Sraffa's Vision of the "Production of Commodities by Means of Commodities" Admits of No Distinct Capital Goods

Classical models, at least on this count, exhibit more similarities than differences when compared with their neoclassical counterparts. Although an explicit period of production is built into the models patterned after classical theory, the collection of capital goods spanning this period is never brought into view. Classical models are completely specified in terms of the initial inputs and the ultimate outputs—which, in turn, become the inputs of the subsequent period. The basic structure of these models requires that the list of inputs and the list of outputs be qualitatively identical. Corn and iron, for instance, are used to produce greater quantities of corn and iron. (The quantitative difference between inputs and outputs, the surplus, becomes the primary focus of classical analysis.)

This vision of the “production of commodities by means of commodities” (Sraffa, 1960) has served as a basis for capital theory in spite of the fact that the vision admits of no distinct capital goods. Capital theory of this type is based on the premise that the time separating the input commodities and the output commodities is a satisfactory surrogate for capital. Insights about the nature of capital goods that come into (and go out of) existence during this time and about the relationships between different kinds of capital goods are simply out of reach.

--Roger W. Garrison, "A Subjectivist Theory of a Capital-Using Economy," in The Economics of Time and Ignorance with a New Introduction, by Gerald P. O'Driscoll Jr. and Mario J. Rizzo, Foundations of the Market Economy (London: Routledge, 2002), 161-162.