federal register - U.S. Government Printing Office

6260 Federal Register / Vol. 62, No. 28 / Tuesday, February 11, 1997 / Noticeshave to account for sufficient totaloutput to give them power over price. 15As far as I can tell, the ‘‘manufacturercartel’’ theory is not relevant to thepresent case. The Commission’sproposed complaint does not allege, letalone provide supporting evidence, thatAmCy has attempted to collude withother agricultural chemical makers,such as DuPont, Monsanto, Ciba-Geigy,or BASF. There is also no evidence thatthese other firms used RPM, as isrequired for the theory to work. Buteven putting aside the absence of suchevidence, it is difficult to imagine anarrangement less suited to cartelstability than that which existedbetween AmCy and its distributors.Specifically, under the terms of AmCy’sC.R.O.P. TM and A.P.E.X. TM programs, adealer’s compensation was tiedexplicitly to the share of chemical salesaccounted for by AmCy’s products.Given that a crucial element of cartelenforcement is the discovery of somemeans by which each member cancommit credibly to maintaining—butnot increasing—its market share, 16 howcould a program that explicitly rewardsmarket share expansion plausibly becharacterized as a cartel enforcementtool?Furthermore, the available evidencesuggests that the C.R.O.P. TM andA.P.E.X. TM programs wereextraordinarily successful in expandingAmCy’s sales and market share, whichgrew substantially while the programwas in use. Certainly, other factors (e.g.,the successful introduction of severalnew product lines) may have accountedfor a portion of this increase; 17nevertheless, it is difficult (if notimpossible) to reconcile the behavior ofAmCy’s output—or of total marketoutput—during this period with anycoherent theory of competitive harminvolving collusion with other chemicalmakers.In the alternative, per se treatmentsometimes is predicated on thecharacterization of RPM as an aid todealer collusion. Under such a scenario,a group of dealers pressures the supplierto adopt RPM to achieve and maintain15 Of course, all of the standard factors used toanalyze market power and the ability to implementand maintain collusive pricing (e.g., ease of entry,heterogeneity of the products, and so forth) wouldalso be relevant to judging the likelihood ofsuccessful supplier collusion.16 As Stigler (supra note 13, at 42) noted, ‘‘[f]ixingmarket shares is probably the most efficient of allmethods of combating secret price reductions.’’17 The likelihood of successfully maintainingcollusion in the face of product innovation (as wasoccurring in this instance) is, of course, quite small.Collusion is more likely to be successful, the greaterthe degree of similarity (e.g., in terms of cost,demand, and product characteristics) among theparties to the agreement.a collusive resale price arrangementamong the dealers. When RPM is usedfor this purpose, we would expect to seecoordinated pressure on themanufacturer to adopt RPM from agroup of dealers with sufficient marketpower to credibly threaten themanufacturer. Moreover, to be effective,the dealer cartel must enter into similararrangements with enoughmanufacturers to be able to affect marketprice; otherwise, the collusive retailprice of price-maintained productswould be undermined by competitionfrom products not subject to RPMagreements. Under such conditions, wewould expect the manufacturer to be areluctant participant in the scheme,though it would enforce the RPMagreement if the dealer threats werecredible. Finally, it is unlikely that thecolluding dealers would carrycompeting products not subject to RPMagreements, as that would be equivalentto cheating on the collusivelydeterminedresale margin.This second anticompetitive theoryfits the facts of this case no better thanthe first. The Commission’s complaintdoes not allege, let alone providesupporting evidence, that AmCy is thevictim of a dealer cartel. As I alreadyhave noted, it does not appear that othermanufacturers had similar arrangementswith the members of any putative‘‘dealer cartel,’’ or that this ‘‘cartel’’eschewed the products of rivalmanufacturers. 18 Had AmCy been thevictim of a cartel, its attitude toward theCommission and numerous stateinvestigations should have been one ofgrateful acquiescence, because theenforcement agencies would be rescuingit from the clutches of its rapaciousdealers. In fact, of course, AmCyunilaterally terminated the challengedprovisions of the C.R.O.P. TM andA.P.E.X. TM programs several years ago.so much for ‘‘dealer coercion.’’ 1918 This is unsurprising, because over 2500 dealersparticipated in the C.R.O.P. TM and A.P.E.X. TMprograms. It is fanciful to believe that a cartel couldhave been formed from among such a large numberof dealers. If such a cartel exists, one mightreasonably ask why the dealers that belong to it arenot also named in the Commission’s complaint.19 In its reply, the majority appears to suggest thatthe existence of a dealer cartel can be inferred fromthe allegation that ‘‘a dealer’s advisory councilvoted to advise American Cyanamid to retain theprogram in order to protect their margins.’’Statement of Chairman Robert Pitofsky andCommissioners Janet D. Steiger and Christine A.Varney in the Matter of American Cyanamid, atnote 5. Even if an advisory council furnished thisadvice to AmCy, communications of this naturebetween dealers and manufacturers do not establishthat the dealers acted collusively. Moreover, the factthat dealers may have communicated this advancesays nothing about the competitive effects ofAmCy’s rebate program. One would expect dealersto provide this same ‘‘advice’’ if AmCy’s programGiven that neither of the twotraditional anticompetitive theories canbe reconciled with the terms of theAmCy program, could the Commission’saction be justified on some other basis?The Commission might attempt to seekrefuge in some unilateral theory ofmarket power, under which amanufacturer with substantial preexistingmarket power is hypothesizedto use vertical restraints because, forsome reason, it cannot extract the fullvalue of its market power simply byraising its wholesale price. Theeconomics literature certainlyacknowledges such possibilities, butthese theories provide a fragile basis forantitrust enforcement. 20 As such modelsshow, vertical restraints often canimprove consumer welfare even whenadopted by firms with substantialmarket power; 21 the models fail,however, to provide empirical criteriaby which enforcers can distinguishanticompetitive from procompetitiveeffects. 22 Thus, the practical utility ofthese theories is questionable even forconduct judged under the rule of reason;their inability to justify a policy of perse illegality appears self-evident.On several grounds, therefore,acceptance of the consent agreement inthis matter represents a poor policychoice by the Commission. From a legalperspective, AmCy’s conduct does notconstitute an illegal agreement tomaintain resale prices; from aneconomic perspective, the evidencepoints to the conclusion that AmCy’sconduct was procompetitive; and from apolicy perspective, the Commission’sdecision hardly delineates a clearerdistinction (and in fact seriously blursthe line) between conduct likely to besubject to per se condemnation andconduct that is not. Instead of reachingfor ways to expand the application ofthe per se rule to conduct that is plainlyprocompetitive, enforcers shouldwere designed to prevent discounters from freeridingon the pre-sale services provided by otherdealers.20 See, e.g., Remarks of Commissioner Roscoe B.Starek, III, ‘‘Reinventing Antitrust Enforcement?Antitrust at the FTC in 1995 and Beyond,’’ beforea conference on ‘‘A New Age of AntitrustEnforcement: Antitrust in 1995’’ (Marina del Rey,California, Feb. 24, 1995).21 As I noted earlier (supra note 2), market poweris a necessary, but not a sufficient, condition forvertical restraints to reduce consumer welfare.22 As Katz (supra note 1, at 713–14) notes,‘‘[m]uch of the literature on vertical restraints hasbeen conducted with the express aim of derivingpolicy conclusions. But in many, if not most,instances there is no widespread agreement onwhether a particular vertical practice is sociallybeneficial or harmful. This unhappy state of affairsis due, in part, to the fact that all of the practicescan be beneficial in some instances and harmful inothers, and it may be extremely difficult todistinguish between the two cases.’’