A Companion to the History of Economic Thought - Chapter 20 ppt

308 R. E. BACKHOUSE CHAPTER TWENTY The Stabilization of Price Theory, 1920–1955 Roger E. Backhouse 20.1 THE NEOCLASSICAL SYNTHESIS IN HISTORICAL PERSPECTIVE In 1955, Paul Samuelson introduced the term “neoclassical synthesis” into his textbook: We shall again and again meet in later chapters what is called the “neoclassical synthesis.” According to this: if modern economics does its task so well that unemployment and inflation are substantially banished from democratic societies, then its importance will wither away and the traditional economics (whose concern is the wise allocation of fully employed resources) will really come into its own – almost for the first time. (Samuelson, 1955, p. 11) In this passage, Samuelson draws a contrast between the Keynesian theory of income determination, described simply as “modern economics,” and “tradi- tional” microeconomics. Samuelson’s rhetoric concerning the neoclassical synthesis made it clear that it was one of the central points, if not the central point, that he wanted his readers to learn. Readers were told that it was import- ant to “insist” on it (p. 659). Nations were everywhere discovering that it worked (p. 624) and problems of international economics could be solved if the world mastered it (p. 676). He even expressed gratitude that the Russians had not discovered it (p. 733). The aspect of the neoclassical synthesis that has received most attention is the implied relationship between macroeconomics and microeconomics, and Samuelson’s role in propagating the Keynesian revolution. Like Alfred Marshall THE STABILIZATION OF PRICE THEORY, 1920–1955 309 before him, Samuelson clearly wanted to establish the scientific credentials of economics and one way to do this was to emphasize consensus within the pro- fession and continuity with the past. This explains his frequent use of adject- ives such as “traditional” and “classical.” It is in this vein that he claimed that “neoclassical economics,” namely the combination of “whatever is valuable in the older economics” and “modern theories of income determination,” was “accepted in its broad outlines by all but about 5 per cent of extreme left-wing and right- wing writers” (Samuelson, 1955, p. 212). It is, therefore, hardly surprising that he presented microeconomics as though it were uncontroversial and settled. One of the remarkable features of Samuelson’s treatment of price theory is that, in complete contrast to the way he treated the theory of income determina- tion, he played down the modernity of the theory that he was describing. He offered only minor hints that price theory had changed. For example, he wrote of the “kernel of truth in the older economics” having been separated from the “chaff of misleading applications” and about preserving “whatever is valuable” in the older economics (Samuelson, 1955, pp. 11, 212). The picture is one of older theories having been polished and refined, without any indication that this refinement might have involved a radical transformation of the subject. It would have been much harder to justify such a claim in or around 1920. Between the 1920s and the 1950s, price theory had stabilized in the sense that a consensus had become established. However, this stabilization involved a series of intense controversies over issues that went to the heart of the subject. Fur- thermore, in all cases important economists remained unconvinced about the outcome. It is a picture with close parallels to what happened in macroeconomics (see Backhouse, 1985, chs. 16, 26; or Laidler, 1999). It is misleading to present the process as just a refinement or perfection of earlier theory. To see this, consider the elements on which the price theory of the neoclassical synthesis rested: 1 The organizing principle was competitive equilibrium of demand and supply. Competition was understood as the inability of individual agents to influence market price combined, in the long run, with freedom of entry and exit. 2 Demand was determined by consumers, who chose their most preferred bundle of goods, subject to their budget constraint. Preferences were assumed to exhibit nonsatiation, transitivity, and convexity. If a utility function was used, it was understood as a purely ordinal representation of preferences. 3 Firms were assumed to behave as if they maximized profit subject to a production function and the prices of factors and products. 4 Where problems clearly involved noncompetitive behavior, imperfections of competition were understood as violations of the price-taking assump- tion: agents were able to influence prices in the markets in which they were trading. 5 Welfare functions should be individualistic, containing as arguments the utilities (in the sense described above) of individual consumers. Given the absence of scientific grounds for comparing different individuals’ utilities, the main welfare criterion was Pareto-efficiency or Pareto-optimality (the two terms being used interchangeably). 310 R. E. BACKHOUSE These assumptions cover several types of theory: “highbrow” theories used in the literature on proofs of the existence, uniqueness, and stability of general competitive equilibrium, as well as “lowbrow” theories used in introductory textbooks. They encompass the Chicago, MIT, and Arrow–Debreu versions of neoclassical economics (for example, as distinguished in Hands and Mirowski, 1998). Samuelson’s textbook brilliantly integrated this diversity of theories, along with Keynesian macroeconomics, into an apparently seamless whole. 20.2 CONSUMERS AND DEMAND In the 1920s, the idea that consumer behavior should be viewed as involving the maximization of utility subject to a budget constraint, with demands being determined by the relevant first-order conditions, was widely accepted. There was, however, no consensus on the way this should be interpreted. The “purest” approach was represented by Pareto. He used terms such as “pleasure,” “utility,” and his preferred term “ophelimity” but he made it clear that, though useful for exposition, they were not necessary to construct the theory (Pareto, 1971 [1906], p. 112). The theory rested on “the determination of the quantities of goods which constitute combinations between which the individual is indifferent” (p. 113). No metaphysical entity was required. He accepted Fisher’s (1892) demonstration that sets of indifference curves could not, in general, be integrated to obtain utility functions. This was the mathematical counterpart of his view that terms such as “utility” were inessential to the theory. In an article that was not widely cited until the 1930s, Slutsky claimed that the merit of Pareto’s theory was “its purely formal character and its complete independence of all psychological and philosophical hypotheses” (1953 [1915], p. 28). He emphasized that, on Pareto’s definition, there was “[no] point of contact whatsoever between economics and psychology” (ibid., emphasis in original). The theory was based solely on observations of behavior, or “facts of economic conduct” (1953 [1915], p. 54). He used it to derive the relationships between prices, quantities, and income. Of particular importance was what has come to be known as the Slutsky symmetry condition: that, if consumers are maximizing utility subject to a budget constraint, The residual variability of the j-th good in the case of a compensated variation of the price p i is equal to the residual variability of the i-th good in the case of a com- pensated variation of the price p j . (Slutsky, 1953 [1915], p. 43) However, having built up his theory on this basis, Slutsky speculated on the implications of going beyond this, to assume that second derivatives of the utility function were negative. Such an assumption would require “internal evidence” relating to the “consciousness of economic conduct,” or the “psychological aspect of utility” (1953 [1915], pp. 54–6). Laws derived on this basis would need to be tested experimentally. Although Slutsky considered this valuable, it would be valuable for the psychic and moral sciences rather than for the economic. In THE STABILIZATION OF PRICE THEORY, 1920–1955 311 short, economics did not require investigation of the psychological aspects of utility (for further discussion of utility theory up to and including Slutsky, see Stigler, 1950). Other economists attached much more importance to the psychological aspect of utility. Wicksell (1934 [1918], p. 221), in a critique of Cassel, made it clear that the psychological interpretation of utility was important in providing an explana- tion of why things were objects of demand. The economist who dispensed with psychology was like a stockbroker who dealt in railway stocks without knowing what a railway was. This was also the attitude of the Cambridge school. Al- though they had moved away from nineteenth-century utilitarianism, which viewed behavior in terms of pleasure and pain, they viewed the utility of goods as indicating their ability to satisfy wants (the relationship of their view to psychological hedonism is clearly discussed in Pigou, 1903, pp. 67–8). Wants provided the “motor-force” or “incentive to action” (Marshall, 1920 [1890], p. 13). Psychology, and hence utility, described the cause of behavior. This was stated even more explicitly by Pigou, who argued that “satisfactions and dissatisfactions” affected behavior through “desires and aversions” (Pigou, 1932 [1920], p. 23). Utility measures the intensity of desire. A further reason why the Cambridge school attached importance to the psy- chological interpretation of utility was that they believed this allowed them to draw conclusions about welfare. Marshall’s welfare economics was based on the concept of consumers’ surplus, and to interpret this as a measure of welfare he needed to talk about the marginal utility of money. He appeared to have no hesitation in saying that a shilling might yield greater satisfaction to the same person at different times, or greater satisfaction to a poor person than to a rich one (Marshall, 1920 [1890], p. 15). For Pigou, the concept of welfare, and hence the psychological interpretation of utility, was even more important. Economics was about welfare, which had to be measured. Psychology provided the link between behavior and welfare. At the other end of the spectrum were economists who wanted to dispense with utility altogether. Cassel (1932 [1923], p. 49) tried to replace the theory of value with a theory of pricing. The pure theory of marginal utility was superflu- ous and failed to extend our knowledge of actual processes (1932 [1923], p. 81). This was also the view of Moore (1914, pp. 66–7), who argued for statistical demand curves derived from regression analysis, and Mitchell (1925, pp. 4–5), who believed that economists would lose interest in nonquantitative models of behavior. In rejecting the idea of consumers who have ready-made scales of bid and offer prices, Mitchell will have been influenced by Veblen’s lambasting of the notion that tastes could be taken as exogenous. These examples illustrate the variety of interpretations of utility theory that existed around 1920. Other theories could be added to the list, such as Fetter’s use of instinct-impulse psychology or Wieser’s “Austrian” theory of value (see Mitchell, 1969). Marshall’s was perhaps the most widely held approach, but there was no consensus on how consumers’ behavior should be analyzed. This situation began to change in the 1920s and 1930s as economists increas- ingly moved away from psychological interpretations of utility. At the London 312 R. E. BACKHOUSE School of Economics, Robbins (1932) attacked Marshallian theory, arguing that there was no scientific basis for the value judgments on which measurable utility rested. His younger colleagues, Hicks and Allen (1934), took up this program, taking Pareto as their starting point (see Hicks, 1983, ch. 31). Pareto had shown that it was possible to go from observable conduct to a scale of preferences, but it was not possible to go from there to a particular utility function. Hicks and Allen (1934, p. 26) argued that this meant that the subjective theory of value was transformed into “a general logic of choice.” Pareto had realized this, but had not carried through the project of completely reworking the theory of value to take account of it. Hicks and Allen saw this as their task. They reformulated the theory of the consumer in terms of marginal rates of substitution and used the theory to analyze the relationship between preferences, income, and demand. After publishing this work, they discovered Slutsky’s article, finding that it anticipated some of their results. Allen (1936, p. 127) argued that, though Slutsky’s results were correct, his method (of using a utility function) could lead to mis- leading results “in the hands of a less sure mathematician.” Robbins (1932, p. 99) was skeptical about whether it would ever be possible to establish quantitative, statistical laws of demand and supply. In contrast, Henry Schultz set up a statistical laboratory in Chicago to help establish such laws. He was a student of Henry Moore, who was known for his discovery of a positively sloped demand curve for pig iron (Moore, 1914) which, he claimed, represented a new type of dynamic demand curve, relevant to the cycle (see Mirowski, 1990; Morgan, 1990, pp. 26–34, 143–5). Schultz carried on this work, but whereas Moore had rejected Marshallian theory, Schultz (1925, 1927b) used Marshall’s elasticities of demand and supply as the framework for his statistical analysis. However, although he used supply and demand analysis, Schultz came to reject the psychology on which the Cambridge economists considered it to rest. He objected to the notion that a change in one price should be seen as “causing” a change in the price of another commodity (Schultz, 1927a, p, 702): by ruling out interdependence of prices, it made a “realistic” treatment of some problems impossible. He referred to the “epoch-making discoveries of Walras and Pareto” (Schultz, 1927a, p. 703). These, however, were only static and needed to be made dynamic. The recent work of Moore (1925, 1926) and Roos (1927) had done this. The following year, Schultz made the methodology underlying his rejection of psychological theories clear: The formulas of geometry enable us to compute distances, areas and volumes; the formulas of economics have no such heuristic properties. This is due primarily to the fact that most economic laws or principles are expressed in terms of the prop- erties of things or persons rather than in terms of operations. Thus we define “utility” as the property which a thing has to satisfy a want, and we talk of “keeping other things constant” without specifying the mental or physical operations by which this may be done. (Schultz, 1928, p. 647) He went on to quote Percy Bridgman’s view that “If a specific question has meaning, it must be possible to find operations by which an answer may be THE STABILIZATION OF PRICE THEORY, 1920–1955 313 given to it” (Schultz, 1928, pp. 647–8). Economics could become an experimental science only by confining itself to concepts that could be made operational. Utility was superfluous. Schultz also pointed out that demands could come from producers as well as consumers, and in this case there was no reason why demand curves should slope downwards. That was why, with the exception of Moore, economists had confined their statistical studies to the demand for foodstuffs. Given these interests, it is hardly surprising that Schultz took an interest in an article, published in the Journal of Political Economy under his editorship, by Hotelling (1932). Hotelling explored three models of demand. One of them was a model of entrepreneurs’ demands, derived from maximization of net profit, u(p 1 ,p 2 , p n ) − p 1 q 1 − p 2 q 2 p n q n , where u denoted sales revenue and the p’s and q’s the prices and quantities of n goods. He showed that the condition under which the function u(.) could be obtained from observations of prices and quantities was that the cross-partial derivatives relating any pair of goods must be equal: that the effect of the price of good i on the quantity of good j must equal the effect of the price of good j on the quantity of good i. This can be called the Hotelling symmetry condition. Although it related to entrepreneurial demands, it would apply to consumers’ demands if the marginal utility of income were constant, as might be the case when aggregating over individuals. This condition could be tested and, even before Hotelling’s article had been published, Schultz used his statistical laboratory to calculate these derivatives for a variety of agricultural goods. The symmetry condition was not satisfied. In the same article, Hotelling related this model of demand to the traditional one, which imposed the condition that total spending be constant, limited by income. For the latter, it was impossible to derive the function u(.) from observed data. Hands and Mirowski (1997) and Mirowski and Hands (1998) argue that Hotelling intended that his model of entrepreneurial demand might serve as a model of consumers’ demand. In contrast, Hurwicz (1997) argues that Hotelling realized it was appropriate only for entrepreneurial demand and did not apply to the individual consumer. Hotelling’s third model, which offered a statistical explanation for the slope and shape of the demand function, was neglected. This marked the beginning of a period of cooperation between Schultz and Hotelling, during which they tackled the problem of demand, Hotelling empha- sizing theory and Schultz empirical work. During this period, they discovered the work of Slutsky, Hicks, and Allen, and the condition that, for an income- constrained consumer, the cross-partial derivatives of the compensated demand function should be equal. The culmination of this line of research (which ended with Schultz’s death in a car accident in 1938) was Schultz’s Theory and Meas- urement of Demand (1938). In the final section of this book, Schultz tested the hypothesis of rational consumer behavior by testing both the Hotelling and Slutsky symmetry conditions for a range of agricultural products. The results were not encouraging – conflicting evidence meant that the demand relations between pork and mutton could not be determined from the data. Schultz tried to find a statistical explanation that saved the theory, but the project of establishing quantitative demand relations appeared unsuccessful. 314 R. E. BACKHOUSE From here, a variety of routes were followed. Knight (1944) defended the use of demand curves, but was very critical of attempts to test the underlying theory. In a dynamic economy, where incomes and psychological factors were changing continually, testing such theories was doomed to failure. Friedman, who had worked with Schultz, agreed with Knight that economists should not seek to go behind the demand curve. He argued that empirical work should move away from indifference curves toward analyzing direct relations between demand and factors such as income, wealth, prices, and personal characteristics. He described the theory underlying such work as a “Marshallian” demand curve (Friedman, 1949; cf., Friedman and Wallis, 1942). Interdependence of demands was played down. During the 1940s, Schultz’s approach to demand was con- tinued at Chicago, at the Cowles Commission under Marschak, who emphasized the interdependence of the economic system. From the late 1940s, this work led into both general equilibrium theory and econometric analysis of demand systems. Indifference curve analysis was also rejected by Samuelson (1938), in his theory of revealed preference. Influenced, like Schultz, by operationalism, he sought to reduce consumer theory to what could be deduced from assumptions about observed choices. It turned out, however, that the differences between Samuelson’s revealed preference theory and the Hicks–Allen ordinal utility theory were not significant. Thus, although there were important differences between the ways in which different economists tackled the theory of demand, there was a consensus on the underlying theory. The attempt to use utility theory as a psychological foundation for the theory of demand had been abandoned, and it was generally accepted that utility functions must be ordinal, describing rather than explaining choice. 20.3 MARKET STRUCTURE AND SUPPLY In 1920, the standard theory of supply was that of Marshall’s Principles. The main characteristic of this book is that Marshall used a formal mathematical structure as a framework for constructing an evolutionary theory. Neither the algebra of the appendix nor the graphical analysis of the footnotes matched the complexity of the text. Industries, the basic unit of analysis, comprised changing groups of heterogeneous firms. Markets were competitive (“free” competition) but not per- fectly competitive – firms had their own special markets. Some industries faced increasing and others faced diminishing returns to scale. Even in the long run, firms and markets were not in equilibrium. And yet Marshall analyzed this com- plexity through supply and demand. Supply curves reflected marginal costs faced by the representative firm in each industry, this being the firm that was judged typical of the market being analyzed. During the 1920s, the Marshallian structure was questioned on both sides of the Atlantic, in very different ways. In England, Clapham (1922) questioned the usefulness of Marshall’s classification of industries according to whether they faced increasing or decreasing returns. No one, he pointed out, had filled in these THE STABILIZATION OF PRICE THEORY, 1920–1955 315 “boxes.” He questioned whether it was even possible to fill in the boxes, due to the difficulty of distinguishing between economies of scale and the results of invention. After a brief controversy (Pigou, 1922; Robertson, 1924), this was followed by an even more influential critique by Sraffa (1926). Sraffa questioned whether returns to scale could be anything other than constant in Marshall’s theory. Increasing returns were consistent with competitive analysis only if they arose from economies of scale that were external to the firm but internal to the industry. It was impossible to find convincing examples. Decreasing returns, on the other hand, were inconsistent with partial equilibrium analysis, because they could arise only because of rising factor prices (assumed constant). Two years after that, Robbins (1928) launched an attack on Marshall’s concept of the representative firm. Economics, he contended, no more needed a representative firm than it needed a representative worker or representative piece of land. It was the marginal firm that was relevant. The effect of the “cost controversy,” as the debate arising from the contributions of Clapham and Sraffa came to be known, led to the development of analytic tools to deal with the problem of monopoly. Following Harrod (1930), econom- ists started to use the term “marginal revenue” to describe the first derivative of the revenue function, a concept understood by Marshall (and Cournot before him) but not identified with a specific name. Harrod pointed out that the firm’s demand curve would not be the same as the industry demand curve and that abandoning perfect competition meant abandoning the supply curve. Supply would depend on the elasticity of demand facing the firm as well as on price. These new conceptual tools received their fullest expression in Joan Robinson’s Economics of Imperfect Competition (1933), which virtually created the modern geometry of the theory of the firm, analyzing perfect and imperfect competition, monopoly, monopsony, and even the kinked demand curve (conventionally attributed to Sweezy, 1939). Although Robinson was responsible for the development of a powerful “box of tools,” to use her expression, to emphasize these is to neglect the most significant aspect of the change that was taking place. Marshall’s dynamic, evolutionary theory was being replaced by static equilibrium analysis. The problems with returns to scale identified by Sraffa were problems with the concept of static equilibrium. Robbins’s strictures against the representative firm were valid only if one was analyzing equilibrium. Robinson, in formalizing Marshall’s reasoning, was making assumptions (such as identical firms and reversible cost curves) necessary to construct formal models of equilibrium. As Shove (1933) pointed out, her theory and Marshall’s did not operate on the same terrain. In the United States, a parallel movement took place. However, its character was radically different. Knight’s Risk, Uncertainty and Profit (1921) spelled out in detail the assumptions needed to ensure perfect competition. It was clear that they were satisfied in very few real-world markets. There was awareness of increasing returns and their implications for competition. J. M. Clark (1923) wrote a book on the implications of “overhead costs,” arguing that such costs, which made increasing returns likely, were an important feature of modern business. A competitive market could not function unless firms that took the initiative in 316 R. E. BACKHOUSE changing prices gained a temporary advantage, which meant that variation in prices was necessary for competition to work. This ability to obtain a temporary advantage, therefore, should not be considered as an imperfection (Clark, 1923, pp. 416–20). Such competition was, however, different from “cut-throat” competi- tion, where temporary price cuts were used to drive rivals out of business in an attempt to increase monopoly power. Problems such as “dumping” were widely discussed. There was thus a much greater emphasis on problems of business and real-world markets in the American literature than in the British. This was the background out of which Chamberlin’s Theory of Monopolistic Competition (1933), submitted as a Harvard Ph.D. thesis (supervised by Young) in 1927, emerged. Marshall’s theory of the firm, had been developed in an age and in a country where it was possible to think of a firm as a family business, with fortunes linked to those of its owner–manager. This was a long way from the business conditions as observed in the United States in the 1920s. Chamberlin and his contemporaries were also aware of many business practices that were ruled out by perfect competition and were not taken into account by Marshall. They were, therefore, concerned with bringing Marshall’s theory up to date, not with overthrowing it. Chamberlin did use marginal revenue and marginal cost, and he derived the tangency condition for monopolistic competition but, unlike Robinson, did not regard it as central to his work. There were even occasions when he used average costs and revenues, closer to the way in which businessmen actually thought, in preference to marginal analysis. However, the radical differences between his work and Robinson’s are evident from his contents page. After discussing value under “pure” competition (competition without monopoly elements), his chapter headings refer to oligopoly and duopoly, product differentiation, and selling costs. His argument was that product differentiation and selling costs such as advertising meant that meant that pure competition no longer described the way in which markets worked. The emphasis on oligopoly explains why Chamberlin chose to adopt the device of using two demand curves facing each firm. One corresponded to the case in which rivals kept their prices constant and the other to that in which they matched any price cut made by the firm concerned (the first would be much more elastic than the second). He used this to analyze the dynamics of price-cutting, entry, and exit, as well as equilibrium. The “tangency solution” referred merely to one special case (equilibrium where the number of firms was large). Chamberlin assumed that firms produced products that were different from each other – product differentiation involved more than simply advertising and brand names. The result was that the concept of an industry became ambiguous and was replaced by that of the group. The logical conclusion of this approach, however, was that every firm produced a unique product – a world of competing monopolies. This meant that it proved difficult to extend Chamberlin’s partial equilibrium theory to encompass general equilibrium (see Triffin, 1940). The most important aspect of Chamberlin’s work was that it opened up the field of market structure. He classified markets according to two characteristics: the number of firms and the degree of product differentiation, for each of which THE STABILIZATION OF PRICE THEORY, 1920–1955 317 he analyzed how firms would behave. In the hands of Bain (1942, 1956), one of his students, this led directly into the structure–conduct–performance paradigm. This involved analyzing the structure of an industry (number of firms, degree of product differentiation, durability of product, and so on) and working out how firms would behave. Having determined this, conclusions could be drawn con- cerning the efficiency with which the industry performed. It was a thoroughly empirical approach to problems of market structure and supply. Although he did not emphasize the fact, Chamberlin, like Marshall, never committed himself to the assumption that firms maximized profits. Others, however, launched more direct attacks on profit maximization. At Oxford, Hall and Hitch (1939), followed by Andrews (1949), undertook extensive surveys of how businessmen actually set prices (a methodological analysis of this work and the ensuing controversy is offered by Hausman and Mongin, 1998). On the basis of this evidence they advocated the “full-cost pricing” hypothesis. Businessmen, they argued, did not know what marginal costs were and instead set prices as a mark-up on average variable cost. They denied that firms maximized profits. In the United States, faced with the apparent inconsistency between Keynesian unemployment theory and the standard theory of factor demands, Lester (1946) surveyed manufacturing plants and came to the conclusion that wages were unimportant in influencing firms’ employment decisions. This provoked a strong response from Machlup (1946) and later Friedman (1953). They made it clear that profit maximization was not intended as a theory of how businessmen thought but was a device by which economists could understand the outcomes of busi- ness decisions. Friedman expressed this by saying that firms behaved “as if” they maximized profits: whether or not they understood this was as irrelevant as whether or not a car driver understood the physics of velocity and acceleration. Machlup also tried to discredit the survey evidence on which criticisms of marginalist theories were based. By the 1950s, the arguments of Friedman and Machlup had become widely accepted. Firms were modeled as being in profit-maximizing equilibrium, and short-run monopoly power was taken to depend on the elasticity of demand. There were differences: the “Harvard school” (including Bain) emphasized the range of market structures, while Chicago economists thought perfect competi- tion and monopoly sufficient (perfect competition as the general case, monopoly for use in individual cases). However, such differences were largely swept into the field of industrial organization, leaving relative consensus in the core of price theory. 20.4 COMPETITION, THE PRICE SYSTEM, AND WELFARE Price theory has always been closely linked to welfare. Walras (1954 [1874], p. 255) had reached the conclusion that “free competition” would, subject to two conditions, give “the greatest possible satisfaction of wants.” This idea was developed by Pareto, who argued that free competition would produce “maximum ophelimity,” which he defined in the following way: [...]... work, because they failed to provide any alternative to the market as a means for creating and disseminating knowledge 20. 5 PERFECT COMPETITION AS THE PARADIGM FOR PRICE THEORY Toward the end of the 1930s, perfect competition emerged as the paradigmatic case in price theory In the same way that Hicks’s work with Allen had 320 R E BACKHOUSE developed Pareto’s consumer theory, his Value and Capital (1939)... the mathematical framework on which the microeconomics of the neoclassical synthesis rested, and virtually provided a manual in price theory for many economists in the 1950s, was Samuelson’s Foundations of Economic Analysis (1947) (This was written in the late 1930s, but publication was delayed because of the war.) Samuelson later wrote that Value and Capital had prepared the ground for his more mathematical... place and did not even take seriously the critiques that had been made of the emerging synthesis The history of consumer theory was presented as though nothing at all was lost in the move to purely ordinal theories The standard account of the Socialist Calculation debate presented Hayek as conceding the market socialists’ argument that central planning was theoretically possible The conflation of Chamberlin’s... about the meaning of welfare The first was that questioning the psychological basis for utility raised doubts about the Cambridge approach to welfare economics If utility could be measured (albeit making assumptions about the marginal utility of money), it was clear what welfare economics was concerned with, but once that was abandoned, the meaning of social welfare became much less clear Pareto, in the. .. focusing on the level and distribution of the national dividend (the value of output) Pigou analyzed welfare in terms of the relationship between marginal private and social products rather than consumers’ surplus, but his conception of welfare was essentially the same as Marshall’s The national dividend (national income) measured the group of satisfactions and dissatisfactions that could be measured in... of socialism The background to this was the Bolshevik revolution in Russia and the attempt to establish a socialist state In the period of “war communism,” the Bolshevik government established a system of central planning in which markets and prices were abolished Influenced by Marx, the idea was to eliminate the waste that was inherent in capitalism and to plan production on a rational basis Von Mises... consumers’ behaviour Economica, 5, 61–72 —— 1947: Foundations of Economic Analysis Cambridge, MA: Harvard University Press —— 1955: Economics: An Introductory Analysis, 5th edn New York: McGraw-Hill Schultz, H 1925: The statistical law of demand as illustrated by the demand for sugar Journal of Political Economy, 33, 481–504, 577–631 —— 192 7a: Mathematical economics and the quantitative method Journal of Political... Mirowski, P 1997: Harold Hotelling and the neoclassical dream In R E Backhouse, D M Hausman, U Mäki, and A Salanti (eds.), Economics and Methodology: Crossing Boundaries London: Macmillan, 322–97 —— and —— 1998: A paradox of budgets: the postwar stabilization of American neoclassical demand theory History of Political Economy, 30 (From Interwar Pluralism to Postwar Neoclassicism), 260–92 Harrod, R F 1930:... the advocates of market socialism had to confront the same question as was raised by the new consumer theory: “What is optimized in a social optimum?” Different answers to this question led to the derivation of different optimum conditions These were eventually developed into the so-called “new welfare economics” and the derivation of the conditions for what became known, in the 1950s, as a Pareto optimum... increasing the ophelimity which certain individuals enjoy, and decreasing that which others enjoy, of being agreeable to some and disagreeable to others (Pareto, 1971 [1906], p 261) In contrast, the Cambridge school adopted an aggregative, utilitarian approach Marshall analyzed welfare using the concept of consumers’ surplus He used this to prove the doctrine that “every equilibrium of demand and supply may . (This was written in the late 1930s, but publication was delayed because of the war.) Samuelson later wrote that Value and Capital had prepared the ground for his more mathematical treatment of the subject the standard theory of supply was that of Marshall’s Principles. The main characteristic of this book is that Marshall used a formal mathematical structure as a framework for constructing an evolutionary. evolutionary theory. Neither the algebra of the appendix nor the graphical analysis of the footnotes matched the complexity of the text. Industries, the basic unit of analysis, comprised changing