“The main thrust of Ricardo’s system is a bold attempt to determine values independent of demand considerations,” concludes Kenneth Arrow near the end of his article “Ricardo’s Work as Viewed by Later Economists” (Arrow 1991, p. 75). The truth, however, is not so clear-cut. Arrow’s position—though by no means a minority, Samuelson, for one, is in agreement—has come under scrutiny as a misreading of Ricardo’s works, with scholars Caravale and Hollander refuting Arrow’s interpretation through their assertions that demand, in fact, constituted a crucial facet of Ricardo’s theory. Frequently at stake in this debate is a historical question of philosophical lineage. In particular, Ricardo’s perceived relation to (and coherence with) the subsequent ‘neo-classical’ turn is implied to hinge on this question: did Ricardo’s labour theory of value negate considerations of demand in the economic process, irredeemably cleaving his thought from the neoclassical school? Indeed, argues Arrow. Hollander takes the opposite tact, seeing Ricardo as a potential progenitor of neoclassical demand-supply analysis (Hollander 1991). Caravale’s interpretation, however, offers the most compelling case, refuting the first clause yet supporting the latter (Caravale 1994). Offering a nuanced and qualified interpretation, Caravale’s conception of Ricardo as a demand-oriented economist nonetheless incompatible with the neoclassicals provides a convincing middle ground in the face of dichotomy.
The neoclassical theory, posits Arrow, was born largely out of a ‘reaction to the limits of Ricardian classicism’ (Arrow 1991, p. 70). In his subsequent exploration of these limits as interpreted by the neoclassicals, he settles on one culprit in particular—the labour theory of value. With value in Ricardo’s system emanating solely from supply costs, that ‘very elementary tool’ of the demand schedule simply could not be conceived. Particularly obstructive was Ricardo’s differentiation of use and exchange value which, while not as explicitly articulated as in Marx, blocked any consideration of utility (and subsequently demand) in determining exchange ratios. As a result, any linkages between demand and price, for Arrow, could not be systematically understood. This renders Ricardo’s thought simplified and restricted, riddled with fragile assumptions crucial to the erection of a “purely supply-based pricing theory” (ibid., p. 75). Crucially, this means that Ricardo’s approach is irretrievably divergent from the neoclassical understanding as there is an implicit assertion that demand, as a tool of economic analysis, is under the mostly-exclusive purview of the neoclassicals. Without the framework of neoclassical demand functions, Arrow suggests, “there is really no way of opposing a theory that demand influences price to a theory where prices are determined purely by technology” (ibid.). As such, through his insistence on value as something cost-determined, Ricardo missed out on demand—that hallowed signifier of neoclassical thought—thus reducing him to an “imperfect and none-too-coherent predecessor” (ibid., p. 70).
Arrows interpretation, upon a cursory reading, offers a compelling-enough case. He is not mistaken in his claims that Ricardo conceived of value as a property emanating exclusively from labour, and that this value would necessarily be reflected in a commodity’s price. To a certain degree then, he is justified in characterizing Ricardo as a theorist none-too preoccupied with questions of utility and demand. Arrow, however, falls victim to his own totalizing language and the kernel of truth clearly present in his analysis becomes overshadowed by his occasionally dogmatic inclinations. In stating that Ricardo possessed no “explicit relation of demand to price” (ibid., p. 75), Arrow leaves himself vulnerable to the more-than-justified rebuttals of Hollander and Caravale. Arrow writes near the close of his article that “the dazzling beauty” of Ricardo’s simplistic argument would understandably “blind the eyes of any investigator to these problems, however obvious they might seem” (ibid., p. 76) and that the “slightest variation” of Ricardo’s sketch leaves the construction vulnerable to collapse. It is somewhat ironic then that his own interpretation may now meet a similar fate, as Caravale and Hollander dually prove.
While it may be accepted that in Ricardo’s scheme, consumer demand does not directly bear upon a commodity’s price, the assertion that the two have absolutely no relation is incorrect. Both Caravale’s “Demand Conditions and the Interpretation of Ricardo” (1994) and Hollander’s “On the Endogeneity of the Margin and Related Issues in Ricardian Economics” (1991) are here largely in agreement. To start, for both scholars Ricardo understands demand as a pre-condition of production. As Caravale quotes, “A commodity is not supplied merely because it can be produced, but because there is a demand for it” (Ricardo 1951-73, I, p. 385 cited Caravale 1994, p. 231). Moreover, this demand varies—subject as it is to the “wants and wishes of mankind” (ibid. p. 88 cited ibid.) and the “caprice of taste” (ibid.). With demand then linked to production, and subject to variation, how may these fluctuations be related to price within the framework of the labour theory? The answer lies in Ricardian rent theory.
Using the example of corn, Hollander explains that in Ricardo’s understanding of rent—where each round of expansion to new land (extensive) or act of intensified cultivation (intensive) brings about a marginal decrease in productivity—the extension of production entails increasing labour costs and decreasing returns. “The magnitude of labour costs is determined (partly) by the degree of land scarcity” (Hollander 1991, p. 163). This land scarcity, in turn, is influenced by demand. Caravale points out that Ricardo makes both an important distinction between ‘market prices’ and ‘natural prices,’ and an explicit connection between demand and market price. If a change in preferences should “increase the demand for silks,” the “market price of silks would rise” (Ricardo 1951-73, I, p. 90-91 cited Caravale 1994, p. 235). The resulting increased rate of profit would subsequently entice a transfer of capital in the direction of silks, resulting in intensified production, as dictated by consumer demand.
Now, remembering Ricardo’s work on rent, in the case of corn this intensified production necessitates that “with any great additional demand for corn land of a worse quality must be taken into cultivation, on which more labour will be required to produce a given quantity, and the natural price of corn will be raised” (Ricardo 1951-73, I, p. 312 cited Hollander 1991, p. 165: emphasis added). The natural price of a good, as assumed to be determined by the marginal labour cost, is explicitly influenced by demand considerations. That the influence of demand on natural prices is not direct and operates instead via the vector of labour costs does not negate the link between the two—Arrow’s interpretation is inadequate.
Despite this, Hollander and Caravale’s interpretations cannot be understood to stand on equal footing. While the two are in general agreement as to the presence of demand in Ricardo’s analysis, the nature of this presence and its implications are disputed. In large part this is a result of the fact that Hollander’s analysis, broadly speaking, is conducted in the pursuit of a specific historiographical objective, namely, the desire to directly oppose the perception that Ricardo’s analysis “goes counter to…the so-called ‘neo-classical’ principles.” In fact, writes Hollander, “Ricardo’s labor theory and the doctrine of rent used in support are part and parcel of [that same] body of analysis” (Hollander 1991, p. 162). Accordingly, much of Hollander’s interpretation is devoted to demonstrating this supposed coherence. Ricardo’s work is thus cast largely in terms of its relation to the neo-classical doctrine, and Hollander may perhaps be prone to an overemphasis on points of agreement at the expense of those more conflictual.
The general thrust of Hollander’s article focuses on refuting a specific comment from Paul Samuelson regarding Ricardo’s failure to understand the endogenous determination of the extensive (and intensive) margin as mutually conditioned by both demand and supply. That is, Hollander is intent on demonstrating that Ricardo did, in fact, account for the endogenously determined margins of production factors with respect to demand and may thus be considered as a bona fide “‘demand-supply’ value theorist” (ibid. n.). After his articulation of demand’s influence on natural and market prices as mediated by marginal labour costs (see above), he comes to a ‘special case,’ offering the insight that Ricardo may, in fact, have implied the existence of a negatively sloped demand curve in the vein of the neoclassical tradition. In the case of ‘capacity output’—when a given plot of land is exploited at maximum capacity and further intensive use is ruled out—Ricardo writes that any tax imposed “will fall on rent, and not on the consumer…[the farmer] cannot raise the price of his corn because it is already at the highest price at which purchasers will or can buy” (Ricardo 1951-73, I, pp. 250-51 cited Hollander 1991, p. 166). Hollander takes this singular passage, and its assertion that consumers’ demand may occasionally be price-elastic, to hint at the hint of the idea that Ricardian thought may be directly compatible with neoclassical demand functions.
Correct though Hollander may be in this specific case, this example is emblematic of his tendency in the article to clutch at connections between Ricardo and the neoclassicals that feel tenuous at best, while ignoring or otherwise underdeveloping potential counterpoints that come to mind. As but one example, Hollander’s disclaimer that Ricardo “usually assumes zero price elasticity of demand” (ibid. p. 166) is subsequently brushed aside despite such a notion’s evident conflict with neoclassical theory. While Hollander provides rigorous and compelling justification as to the fact that demand constituted a crucial and conscious component of Ricardo’s economic considerations, he falls shorter in establishing the link he seeks with the neoclassicals. In large part, this is due to the absence of an interrogation into the concept of demand itself and its possible permutations. After proving that demand influences price in Ricardo’s work, Hollander—to an extent—takes it as given that this version of demand is one shared by the neoclassicals. As Caravale demonstrates, this cannot be the case.
Ricardo, Caravale argues, possessed a particular understanding of demand more in line with Adam Smith than the neoclassicals. Demand, in Ricardo, is to be understood in terms of Smith’s work on ‘effectual demand’—defined simply as the “the demand of those who are willing to pay the natural price of a commodity” (Smith 1973, p. 63 cited Caravale 1994, p. 231: emphasis added). While the demand for any good at the given market price may be higher or lower than this point of effectual demand, linked as it is to the natural price, Ricardo is said to understand the presence of an equilibrating mechanism. A higher (lower) market price will incentivize expansion (contraction) of that particular sector, upon which the market price will converge with the natural price, and the number of quantities purchased will be equal to the effectual demand. The disparity between natural prices and market ones provides the information that allows the competitive adjustment mechanism to balance the market out at the point of effectual demand.
Such an equilibrating mechanism, Caravale argues, is distinctly different from neoclassical-style intersecting supply and demand functions. In Ricardo’s model, the equilibrium relies on knowledge realized “outside the market, on the basis of elements whose magnitude does not depend on…the market mechanism” (Caravale 1994, p. 237)—knowledge of the natural price and its corresponding point of effectual demand must be a precondition for competitive adjustment. In the neoclassical model, alternatively, “all relevant economic magnitudes are simultaneously determined” (ibid., p. 234).
Demand, in Ricardo, is incompatible with the neoclassical notion of the demand curve for this reason. Marshallian demand curves, writes Caravale, can only be defined in analytical terms with three ceteris paribus conditions: that the tastes of consumers, their incomes and the prices of all other good may be taken as given. In Ricardo’s framework however, this is not possible as the latter two values (income and relative prices) can only come into existence iteratively “as part of the solution of the model” derived from the given real wage (ibid., p. 233). The requisite simultaneity of neoclassical analysis indicates a fundamental difference in approaches for Caravale and means that the “notion of the demand function should be recognized as basically incompatible with the structure of classical analysis” (ibid., p. 234). Caravale, in his interpretation, digs deeper than Hollander in analysing the exact nature of demand as understood by Ricardo. While both convincingly refute Arrow’s understanding of the matter, Caravale’s willingness to interrogate the specificities of Ricardo’s alternative approach and highlight the uniqueness of the effectual demand paradigm results in a qualified interpretation convincingly splitting and reframing the Arrow-Hollander dichotomy.
References
Arrow, Kenneth J. “Ricardo’s Work as Viewed by Later Economists.” Journal of the
History of Economic Thought 13, no. 1 (1991): 70–77. doi:10.1017/S1053837200003400.
Caravale, Giovanni A. “Demand Conditions and the Interpretation of Ricardo.” Journal
of the History of Economic Thought 16, no. 2 (1994): 229–47. doi:10.1017/S1053837200001966.
Hollander, Samuel. “On the Endogeneity of the Margin and Related Issues in Ricardian
Economics.” Journal of the History of Economic Thought 13, no. 2 (1991): 159–74. doi:10.1017/S1053837200003540.
Oskar Steiner 