I. INTRODUCTION: THE ISSUE
“It is but recently that, in pure theory, the concept of social value came into prominence.” So wrote Joseph A. Schumpeter (Reference Schumpeter1909, p. 213) in 1909 in the pages of Harvard’s Quarterly Journal of Economics. While the fathers of neoclassicism in Europe spoke only of “individual value,” the Austrian economist continued, the notion of social value owes its introduction mainly to John Bates Clark in the United States and there it quickly met with such a general approval that “today it is to be found in nearly every text-book” (Reference Schumpeter1909, p. 213). Schumpeter was promptly echoed by Benjamin Anderson, then a recent Columbia graduate. The idea of social value, Anderson (Reference Anderson1911, pp. 3–4) noted, “has become important only of late years, chiefly through the influence of Professor J. B. Clark [and] has been found attractive by many other American writers.” Schumpeter’s and Anderson’s words are significant because they reveal an aspect of Clark’s contribution and its consequences for American economics that has received little attention in the literature. Clark had presented an embryonic discussion of the social dimension of economic valuation in his Philosophy of Wealth (Reference Clark1886), but it was only after his later “marginalist turn” that the concept of social value became a distinguishing feature of American neoclassicism. The publication in Reference Clark1899 of the Distribution of Wealth, Clark’s magnum opus, triggered in fact a whole array of reactions, both supportive and critical, centering on Clark’s contention that the value of a good is determined by, and accords with, its marginal utility to society considered as a unitary organism. The discussion was carried on simultaneously at methodological, epistemological, and theoretical levels, and several of the leading figures of the period took a more or less active part in it. The aim of this paper is to provide a reconstruction of this debate, which marks a crucial, and still largely unexplored, episode in the early history of neoclassical economics in the United States.Footnote 1 This will be done in three steps. The first step analyzes Clark’s approach to social value as presented in his Distribution of Wealth; the second step deals with other influential contemporaries who adopted a similar social value perspective, with a main focus on Edwin R. A. Seligman; the third step discusses those critics who, with due differences in style and intent, animated the debate over social value, reviewing the contributions of (among others) Herbert J. Davenport, John Maurice Clark, and George P. Watkins, in addition to the already mentioned Schumpeter and Anderson. The final section presents some conclusions.
II. J. B. CLARK ON SOCIAL VALUE
Although Clark had addressed the broad question of value in one of his earliest essays (Clark Reference Clark1881; see below), his most systematic discussion, and the one that mostly concerns us here, is to be found in a later paper, “The Ultimate Standard of Value” (Clark Reference Clark1892), then reproduced without substantial changes as chapter XXIV of the Distribution of Wealth (to which we will refer here). Clark’s starting point was his familiar assumption that all concrete instruments, including land, dissolve into a “permanent fund” of capital, perfectly mobile and measured in monetary units. Likewise, all different forms of labor can be reduced to a “permanent force,” equally homogenous and fluid. “There is before us the picture of social labor cooperating with social capital,” Clark could write (Reference Clark1899, p. 373), and “both are governed by the law of diminishing returns.” Owing to the pressure of competition, then, “wages conform to the product of the final increment of social labor” and “interest to the product of the final increment of social capital.” The question then arises as to “how may we measure labor, capital, and their products” so as to make certain that each value-productivity unit is remunerated in proportion to its marginal contribution—“a universal unit for measuring economic value is necessary, if the law of final productivity is to have scientific exactness” (Reference Clark1899, p. 375). This universal unit was found in marginal utility, or “effective utility” in Clark’s vocabulary. Effective utility is “the power that a particular unit of a commodity has to change the status of its possessor and to promote him in the scale of well-being” (Reference Clark1899, p. 376).
The problem with utility, however, is that it is measurable only ordinally. In quantifying well-being, Clark affirmed (Reference Clark1899, p. 380), the human mind is subject to the same limitation the eye faces when trying to measure the intensity of light: “it is possible to pronounce two lights equal; but it is not possible to tell, by the mere effect on the eye, how much brighter one light is than another.” By the same token, “it is possible to say that two pleasures are equal, but not to say that one is just twice as great as the other.” This was, however, an obstacle that could be overcome. An isolated Crusoe who produces and consumes his own products, Clark explained, reaches an equilibrium where effective utility and sacrifice are equal, and that is the moment in which he stops to work. If the product of any hour’s labor is taken away from him, he would not suffer the loss of anything more important than the product of the last hour, since he would simply forego that and re-create the product with the higher utility. Now, Clark affirmed, if it is possible to determine when a pain and a pleasure offset each other, and
if we can compare many kinds of pleasure with one kind of pain, we can, as a result, both compare pleasures with each other and obtain a sum total of many different ones. If a man knows that he would walk a mile for one gratification, and that he would do this twice over for another, he has the means of knowing that the good afforded by the second is twice as great as that afforded by the first, and that the gain insured by the two together is an offset for three walks of a mile each. (Reference Clark1899, p. 380)
Clark thus appears to be driven to a labor pain-cost theory of value, through which all commodities—including capital goods—are measured. Quite crucially, however, he left unexplained the psychological principles according to which “labor pain,” differently from utility, could be measured in cardinal terms.Footnote 2 In this connection, it is significant that a sharp critic such as Eugene Bohm-Bawerk (Reference Bohm-Bawerk1907) noted this incongruence in Clark’s thought, also pointing out a parallel between the methods of Karl Marx and Clark, despite the exact oppositeness of their conclusions.
Now, and this is the most relevant aspect, Clark asserted that the same reasoning that he described for an isolated worker applies to society as an organism.
Of a society regarded as a unit the same is true. It produces for itself, and the burden of its final labor measures the utility of its final products, which is the same as the effective utility of any of its products created by the same expenditure of working time. Take away the articles that the society gains by the labor of a morning hour,—the necessary food, clothing and shelter that it absolutely must have,—and to make good the loss it will divert the work performed at the approach of evening, which would otherwise have produced the final luxuries on its list of goods. To society the net importance of the different grades of commodities is equal: take away one variety entire, and terminal labor will be made to replace it. The things otherwise produced by that final labor will be the ones really lost, and their utility is measured by the burden entailed in the creating of them. (Reference Clark1899, p. 386)
This passage needs explanation. Clark thought it possible to merge all individual marginal utility schedules for goods satisfying wants of similar intensity into “social complements” of consumption. These complements are then ranked according to their “importance” to society—from the first, which includes society’s “necessities of life,” to the last, containing all kinds of “luxuries.”Footnote 3 It is also explicitly assumed that in arranging production (and the concomitant consumption), society begins with those complements possessing higher utility, gradually shifting to those of a lower grade. The marginal (or “absolute” in Clark’s varying vocabulary) utility of the last social complement of consumption then fixes society’s aggregate final utility—determining at the same time the measure of the social cost of production. Quoting again from Clark (Reference Clark1899, pp. 387–388): “everything that is produced by one hour of social labor, whether that labor be performed early in the day or late, possesses an effective social utility that equals the absolute utility of the final complement of goods consumed; and this, again, is counterpoised and measured by the sacrifice which all society undergoes in the labor of its final hour.”
But here a problem emerges. To affirm that the final utility of a social complement of consumption is measured by society’s marginal sacrifice does not necessarily indicate that the same reasoning applies to the individual goods included in the complement with respect to the individual labor that contributed to their production. To put it differently, in Clark’s words (Reference Clark1899, p. 390), “though the social organism will work till what it gets offsets what it suffers, will a man also work till what society gets from him offsets what he suffers?” In the case of the individual, Clark answered, the equilibrium is reached in a different way. Here the solution he proposed is that the utility of what an individual produces and offers in the market is measured by the disutility of society in producing what the individual obtains in exchange for it. Clark illustrated this point through a simple example. Let’s imagine a “miniature society” where individual A makes the article W, B makes X, C makes Y, D makes Z, and each receives and uses some portion of each product. The value of W, then, is not measured by the time A employs to produce it but by the time it takes A, B, and C to produce the goods that are exchanged for it. The same argument applies to every product of the society, so that “the price of things corresponds to the pain of acquisition, of which the unit is the sacrifice entailed on society by the work of the final period in each of a series of days” (Clark Reference Clark1899, p. 393)
The equivalence between an individual’s sacrifice and social utility is therefore only indirect, i.e., mediated through the different market transactions that contribute to the fixing of a good’s market value. This implies, Clark conceded, that “the direct burden of labor entailed on a man in the making of an article has no fixed relation to its market value.” He provided the following example: “The product of one hour’s work of an eminent lawyer, an artist or a business manager may sell for as much as that of a month’s work of an engine stoker, a seamstress or a stone-breaker. Here and there are ‘prisoners of poverty,’ putting life itself into products of which a wagon-load can literally be bought for a prima donna’s song” (Reference Clark1899, p. 393).
This was due to a form of market failure. Wherever in fact we find some “exceptional personal power or position” granting to any producer the “advantage of a monopoly,” there appears a divergence between the personal sacrifice incurred in the production of a good and its market value. A similar divergence emerges because of differences in wealth between individual producers, which Clark treated as a special form of monopolistic advantage—“the rich worker stops producing early, while the sacrifice entailed is still small; but his product sells as well as if it cost much more sacrifice” (Reference Clark1899, pp. 393–394). Yet, Clark promptly reassured his readers, for all “monopolized products” at the aggregate the general rule still applies: “they sell according to the disutility of the terminal social labor expended in order to acquire them.” Clark then added a final complication, arguing that the above standard is equivalent to saying that “the price of goods corresponds with the amount and efficiency of the labor that creates them” (Reference Clark1899, p. 394; emphasis added). Efficiency is in turn defined as the “power to draw out labor on the part of society”—and this apparently implies a labor command and not a labor pain-cost theory of value (Henry Reference Henry1995, p. 86). But, nonetheless, he immediately reiterated in a confusing fashion, “goods must sell at rates that are in accordance with the quantity and the efficiency of the work which creates them” (Reference Clark1899, p. 394).
Further aspects of Clark’s notion of social value are discussed below in connection to the debate it triggered among American economists. Before concluding this section, however, a few words should be spent as to what has been referred to as the “Clark problem,” i.e., whether the Social Christian views he had expressed in the Philosophy of Wealth can be reconciled with the strong defense of competition introduced in his later works. This is quite a controversial issue that goes well beyond the aim of this paper.Footnote 4 What is of interest here, however, is that in the Distribution of Wealth, Clark himself indicated in the idea of society as an evaluating organism the element of continuity with his previous works.Footnote 5 Clark had made his position clear as early as in 1881 in the essay “The Philosophy of Value,” then included as chapter V of the Philosophy of Wealth (Reference Clark1886). “Society, as an organic whole,” we find him saying (Reference Clark1886, p. 81), “is to be regarded as one great isolated being; and this being may and does measure utilities like a solitary tenant of an island.” In a following passage, he insisted (Reference Clark1886, p. 83): “market value is a measure of utility made by society considered as one great isolated being.” Such a cohesion was guaranteed by a “sympathetic connection between members of society,” a form of sympathetic solidarity that “moves the social organism more powerfully in the same direction” (Reference Clark1886, pp. 83–84). In Clark’s later works, the idea of society as the ultimate agent who sets values was retained, but his young faith in an altruistic human nature was substituted with an idea of social unity granted by the synchronizing pressure of competition on individual behavior. Competition was thus benevolent in a twofold way—because it led to an ethically just distribution and because it equated exchange value with social value.Footnote 6
III. FROM CLARK TO AMERICAN ECONOMICS
The idea that economic value reflects social utility shortly gained consensus among Clark’s contemporaries, to the extent that both Schumpeter (Reference Schumpeter1910) and Anderson (Reference Anderson1911) spoke of the existence of a “social value school” in American economics (see also Haney Reference Haney1914). Such a school, if we can speak of it as a school, was far from being a monolith. Some of its adherents did not go beyond passing remarks loosely identifying exchange value with social utility. David Kinley from the University of Illinois (Reference Kinley1904, p. 285), for instance, spoke of the “marginal utility of goods to society” being “indicated by the price level” and reflecting “a value proportionate to changing social values.” Harvard’s Thomas Nixon Carver saw diminishing utility as exerting influence “in the social sense” (Reference Carver1904, p. 18), and held that production and consumption are synchronized in “accordance with social utility” (Reference Carver1904, p. 108). Along the same lines, we find Henry Seager (Reference Seager1905, p. 265; see also Seager Reference Seager1913) affirming that “it is not the marginal utility of each good to each consumer that determines its value, but the marginal utility of each good to society as a whole.”Footnote 7 Fred M. Taylor, in his widely used Principles of Economics (Reference Taylor1918), made the point in distinct normative terms. Without ever citing the name of Clark, Taylor (Reference Taylor1918, p. 526) held that the marginal productivity theory of factor pricing and distribution is necessarily just and efficient. Production, therefore, should be regulated as “to provide for the satisfaction of wants according to the scale which is in effect embodied in the system of distribution.” Only in this way, Taylor concluded, would the “general demand schedules made up of the different individual demand schedules” come to express the “social importances of wants.”
Other figures revealed a more direct intellectual debt to Clark, drawing explicitly from his works. One of these was Lucius S. Merriam from Cornell, an American pioneer of marginal analysis largely forgotten today.Footnote 8 In a polemical exchange with Edward A. Ross (Reference Ross1892, Reference Ross1893), Merriam (Reference Merriam1893) contested the idea of a constant quantity of goods as the proper standard of “deferred payments.” According to Merriam, if, as in the recent years, the productivity of labor and the variety of goods have both increased, the same quantity of goods no longer represents the same “social value.” He then proceeded to define social value, following Clark almost verbatim:
Value is a social, not an individual phenomenon. And value is measured by the final utility of consumption or the final disutility of production, not of the individual, but of society. The existence of certain ratios of exchange between a given commodity and other commodities simply means that society as a whole obtains as much pleasure from consuming the commodity as it suffers pain from producing enough of other commodities to buy it from its producer at the ruling price. (Reference Merriam1893, p. 104)
At this point Merriam warned that “there is no assurance that final utility and final disutility are equal in the case of individual producers.” His argument again follows Clark’s, almost word for word.
The mere preservation of life is accounted so high and men are psychologically so constructed that the final disutility of production may never exceed the final utility of consumption, even in the case of the inefficient and overworked. But it may often fall short of it. The rich or capable producer who is already well off or whose labor is above the average in productiveness does not need to suffer the discomforts of long hours. Likewise, there exists no relation of equivalence between the final increments of producer’s sacrifice and consumer’s pleasure in the case of those who have a monopoly either natural or artificial, either whole or partial, and whose products in consequence sell disproportionately high. The same thing is true of all who possess any peculiar advantage in production whatever it is. (Reference Merriam1893, p. 104)
Merriam was echoed by Charles A. Tuttle, another figure deeply influenced by Clark.Footnote 9 Like Merriam, Tuttle proposed a “social value” standard to anchor the value of the dollar rather than a tabular standard as suggested by some writers at the time. “The social value of an article in a given market,” Tuttle (Reference Tuttle1901, p. 231) wrote, referring directly to Clark, “expresses quantitatively the wealth-relation of society to that article.” Therefore, the “closest approximation to social value is the prerequisite characteristic of a money material” (Reference Tuttle1901, p. 233). Any precious metal fulfilling a monetary office must possess an “approximately stable quantum of social ‘effective utility,’” and this “shall be the standard unit of value” (Reference Tuttle1901, p. 235). Tuttle recognized that “there is no material that has perfect stability of value”; yet he thought that this should not impede the government from “making the effort to discover that material which fluctuates the least” (Reference Tuttle1901, p. 235).
No one, however, followed Clark more closely than Edwin R. A. Seligman—Clark’s personal friend, admirer, and Columbia colleague.Footnote 10 Seligman presented his views on social value in a paper significantly entitled “Social Elements in the Theory of Value” (Reference Seligman1901; see also Reference Seligman1904), then absorbed and expanded as chapter XIII of his Principles of Economics (Reference Seligman1905). Like Clark, Seligman advanced a social, utility-based, pleasure-pain approach to value measurement. Value, he wrote (Reference Seligman1905, p. 323), is the expression of social marginal utility, in that “it is the aggregate of individual wants that shapes value in actual life.” Seligman’s position, however, becomes problematic when it comes to the aggregation of individual utilities into an index of social value—an aspect that Clark had left substantially unexplored. As Seligman put it in an earlier paper:
I not only measure the relative satisfaction that I can get from apples or nuts, but the quantity of apples I can get for the nuts depends upon the relative estimate put upon them by the rest of society. Some individuals may prize a commodity a little more, some a little less; but its real value is the average estimate, the estimate of what society thinks it is worth. If an apple is worth twice as much as a nut, it is only because the community, after comparing and averaging individual preferences, finds that the desire unsatisfied by the lack of an apple is twice as keen as that unsatisfied by the lack of a nut. (Reference Tuttle1901, pp. 323–324)
The same passage is exactly reproduced in the Principles (p. 180) except for the sentence in italics, and the conception of social value as an average of individual values is completely withdrawn from the discussion. In its stead we find Seligman vaguely assuming the existence of a “marginal individual to whom pleasure and sacrifice are equal” (Seligman Reference Seligman1905, p. 180). As some critics pointed out (Davenport Reference Davenport1906), there is absolutely no guarantee of the actual existence of such an individual in the real world and, even if the individual did exist, Seligman nowhere clarified the relationship between the balance reached by such an alleged representative individual and the equilibrium obtained at the aggregate level.
Equally problematic is Seligman’s discussion of the equality, in equilibrium, between social utility and social cost. For Seligman, again as for Clark, the social disutility of the last hour of work, determined by the subjective calculations of the aggregate labor, fixes the social cost (or sacrifice) of production. “Social sacrifice,” Seligman wrote,
means the sacrifice which members of society as a whole are willing to make. The exertion of one man is estimated in relation to the exertion of another, and the sacrifice of each is compared with the needs of society as a whole. The standard is social, not individual. It is far easier to be a street-sweeper than an opera singer. Society is more willing to spare the former than the latter; for, to replace the one, society must give up more of its energy than to replace the other. Consequently, although the street-sweeper may work the harder the sacrifice or cost to society is less than in the case of the opera singer. The opera singer saves society more effort. When one commodity is exchanged for another, or when both cost the same, it means that the additional sacrifice imposed upon society to replace either of them is the same. The marginal social cost is identical. (Reference Tuttle1901, p. 332)
The contention that to replace an opera singer, “society must give up more of its energy” than to replace a street-sweeper is at least controversial. This in fact would be true in the case that the opera singer owed her/his position only to superior training and not also (as it seems plausible) to some inborn characteristics that de facto may be assimilated to a natural monopoly. This, we have seen, was the interpretation given by Clark. Seligman must have realized the problem, and in the Principles (Reference Seligman1905, p. 193), he reproduced the same passage but this time substituting the opera singer with a “skilled factory hand.”
IV. THE SOCIAL VALUE DEBATE
The publication of Clark’s Distribution of Wealth and (especially) of Seligman’s Principles of Economics led to a series of reactions, mostly critical, from some of the leading figures of the period. Among those who joined the discussion we find open critics of the “organicistic” approach, like Davenport and Schumpeter, who defended methodological individualism; friendly critics, such as Anderson, who suggested an alternative approach to social value; and constructive critics, like John Maurice Clark and Watkins, who accepted the social dimension of value emphasizing the role of the externalities.
The Critics
Quite interestingly, the first direct attacks on the notion of social value came from Davenport and Frank W. Taussig as commentaries of Seligman’s Principles—although one gets the clear impression that Clark was their main target. Davenport (Reference Davenport1906, p. 159), who unequivocally titled his review article “A New Text; Seligman: ‘Social Value,’” cautioned his readers that “whatever is said here is … general in its reference,” rather than “primarily an examination of Professor Seligman’s doctrine as such.” Taussig, who phrased his criticism in decidedly harsher terms, was far more explicit. In the discussion of value and distribution, he wrote (Reference Seligman and Taussig1906, p. 627) that Seligman follows so closely the “lead of his colleague, Professor J. B. Clark,” that “to express an opinion on this part of his book is … to express it on that scholar’s contributions to economic theory.”Footnote 11
Davenport and Taussig found the very notion of social value to be without content, methodologically and/or theoretically. Davenport admitted that through the possibility of social exchange, “a commodity may come to have a greater indirect utility to its possessor than it has direct utility to him,” but in his view this is a far cry from asserting that such a commodity has a “direct marginal utility to society.” It is a “regrettable fact,” Davenport continued, that throughout Seligman’s volume “the transition over from the individual psychology to an alleged social psychology is nowhere seriously attempted.” Taussig dismissed the whole distinction between individual and social utility along similar lines:
Much is said also as to the distinction between individual and social utility. The gist of this matter is very simple. In a large and complicated society the demands of a great many persons of different tastes and different purchasing power affect the price at which an article will be disposed of. This, and nothing more, is meant when it is said that “society” demands a thing, or “society” sets the price at which a thing will be sold. I have never been able to see that Professor Clark’s elaborated theories do more than set forth this familiar fact. (Reference Taussig1906, p. 629)
Davenport returned to the topic a few years later in his Value and Distribution (Reference Davenport1908), where he devoted a whole chapter to Clark. This time he pointed directly to the organicist conception of society implied in the idea of social value. Even if “no proof has yet been offered … that society is not an organism,” he stated (Reference Davenport1908, p. 467), it is equally true that “no proof has yet been anywhere adduced that it is.” Even the sociologists, once the champions of organicism, “have long since mostly abandoned the doctrine.” But, more importantly, the notion of society as an organism is not just methodologically unsound, it is also irrelevant with respect to the core of Clark’s theoretical contribution. To Davenport (Reference Davenport1908, p. 469), “Clark has overestimated or wrongly located the difficulties of his problem.” Even without any reference to the alleged “psychology of the social organism,” in fact, the whole apparatus of the productivity theory of distribution can still retain its validity if only the “market values of products and of factors may be made homogeneous on some sort of workable basis.” Such a homogeneity, Davenport concluded (Reference Davenport1908, p. 469), is to be found “under the simple, everyday, and commonplace guise of market price”—a uniformity expressed in terms of the “money fact” and “expressive of undifferentiated purchasing power.”Footnote 12
Schumpeter was the only European figure who directly took part in the debate, and he did so with a paper, “On the Concept of Social Value,” published in 1909 in Harvard’s Quarterly Journal of Economics. Schumpeter opened with a straightforward exposition of “methodological individualism” as applied to pure theory or pure economics.Footnote 13 “Almost every modern writer,” he affirmed (Reference Schumpeter1909, p. 214), “starts with wants and their satisfaction,” and “takes utility … as the basis of his analysis.” This approach “unavoidably implies considering individuals as independent units or agencies,” for “only individuals can feel wants.” Standard assumptions about individual wants and the varying of their intensity with their satisfaction, then, provide the foundations for utility curves, and these curves, together with the relative scarcity of each available commodity, determine individual marginal utilities schedules. All this, however, would not be enough for a “theory of utility.” Schumpeter was adamant in stressing that to fix values, it is not sufficient to know the quantity of goods existing in the aggregate. It is necessary to know their distribution among individuals—“marginal utilities do not depend on what society as such has, but on what individual members have.”
This procedure, Schumpeter intimated, cannot be applied to society as a whole. He illustrated his point:
Society as such, having no brain or nerves in a physical sense, cannot feel wants and has not, therefore, utility curves like those of individuals. Again, the stock of commodities existing in a country is at the disposal, not of society, but of individuals; and individuals do not meet to find out what the wants of the community are. They severally apply their means to the satisfaction of their own wants. Theory does not suggest that these wants are necessarily of an exclusively egotistical character. We want many things not for ourselves, but for others; and some of them, like battleships, we want for the interests of the community only. Even such altruistic or social wants, however, are felt and taken account of by individuals or their agents, and not by society as such. For theory it is irrelevant why people demand certain goods: the only important point is that all things are demanded, produced, and paid for because individuals want them. Every demand on the market is therefore an individualistic one, altho, from another point of view, it often is an altruistic or a social one. (Reference Schumpeter1909, pp. 215–216)
Schumpeter was willing to concede that there are ways in which society can be said to “set values on things” (Reference Schumpeter1909, p. 216). The influence of fashion and imitation, for instance, clearly demonstrates the effects of social interaction in modifying individual utility curves. But such an “obvious” interdependence between socialized beings is quite distant from the idea of “all individuals acting as a community consciously and jointly.” This could be realized only in a communistic society. Only in that case, in fact, “society would have direct control of all means of production,” and would act as a Crusoe producing goods up to the point where the marginal utility of the goods equals the marginal disutility of the labor expended. “Production and distribution would,” accordingly, “be ruled by social value and social marginal utilities.” In some cases, Schumpeter specified, economists do combine all the individual demand and supply schedules so to obtain “general utility curves” for a competitive economy. These, however, are by no means the same as the utility curves of a communistic society: “[t]hey resemble them and have about the same shape; but they refer to individual wants and to a given distribution of wealth” (Reference Schumpeter1909, p. 216).
Ultimately, Schumpeter saw the very notion of social value as inherently normative. In his view, its “practical importance” lies in the assertion, which he found implicit in Clark’s and Seligman’s account, that competition and private ownership of productive factors tend to bring about a distributive process similar to one regulated by a “benevolent and intelligent ruler” (Reference Schumpeter1909, p. 222). But in their normative impetus, Schumpeter asserted, the social value theorists are both inconsistent and fallacious. The inconsistency is to be found in the fact that they present an economic system that, on the one hand, admits private ownership of all factors of production but, on the other, regulates and distributes the national product according to a vaguely defined principle of collective efficiency. This is an unrealistic picture of both competitive and communistic societies.Footnote 14 The crucial fallacy, instead, is to be found in the idea that marginal utility, irrespective of total utility, provides an adequate index of social importance. What any individual receives is the product of the “social utility” of the factor he has to offer with the quantity of it actually sold, and this, depending on marginal value, has nothing to do with total value. Everyone, thus, gets less than his contribution is worth to the community, and “even if the total-utility of what he contributes were very great, he might get very little if the marginal utility of it happened to be small.” For this reason, Schumpeter (Reference Schumpeter1909, p. 232) said, “no conclusions as to the justification of the competitive regime can be drawn from this theory.”
A Friendly Critic: Benjamin McAlester Anderson
With respect to the lines of criticism discussed above, Anderson’s contribution to the debate was of a quite different tenor. A graduate student of Seligman’s at Columbia and a self-professed admirer of Clark (Dorfman Reference Dorfman1949), Anderson approached the social value theorists with contrasting feelings. In his doctoral dissertation, published under the title Social Value: A Study of Economic Theory Critical and Constructive (Reference Anderson1911, p. 9), he found the whole idea of social marginal utility, together with the associated notion of social marginal cost, to be “unsatisfactory” and “unilluminating” at the same time. On the one hand, he argued, if social utility is to be considered the main determinant of value, it cannot be an arithmetical average or a mere addition of individual marginal utilities, let alone the “marginal utility of a good to some particular individual who stands out as the marginal individual in society” (Reference Anderson1911, p. 9). On the other hand, Anderson did not hesitate to affirm that the notion of social value is a “necessity for the validation of economic analysis,” and a conception that “present-day psychological and sociological theory abundantly warrant us in accepting.”
Against this background, Anderson turned his attention to Clark, whom he saw as the “leading exponent of the conception that value is a social fact” (Reference Anderson1911, p. 50). Anderson praised Clark’s attempt to reduce all productive services and their products to one comparable social basis, but he was left unconvinced by the labor cost theory of value he had proposed. If one accepts Clark’s contention that the final social utility of a certain composite of goods is equal to the social sacrifice incurred in producing it, Anderson speculated, and granted that for each individual the marginal disutility of labor in making his part of this composite is equal to the final utility obtained from consuming his part of the composite (not the same part he had produced), there is one aspect that still remains unexplained: “the connection of the marginal utility and the value of the particular goods that make up the composite, with social labor” (Reference Anderson1911, p. 52). Specifically, what bothered Anderson was Clark’s treatment of monopolized products. As we have seen above, Clark had recognized that wherever there is a privileged position giving to some individuals the “advantage of a monopoly,” there is no necessary connection between the cost of production of a certain article to an individual and its market value of that article fixed by its cost to society. If this is the case, Anderson objected,
suppose that a monopoly element enters, at some stage or other, into every product of the market, and in varying degrees in each, either in the form of control of raw material, or special native mental or physical aptitude, or patent right, or any other of the innumerable forms that monopoly takes. Can these monopoly products then call forth a definite amount of social labor? Or can they merely call out a definite amount of value? … If an ultimate cause of value is being sought, it is certain that labor does not supply it for the monopolized goods. (Reference Anderson1911, p. 53)
But, to Anderson’s eyes, this was not all. We have seen that Clark had also acknowledged that differences in wealth among individual producers can lead to a divergence between cost of production and social value for certain commodities. The problem, Anderson noted, is that Clark did not realize the circular reasoning into which he had fallen. If individual differences in wealth, which is itself a stock of value, are to affect the operating of personal sacrifice, it is then “difficult to see how that pervasive element can in any way be an ultimate explanation or measure of value” (Anderson Reference Anderson1911, pp. 53–54).Footnote 15 Similar problems emerge when differences in efficiency among workers are allowed to influence value. “The effect that gauges the efficiency of a worker is the value of what he creates,” wrote Clark (Reference Clark1899, p. 394). The circularity here is again evident, Anderson insisted, and lies in the fact that the price of the good depends on the efficiency of the labor that produces it, and the efficiency of the labor depends on the value (of which price is the expression) of the good produced.
What is relevant to point out, however, is that despite these strictures Anderson remained a friendly critic of the social value school. In the end, he conceded, all his remarks in no way invalidate Clark’s main contention that value is “the work of the social organism” and that the value of a good measures its “importance to the social organism” (Anderson Reference Anderson1911, p. 54). The difficulty with Clark’s theory, Anderson stated, is that
it has not been an analysis of an organic process, but rather, a mathematical study of sums. The individuals have been treated, not as interacting in their mental processes, but as isolated atoms, each of whom has a definite individual quantum of pain or pleasure, and the social unit of pain or pleasure has been treated as simply a sum of these. But it is characteristic of an organism that the simple rules of arithmetic do not hold precisely in its activity. The whole is more than the sum of its parts, and something different from that sum. (Reference Anderson1911, p. 55)
Anderson’s dissatisfaction with Clark led him to propose a different approach to social value, a sort of (proto) social behaviorism inspired by Charles H. Cooley’s sociology and John Dewey’s functional psychology.Footnote 16 In the “value experience,” Anderson wrote (Reference Anderson1911, pp. 103–104), we can distinguish a structural side, which includes the three distinct phases of feeling, desire, and reality-feeling, and a functional side. From the functional side, “the test of a value is its influence upon activity.” With her emphasis on individual sensations, the economist has mostly focused on the former, but the latter side is most important from the standpoint of social theory. The “common denominator” of values, Anderson reiterated (Reference Anderson1911, pp. 104–105), “is not feeling, nor desire, but power in motivation, and the expression of this is of course the activity itself.” Value for Anderson is a social construction; it is primarily a property of social behavior and only secondarily of socially constructed and shared “objects.” Anderson’s view of language as communication in cooperative and coordinated partnership in the construction of all values clearly reveals his intellectual debt to Dewey. The “functional significance” of values, he stated, comes in when they “are to be shared consciously by different individuals, when they are to be communicated and discussed, that is to say, are to become objects of a group consciousness” (Reference Anderson1911, p. 105). Only in this way, i.e., through this working of the “social mind,” can values enforce the entire conduct of “every man and woman impelling them in paths which no individual man foresaw or purposed” (Reference Anderson1911, pp. 115–116).
A detailed analysis of Anderson’s social psychology would go beyond the limits of this paper. Two aspects, however, concern us here. First, Anderson (Reference Anderson1911, p. 188) saw his social psychology as complementary, and not as a substitute, to the “work of the Austrian economists,” for which he had “the highest admiration.” According to Anderson, in fact, “power in motivation,” though different from utility, behaves in the same diminishing fashion. In the field of values, he stated (Reference Anderson1911, p. 108), “the Austrian analysis gives us an answer, which … may be accepted with little modification.” Second, Anderson’s contribution gave new vigor to the social value debate. Outside the economists’ camp Anderson’s book was well received, and this may be explained by the non-analytical and overtly interdisciplinary approach he had proposed.Footnote 17 Reactions from some of the leading economists of the time, instead, were far more cautious—and this in spite of the fact that Anderson’s book was awarded the prestigious Hart, Schaffner, and Marx Prize for economic research.Footnote 18 For Harvard’s Thomas Nixon Carver (Reference Carver1912, p. 639), a philosophy of value heavily relying for its validity on some “peculiar notions of the social psychologist” may be interesting but is not necessarily a sign of theoretical advance. Alvin S. Johnson (Reference Johnson1912, p. 322), a devoted student of Clark’s, found Anderson’s “purely logical” doctrine of social value to be a “mere redefining of terms, a restatement of a problem.” Among the several other commentaries that followed the publication of Anderson’s volume (see Cooley Reference Cooley1913; Haney Reference Haney1913, Reference Haney1914; Bilgram Reference Bilgram1915; Perry Reference Perry1916), John Maurice Clark’s was the one that imparted a decisive twist to the social value debate.
From Social Value to Externalities
The younger Clark (J. M. Clark) first presented his general views in a brief paper entitled “The Concept of Value,” which appeared in 1915 in Quarterly Journal of Economics (Reference Clark1915a). In opposition to the notion that value is an absolute quantity of “something homogeneous,” Clark (Reference Clark1915a, p. 665) maintains here that value can be conceptualized only in relative terms, or in terms of the ratio of quantities exchanged. Classical economists, and their contemporary followers, like Carver (Reference Carver1907), erred when they neglected this important point. Similarly erroneous, Clark continued without mentioning the name of Anderson, is the view of those who assert that value may be considered as a quality like strength and called “social marginal utility” or “power in motivation.” When value is measured, he wrote (Reference Clark1915a, p. 673), “no one thinks of using a psychological laboratory for the purpose,” and what is actually estimated “is not utility in general but the power utility has to produce one kind of effect.” This effect, Clark insisted, is known only through prices—i.e., exchange relations—and has nothing to do with social value. Somehow called into question, Anderson (Reference Anderson1915) published a reply defending his views, which was, in turn, followed by a lengthy rejoinder by Clark (Reference Clark1915b), where he further elaborated on his criticism of Anderson and the social value school.
Clark made explicit that he was not opposed to the adoption of a social value perspective in economic analysis. A workable concept of social value is essential, he held (Reference Clark1915b, p. 710), “if we are to have any standard of judgment on economic reforms which are continually overruling the valuations of the market.” The problem was that in the current usage by its proponents, the term seems to convey two distinct meanings—on the one hand, a “social value which varies independently of prices and rests on the facts of social psychology”; on the other hand, an index of social importance deduced from the “idea that price is a ratio between two values” (Reference Clark1915b, p. 711). To Clark’s eyes, Anderson “appears to blend these two concepts,” and it was Anderson’s “overworking of the deductions drawn from the exchange relation” that he challenged in his reply. A central component of Clark’s own approach to social value was a concept of externality not distant from that advanced by Arthur C. Pigou (Stabile Reference Stabile1995, Reference Stabile1996) and defined as the divergence between the marginal social (or net) value and the marginal private (or market) value of a transaction. In Clark’s own words: “the net economic value of a given service may be considered to include not only the familiar marketable ‘utilities,’ but also (1) potentially exchangeable by-products in the way of service or damage, valued at the price they would presumably command in exchange; (2) unmarketables measured by a standard derived from market price” (Reference Clark1915b, p. 712).
However, Clark immediately added (Reference Clark1915b, p. 712), there is a need to go farther than this if “we are studying such fundamental values as might prevail in a socialist state as well as in our own.” Goods or services (including their externalities) may be valued by other standards than that of market price, especially if these other standards are “effective” in society. Many commodities, in fact, command a price only because their negative social value is lower than the costs involved in the suppression of their use. Whiskey, for instance, has at the same time positive and negative social values and it “motivates prohibitionists to much expenditure of time, effort and money” (Reference Clark1915b, p. 713). Yet such a negative power in motivation has no direct effect on the price of whiskey until it leads to the passing of a prohibition law. And even in that case, the price may increase rather than decrease, so that the banned trade may become more profitable than before. To Clark, even the simplest exchange is not free from “these relationships to values outside the market,” and each single transaction should be seen as “a unit in a great social joint product” (Reference Clark1915b, p. 716). In this connection, he added, the theory of social value is “anti-marginal,” in the sense that “the part takes its price from the value of the whole and not vice versa.” Anticipating many of the themes that were to become central in the institutionalist debates of the interwar years, Clark described the main coordinates of his view of social value:
These, then, are some of the elements which must count in a theory of social value. The theory of inappropriables, the conscious social weighing of men and their desires on scales different from that of free exchange, the insistence that institutional valuations and commodity valuations be distinguished and not both attributed to single commodities, and the readiness on occasion to reverse the marginal method of analysis: all have their place in the interpretation. Such studies can be vitally useful, even if they never attain the precision of a yard-stick. (Reference Clark1915b, p. 715)
In the following decade or so, Clark devoted his attention to the role of “unpaid damages” and “inappropriables,” mostly in connection to the sphere of production and to what he termed the “social overhead cost” of labor (Reference Clark1923). Yet, in his reply to Anderson, Clark (Reference Clark1915b, p. 711) had also made explicit reference to the externalities involved in consumption, speaking of the two “great social impulses” governing consumers’ behavior—the impulse to be like one’s fellows, and the impulse to be “different and distinguished.” Now, Clark argued, these “emulative and invidious” utilities pose a problem in terms of their impact on social value. They are in fact of such a peculiar character that “one man’s gain is another’s loss,” so that “the resultant is a social utility far different from the sum of its individually appropriable parts.”
Interestingly, a more systematic discussion of consumption externalities, along the lines sketched by J. M. Clark in 1915, was advanced the same year by Watkins in his little volume Welfare as an Economic Quantity. Watkins was only indirectly concerned with social value, and in many respects he saw his efforts as an attempt to expand on J. B. Clark’s theory of consumption (Reference Watkins1915, p. 47).Footnote 19 In the Distribution of Wealth Clark had proposed an amendment of the Austrian theory that shortly came to be known as “Clark’s Law” (Seligman Reference Seligman1905). According to J. B. Clark (Reference Clark1899, p. 237), most economic goods or services are not single utilities but “bundles” of different utilities and can render “various unlike kinds of service” at once. This in turn implies that individuals, in choosing between different commodities of the same general kind, are valuing not commodities as a whole but qualitative differences between them. Different buyers, thus, are marginal for different incremental qualitative variations in any given commodity bundle—“a bundle, as a whole, is never a final unit of any one consumer’s wealth; but each element in it is a final utility to some class, and it is that class only whose mental estimate of it fixes its price” (Reference Clark1899, p. 241). Clark’s recognition that goods possess multiple qualities was decisive for his contention that commodities are valued by the social organism rather than by individual purchasers.Footnote 20
Watkins accepted Clark’s “bundle of utilities” approach but went a step farther, distinguishing among different categories of utility according to their effect on social welfare. A certain good, for instance, may be desired chiefly for the advantages it confers in terms of social recognition and prestige. This is the realm of adventitious utility, which Watkins defined as “that part of the utility of a good which is attributed to it on account of the distinction that its consumption or enjoyment is felt to confer on the possessor or consumer by comparison with others apparently or constructively not equally well able to pay” (Reference Watkins1915, p. 141). According to Watkins, and this is what matters to the present discussion, an increase in one’s adventitious utility always imposes a corresponding negative externality on others. As he put it in a salient passage: “adventitious utility … cancels out in the social summation of welfare; that is, the enjoyment of it by one member of society is accompanied by actual or presumptive subtraction from the enjoyment of others. The enjoyment of the one is on the whole proportional to the disagreeable feelings of envy and humiliation excited in others” (Reference Watkins1915, pp. 150–151).Footnote 21 By contrast, certain goods and services provide what Watkins called “multiple utility”:
A good or service collectively enjoyed has the character of utility simultaneously in relation to two or more consumers. The appropriate name for this capacity is multiple utility. Such institutions as the theater, the museum, the public park, and the public library possess multiple utility. Most means of instruction and amusement either have or are capable of having this character. (Reference Watkins1915, p. 163)
Adventitious and multiple utility stand in a dichotomous relationship. The former fosters economic waste and injects envy into the social fabric, contributing to its “atomization,” while the latter promotes social cohesion and increases “society’s total utility” (Watkins Reference Watkins1915, p. 176). This was due to the positive externalities brought about by forms of social interaction inherently connected with this kind of “relational” consumption (Reference Watkins1915, pp. 164–165). To put it differently, whereas adventitious utility denotes a zero-sum game, multiple utility leads to a super-additive game, namely a condition in which one’s utility has a positive impact on the utility of others (Fiorito and Vatiero Reference Fiorito and Vatiero2018). It should be noted, however, that despite his recurrent use of terms such as “social utility” and “social consumer’s rent,” Watkins carefully avoided any organicistic metaphor. His conception of society, he specified (Reference Watkins1915, p. 176), is “aggregate” and not “corporate”—“society as such is not a consumer.”
V. FINAL CONSIDERATIONS
The reconstruction of the social value debate proposed in this paper allows us to shed some new light on the development of American economics during the second half of the so-called Progressive Era. First, it shows how Clark’s mature contribution, culminating in his Distribution of Wealth, was as much divisive as consolidating for early American neoclassicism. Clark was certainly the only one among the first-generation American marginalists who “[came] near developing a distinctive school,” as one contemporary observer remarked (Ise Reference Ise1932, p. 390), and his idea of social value had an impact on several figures of the time. Seligman was the one who followed more closely the Clarkian approach and contributed to its diffusion, but others, such as Tuttle, Kinley, and Merriam, just to name a few, also reveal a distinct social value point of view in their writings. At the same time, this paper has documented how Clark’s (Reference Clark1899, p. 397) crucial contention that “a commodity is measured for value on the basis of the social service that it renders” attracted much criticism. With difference in style and emphasis, Davenport, Taussig, Schumpeter, Anderson, and the younger Clark all pointed out the analytical and methodological difficulties implied in the notion of social value as presented by Clark and Seligman. It may well be the case, as later remarked by Schumpeter himself ([1954] Reference Schumpeter1994, p. 916), that these difficulties were “typical of an early stage of analysis” and that, in the end, “not much came from the discussion on social value.” Yet, the relevance of this whole debate should not be underestimated. Our discussion, in fact, indicates that during the first two decades of the last century, American marginalism remained to a large extent a “contested terrain”—highly pluralistic and without a leading authority comparable with that of Alfred Marshall in England (Rutherford Reference Rutherford2011). In this connection it is significant what Taussig (Reference Taussig1906, p. 622) wrote in his harsh review of Seligman’s Principles. “In the present confused state of economic theory there is no possibility of confining one’s self to the mere exposition of accepted principles,” he stated in the opening paragraph, lamenting that there seemed to be no fundamental principles “on which economists are agreed.” The notion of social value did certainly contribute to such a theoretical fragmentation in American economics.
Second, the reconstruction proposed in this paper of the social value debate tells us something about the transition from the Progressive Era to the interwar years. The idea of society as an organism was part of the cultural milieu of the time and represented a distinguishing trait of progressivism (Leonard Reference Leonard2016). Among other things, the organism metaphor conveyed the progressive idea that the social whole always had primacy over the individuals it subsumed. The very notion of social value fitted perfectly into this schema, at the same time supplying economists with an analytical tool and a normative yardstick. This holistic view of society began to lose much of its original appeal after the opening decade of the last century and with it the notion of social value. In this connection, the younger Clark (Reference Clark and Clark1936, p. 62) spoke for much of the profession when he complained in 1917 about the “overworking of the analogy between social organization and biological organisms.” Many factors contributed to the decline of the society-as-organism discourse, including the crucial changes in American philosophy and psychology that took place in the United States during the second decade of the last century. From the narrative presented here, we can affirm that the social value school found itself caught between two fires. On the one hand, the defense of methodological individualism advanced by some critics proved to be successful among the more theoretically inclined minds of the period. Frank Knight (Reference Knight1921, p. 85n1) summed up much of this line of attack, referring to the social value school as a “particularly glaring instance of the organism fallacy.” Both value and value productivity must be defined with reference to individuals and individual firms, Knight wrote, while the social value theorist “palpably confuses social influences back of individual judgments and preferences with social judgments and preferences in any proper sense.”Footnote 22 Quite unsurprisingly, references to the notion of “social value” disappeared from virtually all the neoclassical literature of the interwar years. On the other hand, this paper has shown how a future institutionalist like J. M. Clark and an eclectic figure like Watkins found the way to distance themselves from the notion of social value as proposed by J. B. Clark and his followers, opting for a view of social welfare based on the notion of positive and negative externalities. In general terms, to the emerging institutionalist school, the social value approach was still anchored to a narrow, market-based (“pecuniary” in their jargon) view of the evaluation process. “Instead of a social value to which market prices tend to conform,” explained Walter Stewart (“Economic Theory—Discussion” 1919, p. 319) at the round table that marked the official launch of institutionalism, “institutional theory regards as the fundamental data for analysis such underlying institutions as ownership, inheritance, the standards of consumption, and the development of technology.”
