Price Behaviour and Business Behaviour

Nitzan, Jonathan Working Paper Price Behaviour and Business Behaviour Discussion Papers, Department of Economics, McGill University Provided in Cooperation with: The Bichler & Nitzan Archives Suggested Citation: Nitzan, Jonathan (1990) : Price Behaviour and Business Behaviour, Discussion Papers, Department of Economics, McGill University, The Bichler and Nitzan Archives, Toronto, http://bnarchives.yorku.ca/160/ This Version is available at: https://hdl.handle.net/10419/157851 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by-nc-nd/4.0/ Department of Economics Discussion Paper 1990 PRICE BEHAVIOUR AND BUSINESS BEHAVIOUR Jonathan Nitzan This paper is a draft of work in progress Department of Economics McGill University 855 Sherbrooke St. West Montreal, Quebec H3A 2T7 Contents Abstract 1 Introduction 2 1. The Administered Price Controversy: Beginnings 4 2. Price Flexibility: Fact or Fancy? 8 3. ‘Full-Cost’ Pricing 14 4. The Marginalists’ Counterattack 18 5. The ‘Target’ Rate of Return 24 6. The Anthropology of Business Behaviour: An Interpretation 29 References 31 Abstract The present essay is the second in a series of three papers which examine alternative approaches to inflation. Here we identify some of the principal criticisms expressed against neoclassical views on price behaviour and business behaviour. These challenges grew from the early discovery of ‘administered prices’ by Means and the subsequent findings by Hall and Hitch regarding ‘full-cost’ pricing. The notions that industrial prices were relatively inflexible and that businessmen set those prices by imprecise rules-of-thumb stood in sharp contrast to the pristine simplicity of neoclassical models. Yet these attempts for greater realism seemed to undermine the prospects of constructing a coherent theory for prices. 1 Introduction The economic and political turbulence of the 1930s spawned a number of serious challenges to the hegemony of classical economic doctrines. Of these challenges, only Keynes’ ‘new economics’ was broadly accepted and assimilated into the mainstream of economic thinking. Keynes was successful partly because his policy propositions sought to reform capitalism while preserving its underlying structure. According to Keynes, the malfunctioning of the system stemmed primarily from a chronic lack of synchronization between the ‘propensities’ of consumers and the ‘animal spirits’ of investors. The ultimate problem was rooted not in the structure of capitalism but in fundamental psychological tendencies stemming from human nature itself.1 In this context, his call for government intervention appeared to be fairly conservative: policies were needed not to alter basic power relationships among specific economic groups but merely to overcome an unfortunate gap between abstract saving and investment ‘tendencies.’ As an orthodox student of Marshall, Keynes rarely questioned the basic structural tenets of neoclassical microeconomics and, indeed, he saw no apparent reason to do so. In his opinion, the macroeconomic problem of unemployment arose despite the efficiency of individual markets and, furthermore, the solution for the problem could be achieved by broad policy measures which need not interfere with the functioning of these individual markets.2 The apparent success of early Keynesian policies during and after the Second World War further strengthened the conviction that macroeconomics was quite independent of underlying microeconomic structures. This legacy of Keynesian macroeconomics has proven more powerful then Keynesian theory itself, for while the primacy of Keynesianism has been subsequently challenged by competing schools, macroeconomics as a whole continues to neglect significant aspects of real structures and institutions. The eventual divorce of mainstream macroeconomics from the dynamics of real economic structure was established only in the post-war era, however. During the 1930s, before the apparent triumph of Keynesian policies, economists were seeking answers also in alternative directions. While Keynes was elaborating the psychological reasoning for his General Theory, some of his contemporaries were trying to identify structural causes for the general economic distress. Their subject of 1 See for example Keynes (1936), p. 97 and p. 161. Keynes was of course very much aware of contemporary structures and institutions but these were significant for his General Theory only in so far as they enhanced the tendency for stagnation or instability. The primary cause for these tendencies remained human nature. 2 Being aware of contemporary research, Keynes (1936, pp. 268, 270-1) was careful to stress that his theory abstracted from ‘administered’ or ‘monopoly’ prices. Half a century later, Tobin (1983, p. 299) expressed retroactive regret for this turn of events: ‘It is unfortunate that Keynes, in spite of the Chamberlin-Robinson revolution that was occurring in microeconomics at the same time he was making his macro revolution, chose to challenge orthodoxy on its own microeconomic grounds of competitive markets.’ 2 inquiry concerned basic convictions about ‘price behaviour’ and ‘business behaviour.’ First, the pioneering work by Means (1935a) and by the National Resources Committee (1939) under his direction questioned the monolithic approach to price dynamics. Means suggested that there were in fact two types of prices – those which were relatively flexible and those which were relatively inflexible. More importantly, he argued that this basic difference was rooted in the structure of modern capitalism. Second, the research by Hall and Hitch (1939) challenged accepted assumptions regarding pricing decisions by firms. Their interviews with businessmen indicated that the latter determined their prices by imprecise rules-of-thumb and were quite indifferent to the notion of ‘profit maximization.’ These studies launched a prolonged controversy which has not yet been ‘resolved’ and, because it involves basic methodological issues, perhaps could not be resolved. The purpose of this essay is not to provide a review of this literature but rather to examine key methodological questions arising from it. Given our limited goal and the availability of numerous surveys, we find it appropriate to focus only on some of the important contributions to the debate. Briefly, the link between ‘price behaviour’ and ‘business behaviour’ involves questions of ‘structure’ and economic or business ‘power.’ The neoclassical notion of ‘pricing power’ suggested that a firm could set its own price but, since the firm was assumed to maximize profit, economists could still ‘determine’ what that price would be. The increasing emphasis since the 1930s on the significance of oligopolistic interdependency did not prove to be detrimental for price theory. With sufficiently restrictive assumptions and a complicated mathematical reasoning, economists often succeeded in finding an ‘optimal solution’ for their game theory. The literature following Means and Hall and Hitch undermined this logical simplicity. The existence of relative price inflexibility in markets other than pure competition did not imply that such prices were ‘optimal’ for firms. It only suggested that prices were ‘administered’ and this was precisely the problem. If these were ‘monopoly prices’ in the neoclassical sense they should have been perhaps higher than comparable competitive prices, but there was no reason to expect them to be less flexible. The fact that administered prices were relatively inflexible implied that firms might not have been acting ‘optimally.’ The writings on business behaviour strengthened this doubt when they pointed to substantial ambiguities and considerable discretion in the way firms set their pricing policies. Ironically, by emphasizing the significance of structure for actual pricing, the new empirical literature operated to undermine the methodological basis for price theory itself. It was implied that firms operating in non-perfectly competitive markets had the privilege not only to determine their own prices, but also to set these prices in a rather ‘arbitrary’ manner. Prices were still influenced by ‘objective’ conditions such as cost, demand, the specific structure of the industry, or the intensity of competition. However, since firms enjoyed substantial discretion over their own goals, the ‘mapping’ of these objective conditions into prices was obscured from the economist. 3 Obviously, these issues have considerable bearing on ‘structural’ theories of inflation which we explore in an accompanying paper (Nitzan, 1990). The first and second sections of this paper deal with the early contributions to the administered-price controversy and the criticisms they elicited. The third section explores the early literature on ‘full-cost’ pricing while, in the fourth section, we deal with the marginalists’ counterattacks against that literature. The fifth section examines the aspects concerning the ‘target’ rate of return and the last section offers some observations on the anthropology of business behaviour. 1. The Administered Price Controversy: Beginnings The controversy over the relationship between market structure and price behaviour was triggered in 1935 by the work of Means on Industrial Prices and Their Relative Inflexibility.3 Means raised two basic questions concerning (1) the apparent anomaly in the behaviour of numerous industrial prices, and (2) the causes behind this behaviour. First, he argued that comprehensive price indices, such as the Wholesale Price Index published by the Bureau of Labor Statistics (BLS), were potentially misleading because they failed to distinguish between ‘market prices’ and ‘administered prices.’ Market prices were defined as prices which were ‘made in the market as a result of the interaction of buyers and sellers.’ Administered prices, in contrast, were ‘set by administrative action and held constant for a period of time’ while sales fluctuated with demand at the rigid price (Means, 1935b, p. 401). This distinction was highly significant because market and administered prices ‘behaved’ quite differently in terms of both frequency and amplitude of change. The evidence for such divergent behaviour was based on an analysis of monthly prices for individual commodities included in the BLS Wholesale Price Index. Means classified 747 such items according to the number of times their price changed during the eight-year period between 1926 and 1933 and demonstrated that prices for the majority of items changed either very frequently or very infrequently.4 His inference that these were in fact ‘quite different types of prices’ was further enhanced by illustrating that ‘items which changed frequently in price showed a large drop 3 Although he initiated the debate, Means was not the first to draw attention to price inflexibility and to discuss its potential causes. Stigler and Kindahl (1970, pp. 11-12) cited earlier works by Berlund and by Jones on the rigidity of steel prices during the early 20th century. In 1927, Mills published a comprehensive study on The Behavior of Prices where he found, much like Means’ later discovery, that industrial prices appeared to be either flexible or inflexible in their frequency of change. Another study by Tinter (1935) on price behaviour in Germany, England and the United States, suggested that the frequency of price changes in monopolized industries was appreciably smaller than in competitive ones. 4 Of the 747 item prices, 50 percent changed very infrequently (between 0 and 24 price changes during a period of 96 months), 24 percent changed very frequently (between 80 and 94 times), while 26 percent fell in the intermediate range (between 25 and 79 times over the period). See Means (1935b) Chart I, p. 402. 4 during the depression while those having a low frequency of change tended to drop little in price’ (Means, 1935b, p. 402 and p. 403). Additional evidence published 4 years later by the National Resources Committee under the direction of Means indicated that, as prices recovered between 1933 and 1937, frequency and amplitude of price changes were again positively related. Writing during the depression of the 1930s, Means was primarily attentive to the broad economic implications of this distinction between market and administered prices. Based on the observation that a substantial number of commodities (over one half) had administered prices, he argued that relative price inflexibility became a major disruptive factor in the American economy: We have always relied in the past on the automatic balancing of economic activity through price changes. This is all right where prices are flexible, since a general drop in demand such as occurred in the depression would result in a drop of prices and maintained production. If all prices had been flexible it is doubtful if we would have had a serious depression after the stock crash of 1929. Where prices are rigid, however, a general drop in demand has quite different and most disastrous result. Instead of producing lower prices, the drop in demand produces a drop in sales and in production. Workers have less to spend, thus amplifying the original drop in demand. In this manner, rigid prices can expand an initial small fluctuation of industrial activity into a cataclysmic depression. (1935b, p. 405) Means went on to illustrate that between 1929 and the spring of 1933 there was a marked inverse relationship between the relative drop in prices and the relative decline in production for a sample of ten major industries. When prices fell substantially, like in the case of agricultural commodities or petroleum for example, the decline in output was below 20 percent, while when prices remained stubbornly rigid like in agricultural implements or motor vehicles, production levels dropped by as much as 80 percent! Given the prevalence of administered prices and given the disruptive effect their relative inflexibility had on macroeconomic performance, Means set to address the second issue, namely the cause behind the phenomenon. In his opinion, administered prices emerged primarily (though not exclusively) as a consequence of industrial concentration. Although he expressed this conviction forcefully already in 1935, empirical support for his ‘concentration thesis’ was first provided only in the National Resources Committee monograph published in 1939. There Means examined price changes between 1929 and 1932 for a subset of 37 out of the 282 manufacturing industries included in the Census’ universe, and contrasted them with four-enterprise concentration ratios associated with each individual industry. In selecting the sample, Means sought to eliminate the possible influence factors other than concentration might have had on price changes. Consequently, he excluded 5 industries where (1) products were not relatively homogenous, (2) more than 2/3rds of the product value originated outside of manufacturing, possibly in demand-sensitive industries such as agriculture or some raw materials, (3) products were not produced for national or international markets, so national concentration ratios were misleading, and (4) reasonably reliable price data were not available. Based on a scatter diagram between percent change in price and concentration ratios for the 37 industries, Means concluded that ‘When the depression drop of prices in these industries is compared with the proportion of value of product which in each was produced by the four largest enterprises, a rough relation is apparent between concentration and price insensitivity’ (National Resources Committee, 1939, p. 142). Means repeatedly emphasized that the existence of administered prices was not synonymous with ‘monopoly profits’ and that the process of industrial concentration did not necessarily mean a growing ‘monopolization’: It is . . . abundantly clear that a considerable degree of administrative control is inherent in the narrowing of markets and the willingness of buyers to accept the one-price system of American merchandising. Further administrative controls is implicit if the efficiencies of modern technology are to be realized. Only to the extent that administrative controls arise from collusion between enterprises or through the bringing of production under common control beyond the extent necessary for efficient operation is there an opportunity to reduce the existing degree of administrative control without incurring a cost of decreased efficiency in the use of resources. (National Resources Committee, 1939, p. 145) In this context, economic ‘power’ was perceived not so much as an attribute of broader social relationships but more as a facet of industrial organization. The power to determine prices did not denote the ability of one group to redistribute income from another, but rather the ability of ‘organizations’ to overcome the ‘market.’ Thus, the apparent link between industrial concentration and the administration of prices was rooted primarily in the growing ‘bureaucratization’ of economic activity: . . . the last century has seen a steadily increasing shift from market coordination to administrative coordination. Gradually, as our great corporations have been built up, more and more of the coordination of individual economic action has been brought about administratively. . . . As a result of this shift from market to administration, the area of coordination remaining to the market has been greatly reduced while the increased bargaining power of the big administrative units has induced the counter concentration in the form of cooperative bargaining organization, farm cooperatives, labor unions and to a small extent consumer cooperatives, 6 thus further reducing the number of separate units interacting through the market. (Means, 1935b, p. 407) To a significant extent, then, the adverse consequences of administered prices were the inescapable price we must all pay for technological progress: Thus a considerable degree of administrative control over prices appears to be inherent in the modern economy. Administrative prices and their depression insensitivity seem to be an integral part of the structure of economic activity. With the century-long transition of this country from a predominantly agriculture to a predominantly industrial country, the administration-dominated prices of industry have gradually displaced the market-dominated prices of agriculture as the more characteristic form of prices. (National Resources Committee, 1939, p. 145, emphases added) Although Means pointed to a structural cause underlying the overall economic depression, he recommended not to alter but accommodate that structure. Breaking up large scale enterprises in order to revive price flexibility, he argued, would be immensely wasteful in terms of foregone output and hence he suggested we ‘accept inflexible prices as inherent in our modern economy and build our economic institutions around them in such a manner that inflexible administered prices will cease to be a disruptive factor’ (Means, 1935b, p. 408). To that end, he proposed we adopt expansionary monetary policies, but although his recommendations were macroeconomic in nature, his analytical framework was too controversial to be unanimously accepted as a basis for such policies. The idea that firms administered their prices with a considerable disregard to ‘market signals’ challenged basic theoretical convictions about ‘optimal’ behaviour. Furthermore, Means’ presupposition that such administrative control was largely unrelated to conventional notions of economic ‘power’ and ‘monopoly profits’ was not sufficiently persuasive to defuse public concern. A series of Congressional hearings on administered prices in general and on steel prices in particular began in the late 1940s and extended through the 1950s. Guidelines on wage and price policies were issued by the Council of Economic Advisors in 1962 and were aimed particularly at concentrated industries such as steel, copper and aluminum. The announcement of these Guidelines was followed by heightened confrontation between the subjected industries and the Presidential office and the debate over administered prices received considerable public attention. Means’ interpretation that price inflexibility was predominately a ‘technical’ outgrowth of modern ‘industrial organization’ and his suggestion that public policy could ‘overcome’ the problem of administered prices presupposed that, in itself, the administration of prices served no particular interests. Given the public turmoil over the issue, this was not a very convincing assumption. There was a growing atmosphere of crisis among economists 7 In general, attempts to deny the existence of administered prices or their ‘perverse’ behaviour were more reassuring than convincing.10 The criticisms, particularly when voiced by eminent economists, helped to reduce the anxiety and justify the continued theoretical neglect of the issue, but never succeeded in eradicating it.11 Continued concern with administered prices was also fuelled by a related debate which began at about the same time and which focused on how individual firms actually set their prices. 3. ‘Full-Cost’ Pricing While Means (1935a) initiated a controversy over ‘price behaviour,’ the Oxford Economists’ Research Group, and in particular Hall and Hitch (1939), helped to launch a related debate over ‘business behaviour.’ The conventional theory of the firm, argued Hall and Hitch, stipulated that firms attempted to maximize their profits and that they did so by choosing the output-price combination (or output in the case of perfect competition) such that marginal revenue was equal to marginal cost. This approach yielded theoretical solutions for equilibrium in the case of pure competition, pure monopoly or monopolistic competition, but when the structure was oligopolistic or when monopolistic competition was mixed with oligopoly, the theoretical method broke down. In those latter instances, interdependency between firms meant that individual demand and marginal revenue curves were indeterminate and, hence, could not be used to determine the output-price combination for maximum profits. Economists commonly chose to either ignore the difficulty by considering oligopoly as an ‘exception’ or to bypass it by using some ad-hoc explanations. According to Hall and Hitch, these two solutions were directed facts of life in the steel industry, insist on raising a fictitious price? Did he not know that a simple revision of transaction prices would have served his purpose and also saved him from detection by the B.L.S. (and its henchmen)? In short, given Stigler’s model, Mr. Blough was either a fool or a provocateur, hankering for a joust with the President of the United States. Both these interpretations of Mr. Blough’s behavior tax credulity’ (88the Cong., 1st Sess., Senate Subcommittee on Antitrust and Monopoly, Senate Committee on the Judiciary, Administered Prices: A Compendium on Public Policy, 1963, pp. 6-7). Quoted in Blair (1972, p. 436). 10 For example, Weiss (1977) concluded that over long period of times, the Wholesale Price Index of the BLS, the ‘realization’ price index based on the Census of Manufacturing and the buyers’ index developed by the NBER were highly correlated and conveyed the same general movements. Coutts, Godley and Nordhaus (1978) compared list and transaction prices for non-food manufacturing industries in the U.K. and concluded that ‘There was little evidence found to support the view that the wholesale price indices, being composed of listed quotations, do not accurately measure transaction prices’ (p. 138). 11 Commenting on the title of Blair’s article, ‘Administered Prices: A Phenomenon in Search of a Theory,’ Bailey (1959, p. 460) brushed aside the entire debate as irrelevant and suggested it was in fact ‘A Theory in Search of a Phenomenon.’ Since then the phrase has been often cited as a summary statement on the insignificance of administered prices. 14 toward the wrong problem. In their opinion, the interesting question was not so much how firms should set their price and output in order to maximize profit, but whether firms indeed set prices and output in order to maximize profit. Their concern was not with what firms ought to be doing but rather with what they were actually doing. In an attempt to address this latter question, Hall and Hitch conducted interviews with 38 British entrepreneurs of which 33 were involved in manufacturing, 3 were retailers and 2 were builders. Based on these interviews, they pointed to a wide gap between the presumptions of conventional analysis and the reality of business practices: For the above [neoclassical] analysis it is necessary that entrepreneurs should in fact (a) make some estimate (even if implicitly) of the elasticity and position of their demand curve, and (b) attempt to equate estimated marginal revenue and estimated marginal cost. We tried, with very little success, to get from the entrepreneurs whom we saw, information about elasticity of demand and about the relation between price and marginal cost. Most of our informants were vague about anything so precise as elasticity, and since most of them produce a wide variety of products we did not know how much to rely on illustrative figures of cost. In addition, many, perhaps most, apparently make no effort, even implicitly, to estimate elasticity of demand or marginal (as opposed to average prime) cost; and of those who do, the majority considered the information of little or no relevance to the pricing process save perhaps in very exceptional conditions. (p. 18) It seemed that the theoretical distinction between monopoly or monopolistic competition (where the demand curve facing the firm was assumed to be known) and oligopoly (where the individual demand curves were indeterminate) was not very important for the issue of practical price determination. In reality, businessmen operating in all of these markets simply did not ‘know’ their demand curve and, furthermore, they did not care to ‘discover’ this demand curve even when they could have done so: Only where oligopoly elements are present is the demand curve ‘indeterminate’ in the economist’s sense, but in the other cases it is unknown to the entrepreneur, and this seems to be the essential point. It is true that in the case of monopoly or monopolistic competition the possibility of finding his demand curve by experimenting is open to the entrepreneur; but there are objections to experimentation, and the prospect of a quiet life seems in many cases to have a greater appeal. (pp. 30-1, emphases added) 15 The revelation that firms neglected their demand led to an even more ‘stunning’ conclusion, namely, that firms did not try to maximize their profits as suggested by standard theory: The most striking feature of the answers was the number of firms which apparently do not aim in their pricing policy, at what appeared to us to be the maximization of profits by the equation of marginal revenue and marginal cost. (p. 18, emphasis added) Instead of equating marginal revenue and marginal cost in an attempt to maximize profits, Hall and Hitch (p. 18) suggested that businessmen were ‘thinking in altogether different terms.’ While under certain circumstances, pricing behaviour could be explained by reference to ‘long-term’ profit maximization, in most cases businessmen applied a simple ‘rule-of-thumb’ which Hall and Hitch called ‘full-cost’ pricing: The formula used by the different firms in computing ‘full cost’ differ in detail . . . but the procedure can be not unfairly generalized as follows: prime (or ‘direct’) cost per unit is taken as the base, a percentage addition is made to cover overheads (or ‘oncost’, or ‘indirect’ cost), and a further conventional addition (frequently 10 per cent.) is made for profit. Selling costs commonly and interest on capital rarely are included in overheads; when not so included they are allowed for in the addition to profits. (p. 19) Firms justified their submission to the practical norm of ‘full-cost’ pricing in a variety of different ways. Some argued it was the ‘right price,’ other considered its application as a ‘fair’ practice toward their competitors, while still others noted that experience ‘proved its advisability.’ When asked why they did not charge a price higher than that implied by the ‘full-cost’ principle, most entrepreneurs cited their uncertainty regarding the response of competitors. When requested to explain why they would not charge a price lower than ‘full-cost,’ the businessmen mentioned primarily the fear that competitors would match the lower price, the unresponsiveness of demand and moral objections to selling below costs. As reasons for not changing prices (however fixed), businessmen explained that they wished not to ‘disturb’ the stability of market prices and also that buyers had a ‘conventional’ price in mind and ‘disliked’ price changes. Hall and Hitch (p. 22, emphasis added) felt that ‘All of these reasons militate against changing the price from the conventional level,’ yet they stressed that the ‘full-cost’ principle was insufficient to explain this ‘conventional level’ itself. The simplicity of the ‘full-cost’ principle was potentially deceiving. ‘It would be useful for economic analysis,’ Hall and Hitch (p. 19-20) wrote, ‘if the magnitude of “full cost” in any case could be deduced from the technical conditions of production 16 and the supply prices of the factors,’ but in practice this was impossible for four principal reasons. First, costs varied with the size of the firm but firms were rarely operating at an ‘optimal’ size which economists could presumably determine; instead, their size apparently was the consequence of a ‘historical accident’ which economists found very difficult to ‘predict.’ Second, overhead cost per unit depended on the ‘normal’ output level used as a divisor in the ‘full-cost’ formula, but this benchmark for output was set by arbitrary accounting conventions. Third, selling expenses were included in costs but were often depended on demand. Fourth and most importantly, the way in which entrepreneurs set the magnitude of ‘conventional’ profit, or the reasons why they changed it were not at all clear. For the businessmen, the ‘full-cost’ principle was a straightforward technical matter yet, because of the many ‘arbitrary’ factors involved, the economist could not anticipate the final price with any reasonable accuracy. Surprisingly, then, getting closer to reality did not seem to enhance our understanding of the pricing process. Hall and Hitch questioned the usefulness of neoclassical price theory because its preoccupation with what firms ought to be doing turned this theory into a normative doctrine. They suggested we explain prices by embarking on a positive scientific inquiry into actual pricing decisions made by real businessmen but, unfortunately, substituting the businessman’s practice for the economist’s postulate did not seem to solve the price question. The explanation provided by businessmen appeared ‘arbitrary’ and were hardly more revealing than the theories of neoclassical economists. Instead of adhering to rigid pricing procedures shaped by necessity, entrepreneurs seemed to follow loose ‘conventions’ and ‘norms of conduct’ which did not appear to have a solid ‘objective’ rationale. Hall and Hitch discarded the normative approach embraced by economists, but their own ‘full-cost’ principle seemed to reflect the normative ethic adopted by businessmen. One could have removed the deadlock by seeking psychological explanations for the behaviour of businessmen but this, of course, would have constituted a retreat from the empirical road into the normative twilight. Instead, Hall and Hitch (p. 33) emphasized that ‘There is usually some element in the prices ruling at any time which can only be explained in the light of the history of the industry.’ The rule-of-thumb for pricing included conventions on what constituted ‘normal output,’ conventions on how to estimate costs, conventions on how to react or cooperate with competitors and, most importantly, conventions on how to set ‘adequate’ profit margins. Yet these conventions were shaped by history, not by the erratic fancy of businessmen and only by accounting for the specific historical evolution of these conventions could one hope to shed some light on current prices. The totality of beliefs and conventions prevailing in any one time were encompassed in what Hall and Hitch (p. 28) called the ‘community of outlook’ of businessmen, and it was within this context that ‘full-cost’ pricing reinforced a tendency toward price stability: 17 We cannot say precisely what this price will be, for reasons already explained; if it is set anywhere over a fairly wide range it will have a tendency to stay there. The nearest that we can get to an exact statement is that the price ruling where these conditions obtain is likely to approximate to the full cost of the representative firm; and that this price is reached directly through the community of outlook of business men, rather than indirectly through each firm working at what its most profitable output would be if competitors’ reactions are neglected, and if the play of competition then varied the number of firms. (pp. 27-8, emphasis added) In a similar way, price instability was not a direct consequence of changes in underlying conditions but was rather created indirectly when such changes led individual entrepreneurs to question the prevailing ‘community of outlook’: Prices in an industry become ‘unstable’ as soon as any of the competitors form an idea of a profitable price which is markedly different from the existing prices. (p. 28) ‘Full-cost’ pricing implied that prices would likely be altered in response to significant changes in the cost of labour or raw material but that, normally, businessmen would not question the existing price structure as a result of moderate or transitory changes in demand. As Heflebower (1955, p. 361) indicated, the new heresy of ‘full-cost’ pricing provided an appealing explanation for relative price stability during the Depression, especially after the findings of Hall and Hitch were supported by subsequent studies like Saxton (1942), Lester (1946), Dean (1951) Oxenfeldt (1951), Fog (1960), Cyert and March (1963) and Skinner (1970).12 Nevertheless, the imprecise nature of the new approach left it open to criticism from mainstream economists who were quick to respond. 4. The Marginalists’ Counterattack The proposition that businessmen did not try to maximize their profits but rather were content with the quiet life of ‘full-cost’ pricing was not universally accepted by economists. Leading the neoclassicists’ counterattack, Machlup (1946) argued that the rejection of marginal analysis by empirical researchers such as Hall and Hitch (1939) and Lester (1946) was in fact baseless.13 In his opinion, Hall and Hitch and their followers erred because their research suffered from one or more of the following shortcomings: (1) a failure to properly understand the essence of marginal 12 For surveys of ‘full-cost’ pricing, see Heflebower (1955) and Silberston (1970). 13 Similar criticisms of ‘full-cost’ pricing appeared in Robinson (1939) and Kahn (1952). 18 analysis, (2) faulty research techniques, and (3) mistaken interpretations of empirical ‘findings.’ Let us consider these criticisms in some detail.14 According to Machlup (p. 521), the emphasis Hall and Hitch put on the ‘history of the industry’ in determining current conditions and in shaping behaviour was ‘by no means denied by marginal analysis.’ Contrary to common beliefs, he insisted, neoclassical theory recognized the role of history and, hence, did not really seek to explain how an individual firm determined the levels for its output, prices and employment. Rather, the theory focused on how the firm altered these variables in response to changing conditions. The overriding principle which guided the firm in its actions was the aim of maximum profit and marginal analysis was merely a technique used to achieve this goal. Machlup emphasized that the procedure whereby the firm equated marginal revenue and cost must be interpreted with great care. First, the magnitudes for the relevant variables were ‘subjective estimates, guesses and hunches.’ They reflected the perceptions, opinions, and beliefs of the businessman and were not necessarily equal to the corresponding ‘objective’ magnitudes as they might be observed by ‘outside’ parties. Second, the businessmen need not be engaged in tedious data collection and complicated calculations in order to equate marginal revenue and cost. In most cases he could rely on his intimate knowledge of his own business and follow an imprecise ‘routine’ which nevertheless accounted for all crucial factors: The business man who equates marginal net revenue productivity and marginal factor cost when he decides how many to employ need not engage in higher mathematics, geometry, or clairvoyance. Ordinarily he would not even consult with his accountant or efficiency expert in order to arrive at his decision; he would not make any tests or formal calculations, he would simply rely on his sense or his feel of the situation. There is nothing very exact about this sort of estimate. On the basis of hundreds of previous experiences of a familiar nature the business man would “just know,” in a vague and rough way, whether or not it would pay him to hire more men. The subjectivity of his judgements is obvious. (p. 535) Thus, contrary to the inference of Hall and Hitch (1939) and others, the observation that businessmen could not or simply did not know all the objective data, and the fact that they did not perform complicated computations failed to demonstrate that firms did not seek to maximize profit. Hall and Hitch further suggested that entrepreneurs did not make use of concepts such as ‘demand elasticity,’ ‘marginal revenue’ and ‘marginal cost,’ and in many cases did not even understand them but, according to Machlup, this also did not 14 For further replies and rejoinders see Lester (1947), Machlup (1947) and Stigler (1947). Later comments can be found in Machlup (1967). 19 invalidate the standard theory. While entrepreneurs might have failed to understand the marginal concepts as presented to them by Hall and Hitch, they have not necessarily failed the crucial test of marginalism. The marginal theory did not stipulate that businessmen must use the jargon of marginal analysis as developed by economists, only that they follow the marginal principles. Businessmen had no interest in the equality of marginal cost and revenue per se but only insofar as it helped them evaluate how their action might affect total profit. For that purpose they could also use many other guidelines which, although expressed in a different language, had practically the same meaning. For instance, a firm might decide to raise its price because it expected unit profit to rise by a greater percentage than the fall in quantity sold. The decision was based on ‘averages’ and ‘totals’ yet the logic was marginal for the focus was on the expected change in profit. Given that Hall and Hitch misunderstood the thrust of marginal analysis, and given that they baffled the entrepreneurs with academic jargon, it was hardly surprising that the two researchers also derived erroneous conclusions from their data. To explain this latter point, Machlup (p. 545) summarized the findings of Hall and Hitch in their own words: “A large majority” of them [of businessmen] explained that they charged the “full cost” price. Some, however, admitted “that they might charge more in periods of exceptionally high demand”; and a greater number reported “that they might charge less in periods of exceptionally depressed demand.” Competition seemed to induce “firms to modify the margins for profits which could be added to direct costs and overheads.” Moreover, “the conventional addition for profit varies from firm to firm and even within firms for different products.” According to Machlup (ibid.) these findings, which apparently ‘shook the researchers’ confidence in the marginal principle and convinced them that business men followed the “full cost principle” of pricing regardless of profit maximization,’ were exactly what one would expect to hear on the basis of marginal analysis! Indeed, in the neoclassical framework: we should expect for most industries that price in the long run would not deviate too much from average cost, yet that the firm would attempt to get better prices when it could safely get them and would not refrain from cutting prices when it believed that this would increase its profit or reduce its losses. (ibid.) The observation that different firms behaved differently and that their experience also vary over time proved, in Machlup’s opinion, that firms paid close attention to 20 variables other than average cost and, in particular, to those variables which affected their demand. In general, he summarized, there is little or nothing in the findings of this inquiry [by Hall and Hitch] that would indicate that the business men observed an average-cost rule of pricing when such observance was inconsistent with the maximization of profit principle. On the other hand, there is plenty of evidence in the findings that the business men paid much attention to demand elasticities – which to the economist is equivalent to marginal revenue considerations. (p. 546) Marginalists attacked ‘full-cost’ pricing on methodological grounds and hence it is interesting to note that their own criticisms suffered from similar methodological shortcomings. One important complaint against Hall and Hitch was that views of businessmen were no substitute for economic theory. Kahn (1952, p. 126), for instance, stated that the fundamental doubt is whether these business men, and other business men in similar predicaments, did not feel called upon to devise and present to the Oxford intellectuals, a theory of business behaviour which is primarily a rationalization and, in considerable measure a false rationalization of behaviour based on instinct rather than reasoning. It is with business men’s behaviour not with their thoughts, that we have to reckon. The economic theory of a business man may be based on the concept of a fair price, which is the price which, it is believed, in the absence of special circumstances, ought to rule. But very often this theory is a theory of ethics rather than of economics, and the business man takes the best price that he can get (through if this is higher than the fair price he may be reluctant to extort it to the full). The marginalists rejected the explanations of businessmen for ‘full-cost’ pricing as mere ‘ethics,’ ‘rationalization’ and even ‘false rationalization,’ yet their dismissal of evidence appeared to be quite selective. When the same businessmen reported on deviations from ‘full-cost’ pricing, Kahn and Machlup were only too eager to cite them as decisive confirmation of profit maximization. The basis for this selective use of evidence is not clear. Machlup (p. 538) wrote that ‘It takes an experienced analyst to disentangle actual from imaginary reasons and to separate relevant from irrelevant data and essential from decorative bits of information furnished,’ but he failed to enumerate the criteria he himself followed in screening the evidence provided by Hall and Hitch. If, as Kahn so forcefully asserts, we have to reckon with ‘behaviour’ rather than ‘thoughts’ then every interpretation provided by businessmen – whether it is consistent or inconsistent with the economic theory under examination – is simply extraneous for our purpose. 21 Beyond this double standard toward evidence, the citation from Kahn raises an even more serious difficulty concerning our ability to prove or refute the norm of profit maximization. A ‘historical’ approach to economic theory could emphasize forces beyond the particular inclinations of individuals and claim that, to a large extent, individual opinions and convictions are shaped by these forces. Hence, the empirical basis for testing such a historical theory for prices can indeed be independent from the ‘business creed.’ This conclusion does not hold for neoclassical price theory, however. The latter is a theory based on motivation and as such can be tested only by resorting to direct evidence on motivation. To say that business behaviour is governed by the aim of maximum profit and then to argue that the stated goals of businessmen cannot be used as evidence in testing the theory seems to us quite inconsistent. Without such direct evidence on motivation the neoclassical theory of profit maximization amounts to either a normative recommendation for businessmen on how they should act, or else it is simply an axiomatic construct. The marginalists could of course claim that, while they did not have direct support for the motivational theorem of profit maximization, the observation of business performance could provide an indirect test for this basic neoclassical postulate. This, however, is easier said than done. For example, Kahn (1952, p. 127) concludes that observed performances do not lend clear support to either ‘full-cost’ pricing or profit maximization: The actual behaviour of prices and profits – as revealed by comparisons of different firms and products and of different points of time – fails to support the “full-cost” principle in its undiluted form. But it fails equally to support, in its undiluted form, a narrow interpretation of the operation of the profit motive. (emphasis added) Yet the bases for such conclusions are not clarified by Kahn. We do not have an empirical yardstick for ‘maximum profit’ so we cannot really determine whether firms obtained this maximum or not. Furthermore, we cannot use business performance as evidence for business motivation. Even if we somehow knew what maximum profit were and even if we observed that firms indeed obtained this maximum, there would be nothing in this observation to demonstrate that firms sought maximum profit. Firms could obtain maximum profit by accident or even despite their efforts to attain another goal. Alternatively, firms could strive toward maximum profits but persistently fail to achieve them. In short, the goal for maximum profit can be demonstrated by interviews with businessmen or can be simply stipulated by the economist, but it cannot be proven or refuted by business performance. The second important criticism against Hall and Hitch was that businessmen acted not on the basis of objective circumstances, but rather on the basis of their own subjective interpretations of these conditions. In particular, it did not matter that 22 entrepreneurs did not know the objective demand curve as long as they acted on the basis of their subjective notion of that curve. This explication of the neoclassical theory is also problematic because profit maximization becomes consistent with every course of action. If, facing an increase in demand, businessmen increase their prices we can argue that profit maximization is vindicated, but we can derive the same conclusion if businessmen lower their prices instead! In this latter case, we can simply argue that businessmen attempted to maximize profits on the basis of erroneous interpretations of current conditions. If for some reason they believed that demand fell or was just about to fall, a policy to reduce prices would have been quite consistent with profit maximization, despite the ‘objective’ increase in demand. Thus, it would appear that when profit maximization is based on subjective perceptions of businessmen and when these perceptions cannot be accurately observed because we cannot rely on what businessmen tell us, the theory becomes irrefutable. Both adherents of ‘full-cost’ pricing and advocates of profit maximization argued that their theories explained business behaviour. They also acknowledged that these theories could not be used to predict prices. According to Robinson (1966), the two doctrines faced the same barrier mainly because they were unable to explain the profit margin. In the ‘full-cost’ approach, price was determined by adding to observed unit cost a certain profit markup but this addendum was admittedly ‘arbitrary’: The gross profit margin, or rake-off on price cost . . . probably depends very much upon historical accident or upon conventional views among business men as to what is reasonable. And any conventional pattern of behaviour which established itself amongst an imperfectly competitive group provides a stable result. So long as all adhere to the same set of conventions each can enjoy his share of the market, and each can imagine that he is acting according to the strict rules of competition, though in fact the group as a whole, by unconscious collusion, are imposing a mild degree of monopoly upon the market. . . . Where outright monopoly rules, or where a group of commodities is produced by a few powerful firms, there is great scope for individual variations in policy, and it is hard to make any generalization at all as to what governs the margin of profit per unit of output. (Robinson, 1966, pp. 78-9) For the neoclassicists, on the other hand, the price was determined when the businessmen attempted to maximize his profit by equating marginal revenue and marginal cost. The magnitudes for marginal revenue and cost, however, were not as clear in practice as they were in theory. The businessmen did not use the ‘true’ value from marginal revenue but rather his subjective interpretation for it. Furthermore, marginal cost included, in addition to observed expenses, an unspecified figure of ‘normal’ profit which the entrepreneur presumably added to cover his ‘opportunity 23 could not be described by simple mechanisms. This presented a serious methodological difficulty for the theory of price movements. The success of neoclassical price theory was contingent, to a large extent, on the ability of this theory to abstract from underlying dynamics of economic and other social relations. The focus on price as the ultimate variable of interest was required to reduce social relations and aspects of economic structure into a simple static framework. Firms are commonly assumed to operate in one of 4 possible market structures, which are fixed for the purpose of analysis. The structure affects the way firms set their prices but, since this structure is assumed to be fixed, it cannot be altered by price behaviour. Note that the static framework is not merely the first step toward a broader dynamic theory as neoclassicists often like to stress. If we allow price behaviour by firms such as IBM, General Motors or Exxon to alter the economic environment in which they operate, we introduce a fundamental ‘non-stationarity’ that is likely to undermine our ability to ‘predict’ such price behaviour. For this reason, the assumption that structure affects prices but prices do not affect structure is quite fundamental to neoclassical theory. This stationarity requirement also explains why it is necessary to assume that a businessman is a slave to a single fixed goal such as ‘profit maximization.’ Without a clear goal, the functional link between objective conditions and price behaviour is severed and prices become ‘arbitrary.’ The new ambiguities regarding the autonomy and diversity of business behaviour in modern capitalist economies removed much of the stationarity necessary for a solid price theory. In the neoclassical paradigm, the theorist could ignore the axiomatic nature of ‘profit maximization’ because this assumption was deeply embedded within the model and was rarely questioned. With the enlarged menu for potential patterns of business behaviour, things became more complicated. The observation that pricing goals and practices were not really fixed and changed with business conditions suggested that we could not ignore structural dynamics in our explanation for prices. It also insinuated that the behaviour of prices could operate to affect underlying structures. Many economists, it must be noted, failed to realize that the expanding field of ‘business anthropology’ created a methodological minefield. Instead of shying away from arbitrary assumptions about ‘business motivation,’ many preferred to ignore the potential hazard and actually welcomed their new freedom to chose. The result has been a flood of alternative models for inflation which could be distinguished mainly on the basis of their arbitrary behaviourial assumptions. We consider some of these models in Nitzan (1990). 30 References Bailey, M.J. (1959) ‘Administered Price Reconsidered: Discussion,’ American Economic Review, Papers and Proceedings, Vol. 49, No. 2, May, pp. 459-61. Blair, J.M. (1956) ‘Means, Thorp and Neal on Price Inflexibility,’ Review of Economics and Statistics, Vol. 38, November, pp. 427-435. —. (1959) ‘Administered Prices: A Phenomenon in Search of a Theory,’ American Economic Review, Papers and Proceedings, Vol. 49, No. 2, May, pp. 431-50. —. (1964) ‘Administered Prices and Oligopolistic Inflation: A Reply,’ Journal of Business, Vol. 37, No. 1, January, pp. 68-83. —. (1972) Economic Concentration. Structure, Behavior and Public Policy (New York: Harcourt Brace Jovanovich, Inc.). Brown, D. (1924) ‘Pricing Policy in Relation to Financial Control,’ Management and Administration, February, pp. 195-98, March, pp. 283-86 and April, pp. 417-22. Cyert, R.M. and J.G. March (1963) A Behavioral Theory of the Firm (Engelwood Cliffs, N.J.: Prentice-Hall). Dean, J. (1951) Managerial Economics (New York: Prentice-Hall). Fog, B. (1960) Industrial Pricing Policies: An Analysis of Pricing Policies of Danish Manufacturers, translated by I.E. Bailey (Amsterdam: North Holland). Hall, R.L. and C.J. Hitch (1939) ‘Price Theory and Business Behaviour,’ Oxford Economic Papers, No. 2, May, pp. 12-45. Heflebower, R.B. (1955) ‘Full Costs, Cost Changes and Prices,’ in Business Concentration and Price Policy, A Conference of the Universities-National Bureau Committee for Economic Research (Princeton: Princeton University Press), pp. 361-92. Kahn, R.F. (1952) ‘Oxford Studies in the Price Mechanism,’ Economic Journal, Vol. 62, March, pp. 119-30. Kaplan, A.D.H, J.B. Dirlam and R.F. Lanzillotti (1958) Pricing in Big Business. A Case Approach (Washington, D.C.: The Brookings Institution). Keynes, J.M. (1936) The General Theory of Employment, Interest and Money, (New York: Harcourt, Brace and Company). Lanzillotti, R.F. (1958) ‘Pricing Objectives in Large Companies,’ American Economic Review, Vol. 48, No. 5, December, pp. 921-40. Lester, R.A. (1946) ‘Shortcomings of Marginal Analysis for Wage-Employment Problems,’ American Economic Review, Vol. 36, No. 1, March, pp. 63-82. —. (1947) ‘Marginalism, Minimum Wages, and Labor Markets,’ American Economic Review, Vol. 37, No. 1, March, pp. 135-48. Machlup, F. (1946) ‘Marginal Analysis and Empirical Research,’ American Economic Review, Vol. 36, No. 4, September, pp. 519-54. —. (1947) ‘Rejoinder to an Antimarginalist,’ American Economic Review, Vol. 37, No. 1, March, pp. 148-54. 31 —. (1967) ‘Theories of the Firm: Marginalist, Behavioral, Managerial,’ American Economic Review, Vol. 57, No. 1, March, pp. 1-33. Mcallister, H.E. (1961) ‘Statistical Factors Affecting the Stability of the Wholesale and Consumers’ Price Indexes,’ in Price Statistics Review Committee, Staff Paper 8, pp. 372-418. Means, G.C. (1935a) Industrial Prices and their Relative Inflexibility, Senate Document 13, 74th Congress, 1st Session (Washington D.C.: GPO). —. (1935b) ‘Price Inflexibility and Requirements of a Stabilizing Monetary Policy,’ Journal of the American Statistical Association, Vol. 30, June, pp. 401-413. —. (1936) ‘Notes on Inflexible Prices,’ American Economic Review, Papers and Proceedings, Vol. 26, No. 1, March, pp. 23-35. —. (1972) ‘The Administered-Price Thesis Reconfirmed,’ American Economic Review, Vol. 62, No. 3, June, pp. 292-306. Mills, C.F. (1927) The Behavior of Prices (New York: National Bureau of Economic Research). Moore, M. (1972) ‘Stigler on Inflexible Prices,’ Canadian Journal of Economics, Vol. 5, No. 4, November, pp. 486-93. National Resources Committee (1939, 1966) The Structure of the American Economy, directed by G.C. Means, Part I, Basic Characteristics, June (New York: Augustus M. Kelley Publishers, Reprints of Economic Classics). Nelson, S. (1939, 1966) ‘A Consideration of the Validity of Bureau of Labor Statistics Price Indexes,’ in National Resources Committee (1939), Appendix 1, pp. 173-185. Nitzan, J. (1990) Inflation and Market Structure, Discussion Paper, Department of Economics, McGill University, Montreal. Oxenfeldt, A.R. (1951) Industrial Pricing and Market Practices (New York: Prentice Hall). Price Statistics Review Committee (1961) The Price Statistics of the Federal Government. Review Appraisal, and Recommendations (New York: National Bureau of Economic Research). Robinson, A. (1939) ‘Oxford Economic Papers,’ Economic Journal, Vol. 49, September, pp. 538-43. Robinson, J. (1966) An Essay on Marxian Economics, 2nd Edition (London: Macmillan). Ross, H.N. (1964) The Theory and Evidence of Price Flexibility, Doctoral Dissertation, Columbia University. Ross, S.A. and M.L. Wachter (1973) ‘Wage Determination, Inflation, and the Industrial Structure,’ American Economic Review, Vol. 63, No. 4, September, pp. 675-92. Saxton, C.C. (1942) The Economics of Price Determination (London: Oxford University Press). 32 Silberston, A. (1970) ‘Surveys of Applied Economics: Price Behaviour of Firms,’ The Economic Journal, Vol. 80, September, pp. 511-82. Skinner, R.C. ‘The Determination of Selling Prices,’ Journal of Industrial Economics, Vol. 18, No. 3, July, pp. 201-17. Stigler, G.J. (1947) ‘Professor Lester and the Marginalists,’ American Economic Review, Vol. 37, No. 1, March, pp. 154-57. —. (1962) ‘Administered Prices and Oligopolistic Inflation,’ Journal of Business, Vol. 35, No. 1, January, pp. 1-13. —. (1971) ‘Editor’s Note,’ Journal of Economic Literature, Vol. 9, September, p. 852. Stigler, G.J. and J.K Kindahl (1970) The Behavior of Industrial Prices (New York: National Bureau of Economic Research). —. (1973) ‘Industrial Prices, as Administered by Dr. Means,’ American Economic Review, Vol. 63, No. 4, September, pp. 717-21. Thorp, W.L. (1936) ‘Price Theories and Market Realities,’ American Economic Review, Papers and Proceedings, Vol. 26, No. 1, March, pp. 15-22. Thorp, W. and W. Crowder (1941a) ‘Concentration and Product Characteristics as Factors in Price-Quantity Behavior,’ American Economic Review, Papers and Proceedings, Vol. 30, No. 5, February, pp. 390-408. —. (1941b) The Structure of Industry, Monograph No. 27, Temporary National Committee. Tintner, G. (1935) Prices in the Trade Cycle (Vienna: Springer). Tobin, J. (ed.) (1983) Macroeconomics, Prices, and Quantities: Essays in Memory of Arthur M. Okun (Washington D.C.: Brookings Institution). —. (1983) ‘Okun on Macroeconomic Policy: A Final Comment,’ in J. Tobin (ed.), pp. 297-300. Weiss, L.W. (1977) ‘Stigler, Kindahl and Means on Administered Prices,’ American Economic Review, Vol. 67, No. 4, September, pp. 610-19. 33