Price Predation: Legal Limits and Antitrust Considerations

Price Predation: Legal Limits and Antitrust Considerations

Gregory T. Gundlach

Competition centered strategies in the form of predatory pricing directed toward weakeningor destroying a competitor are receiving increasing emphasis in the Courts and amongantitrust theorists and policymakers. Recently, the Supreme Court found little evidence ofantitrust injury e.xtending from price predation in the case fj/Brooke Group v. Brown &Williamson Tobacco Corporation (1993). The author examines the current legal .standard forpredatory pricing and juxtaposes it against emerging insights on this competitive practice.

I n markeling. cotnpetition-centered strategies, includingthe manipulation of price, have received increasing em-phasis (cf. Weitz 1985). According to Kotler and Singh

(1981, p. 30). "successful marketing ... require[s] devisingcompetition-centered strategies, not just customer-centeredand distribution-centered strategies." Faced with decliningmarket growth, resource scarcities, and the proliferation ofnew technologies, companies are increasingly pursuingprofit gains at the expense of their rivals through marketshare rather than market growth.

Strategies directed toward competitors can involve differ-ent objectives. On the one hand, a company may employcompetition-centered strategies to create a "better state ofpeace" hetween rivals (Kotler and Singh 1981). Such an ap-proach involves a business entrenching itself in some part ofthe market in which it has a natural and comparative advan-tage. This advantage discourages aggressive actions by ri-vals and favors peaceful coexistence. Rivalrous behaviorthat does occur is usually the result of one firm poorly serv-ing its market niche, a new competitor attempting to bringnew advantages to the market, or changes occurring in theenvironment. Strategies of this kind are welfare enhancing,because they facilitate efficient competition.

On the other hand, competitive strategies may also pos-sess the objective of weakening or destroying a competitor(i.e.. cutthroat competition). These strategies can be injuri-ous to competition and may potentially reduce consumerwelfare. Such strategies are commonly referred to as preda-tion OT predatory strategies and can involve both price andnonprice tactics (cf. ZeithamI and ZeithamI 1984). Nonpricepredation involves a competitor's attempt to increase arival's costs through such tactics as the acquisition and"sleeping on" of patents, predatory product preannounce-ments, useless product modifications, exclusionary marketchannel arrangements, disparaging advertising, and shamlitigation (see Gundlach 1990). An aggressive firm may alsoattempt to lower a rival's profits through price predation. Inthis case, a firm sets its prices low enough and for a suffi-cient time period to disadvantage its rivals in some way(Sheffet 1994). Tactics can include sustained price cuts lo

GREGORY T. GUNDLACH is an associate professor. Department ofMarketing. University of Notre Dame. The author thanks the edi-tor, four JPP&M reviewers, the members of the Department ofMarketing at the University of Notre Dame, and especially JosephP. Guiltinan and Debra J. Ringold for their helpful comments.

below-cost levels, discriminatory pricing, and temporaryprice warring.

A primary objective of the antitrust laws is to distinguishcompetitive strategies (and more particularly, predatorystrategies) that are welfare enhancing from those that reducethe welfare of consumers. The difficulty of this task is madeeven more complicated when predatory pricing is involved.The Supreme Court states:

[TJhe mechanism by which a firm engages in predatory pric-ing—lowering prices—is the same mechanism by which a firmstimulates competition; because cutting prices in order to in-crease business often is the very essence of competition ...[;]mistaken inferences ... are especially costly, because they chillthe very conduct the antitrust laws are designed to protect{Brooke Group Ltd. v. Brown & Williamson Tobacco Corp.1993. p. 4703).

Because of the possibility of mistaking welfare enhancingprice competition and extant theory that suggests the irra-tionality of predatory pricing, policymakers and the Courtshave long held a skeptical view of this form of (anti)compe-tition. Their thesis is that a predator would lo.se so muchmoney attempting to drive out a rival through lt)wering pricebelow cost that it would never eam enough profits later torecoup its losses. The argued irrationality toward predatorypricing and resultant enforcement posture taken by the De-partment of Justice (DOJ) and the Federal Trade Commis-sion (FTC) (see Rule 1988) is summarized by a former di-rector of the Bureau of Competition: "Many argue that pre-dation is impossible, and hence, enforcement of the antimo-nopolization provisions of the Sherman Act should not bebased on this empty concept" (Campbell 1987, p. 1625).This perspective also underlies key decisions of theSupreme Court, which has stated, "predatory pricing ... is bynature speculative ... landl ... rarely tried, and even morerarely successful" (Matsushita Electric industrial Co. v.Zenith Radio Corp. 1986. pp. 588-89).

Recently, the Supreme Court again addressed predatorypricing in Brooke Gmup v. Brown & Williamson Tobacco Cor-poration (1993). In announcing a two-part standard, the Courtfound little evidence of predation in a price war between twomajor tobacco companies. The Court's decision establishes avery difficult threshold for firms threatened by predatory pric-ing. Some commentators have even suggested that the impactof the ca.se as a precedent may prove to be "fatal to all futurepredatory pricing claims" (Glazer 1994. p. 606).

In contrast to the Supreme Court's holding and thosecommentators espousing such a view, a number of theorists

278 Joumal of Public Policy & MarketingVol. 14 (2)

Fall 1995, 278-289

Journal of Public Policy & Marketing 279

have begun to reconsider the nature and occurrence of pre-dation as an anticompetitive practice. Among these scholars,"il is now generally agreed that the argument that predationis impossihle is incorrect at least as a matter of theory" (Hay1990. p. 913). These theorists propose numerous models ofpredation that rely on nonstatic notions of competition in-volving multiple markets, asymmetries of information, andnonprofit maximizing strategies (see Schmalensee andWillig 1989). According to this "new learning." in theory,even above-cost pricing can be employed to disadvantageequally or more efficient rivals and result in injury to con-sumer welfare (Hay 1990. p. 914). Of course, the desirabil-ity of adopting and implementing such an antitrust standardis questionable. (For a discussion of this issue, see. for ex-ample. Barry Wright Corp. v. ITT Grlnnel Corp. 1983.)However, tbe insight offered by scholars of this paradigmbas prompted some to reconsider prior antitrust thinking.

I examine the current law and theory of predatory pricing,while emphasizing the Supreme Court's two-part standarddeveloped in Brooke Group (1993). This standard is juxta-posed against recent advancements in industrial organiza-tion theory and marketing that describe how predatory con-duct involving price may be a rational strategy for tbe ag-gressive competitor and may be employed in a variety ofways to disadvantage rivals, some of which create antitrustinjury. I then discuss implications for antitrust policy devel-opment, marketing research, and the practice of markeling.

Background and Current LawStatutory LawFederal antitrust law attempts to control predation throughSection 2 of the Sherman Act (1890). Section 2 of the Clay-ton Act (1914) as amended by the Robinson-Patman Act(1936). and Section 5 of the FTC Act (1914). The ShermanAct (Section 2) states:

Every person who shall monopolize, or attempt to monopolize,or combine or conspire with any other person or persons, to mo-nopolize any part of the trade or commerce among the severalStates, or with foreign nations, shall be deemed guilty of afelony....

As outlined in United States v. Grinnell Corporation (1966,pp. 570-71). monopolization requires monopoly power andconduct indicating a "willful acquisition or maintenance ofthat power, as distinguished from growth or development asa consequence of a superior product, business acumen, orhistoric accident." An attempt to monopolize, alternately re-quires (1) specific intent to control prices or destroy compe-tition, (2) predatory or anticompetitive conduct directed to-ward that end, and (3) a dangerous probability of success{Spectrum Sports, Inc. v. McQuillan 1993). Often a specificintent to monopolize is inferred from predatory conduct,whicb thus combines tbe first and second elements.

The Clayton Act (1914, Section 2), as amended, makes itunlawful for any seller to discriminate "in price between dif-ferent purchases or commodities of like grade and quality ...where the effect of such discrimination may be substantial-ly to lessen competition or tend to create a monopoly in anyline of commerce." The Act forbids exclusionary conduct ofa discriminatory nature (Sullivan 1977). In the case of

predatory pricing, a firm is prohibited from selling at a lowprice when it encounters competition and at a high price oth-erwise if its purpose is to drive rivals out of the market. Inaddition, other discriminatory practices aimed at forcing acompetitor to consolidate or disciplining a rival that reducedits prices or competed with the firm are forbidden.

Section 5 of the FTC Act (1914) autborizes ibe Commis-sion to cballenge predatory practices deemed to be "unfairmetbods of competition." Tbe Act permits (I) challenges topredatory conduct violating the Sherman and Clayton Actsand (2) transgressions of the spirit of these Acts.

Besides the federal laws, many states have sales-below-cost statutes or constitutions prohibiting monopolies thatcan be employed to cballenge predatory pricing (see Haynes1990). Often, these statutes require evidence of below-costpricing and a specific intent to injure a competitor. For ex-ample, in a recent well-publicized case. Wal-Mart. Ameri-ca's leading discounter, was found guilty of violating anArkansas state unfair pricing statute: Its stated pricing strat-egy was to "meet or beat tbe competition without regard tocost" (American Drugs. Inc. v. Wal-Mart Stores. Inc. 1993).Courts have observed these unfair pricing laws to ease thehardships of cutthroat competition (Quinn 1990).

Federal StandardsAt the Federal level, differences among the statutory lan-guage of botb the Sherman and Clayton Acts bas led to dif-ferent standards for injury to competition and the nature ofconduct considered anticompetitive. Tbe Sbennan Act con-demns predatory pricing when it poses a "dangerous proba-bility of actual monopolization" (Spectrum Sports. Inc. v.McQuillan 1993. p. 8). In contrast, tbe Robinson-PatmanAct requires only "a reasonable possibility" of substantialinjury to competition before its protections are triggered(Brooke Group 1993, p. 4702). Thus, for predatory pricingto be found anticompetitive, an actual determination of mo-nopolization or injury to competition need not be found.Rather, only a reasonable prospect (Robinson-Palman Act)or dangerous possibility (Sherman Act) is required.

Supreme CourtOver the years, the Supreme Court has addressed predatorypricing on various occasions. Tbese decisions suggest a nar-rowing of the Court's stance from relying on predatory in-tent derived from extensive analysis of circumstances in-cluding below-cost pricing, to relying on a more limited, butobjective, below-cost price and reasonable expectation ofrecoupment test.

One key early decision involving intent is Utah Pie Com-pany V. Continental Baking Company (1967). Analyzingpredatory discrimination in the pie market, the Court heldthat, "the existence of predatory intent might bear on thelikelihood of injury to competition" (Utah Pie 1967. p. 702).Such intent could be inferred from economic circumstances,such as "persistent unprofitable sales below cost and drasticprice cuts" (p. 702). Little introspection regarding tbe basisof employing intent as an element of injury or clarity for de-veloping a useful standard is provided, however.

With the increasing application of economic analysis inantitrust during the 1980s, predatory intent was replaced as

280 Price Predation

an element of inquiry with more objective standards. Twokey decisions include Matsushita Electric Industrial Com-pany V. Zenith Radio Corporation (1986) and Cargill v.Monfort of Colorado., Inc. (1986). In Matsushita, predatorypricing was defmed as "pricing below some appropriatemeasure of cost" (p. 584, n. 8) and "a reasonable expectationof recovering, in the form of later monopoly profits, morethan tbe losses suffered" (p. 524) from a predatory episode.In Cargill, a similar conclusion was reached: "Courts shouldnot find allegations of predatory pricing credible when thealleged predator is incapable of successfully pursuing apredatory scheme" (pp. 119-20. n. 15). Togetber. thesecases reflect tbe Court's movement toward adopting eco-nomic cost-price comparisons and recoupment-based analy-ses for addressing predatory pricing.

Brooke Group v. Brown & Williamson TobaccoCorporation (1993)The Court's most recent case involves an allegation ofpredatory price discrimination. In 1980. Liggett pioneeredthe economy segment of the cigarette market through the de-velopment of a line of generic cigarettes offered at a listprice 30% lower than branded cigarettes. By 1984, genericsheld 4% of the market, primarily at the expense of marketshare for branded cigarettes. Brown & Williamson, feelingthe loss of share in its branded cigarettes, entered the gener-ic market and beat Liggett's price. The ensuing price war re-sulted in Liggett's complaint that Brown & Williamson'sstrategy (volume rebates to wbolesalers) amounted tobelow-cost pricing and an attempt to pressure Liggett toraise prices on generics. which would, thus, restrain the seg-ment's growth and preserve Brown & Williamson's supra-competitive profits on branded cigarettes (for a more exten-sive discussion of this ca.se, see Sheffet 1994). In findingtbat predatory discrimination had not occurred, the Court ar-ticulated a two-part standard:

First, a plaintiff seeking to establish competitive injury resultingfrom a rival's low prices must prove that Ihe prices complainedof are below an appropriate measure of its rival's costs (p. 4702).

The second prerequisite to holding a competitor liable under theantitrust laws for charging low prices is a demonstration that thecompetitor had a reasonable prospect lunder Robinson-Patman],or under Section 2 of the Sherman Act. a dangerous probability,of recouping its investment in below-cost prices (p. 4703).

Tbe first prong of tbe standard requires tbat the predator'sprices are below an appropriate measure of cost. Accordingto the Court, "only below-cost prices should suffice" and"above-cost prices tbat are below general market levels orthe costs of a firm's competitors inflict injury to competitioncognizable under the antitrust laws" (p. 4702). The rationalefor this requirement is that,

Ia]s a general rule, the exclusionary effect of prices above a rel-evant measure of cost either reflects the lower cost structure ofthe alleged predator, and so repre.sents competition on the mer-its, or is beyond the practical ability of a judicial tribunal to con-trol without courting intolerable risks of chilling legitimateprice-cutting (p. 4703).

Accepting a standard of below-cost pricing, the Court de-clined to resolve the appropriate cost measure to be consid-

ered. Instead, botb parties agreed that tbe relevant measureshould be average variable cost.

Tbe standard's second part involves assessing the reason-able prospect (Robinson-Patman Act), or dangerous proba-bility (Sberman Act) of a predatory firm recouping its in-vestment through supracompetitive prices. A further aspectof this inquiry is determining the likelihood of injury tocompetition if recoupment were to occur. The Court indicat-ed that such a determination should be similariy analyzedunder botb tbe Robinson-Patman and Sherman Acts andmust include "an understanding of the extent and duration ofthe alleged predation, the relative financial strength of thepredator and its intended victim, and their respective incen-tives and will" (p. 4703). With this understanding, "if mar-ket circumstances or deficiencies in proof would bar a rea-sonable jury from finding that the scbeme alleged wouldlikely result in sustained supracompetitive pricing, the plain-tiff's case has failed" (p. 4703. emphasis added).

The market circumstances identified include markets thatare highly diffused or extremely competitive, markets inwhich new entry is easy, or in which a defendant would beincapable of taking on additional capacity because of theexit of a rival. In sbort. the court indicated that the inquirymust evaluate whether "given the aggregate losses caused bythe below-cost pricing, the intended target v^ould likely suc-cumb" (p. 4703. empbasis added) and whether the predatorcould benefit from tbe rivals sunender.

In discussing tbeir standard, tbe Court observed. "Itlheseprerequisites are not easy to establish, but they are not arti-ficial obstacles to recovery; rather they are essential compo-nents of real market injury." The Court also reiterated its as-sertion from Mat.sushita (1986) that, "predatory pricingschemes are rarely tried, and even more rarely successful"{Brook Group 1993, p. 4703). Together, the Court's com-ments suggest a skeptical outlook toward predatory pricingand one that arguably disfavors legal claims that rely onsuch conduct.

Department of Justice and Federal TradeCommissionClaims of predatory pricing typically originate betweencompetitors, and thus tbe majority of cases involve privatelitigation. However, actions against predation are also with-in the authority of both the DOJ and FTC. In general, theseagencies have adopted a similar posture to that of theSupreme Court. According to the DOJ,

It is almost always very difficult, if not impossible, to tell thedifference between beneficial, efficient competition and objec-tionable. inefTicient predation-... Therefore, although we may as-sume that some anticompetitive strategic behavior does occur.Ihe appropriate role of antitrust law in regulating that conduct isnevertheless very limited (Rule 1988, pp. 426-427).

The DOJ has noted that, even if violations were found, thelack of a workable remedy poses problems for pursuing pre-dation. Any remedy requires limiting a predator's ability tocompete and could result in additional concems.

Similar to tbe DOJ. the FTC has taken a cautious view.When asked about the prospect of the FTC initiating a com-plaint for predatory pricing. Commissioner Oliver stated:

Journal of Public Policy & Marketing 281

Figure 1. Predatory Pricing: Analysis of Federal CourtCases

70 71 72 73 74 75 76 77 76 79 60 81 B2 B3 04 85 86 87 8B 89 90 91 92 93 94

^ 1 Cases per year

tJ.S. Supreme Court. Court of Appeals and District Court (total = 787)

[TIhe Supreme Court has said, correctly we think, that predato-ry pricing is rarely tried and rarely successful. The complaintsthat we get, and we do get some complaints on it, when we in-vestigate we discover that basically it's a competitor upset at thesuccess that his competitor is having {Antitrust & Trade Regu-lation Report 1987, p. 340).

Moreover, in response to an inquiry conceming a reported42 complaints conceming predatory pricing and only oneinitial phase investigation, the Director of the FTC's Bureauof Competition. Jeffrey I. Zuckerman, commented that theFTC does not "want the message to be out there that cuttingyour price is going to bring the govemment down on you"(Antitmst & Trade Regulation Report 1989. p. 127; see alsoScbildkraut et al. 1992). More recently, reauthorization leg-islation now requires the FTC to report to Congress the ac-tivities of the Commission relating to predatory pricing (An-titrust & Trade Regulation Report 1993).

Private Case LawIn contrast to the stated opinions of the Supreme Court andenforcement agencies regarding the plausibility of preda-tion. various indices suggest a steady concern for this prac-tice. An examination of Federal cases beginning in 1970 andextending to the present indicates a steady (if not increasing)number of cases discussing predatory pricing. Figure 1shows a summary analysis of the frequency with which Fed-eral cases have referenced price predation. Archival exami-nation of written opinions of the Federal judiciary contain-ing reference to "predatory pricing" or its derivatives wasemployed for this analysis.'

The number of cases discussing predatory pricing acrosstime suggests continued concem for this form of (anti)com-petition. Although the cases are not consistent or dispositive

'A computerized dala base (WESTLAW) search of Antitrust and TradeRegulation Cases (Data base: FATR-CS) before the U.S. Supreme Court,U.S. Court of Appeals, and U.S. District Couns was conducted. Initial re-view of key case.s involving predatory pricing aided the development of thesearch terms employed for the analysis. The search wa.s conducted on De-cember 6. 1994. Overall. 787 cases were idenilfied. Each ca.se was exam-ined to ensure its proper reference.

regarding the actual frequency of predatory pricing, thesefindings parallel results found by others, which suggest thatconcem for these practices exists on the part of attomeys(Beckenstein and Gabel 1983), corporations (Beckensteinand Gabel 1986), and states (i.e.. State Attorneys Generals).

Emerging Perspectives^Current legal assessments of the antitrust effects of predato-ry pricing, as reflected in the Supreme Court's decision inBrooke Group (1993) and the sentiments expressed by boththe DOJ and FTC, rely heavily on neoclassical price theo-ry—a static model of economic behavior in which partiesare assumed to possess perfect information regarding eachother's behavior.

The dominant form of predation considered according tothis model is of a firm pricing merchandise at below-costlevels to promote the exit of a competitor, with the intent ofraising its prices after the competitor leaves the market. Al-though yielding insights for predation. the model's limitingassumptions restrict its inclusiveness for other forms of he-havior that might be considered predatory, involve price, andpossess negative implications for consumer welfare. Guidedby this understanding, several theorists have recently at-tempted to examine the nature and plausibility of predatorypricing under less constrained circumstances than were pre-viously assumed. In industrial organization theory, this bodyof growing scholarship is known as part of the new learningof industrial organization (see Schmalensee 1982). This per-spective has also been investigated and advanced hy mar-keting scholars (Heil and Langvardt 1994).

In contrast to previous focus on structure and methods ofstatic analysis for understanding Ihe behavior of market ac-tors, the new leaming approach emphasizes the evolving,strategic, and dynamic nature of economic behavior. It in-corporates the richness of microeconomic theory with dif-fering models of imperfect information and competition,and cooperative, as well as noncooperative. game theory todevelop a range of models that characterize competitive con-duct—including predatory pricing (Schmalensee and Willig1989). Through the approach's methodologies, the conceptsof purpose and intent are embedded in these models andstrategy decision sets and information available to presentand future market participants are identified explicitly (Or-dover and Saloner 1989).

The product of new leaming is the conclusion that wel-fare-reducing, aggressive, and exclusionary conduct is morelikely and potentially more rational than was indicatedthrough the application of classical models of cotnpetitionand monopoly. In addition, because of its methodologiesthis conduct is able to be more precisely characterized thanthrough prior methods of static analysis.

For understanding predatory pricing, the new leaming ap-proach yields considerable insight pertaining to the natureand feasibility of this practice for the aggressive competitor.A variety of models that characterize predation and theirwelfare effects have been developed. These models go be-yond suggesting that the rationality of predatory pricing is

^Emerging, here, refers to the comparative quality of new insights re-garding predatory pricing relative to extant neoclas.sical price theory.

282 Price Predation

lowering prices to below-cost levels and then subsequentlyattempting to recoup losses through monopoly profits (e.g.,the current view) to embrace altemative conceptions of pre-dation involving price.^

Indeed, many of the predatory behaviors identified inthese models do not involve prices that are below marginalcost. The models only require that a predator incur lowerprofits, and not necessarily losses, from its predatory strate-gies. As Baker (1994, p. 590) notes, a firm can deter aggres-sive competition with a low price, even if the low price ex-ceeds the price-cutter"s average cost, as long as the price issufficiently lower than its rival's costs.

Although debate may center on the desirability of estab-lishing an above-cost standard because of the potential offalse positives,** the importance of these models is to showthat predatory pricing can make excellent theoretical senseand that a criteria of below-cost pricing as an antitrust stan-dard may not address the full range of predatory possibilitiesinvolving price. The incorporation of these models in cur-rent antitrust analysis of predation, however, remains to heseen.^

Predatory Pricing ModelsThe common perspective underlying the new leaming mod-els is their realization of the informational asymmetries pre-sent in most markets and their consideration of the dynam-ics of strategic hehavior among actors in these markets. Thepresence of informational asymmetries enables a predatorfirm to influence the beliefs and expectations of a rival insuch a way (e.g., entry, exit, setting price, determining out-put) that a predatory outcome results.

Becau.se of price's relative ease of change compared toother marketing mix variables; its ability to be fine-tuned todifferent regions, channels, or segments; and its ease in re-versing changes, price provides one mechanism for a fimi toachieve predatory outcomes (Heil and Langvardt 1994).Furthermore, the added dynamic dimension of these modelstakes into account the understanding that the prospect of areputation for predatory behavior could affect other rivals inother markets and across time. These include predatory out-comes extending from signaling, signal-jamming., and repu-tation models.^

'Regarding the "traditional" view of predation as a predator pricingbelow cost in an effort to drive out a rival, il" imperfections in the market forcapital cause the rival lo have relatively less access to financial capital (bothinternally and externally) and entry barriers exist in the preyed upon mar-ket, a predator may reasonably be expected to be capable of using its "deeppiK-kets" and greater access to capital to drive out its rival.

*A false positive diagnosis of predatory pricing occurs when a test ac-cuses a finii of predatory pricing when in fact there is nnne. A false nega-tive occurs when a test exonerates a Firm engaging in predation.

''Aspects of the new learning approach have been applied acro.sK someareas of antitrust (for a review of marketing related applications, see Heiland Langvardt 1994).

''Other models that rely on experience-curve pricing and the dynamics ofdemand suggest that under the classic below-cost pricitig model of preda-tion. losses incurred may be recovered in ways that do not require supra-competitive pricing. The anomaly of the.se models shows that under partic-ular circumstances, some cla.ssic predatory price conduct may actually bewelfare enhancing. For a discussion of these models, see Handler et al.(1990). My subsequent discussion extends Milgrom and Roberts's (1990)research.

Signaling ModelsThe signal-based models of predatory pricing rely on apredator being better informed than its "prey" (i.e., rival)about a market or demand characteristic related to price andrelevant to the prey's market entry and output decisions. Avariety of markei and firm characteristics provide this po-tential, such as different knowledge regarding demand andmarket characteristics, technology, and production.

At a basic level, a firm's costs yield an excellent illustra-tion. This information is often private and is. thus, heldasymmetrically among fimis. Moreover, cost information isof considerable value to competing firms for predicting arival's strategy, as well as its potential response to the firm'sown strategies. In this way, competing firm's possess signif-icant incentives for underslanding each other's costs. Giventhese incentives, if a rival fimi were lo successfully alter thecompetitor's knowledge of these costs, it stands the poten-tial of influencing that competitor's hehavior.

One way to accomplish this outcome is through pricingstrategies. Price is one observable indicator of a finn's costs,because these variables often correlate. Setting a low (orhigh) price often suggests associated low (or high) costs.When a firm sets a price it would not otherwise have set (ei-ther high or low), the rival's inferences regarding the finn'scosts may he biased. For example, if a finn sets a lower pricethan normal, a competitor attempting to make strategic in-ferences regarding the firm's costs may bias its estimatesdownward.^ Inferring lower costs, the competitor may con-clude the firm to be a much more efficient firm and thereforetougher competitor. This belief may alter the finn's strate-gies for entering a market. Predatory outcomes can arisewhen these beliefs result in equally or more efficient firmsaltering their behavior in ways that erode consumer welfare.

Consider a hypothetical example involving a small retail-er pursuing entry in a market occupied by a larger more re-sourceful retail rival. In an attempt to influence the small re-tailer's inferences regarding the larger rival's costs, the rivalmay price its products accordingly. It is not difficult to imag-ine a scenario in which the smaller retailer, on the hasis ofsuch conduct, concludes the larger retailer to be a more effi-cient rival (even if it is not) and as a result decides not toenter the market. If the smaller firm is a more efficient com-petitor, its deterrence from entering the market can result inreduced consumer welfare.

The prospect of infomiational asymmetries being thusemployed is plausible when a person considers the rangeand circumstances in which these conditions manifest them-selves. Consider, for example, the asymmetries that exist tonew or recent entrants in a market, in which firms adopt newtechnologies, open production facilities, change manage-ment/ownership, and incur changing market conditions. In-fomiational asymmetries may exist simply by virtue of mar-ket research, multiple market circumstances, market size,and strategy. When these asymmetries involve price related

'An imponant caveat to this assertion is that under a rational expecta-tions equilibrium this bias cannot occur, because in equilibria, no parties*beliefs can be systematically bia.sed. even if another party (i.e.. u predatitr)is attempting to act in ways designed to create such bias, tJnder these strictconditions, both the signaling and signal-jamming models of predation arerestricted to those circumstances surrotinding potential market entt^.

Journal of Public Policy & Marketing 283

variables, they can be employed to alter the perceptions ofcompeting firms' costs and potentially result in predatoryoutcomes. When these outcomes result in reductions to con-sumer welfare, they become antitrust concems.

Various extensions of the signaling models have been de-veloped. Milgrom and Roberts (1982a) first developed thenotion of signaling as a theory of limit pricing, that is. amethod of deterring entry through biasing prospective en-trant's cost estimates. Roberts (1986) has also developed amodel involving the bias of demand estimates to yield a the-ory of predation aimed at inducing exit or, failing this, re-straining a rival's future output. In this model, a more in-formed predator attempts to make demand appear weak inan effort lo induce a rival to exit the market or lower its lev-els of output. Extensions of the basic model have also beenemployed to show how firms can be induced to "soften up"a merger candidate through biasing the firm's estimates ofits profits, which, in tum, lead to adjustments in the terms ofthe acquisition (Saloner 1987).

In each of the previous cases, the predatory outcomes thatcan result are not dependent on lowering price to below-costlevels as required under the current antitrust standard. Allthat is needed is an informational bias to occur that resultsin a less infonned firm altering its conduct. If that conductresults in diminished consumer welfare, anticompetitivepredation has occurred. Although these outcomes can alsooccur for below-cost pricing, the signaling models show thatthese outcomes can occur at other than below-cost pricelevels.

Signal-Jamming ModelsThe signal-jamming models of predation incorporate manyof the same aspects of the signaling approaches. In thesemodels, however, the asymmetry of information does not re-late to a strategic variable of interest, such as costs or de-mand, but to the (un)observahility of the predator's actionsregarding one of these variables.

For example, a strategic variable of competitive interest isa firm's costs. When a predator firm engages in conduct thatalters perceptions regarding its costs without a competitor'sknowledge or observation of these actions, the firm can ob-tain information asymmetries. The unobservability assump-tion enables the predator to influence the information pos-sessed by its prey and potentially its behavior. In addition,the predator and prey are assumed to possess asymmetries inposition; that is, one is contemplating exit from or entry intoa market: the other is an incumbent.

A marketing related example of signal jamming involvesa firm test-marketing a product in a predator's market. Thatan incumbent finn pos.sesses an asymmetric advantage ofposition and there exist infonnation asymmetries (in tennsof predator actions) under such conditions is plausible.Under these circumstances, a predator firm may engage inactions unobservable to the firm test-marketing the product,which can cause the finn to underestimate the profitabilityof the new product. This can occur when the predatory firmprovides price breaks or other actions (e.g., information,product preannouncements. incentive promotions) to itsmarkets sub ro.sa. which disrupt the finn's test market re-sults. If severe enough, the firm may decide not to enter the

market. In cases in which the entering firm is as efficient ormore efficient than the incumbent, welfare reducing preda-tion may occur.

As with the signaling models of predation, the signal-jam-ming models neither require nor depend on below-cost pric-ing. All that is needed is conduct unobservable to a prey oc-cupying an asymmetrical position that results in the finn al-tering its conduct. If this occurs to the diminution of con-sumer welfare, this form of predation becomes of antitrustconcem.

Reputation ModelsThe reputation-based models of predation extend the singlepredator, single prey models to include multiprey consider-ations to achieve a reputational effect (Posner 1976; Scher-er 1980; Williamson 1972. 1977). In these models, a firmoperating in several markets attempts to prey upon early en-trants by whatever means is available (e.g.. signaling, signal-jamming, below-cost) to establish a predatory reputation(Kreps and Wilson 1982; Milgrom and Roberts 1982b). Thisreputation, in tum, deters other potential entrants, because itleads them to believe that their entry will meet the samepredatory response (Masson and Reynolds 1980), Baker(1994, p. 590) provides an example of multimarket preda-tion in which informational asymmetries are present:

Suppose a chain store faces a non-chain rival In each of a laigenumber of |distant] towns. The chain cuts ils prices dnisticallyin a few towns. When the chain's rivals in those towns either exitor begin to compete less aggressively with the chain, the pricewar ends and high prices are restored. In addition, the chainstore's rivals in all ihc other towns, in which the chain did notcut prices, also respond by avoiding aggressive competition withthe chain. As a resuh, prices also increase in the towns in whichpredation did not occur.

Informational asymmetries within the reputation modelsare similar to those found within the signaling and signal-jamming models. Preying in this context may be worthwhilein a dynamic sense, even when losses are incurred over theshort-term. These profits derive from the economic effectsobtainable across time and other preyed upon fimis. Beyondthese outcomes, a firm may employ these repuiational prac-tices to keep rivals from breaking away or follow a particu-lar price system or other tacitly agreed-to policy (McCal!1987). Such "disciplining" may be enough to keep rivals "inline" and maintain price levels at other than competitive lev-els, thereby disadvantaging consumers.

A key aspect of these models is the necessity of reputationbeing establishable. This prospect is most likely when firmsare operating within multimarket situations in which effectsare easily observable to other firms. The greater the numberof markets involved also tends to be a factor in this form ofpredation. Reputational effects might achieve economies ofscale in this context.

Similar to both the signal and signal-jamming models, thereputation models do not rely solely on pricing at somebelow-cost level for their outcomes to he realized. In thisway, they contrast markedly from the traditional forms ofpredatory pricing considered in antitrust. Together, the im-plication of the newer predation models is that predatorypricing, which involves imposing losses on a current rivalthrough below-cost pricing in hopes of subsequently re-

284 Price Predation

couping losses (i.e., cta.ssic predatory pricing), is the onlymodel of predation in which below-cost pricing plays anycentral role (Milgrom and Roberts 1990). Thus, such a stan-dard for detecting predatory pricing is of limited scope inobserving altemalive predatory strategies.

However, there is no consensus on what standard shouldbe applied for observing these varying strategies of preda-tion. Because of the potential of mistaken inferences ex-tending from evaluating price reductions by competitors(i.e., false positives), considerable debate centers on thisissue. In this respect, some scholars of the new leaming ap-proach, while acknowledging the social costs of such prac-tices, suggest "it may he hest simply to give up on attemptsto control predation, even if one believes that it can and doesoccur." until such prohlems can he addressed through con-tinued research (Milgrom and Roberts 1990. p. 134). Fornow. bowever, the importance of the new leaming modelsresides in their dehunking previously held beliefs about theimplausibility of predation through identification of a vari-ety of predatory actions that firms can take through otherthan below-cost pricing.

Antitrust ConsiderationsCunent assessments of predatory pricing, as outlined by theSupreme Court in Brooke Group (1993) employ a bench-mark of below-cost pricing and a reasonable prospect of re-coupment for detecting pricing conduct that may result indiminished consumer welfare. These standards are juxta-posed against the insight offered by the new teaming ap-proach, with considerations for antitru.st policy discussed.

Injury From a RivaPs Price Below CostAlthough the Court in Brooke Group (1993) did not an-nounce the appropriate measure of cost to be employed inassessing claims uf predatory pricing, litigants in the caseagreed on the use of average variable cost. This measure ofbelow-cost pricing has been employed extensively in theCourts since first heing proposed by Areeda and Tumer(1975). According to these authors, average variable costrepresents a surrogate measure of marginal-cost pricing as"the economically sound division between acceptable com-petitive behavior and *below-cost' predation" (p. 716).

Adopting a perspective of predaiion as that of a rival al-tempting to impose severe losses on a competitor throughbelow-cost pricing in hopes of subsequently recouping theselosses, Areeda and Tumer believe rules against predatorypricing should be underinclusive, rather than overinclusive,to avoid deterring legitimate, competitive pricing. Accord-ing to the authors (p. 699), "the prospect of an adequate fu-ture payoff should "seldom he sufficient to motivate preda-tion." As a result, their test bears an inherent skepticism to-ward predatory claims.

In this context, Areeda and Tumer (1975, p. 733) observethat "a price at or above reasonably anticipated [marginalcost] should be conclusively presumed lawful." Further-more, "|a| price below reasonably anticipated [marginalcost] should be conclusively presumed unlawful" (p. 733)(subsequently amended to "rebuttable presumption" [Aree-da and Tumer 1978]). Because marginal cost is difficult todetermine, average variable cost is suggested as a surrogate-

Viewed against the insight of the new leaming approach,a standard of below-cost pricing is indeterminate in its abil-ity to detect pricing behavior that is predatory and injuriousto consumer welfare. Such a standard is at once too permis-sive of some forms of predatory conduct as identified by thesignal, signal-jamming, and reputation models, and issimultaneously overly restrictive of other forms of pricingconduct.

From a markeling standpoint, this last point can be illus-trated through acknowledging the multitude of nonpredato-ry motivations underlying a pricing strategy that may resultin a price below (or above) cost. Consider, for example, thatsome suppliers might price their products (e.g., sparkplugs)to manufacturers (i.e., automakers) below cost to attractbuyers of replacement parts (i.e., aftermarket spark plugs) ifmany replacement buyers continue to purchase the originalbrand even when they are charged a higher price (see StittSpark Plug Co. v. Champion Spark Plug Co. 1988). Pricingat below-cost in such a case involves not a predatory intent,but a desire to maintain long-term customer relationships.

Similariy, consider that competing copier manufacturersmight price their copiers below cost to attract buyers whocan liiter he charged higher prices for parts and services. Or,competing firms producing hoth cameras and film might setcamera prices below cost to sell more film at high prices.Competing sellers of some types of computer softwaremight price their product below cost when high switchingcosts inhibit migration of the installed customer base to rivalsoftware, to increase the number of customers to whom theycan subsequently sell high-margin upgrades. Competingtruckers might also price certain freight movements belowtheir stand-alone cost when those movements pcmiit thetrucker to obtain high-margin back-haul business. Finally,competing computer manufacturers, or those of any highvolume product, might price a new product below its initialvariable cost to generate, through increased production andsales, the scale economies and cost reductions from the ex-perience curve that would justify the low price. In each ofthese instances, the reduction of below-cost prices may re-flect a procompetilive marketing strategy rather than preda-tion (Baker 1994).

Although it provides a '•bright-line" from which to assesspricing conduct. Areeda and Tumer's( 1975, 1976) approachhas shortfalls that extend from its focus on short-temi effi-ciency versus long-term considerations and inherent prob-lems in employing average variable cost as a surrogate formarginal cost (tbese two constructs equal one another atonly one point—minimum average variable cost). Imple-mentation issues in the calculation of variable versus fixedcosts in detennining average variable costs are also of con-cem. Set standards for the allocation of costs to either cate-gory are not uniformly followed and may vary (Shimer1981; tor equipment depreciation as fixed or variable cost,see Kelco Disposal. Inc. v. Browning-Ferris Industries, Inc.1988).

A key aspect of the Court's holding in Brooke Group(1993) is its addressing the below-cost standard. In declin-ing to adopt explicitly average variable cost as ils standardand. instead, relying on an appropriate measure of cost asits benchmark, the Court appears to have left open the

Joumal of Public Policy & Marketing 285

prospect of further clarifying this standard, ln this respect,incorporation of knowledge gamered from the new leamingapproach would seem particularly important for thisrefinement.

That predation may occur at both above- and below-costlevels, combined with the assumed potential of false posi-tives being an above-cost standard set, a low-cost standardmay seem justified. However, such a conclusion implies anempirical judgment regarding the relative incidence andharm of false positives and false negatives. In the eventpredatory conduct, through actual empirical study, werefound to be more common and costly than currently as-sumed, a more inclusive standard may he desirable. To besure, it is too early to tell whether instances of predafionoccur with greater frequency than is assumed under the cur-rent standard. The answer to this question ntust await em-pirical study and application of new leaming insight inlitigation.

Another and equally important aspect of the Court's hold-ing is its declaration that though a standard of below-costpricing is a requirement for showing competitive injury, it isnot the determining factor (Brooke Group 1993, pp.4702-3). Instead, the Court maintains that a reasonableprospect of recoupment is the dispositive factor in which tomeasure injury to competition. By relegating the below-costcriteria to .secondary status, the Court further distances itselffrom previous reliance on this standard (Elzinga and Mills1994). This shift appears, at least indirectly, to acknowledgethe emerging insight of the new leaming approach, whichshows such a criteria to be indeterminate for some forms ofpredation.

Together, Ihe Court's refusal to announce an appropriatemeasure of cost (if intended) and its relegation of the below-cost standard to secondary status suggests its acknowledg-ment of the anticompetitive aspect of predatory pricing de-riving from the injury to competition that results from afirm's subsequent recoupment rather than from the initial actof lowering prices. In other words, the decision in BrookeGroup (1993) indicates that the Court is no longer con-cemed with low prices in and of themselves, but rather, withthe possibility of subsequent monopoly profits that may beanticompetitive.

This shift has dual implications in the antitrust treatmentof predation. First, it enables the Court potentially to assessthe many variant forms of pricing conduct identified underthe new leaming approach and described by others, whichmay result in anticompetitive consequences. Second, it ben-efits marketers in their pricing decisions by not prescribinga particular form of pricing conduct to which managers mustconfonn. As to this last implication, the Court's shift seemsto address the concem noted in Matsushita (1986, p. 594),"lest a rule or precedent that authorizes a search for a par-ticular type of undesirable pricing behavior end up [by] dis-couraging legitimate price competition." A caveat to this dis-cussion, however, is that the Court may have simply de-clined to address the appropriate measure of cost to be em-ployed in evaluating the claim of predation, because the lit-igants agreed on the use of average variable cost. It is possi-ble to interpret the Court's holding as accepting the use ofaverage variahle cost as the appropriate measure of cost.

Dangerous Probability of RecoupmentStatements by ihe Supreme Court in Matsushita (1986) andCargill (1986) elude to the eventual standard adopted inBrooke Group (1993), which requires a reasonahle expecta-tion, based on market circumstances of recovering lossesthrough supracompetiiive pricing, to sustain a predatorypricing claim, This requirement was fomially adopted inBrooke Group as the second part of its two-prong standard.Quoting their earlier decision in Matsushita (p. 590-91), theCourt states:

...iti order to recoup their losses, [predators! must obtain enoughmarket power to set higher than competitive prices, and thenmust sustain those prices long enough lo eam in excess profitswhat they earlier gave up in below-cost prices {Brooke Group1993. p. 4703).

The Court observes that a i^easonable prospect of recoup-ment is the detemiinative factor in assessing injury to com-petition from predatory pricing conduct.

The insight provided by the recoupment standard is thatlosses by a predator, in whatever amount, must be recoveredfor predation to make sense and for injury to competition(defined in terms of efficiency) to occur. In this way, re-coupment is analogous to competitive injury. Without somefomi of recoupment, price cutting hy an alleged predatorbenefits consumers. With an expansive definition, a slandardof recoupment applies to those models developed under thenew leaming approach. In other words, though a key contri-bution of these models is the idea that predatory outcomescan be achieved through incurring more marginal lossesthan those resulting from low-cost pricing, these losses (orlost profits) must still he recovered in some economic fash-ion for predation to be a plausible strategy.

The insight to be gained from the newer models, howev-er, is that the amount necessary to be recouped may be muchless than previously thought. Moreover, compared to the po-tential benefits of predation. these losses may be low enoughto suggest the plausibility of this practice as a rationalstrategy.

The implication of the new leaming approach is thai thejudicial criteria set forth in Brooke Group (1993) for estab-lishing recoupmenl. namely, high market share and highentry barriers, though applicable to predation in a classicsense (e.g., steep price cuts below cost), may not be neces-sary for the anticompetitive effects of other forms of preda-tory pricing conduct to manifest themselves (Rapp 1991).As Goldstein (1991, p. 1772) suggests.

Even firms that have no chance to obtain real monopoly power[i.e., do not possess the requisite high tnarkei sbare and entrybarriersi might find certain predalory tactics valuable tor tbeirreputalional effects or as a way to deter new entrants lo the mar-ket wbo migbi become successful competitors, because sucbtactics migbt result in future gains to the predator greater thanthe losses incurred....

Overemphasis on whether conditions exist for obtainingmarket power enables some firms to engage in predatoryconduct that might not result in such power, but neverthelessbe hannful to competition and consumer welfare.

An important aspect of the Court's holding is their re-fusal, as a matter of law. to rule out the possibility that re-coupment could take place tbrough supracompetitive

286 Price Predation

oligopoly prices. Moreover, the Court accepted the possibil-ity of recoupment through coordinated oligopoly pricing, lnthis resprect, the Court appears to have acknowledged thatpredalory pricing can occur under the appropriate factualsettings prescribed by the new leaming approach.

However, undermining this acknowledgment to some de-gree is the high threshold of proof the Court requires toshow recoupment. The Court's holding in Brooke Group, ineffect, requires proof surpassing the reasonable prospect anddangerous probability elements established legislatively forthe Robinson-Patman and Sherman Acts. According to thethree dissenting Justices. "[T|he Court's [holding of the ma-jority] most significant error is Ihe assumption that... Liggethad the hurden of proving either the actuality of supracom-petitive pricing, or the actuality of tacit collusion [in 'preda-toriaily' setfing price]" (p. 4712).

Establishment of this criteria requires that a predatoryepisode be nearly complete before satisfying the necessaryburden of proof. Thus, the element further amplifies the em-phasis on evaluating the conditions in existence for estab-lishing the probability (or actuality) of obtaining marketpower

Implications and DiscussionJuxtaposing the theoretical potential of anticompetitive con-duct involving predatory pricing by rational firms againstthe current judicial standards (in particular, ihe recent testdeveloped in Brooke Group), I examine the prevailing stan-dard for predatory pricing. This analysis suggests that theCourt's new standard embraces a core theoretical tenet ofthe new learning approach to predatory pricing; that is, astandard of below marginal cost is indeterminate regardinganticompetitive predatory pricing. The Coun also appears tohave advanced beyond prior antitrust thinking in its require-ment of and emphasis on a showing of recoupment prior todetennining that pricing conduct is predatory. In theory, re-coupment represents a necessary outcome for predatorypractices to be rational.

Overemphasis on conditions necessary for recoupment tooccur (i.e., market power) and establishment of an extreme-ly high burden of proof for its finding, however, hinders theeffectiveness of the Court's holding for some victims of pre-dation. Thus, the Court's finding in Brooke Group representsonly an initial step toward embracing emergent thinking re-garding predatory pricing.

AntitrustVarious scholars have suggested supplementing the extantapplication of economic analysis in antitrust with other per-spectives to improve its explanatory power and enhance itsnomiative prescriptions. For example, Williamson (1979, p.991) observes, "[A|nlitrust is an interdisciplinary field thatis best served by acknowledging that a deeper understandingof the issues will result by addressing the subject from sev-eral points of view." Others have also suggested that eco-nomic analysis in law could benefit from knowledge gar-nered within other branches of the social sciences (see, e.g.,EUickson 1989).

Judicial acknowledgment of the benefits of supplement-ing cunent analysis with altemative insights as offered hy

the new leaming approach and other perspectives is de-tectable in at least one recent holding of the Supreme Court.In Eastman Kodak Cotnpany v. Image Technical Services(1992), a doctrinal .shift in how economic analysis is em-ployed in antitrust cases is identifiable. By refusing to up-hold a lower court's grant of Summary Judgment, ihe Courtheld that an eariier determination of an importanl issue (i.e.,market power) based on mere economic presumption wasimproper. In this way, the Court shifted away from its use ofeconomic analysis as a basis for delennining the economicplausibility or implausibiiity of an evenl—and, therefore,the presumptions regarding its occurrence—and toward theu.se of economic theory as only one tool among others forexplaining market realities.

The Couri's requirement that the theories brought beforeit should reflect accurately business reality suggests its will-ingness to consider perspectives that provide an enrichedview of the marketplace (Judson 1993). Those models pro-vided hy the new leaming approach yield such a perspective.It remains to be seen, however, how many plaintiffs, prose-cutors, and antitrust jurists will attempt to apply these de-veloping insights.

The eventual insight to be gained from supplementingcurrent analysis of predatory pricing with altemative in-sights is to understand hetter the nature of this practice andthe factors contributing to its occurrence. Only in this waycan an informed antitrust policy be developed. Althoughconsiderable challenge resides in the development of thispolicy, extant approaches, are. in many ways, inconclusivein their ability to detect predatory conduct.

Marketing ResearchAn important implication of my assessment involves the ne-cessity of continuing to develop our understanding of preda-tion through further research. To date, the majority of in-sights regarding the nature of predatory conduct has beenadvanced through economists applying microeconomic the-ory and static-based conceptions of competition. Newer in-sights, illustrated here, but not yet explicitly incorporated incurrent antitrust analysis, have been obtained through con-sideration of asymmetries in information, dynamic competi-tion, and rationality of decision makers.

Such an inquiry, at its core, investigates the business de-cision making involving pricing and other strategies at lowerlevels of abstraction than has been previously conducted.Because the focus of markefing involves strategies of thiskind, as well as inquiry at the level of the finn and its deci-sion makers, theory and research from this discipline shouldhe considered an important resource and a potentially usefulsupplement to current investigations of predation.

Marketing's focus on the nature of pricing strategies andcompetition, as well as its empirical approach to manageri-al issues, could yield considerable insight to predation anal-ysis. For example, Heil and Langvardt (1994) focus on cotn-petitive market signaling and its implications for antitrust.Reviewing an extensive body of research in marketing andstrategic management that shows that managers interpretmarket actions hy others and use these interpretations intheir reactions (e.g., the essence of market signaling), theydetail a variety of implications for antitmst policy. In partic-

Journal of Public Policy & Marketing 287

ular, they provide evidence that "a market action may signalsubstantial aggressiveness and hostility on the part of thesignal-sender" (p. 92). Heil and Langvardt also note the po-tential of relying on this signaling behavior as the founda-tion for an allegation of predatory pricing, but observe theearly stages of such development. In this regard, researchthat provides a precise delineation of which forms of signal-ing under which circumstances are suggestive of predatoryintent would be useful.

Continued research toward improving our understandingof the extent to which market signaling is employed as a de-structive mechanism in competitive interaction would alsobe fruitful. Eventually, research on signaling and predationcould yield a basis for identifying and distinguishing be-tween anticompetitive predatory signaling involving priceand price signaling of a procompetitive nature.

One immediate avenue for extending the insight intopredatory pricing involves further examination of the strin-geiU theoretical assumptions underlying current predatorypricing policy. Although emergent theory, provided by tho.sestudying the new leaming approach, addresses assumptionsregarding the nature of information held by competitors andfavors dynamic analysis over static considerations, both newleaming and extant neoclassical price theories maintain theassumption of strict rationality on the part of rivals. Both ap-proaches assume that actors act calculatedly and rationallyto maximize their expected utility. Those who subscribe tothe neoclassical view argue for the irrational nature ofpredatory conduct. Altematively, new leaming economistssuggest ways in which a firm acting rationally may engagein such conduct.

Both the neoclassical and new leaming perspectives,however, ignore important issues: (1) Firms, as entities, can-not act on their own; (2) any actions by firms are thereforeactions by the human beings who manage tbe firms; and (3)the actions of human beings are not necessarily rational. Ac-cepting this proposition, further insight about predatorypricing may emerge.

Considerable research, both in economics and its relateddisciplines, demonstrates that neither the choices peoplemake nor the ways in which they make them meet the re-quirements of strict rationality. In economics, Simon's(1955, 1956, 1964, 1978) and March and Simon's (1958)works are well-known for their exploration of the ways inwhich rationality as maximization is descriptively inaccu-rate. Indeed, their work resulted in the concept of "'houndedrationality." Considerable social science now focuses on theidea of rationality as a variable concept.

These research traditions (e.g., psychology, organizationtheory, marketing) emphasize the extent to which the deci-sion-making process reflects a desire to make the best deci-sion possihle under specific circumstances, not whether util-ity is maximized. Application of this perspective may yieldadditional insights regarding predatory pricing.

Consideration of nonprofit maximizing behavior in thecontext of predatory conduct has not been totally ignored inantitrust analysis. For example, Sullivan (1977, p. 110)observes.

The fact that predatory activity is costly to the predator and thatihere is only an uncertain prospect of adequate supracompetitive

returns after others are excluded surely must reduce the fre-quency of predatory forays, it hardly follows thai they neveroccur or can be safely ignored. Man's capacity for destructiveconduct has never been totally inhibited merely because hestands himself in the target area along witb his wotild-be victim.

In a more recent work, Gerla (1985) explores the insight thatcan be gained from applying the psychology of risk-takingto the issue of predatory pricing. Gerla demonstraies thatpredatory pricing can he a serious problem when manageri-al hehavior is viewed in light of the psychological theory ofrisk-taking, instead of in terms of extant neoclassical pricetheory.

Similarly, addressing the prospect of less than rational be-havior, a growing number of scholars in marketing are be-ginning to explore the procedural rationality of decisionmaking in firms. Accepting the proposition that managersact in ways other than to maximize utility, these researchershave hegun to examine the extent to which their decisionprocesses involve (I) the collection of information relevantto the decision and (2) their reliance on analysis of this in-formation in making a decision choice.

Such an approach varies distinctively from the assump-tion of decision makers as profit maximizcrs acting abso-lutely rationally and embraces the notion that managers actvariably rationally. A variable conception of rationality en-ables researchers the potential of further extending the cur-rent understanding of predatory pricing. Applying such re-search, for example, could suggest altemative motives un-derlying pricing decisions or provide a basis for understand-ing the circumstances that lead a manager to engage in suchconduct.

Beyond characterizing interfirm pricing decision makingand hehavior and providing explanations for the basis forconduct such as predatory pricing, research in marketing canalso provide insight about the effects of this practice. Al-though marketing has not focused to a great extenl on thewelfare implications of marketing strategies as defined inantitrust law, constructs relating to consumer welfare arecommon to marketing scholarship (e.g., satisfaction,choice). Continued debate centers on the legislative correct-ness or desirability of economic efficiency providing thesole definition of consumer welfare.

Even when using purely economic definitions of con-sumer welfare (i.e., efficiency), marketing scholars possessthe potential of infonning current antitrust analysis. As cur-rently defined, productive and allocative efficiency are mostoften identified as the economic standards against whichwelfare implications for consumers should be judged (Bork1978). Efficiency in this context refers to the maximizationof societal resources through efficient use of inputs (i.e..productive efficiency) and efficient quantity of outputs (i.e.,allocative efficiency). Productive efficiency reflects efficien-cy within a single exchange, market channel, or industry andentails the maximization of input resources and the mini-mization of process costs associated with these resources.High productive efficiency is achieved through producingand distributing a product with a minimum consumption ofresources.

Much of marketing entails development of strategies di-rected at achieving these outcomes. For example, acrossmarketing functions, just-in-time inventory systems, inter-

288 Price Predation

firm partnering relationships, and interfirm influence or con-trol mechanisms are directed at achieving efficiency in thedistribution of goods to consumers. Efficiency-based depen-dent variables examined in marketing research include per-formance indices within the fimi and across exchange rela-tionships (e.g., intra- and interfirm relationships) and relateto productive efficiency.

Allocative efficiency is the efficiency of output across in-dustries and involves the question of whether societal re-sources should be invested in one industry as opposed to an-other. Allocative efficiency is a measure of the success of aneconomic system in maximizing consumer preferences fromthe limited resources available.

Marketing has not examined the question of allocative ef-ficiency directly, because its unit of analysis has emphasizedthe dyad. Recently, however, scholars have begun to expandtheir focus beyond dyadic relationships and have addressedissues underlying broader networks of exchange relation-ships (e.g., Iacobucci and Hopkins 1992). In this respect,market channel relationships have been viewed as valuechains or systems of interrelated organizations involved inmarketing to end-users. Truer network conceptualizationshave also been advanced. Tbis expansion holds the potentialfor examining the relative efficiency of differing exchangesystems and, thus, addressing issues regarding allocativeefficiency.

Marketing PractitionersFor marketing managers engaged in pricing decisions, cur-rent judicial and antitrust enforcement standards suggestthat aggressive pricing strategies—including those involvingbelow-cost pricing—will he considered violative of the an-titrust laws only under limited situations (e.g., those that sat-isfy the Supreme Coun's two-part test). Firms contemplat-ing such strategies in circumstances in which the probabili-ty of recoupment is not a reasonable prospect or a dangerousprobability (according to the Court's test) do not violate thelaw. Of course, before their implementation, such strategiesshould be scrutinized at length for their individual antitrustimplications.

Marketers who find themselves the target of predatorypricing possess limited judicial recourse in the absence ofcircumstances that meet the standards set forth by theSupreme Court. These standards reflect a suspicious posturetoward predatory claims. A review of cases that have beenbrought before the court suggest few circumstances in whichvictims of predation have been successful in their claims.

ConclusionPredatory pricing represents a particularly troublesome an-titrust concept. On the one hand, a rival's lowering of pricecan enhance consumer welfare. On the other hand, such be-havior can also disadvantage consumers under particularcircumstances. Extant antitrust thinking, driven primarily byneoclassical price theory, envisions predatory pricing as cut-ting prices to below-cost levels in hopes of driving a rivalfrom the market and subsequently raising prices to supra-competitive levels. Those who still view predation as suchcontend that because it does not provide advantages to afirm, given the extreme losses that must be incurred, it is ra-

tionally implausible and therefore rare. Emergent theory,through inclusion of additional factors and altemative per-spectives, however, indicates that the plausibility of predato-ry strategies may be greater than once thought. Incorporat-ing strategic considerations and relaxing prior assumptionsregarding perfect infonnation among competitors, a varietyof models of predation have been developed that do not de-pend on extreme below-cost pricing to achieve their anti-competitive outcomes. At its center, this paradigm empha-sizes the imperfect infomtation and dynamic qualities ofeconomic interaction. In this way, the new leaming ap-proach provides an enriched framework for understandingpredatory practices.

My analysis indicates the necessity of further inquiry andresearch into this perspective to aid in the development of amore informed antitrust policy. In this respect, the market-ing discipline should be considered a potential source ofknowledge and research for furthering our understanding ofthis phenomenon and for developing such a policy,

ReferencesAmerican Prugs, Inc. v. Wal-Mart Stores, Inc. (1993), 62 USLW

2279. I993-2Trade, 70, 382.

Antitrust & Trade Regulation Rcptm (1987), "Oliver RecommendsAgainst Changes in Resale Price Maintenance Standard," TheBureau of National Affairs. 1329 (August 20), 340.

(1989), "Practitioners Must Sift Conflicting Trends in Fed-eral, State, and Foreign Enforcement," The Huremi of NationalAffairs. 1400 (Jantiary 26), 127.

(1993), "Senate Commerce Committee Cleans FTC Reau-thorization Legislation." The Bureau of National Affairs. 1626(August 5), 197.

Areeda. Phillip and Donald F. Tumer (1975), "Predatory Pricingand Related Practices Under Section 2 of the Sherman Act," Har-viint Law Review. 88 (Febrxiaiy), 697-732.

and (1978), Antitrust Law. Boston: Little, Brownand Co.

Baker, Jonathan B. (1994). "Predatory Pricing After Brooke Group:An Economic Perspective," Antitrust Law Joumal, 62 (Spring),585-603.

Barry Wrif-ht Corp. v. TTTGrinnell Corp. (1983), 724 F. 2d 227 (1stCir. 1983).

Beckenstein, Alan R. and H. Landis Gabel (1983), "The Economicsof Antitrust Compliance," Southem Economic Joumal, 52 (3),673-92.

and (1986), "Predation Rules: An Economic andBehavioral Analysis," The Antitrust Bulletin, 31 (I), 29-49.

Bork. Roben H. (1978), The Antitrust Paradox. New York: BasicBooks.

Brooke Group Ud. v. Brown & Williamson Tobacco Corp. (1993),61 L.W. 4699.

Campbell, Thomas J. (1987). "Predation and Competition in An-titru.st: The Case of Nonlundable Goods." Columbia Uiw Review,87(8), 1625-75.

Cargiti Inc. v. Monfon of Colorado, Inc. (1986). 479 U.S. 104.

Clayton Act (1914), 15 U.S.C.A § 12-27.

Eastman Kodak Co. v. Image Technical Services, Inc. (1992), 112 S.Ct. 2072.

EUickson, Robert C. (1989). "Bringing Culture and Human Frailtyto Rational Actors: A Critique of Classical Law and Economics,"

Journal of Public Policy & Marketing 289

Chicago-Kent Law Review, 65 (Winter), 22-55.Elzinga, Kenneth G. and David E. Mills (1994), "Trumping the

Areeda-Tumer Test: The Recoupment Standard in Brooke^ An-titrust Law Joumal. 62 (Spring), 559-84.

FederalTradeCommis.sion Act (1914), 15 U.S.C.A. §41-58.

Gerla, H. (1985), "The Psychology of Predatory Pricing: WhyPredatory Pricing Pays." Southwestern Law Joumal, 39 (Septem-ber). 755-80.

Glazer, Kenneth (1994), "Predatory Pricing and Beyond: Life AfterBrooke Group," Antitrust Law Joumal. 62 (Spring), 605-33.

Goldstein. Jessica L. (1991). "Single Firm Predatory Pricing in An-titrust Law: The Rose Acre Recoupment Test and the Search foran Appropriate Judicial Standard," Columbia Law Review, 91 (7),1757-92.

Gundlach, Gregory T. (1990), "Predatory Practices in CompetitiveInteraction: Legal Limits and Antitrust Considerations," Joumalof Public Policy & Marketing, 9, 129-53.

Handler, Milton. Harlan M. Blake, Robert Pitofsky, and Harvey J.Goldschmid (1990), Cases and Materials on Trade Regulation.Mineola. NY: Foundation Press.

Hay, George A. (1990). "Predatory Pricing." Antitrust Law Joumal,58 (Fall). 913-19,

Haynes, William J. (1990). State Antitru.st Laws. New York: BNA.Heil. Oliver P. and Arlen W. Langvardt (1994), "The Interface Be-

tween Competitive Market Signaling and Antitrust Law." Joumatof Marketing, 58 (July), 81-96.

Iacobucci. Dawn and Nigel Hopkins (1992). "Modeling Dyadic In-teractions and Networks in Marketing." Joumal of Marketing Re-search, 29 (February), 5-1.

Judson, Lisa M. (1993). "Kodak v. Image Technical Services: TheTaming of Matsushita and the Chicago School," Wisconsin LawReview, (November/December), 1633-73.

Kelco Disposal, Inc. v. Browning-Ferris Industrie.^, Inc. (1988), 845F2d 404 (2d Cir), affd on other grounds. 109 S.Ct. 2909 (1989).

Kotler Phillip and Ravi Singh (1981), "Marketing Warfare in thei9%0s," Journal of Business Strategy, 1 (Winter), 30-41.

Kreps, D. and Wilson, R. (1982). "Reputation and Imperfect Infor-mation," Joumal of Economic Theory. 27 (August). 253-79.

McCall, Charies W. (1987), "Predatory Pricing: An Economic andLegal Analysis," Antitrust Bulletin, 32 (Spring), 1-59.

March, J, G. and Simon, H. A., (1958). Organizations. New York:John Wiley & Son.s.

Masson R. and R. Reynolds (1980), "Predation: The 'Noisy' Pric-ing Strategy." U.S. Department of Justice.

Matsu.shita Electric Industrial Corp. v. Zenith Radio Corp.(\986),475 U.S. 574.

Milgrom. Paul and John Roberts (1982a), "Limit Pricing and EntryUnder Incomplete Informarion: An Equilibrium Analysis,"Econometrica, 50 (March). 443-59.

and (1982b). "Predation, Reputation and Entry De-terrence," Joumal of Economic Theory, 27 (August). 280-312.

and (1990), "New Theories of Predatory Pricing,"in Industrial Structure in the New hulu.striat Economics, Giaco-mo Bonanno and Dario BrandoUni, eds. Oxford: ClarendonPress. 112-37.

Ordover, J. A. and Saloner, G. (1989), "Predation, Monopolizationand Antitrust." Handbook of Industriat Organization, in R.Schmalensee and R. Willig. eds. New York: Elsevier SciencePublishing Co., 537-89.

Posner, R. (1976). Antitru.st Law: An Economic Perspective. Chica-go: University of Chicago Press.

Quinn, Michael C. (1990). "Predatory Pricing Strategies: The Rele-

vance of Intent Under Antitrust. Unfair Competition, and TortLaw." Sr. John's Law Review, 64 (Spring/Summer). 607-28.

Rapp, Richard T. (1991). "Predatory Pricing Analysis: A PracticalSynthesis." Antitrust Law Joumat, 60 (2), 595-607.

Roberts. J. (1986), "A Signaling Model of Predatory Pricing," Ox-ford Economic Papers, 38 (Supplement), 75-93.

Robinson-Patman Act (1936), 15 U.S.C.A §13.

Rule, Charles F. (1988), "Claims of Predation in a CompetitiveMarketplace: When is an Antitrust Response Appropriate?" An-titru.st Law Journat, 57 (Summer), 421-34.

Saloner, G. (1987), "Predation, Mergers, and Incomplete Informa-tion,"/ia/«/7ourna/o/Economics, 18,165-86.

Scherer. F M. (1980). Industrial Market .Structure and EconomicPerformance, 2d ed. Chicago: Rand McNally Publishing.

Schildkraut, Marc G., Dennis A. Yao, Mary A. Azcuenaga. andRoscoe B. Starek III (1992). "FTC Policy on Predation: No Prob-lem if There is at Least One Surviving Cotnpetitor." AntitrustLaw & Economics Review, 23 (2). 17-32

Schmalensee. Richard (1982), "The New Industrial Organizationand the Economic Analysis of Modem Markets," in Advances inEconomic Theory. W. Hildenbrand, ed. Cambridge: CambridgeUniversity Press, 253-85.

and Robert D. Willig (1989), Handbook of Industrial Or-ganization. New York: El.sevier Science Publishing Co.

Sheffet. Mary Jane (1994), 'The Supreme Court and Predatory Pric-ing," Journal of Public Policy & Marketing, 13(1), 163-67.

Sherman Antitmst Act (1980). 15 U.S.C.A. {?§l-7.

Shimer. Charles P. (1981). "Predatory Pricing: The Retreat from theVAC Rule and the Search for a Practical Altemative," BostonCollege Law Review, 22 (March), 467-510.

Simon, H. A. (1955), "A Behavioral Model of Rational Choice,"Quarterly Journal of Economics, 69. 99-118.

(1964), "On the Concept of the Organizational Goal," Ad-ministrative Science Quarterly, 9 (1) , 1-22.

(1978). "Rationality as Process and Product of Thought,"Joumal of the American Economic A.tsociation, 68 (1), 1-16.

Spectrum Sports. Inc. v. McQuillan, (\993), 506 U.S. 8, slip opin-ion.

Stitt Spark Plug Co. v. Champion Plug Co. (1988), 840 F.2d 1253(5th Cir.). cert, denied. 488 U.S. 890.

Sullivan. Lawrence A. (1977), Handbook of the Law of Antitrust. St.Paul: West Publishing Co.

United States v. Grinnetl Corp. (1966), 384 U.S. 563.

Utah Pie Co. v. Continental Baking Co. (1967), 386 U.S. 685.

Weitz, Barton A. (1985). "Introduction to Special Issue on Compe-tition in Marketing." Joumal of Marketing Research. 11 (Au-gust), 229-36.

Williamson. Oliver (1972), "Dominant Firms and the MonopolyProblem: Market Failure Considerations," Harvard Law Review,85(8), 1512-31.

(1977). "Predatory Pricing: A Strategic Welfare Analysis,"Yale Law Journat, 87 (December), 284-340.

(1979) "Assessing Vertical Market Restrictions: AntitrustRamifications of the Transaction Cost Approach," University ofPennsylvania Law Review, 127 (April), 953-93.

ZeithamI Carl P and Valeria A. ZeithamI (1984), "EnvironmentalManagement: Revising the Marketing Perspective," Joumat ofMarketing, 48 (Spring), 46-53.