![Sample our Economics, Finance,Business & Industry journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5931/TOC-Subject-Sample-2021-Economics-Finance-Business-Industry.jpg"})
Abstract
Unlike the common interpretation and consequent general rejection of the kinked demand curve, J.M. Clark developed a different, and more useful, interpretation of the kinked demand curve. While his approach was not generally understood, for reasons we will discuss (including the fact that his discussion of related issues was scattered among multiple books and articles, many of which primarily focussed on other issues), Clark’s approach to the kinked demand curve offers valuable insights into typically overlooked properties of imperfect, price-quoting markets, as opposed to the standard implicit assumption of perfect markets.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 For a more extensive discussion of the early evolution and use of the kinked demand curve, see Spengler (Citation1965).
For more extensive evaluations of J. M. Clark’s body of work, see Shute (Citation1997), Hickman (Citation1975), Rutherford (Citation1997, Citation2000, Citation2001, Citation2003, Citation2013), Markham (Citation2008), Gruchy (Citation1947, Citation1972), Hutchinson (Citation1953) and Dorfman (Citation1969).
2 Alternatively, it could be referred to as a modified monopolistic competition model, using one of E.H. Chamberlin’s (Chamberlin Citation1933) firm demand curves for price increases and another for price decreases. Both of Chamberlin’s curves are modified monopoly demand curves, however, so that the monopoly characterization holds.
3 Since there is a gap in marginal revenue at the demand kink, so that it is not defined at the solution quantity, the standard rule of equating marginal revenue to marginal costs must here be more accurately stated as producing every unit for which marginal revenue exceeds marginal cost.
4 The demand curve is more inelastic in the downward direction, though not necessarily relatively inelastic. Given estimates of relatively price inelastic short-run demand facing many industries, however, marginal revenue would actually be negative for price decreases in such cases, implying that a firm would never lower price in such circumstances if short-run profit maximization (as presumed in the model) was the goal.
5 Note that the kink is based on changes in price, but marginal revenue is defined in terms of changes in quantity, so that standard quantity determination rules can be applied to determine when prices will or won’t change.
6 The kinked demand curve is in fact the only market model that can address price stability with the standard economic tools. There is nothing in the other standard market models—perfect competition, monopoly, monopolistic competition–which predicts this result, leading users of those models to invoke reasons for such stability which are external to the models themselves, such as costs of changing prices or long run effects (Stigler Citation1947, Citation1978). However, this resort to external rationalizations for price stability which is not implied by the models is entirely ad hoc and is little more than an admission that those models cannot explain price stability. Further, the reasons given for price stability, which are related to price-setting behavior, are inconsistent with the implicit perfect market assumption (in which a seller is assumed able to determine the profit maximizing quantity he will sell) of quantity setting models.
7 A careful reading of both Hall and Hitch (Citation1939) and Sweezy (Citation1939) indicates that they were not trying to construct a model cast in the standard mold, from which price stability could be derived, as neoclassical economists have treated it, but rather that they were trying to illustrate observed cases of price stability within the conventional model context by using the device of a kinked demand curve. For example, Sweezy’s opening line includes the statement that “…the ordinary concept of a demand curve is inapplicable to the study of oligopoly” (Sweezy Citation1939, 404), while Hall and Hitch’s opening page includes the statement that their analysis “casts doubt on the general applicability of the conventional analysis of price and output policy in terms of marginal cost and marginal revenue” (Hall and Hitch Citation1939, 107). This indicates that current presentations of the kinked demand curve as part of the textbook canon of marginal revenue equals marginal cost models is far different from the intent of the kinked demand curve’s originators. Clark’s treatment is closer to the original approaches to the kinked demand curve than the later neoclassical treatment.
8 This is an empty criticism in an important sense, because none of the standard economics market models actually addresses price determination at the firm level; all of them address quantity determination, assuming “the market” will automatically generate a price that will lead their sales to equal what they intended. It is this presumption of an institution that automatically generated market-clearing prices in such models that is the core reason for Clark’s deviation from Stigler and others with respect to the kinked demand curve, because he argued that price-quoting markets have very different properties and implications.
9 While he is well known, J.M. Clark’s entire body of work has not been well-studied. That is illustrated by the fact that Shute’s book on Clark (Shute Citation1997), which was published long after Clark’s death, was promoted as “the first comprehensive study of the life and works of John Maurice Clark.” And even then, Niehans’ review of the book (Niehans Citation1998) complained of a “paucity of critical analysis and historical evaluation of Clark’s contributions to economics.” Clark himself suggested a possible reason when he wrote (Clark Citation1927), that “The key to statics [is]…a normal level toward which the economic forces of gravity exert their pull. The key to dynamics is…processes which do not visibly tend to any complete and definable static equilibrium,” This theme was still being echoed in his last major work, Competition as a Dynamic Process (Clark Citation1961). He wrote that “the threat of failure looms large, in that readers whose conception of theory is identified with models of determinate equilibrium are likely to decide that no theory has been produced.” (x) That, in turn, reflected Markham’s description of Clark’s approach: “Probably more than any of his contemporaries, he concerned himself explicitly with bridging the enormous gap between the models of static theory and the dynamic realities of a market economy,” in which “Clark digested and skillfully built upon the logic underlying the formal models of others,” but “fashioned his own methodology from the written word.” (Markham Citation2008, 2) .
10 Note that Sweezy (Citation1939), and Hall and Hitch (Citation1939), treat such demand curves as subjective, not objective, leaving open the possibility of investigating when such expectations would be held and when they would not. See also Efroymson (Citation1943, Citation1955).
11 The closest to a single clear statement of what follows is in Clark (Citation1961), but elements of Clark’s approach can be found sprinkled throughout his other works. As we will discuss later, a good deal of that discussion appeared in Clark’s work on basing point pricing, which all but disappeared as a topic in economics literature after major antitrust rulings on basing point pricing in cement and steel in 1949.
12 This is in sharp contrast with Niehans’ (Citation1998) conclusion that “Clark belonged to a generation of problem raisers, not problem solvers. He thought he had achieved something if he could describe the solution to a problem as ‘thoroughly indeterminate’.” In fact, Clark was not satisfied with that. Finding something “thoroughly indeterminate” provided the impetus to try to understand it in dynamic terms, when it did not lend itself to solving for a determinate equilibrium using the static tools typically employed. As Gruchy (Citation1947, 358) put it, Clark wanted to come “to grips with the dynamic movements and resistances to movement,” which characterize modern industry. Or as Dorfman (Citation1969, 443) cited Clark, “If competitive price tends to no definite level, much of one’s old economics needs revising.”
13 In terms of the conventional kinked demand curve model, this would be a case where a price reduction would shift a firm’s share of the market demand curve to the right, rather than simply moving the firm down along a fixed share of the market demand curve.
14 Better, it could be said that Clark believed the prevalence of overhead costs in quoted-price markets for standardized or homogeneous products led to fewness of sellers, who are then necessarily in an interdependent or oligopolistic situation, where price is the critical variable under each seller’s control, because of quoted-price marketing.
15 See Clark & Clark (Citation1914, 279, 288–292, Citation1923, 439–443, Citation1934, 60–61, Citation1949b, 440–441, Citation1955, 457, 460–461, Citation1958, 474–482, Citation1961, 50–51, 94–95, 118–119, 156–161, 170, 271–297, 429–438). This reflects what Clark wrote in his “The Socializing of Theoretical Economics” (Citation1924, 75): “the core of scientific method lies…in taking account of all relevant facts and excluding none.”
16 Any model whose equilibrium solution depends on the equation of short run marginal revenue and short run marginal cost implies that firms determine the quantity they succeed in selling and that price will change whenever the underlying demand and cost conditions change.
17 These are the “factors making for price stability [that] were generally incorporated (without emphasis) into the neoclassical theory” (Stigler Citation1947, 418) that Stigler cited in favor of conventional theory over the kinked demand curve, reasoning that “There is…no purpose in adding it to neoclassical theory unless it explains price behavior in areas where the other explanation is silent or contradicts the implications of the neoclassical theory in the areas where both apply” (Stigler Citation1947, 419–420). In fact, the models of neoclassical theory are silent on price flexibility in price setting markets, as the quantity-setting sellers of those models never face direct pricing decisions.
18 Since those models are all short run quantity clearance models, they assume away any consideration of pricing rivalry, and it is pricing rivalry in quoted-price markets that makes these concerns relevant.
19 Clark & Clark (Citation1914, 275), Clark (Citation1914, 275, Citation1923, 459, Citation1936, 9, Citation1946, 347) argued that while determinate equilibrium results could be reached via the conventional standard modeling approach, an accurate, though less than completely determinate, analysis of the factors actually at work was preferable to a precisely determinate analysis that omitted those factors. But Clark was also concerned that as a consequence, some for whom economic theory is about determinate equilibrium solutions would not even recognize his “different way of thinking” as theory, which seems to have been justified by the subsequent lack of serious treatment in the mainstream literature of economics.
20 For example, a firm not currently in financial trouble might not wish to shade prices and risk starting a price war whose outcome may be disastrous to it, while a firm in immediate financial distress may not have the luxury of refraining from shading prices even if a price war may start as a result. However, no such price shading scenario can be considered in the standard quantity determination framework, since firms do not directly change prices in that approach.
21 This is especially true of smaller firms who may hope that larger firms will not find their inroads worth responding to. However, such inroads may lead to price wars, if and when bigger firms do choose to respond. Further, once a price war starts, no small firm could unilaterally end their “punishment” by forcing rivals to once again raise their prices.
22 Clark (Citation1923, 439–443) includes several considerations under this heading. Price reductions may lead buyers to expect still further price reductions, and their holding off on purchases increases the pressures for such further reductions to be made. Sellers must weigh those possibilities before lowering prices. Price reductions may also shake buyer confidence in the fairness or necessity of the old prices, making it hard to raise them back up later in better times, adding incentives for price stability. In addition, price cuts may primarily move more profitable sales from higher price periods to low price periods, and this prospect of regularization spoiling the peak tends toward price stability. There is also the prospect of retaliation, with the potential for severe future losses from current aggression, at their rivals’ discretion, to consider in choosing to lower price.
23 As opposed to the traditional interpretation that the kinked demand curve causes price stability, a better interpretation would seem to be that it illustrates how quoted-price marketing generates resistance to price changes, whileother conditions generate forces leading some firm or firms to consider price changes. It does not mean prices never change; only that pressures need to be great enough to overcome quoted-price marketing’s pressures toward price stability—the potential risks from initiating increases or decreases in prices–before prices would tend to change.
24 The consistency of the evidence with Clark’s modified kinked demand curve analysis, where various quoted-price considerations can lead to price changes inconsistent with the conventional kinked demand curve model, stands in sharp contrast not only to Stigler’s conclusion that the model is not useful, but also to his statement that “I would like to think that this type of criticism is no longer publishable” (Stigler Citation1978, 191).
25 This is a great strength in Clark’s institutional analysis, because he makes an extremely persuasive case for the view that the standard modeling approach necessarily leaves out of consideration important determinants and implications of quoted-price market behavior. But those who approached the subject from a neoclassical viewpoint, such as Stigler, asserted that its failure to generate a determinate optimization solution meant that it was valueless. Therefore, where Clark’s approach of incorporating more of the complex reality, in pursuit of a new type of “dynamic” theory (Clark Citation1914, Citation1961, ix), even though it is not based on a well-defined optimization problem with an unambiguous answer, offers superior predictions, the evidence must be seen as on Clark’s side. Despite that, however, the development of economic theory since has been sharply in the opposite direction from Clark’s institutional approach, leading analysis away from the considerations in his approach.
26 They can be considered in an ad hoc manner, as Stigler does in his critique of the kinked demand curve, but since they do not fit within the standard model framework, they play no active role in the typical model-based analysis. As Clark stated in response to a similar criticism by Frank Fetter, “Professor Fetter claims that I do not understand static theory in that static theory emphasizes its own limitations. Nobody realizes more than I do that as compared to theory which is unconsciously static, any theory which announces its static character takes a long step in the right direction. But a sign which announces: ‘This happy valley is not the whole world’ does not take the place of explorations outside the happy valley, no matter how many exclamations there are on the sign” (Hickman Citation1975, 46).
27 See Clark (Citation1914, 275, Citation1927, 197–203, Citation1936, 1–8, Citation1947, 9–10).
28 See Clark (Citation1927, 199), where, in contrast to any sort of adequate incorporation of dynamic elements in an analysis of price stability, he describes the developments of static theory as refinements “whose accuracy is hardly justified in view of the wide gap between the assumed conditions on which the whole structure rests and the reality…”
29 Clark (Citation1948, 66, 69) discusses why monopoly is a far more comprehensive concept than competition within the perfect market model framework, because perfect competition leaves firms no decision-making power over prices or marketing strategy. Since quoted price market rivalry involves discretionary power, such behavior thus tends to be characterized as monopolistic (or oligopolistic, where such behavior is interpreted to be monopolistic in nature). Even for real monopolists, the standard analysis is too short run to capture all the relevant issues.
30 In this sense, the criticism that the kinked demand curve doesn’t explain the level of prices has no adequate foundation. No other market model can adequately explain the history of prices in an industry, because none of them directly consider the price setting decision. If anything, the kinked demand curve is superior in this regard because it at least tries to address the question of when a firm would choose to change its quoted prices, which cannot even be asked in quantity determination models.