital will be left with an empty feeling. There are no distinct capital goods, much less a capital structure in the classical vision.
        As Walsh and Gram demonstrate, the classical model can be expanded to represent a two-sector economy or even an n-sector economy. In these more complex models the surplus has to be expressed as a vector of goods. But still, the list of inputs is identical to the list of outputs. The authors focus on an economy in which iron and corn are used to produce iron and corn. This is one member of a class of models that conform to Pierro Sraffa's vision of the "production of commodities by means of commodities." Again, there are no distinct capital goods. The sympathetic reader can imagine all sorts of capital-using production processes that transform the pile of inputs into a larger pile of outputs, but these processes are no part of the classical analysis.
        The neoclassical model (as outlined by Walras) differs markedly from the classical model, but deserves no higher mark on the question of capital. Here, the subscripts on the input quantities have to be altered to indicate that the inputs consist of land and labor rather than iron and corn. In the neoclassical vision these distinct factors of production are transformed into an array of consumption goods--wheat and rice in the two-commodity model presented in the book. Whatever capital goods may come into existence during the transformation phase are assumed to go out of existence just before the output emerges. The authors recognize that the Walrasian model can be generalized to accommodate an intertemporal equilibrium. The output variable is double subscripted with the first subscript representing the particular good and the second representing the period in which its production is completed. This formulation allows for a variable period of production but constitutes--in the eyes of one Austrian writer--capital theory without capital. In any case Walsh and Gram mention the intertemporalized Walrasian model only to deny that it can be found in Walras' own writings.
        The issue of capital theory is not tangential to the understanding of this book. It bears directly on the validity of the book's central message. For theories that allow for the production of distinct capital goods (factories, machines, and tools) there is no sharp distinction between classical accumulation and neoclassical allocation. The allocation of resources to the production of a fish net, for example, can be modeled with a neoclassical construct, and yet the fish net itself constitutes accumulation in the classical sense. The distinct capital good is produced with a given amount of resources available at a particular point in time; it represents, at the same time, an increase in the capital stock and hence an increased capacity to produce in the future. The concept of capital goods bridges the gap between the notion of allocating resources at a point in time and the notion of allocating resources over time. It is only with this bridge out that Walsh and Gram can draw the sharp distinction that serves as the thesis of their book.
        If the reader finds too little discussion about capital goods, he will find more than enough discussion about the economic surplus. This concept, which defines the theme of the classical writings, is alien to the writings of the Austrian school. But a careful reading of Walsh and Gram's treatment of the concept of a surplus can help to reconcile the ideas of Quesnay and Smith with those of Menger and later Austrians. The issue of whether or not a particular economic activity yields a surplus is the same as the issue of whether or not that activity is productive (in the classical sense). What is it about the classical models that causes a particular factor to appear to be productive of a surplus? An answer to this question, which can be found "between the lines" in Walsh and Gram's discussions, can explain much.
        In the vision of the Physiocrats, agricultural production is productive; industrial production is not. Only land yields a surplus. And inspection of the model that yields this conclusion reveals that land is the only significant factor of production that is not explicitly taken into account. Iron and corn (and land) are used to produce iron and corn. The unaccounted-for input (land) yields an unaccounted-for output (a surplus of iron and corn). Since land is dominant in agricultural production and trivial in industrial production, the former activity appears to be productive while the latter does not.
        The vision of Adam Smith can be analyzed in a similar way. Smith accounted for land in his formulation, but believed that both the agricultural sector and the industrial sector were productive of a surplus. The thing that allows for this productivity is time--the factor that is not explicitly counted as an input in Smith's formulation. Smith saw labor as being unproductive if it is aimed at producing a service for present consumption and productive if directed at producing goods and services for the future. That is, only if labor is combined with the unaccounted-for input, time, will it yield an unaccounted-for output. Had time been taken into account explicitly, or--what amounts to the same thing in this context--had future consumption been appropriately discounted, the Smithian surplus would have disappeared.
        In Walras' neoclassical vision there is no unaccounted-for input. This follows from the very structure of the model, in which the value of all goods is fully imputed to the corresponding factors of production. There is no possibility of a surplus because the value of the output is always equal to the value of the input. The contrast between Smith and Walras is instructive. Smith, in effect, discounted the present rather than the future. This caused him to emphasize the importance of maximum growth. Consumption deferred is consumption increased. Walras constructed a model in which no growth is possible. The total value of goods is always confined to the value of preexisting resources. The economy is forever trapped in what Murray Rothbard has recently called the "Walrasian Box."
        Focusing on the unaccounted-for input can cast Austrian theory in an interesting light. The different models (Physiocratic, classical, neoclassical, Austrian) can be identified in terms of the factor that lies outside the economic calculus. It is land for the Physiocratic model; time for the classical model. In the neoclassical model, nothing lies outside the economic calculus. For Austrian theory the unmodeled factor is entrepreneurship. The Austrians account for land, labor, time and existing capital goods, but they escape from the Walrasian Box by allowing for entrepreneurs to combine these factors in new ways, giving rise to an output which in (potentially) more valuable than the previous collection of inputs. This perspective is particularly favorable to Austrian theory. There is no justification for failing to incorporate land or time into the economic calculus, but entrepreneurship is inherently unmodelable. This unique aspect of entrepreneurship has been emphasized by the Austrian school from Menger down to the present-day Austrian theorists.
        A final perspective can be offered in terms of economic classes and institutions. Most modern theorists are critical of classical theory to the extent that it turns on behavioral differences between broadly defined economic classes. The assumption, for instance, that capitalists save and workers consume just makes for bad theory. Neoclassical theory is classless theory and, in this respect, has to be judged an improvement. But neoclassical theory is also institutionless. None of its conclusions depend upon the existence of any particular set of institutions. As Walsh and Gram explain, neoclassical theory can be adapted to apply to any economic system and any pattern of resource ownership. In principle neoclassical theory can predict how a market economy will allocate the economy's resources, or, alternatively, it can instruct the government how to allocate them. The analytical equivalence of the decentralized direction and the centralized direction of resources accounts for the Lange-Lerner type endorsement of socialism and for the tendency of some modern neoclassicals to be virtually indifferent between market solutions and governmental solutions to allocation problems.
        The theorists of the Austrian school do not conform to either the classical or the neoclassical view. Austrians theory is classless but not institutionless. The knowledge about resource availabilities, production techniques, and consumer preferences is, by its very nature, decentralized knowledge. This requires, then, that the economic decisions that make use of this knowledge be similarly decentralized. The spontaneous order that can arise only from decentralized decisionmaking is the central theme in the writings of Hayek and several other Austrian theorists. The failure to recognize this institutional requirement can lead to bad theory and bad policy as is evidenced by the propositions of "welfare economics" and the attempt to use government to correct for so-called market failures.
        The reader of Walsh and Gram's book will gain many insights. He will gain even more if he keeps in mind that the authors' thematically defined categories of theories are not jointly exhaustive. The very identification of the two themes requries that capital theory be suppressed; neither theme recognized the unique role of the entrepreneur; and the fundamental importance of economic institutions is overlooked throughout the book. When these considerations are taken into account, the reader gains a healthy appreciation for the Austrian school on the basis of this comparison of classical and neoclassical theories.

    Roger W. Garrison
    Auburn University